(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "A treatise on masonry construction"

Google 



This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 



jvGooi^le 



jvGooi^le 



^*. 



jvGoo'i^le 



/y.C.l^a///'er. 



jvGooi^le 



jvGooi^le 



jvGooi^le 



WORKS OF PROF. I. 0. BAKER | 




PUBUSBED BY 


JOHN 


WILEY & SONS. 




Ntnih^'Edii 
600 paga, 


udHKH^TcMini Ihs Bsrlni fowa 
MuonrrSlnicliirc-Subiiilv AplfM 

(o Egim uid 6 [aJ^Di plate*, cloth. 


Editlan. R«viKd and Gieallf Enluied. umo, U + 


ATrMtlMVO 


^^^. . 





jvGooi^le 



A TREATISE 



MASONRY CONSTRUCTION. 



IBA 0. BAKEE, 0. E., D. Ehq'o, 

FBOTKSeOB OJ OiriL BNfllNBSiKISa, UKtVKBHITT Of ILLQIOIB. 



J/INTS BDITIOS, REVISED ASD PABTIALLT REWRITTSJU. 

TWELFTH THOUSAND. 



NEW VORK: 

JOHN WILEY & SONS. 

London: CHAPMAN ft HALL, LiumD. 

1906 



ovGoQi^lc 



mjt o. BAXOL 



ovGoQi^lc 



234481 

m 20 1320 t^o.3-icc. 

.- .•'^ 

PREFACE. 



The present Tolnme is an outgrowth of the needs of the snthor'B 
own claBB-room. The matter is eaeentially that presented to hi& 
classes for 3 namber of years past, a considerable part having been, 
used in the form of a blue-print manuscript text-book. It is now 
published for the greater convenience of his own stadents, and with 
the hope that it may be useful bo others. The author knows of no 
work which treats of any considerable part of the field covered by 
this volume. Nearly all of the matter is believed to be entirely new. 

The object has been to develop principles and methods and to 
give Boch esamples as illustrate them, rather than to accumulate 
details or to describe individual structures. The underlying prin- 
ciples of ordinary practice are explained ; and, where needed, ways 
are pointed out whereby it may be improved. The common theo- 
ries are compared with the results of actual practice ; and only 
those are recommended which have been verified by ezperimeDts 
or experience, since true theory and good practice are always 
in accord. The author has had the benefit of suggestions and 
advice from practical masons and engineers, and believes that the 
information here presented is reliable, and that the examples cited 
represent good practice. The general prices are the average of a 
large number actually paid ; and the special prices are representa- 
tive. The structures illustrated are actual ones. The accredited 
illustrations are from well-authenticated copies of working drawings, 
, and are presented without any modification whatever ; while those 
not accredited are representative of practice so common that a single 
name could not properly be attached. 

In the preparation of the book the endeavor has been to observe 
a logical order and a due proportion between different parts. Groat 
care has been taken in classifying and arranging the matter. It 
will be helpful to the reader to notice that the volume is divided 
successively into parts, chapters, articles, sections having small-cap- 
ital black-face side-heads, sections having lower-case black-face side- 
heads, sections having lower-case italic side-heads, and sections hav- 
ing simply the aerial number. In some cases the major snbdivis> 



ovGoQi^lc 



iona of the sections are indicated by small numerals. The constant 
aim has been to present the subject clearly and concisely. 

Every precantion has been taken to present the work in a form 
ior convenient practical use and ready reference. Nnmeroas cross 
references are given by section number; and whenever a figure or a 
table is mentioned, the citation is accompanied by the number of 
the page on which it may be foond. The table of contents shows 
the general scope of the book ; the running title assists in finding 
the different parts ; and a very full index makes everything in the 
book easy of access. There are also a number of helps for the 
gtudent, which the experienced teacher will not fail to recognize 
and appreciate. 

Although the book has been specially arranged for engineering 
and architectnral students, it is hoped that the information con- 
cerning the strengths of the materials, the data for focilitating the 
making of estimates, the plans, the tables of dimensions, and the 
costs of actual stmctures, will prove useful to the man of experience. 
'Considering the large amonnt of practical details presented and 
the great difference in the methods employed by various construc- 
tors, it is probable that practical men will find much to criticiaa 
The views here expressed are, however, the results of obserration 
thronghoot the entire country, and of consultation and correspond- 
ence with many prominent and practical men, and represent average 
good practice. The experienced engineer may possibly also feel 
that some subjects should have been treated more fully ; but it is 
neither wise nor possible to give in a single volume minute details. 
These belong to technical journals, proceedings of societies, and 
special reports of particular work. 

No pains have been spared in verifying data and checking re- 
sults. The tables of cubic contents have been computed by differ- 
ent proceeses by at least two persons, and to at least one more place 
than is recorded. Should any error, either of printer or author, 
be discovered — as is very possible in a work of so much detail, 
despite the greiit care used,— the writer will be greatly obliged by 
;prompt notification of the same. 

The author gratefully acknowledges his indebtedness to many 
■engineers for advice and data, and to his former pupil and present 
eo-laborer, Prof. A, N. Talbot, for many valuable su^estiona. 

OsAHFAiOK, III., Julj 9, 1889. 



ovGoQi^lc 



PREFACE rOR NINTH EDITION. 



Thb order of the snbdiTiBJons of Art. 3, Chap. I, has been 
changed, and pages 7-11 have been rewritten. Chapter III — Cement 
and Lime — and Chapter IV — Mortar and Concrete— have been 
entirely revrittea. Chapter IIIa — Sand, Gravel, and Broken 
Stone — has been added. The Definitions of Kinds of Masonry— 
pages 136 and 137 — have been rewritten and new illustrations hare 
been prepared. The specifications for the different classea of stou« 
masonry — paged 142, 144, and 147 — have been rewritten. Many 
minor changes have been made in varions parts of the book. 
CsAxrueif, lu., June 27, ISMi 



ovGoQi^lc 



jvGooi^le 



TABLE OF CONTENTS. 



f PART I. THE MATERIALS. 

CHAPTER I. MATUBAL STONE. 
IimiODCOTioii 1 

ABT. 1. REQinBITBS TOR GoOD BCILDINO Btohb. S 

Aki. 3. TxBTDio Bdildins Btoss. (L 

Waight -«. 

BardnMi and Tougkneu. T 

Strength. Crushing Strength. TransverM Slraogtb. Elaitlcltf . . & 
DurtMlitj/, Destructive Agents : mecbanlcal, cbemical. ResiBting 
Agenia : chemical composition, physicnl structure, Beasouing. Meth- 
ods of Testing Durability: absorptiTe power, methods and results; 
effect of frost, methods and results; eflect of atmoephere, methods 

uid results. Methods of Preserving 14 

^KT. S. Ci.AsaiFiCATioK AKD DKscRrt-noN of Bcildiko Stones. . . Si& 
Classlflcatlon : geological, chemical, physical. Description of Trap, 
Or(uiit«, Marble, Limsstose, and Sandstone. Location of Quarries. 
Weight of Stone. 

CHAPTER n. BRICK. 
Process of mannfacture. Classillcatlon. Requisites for good Brick. 

Uethods of Testing : absorbing power, tiaosverge strength, crashing 
Strength; results. Size. Cost SS. 

CHAPTER in. LIME AND CEMENT. 

CliABBIFlCATIOIt 48 

Abt. 1. CoMHOH Limb. 4IK 

Heibods of manufacturing, t«tlng, and preserving. Cost, 

Art. 2. Htdbaouc Limb. 61 

Art. 8. Htdraulig Cembnt. 61 

Description : Portland, Natural, Pozzuolana. Weight. Cost. 

Art. 4. Ubthods or Tstmna Htdbauwc Cekekt 0& 

Color, Thoroughness of Buniiug, Aclivllj, Soundness, Finsneas, 
Strength. 



ovGoQi^lc 



TABLE OP CONTENTS. 



Abt. 6. Speoifioatiohb tob Ceuent 07 

Quality : Oermau, EnglUb, French, Americui, Phllkdelphla- De- 
liverj Bud Stonge. 

CHAPTER niA. SAND, GRAVEL, AND BROKEN STONE. 

Art. ]. Band. ' 7t>a 

Requisites for Good Sand : Durabllitj, SharpDea*. Cleanaew, Fine- 
ness, Voids. Btone ScreeolDgs. Cost, Welgbl. 

Art. 2. Qbatbl and Bbokbn Stokb 7M 

Qrarel. Broken Stone. Voids. Weight. Cost. 



PART II. PREPARING AND USING THE MATERIALS. 

CHAPTER IV. MORTAR. CONCRETE, AND ABTinCIAL STONE. 

.Abt. 1 MoBTAR. 81 

Lime HoTtar. Cement Hortar : proportions and preparation. 
Data forEitlmal«8. Strength : tensile, compresBlve, adhesire. Cost. 
ESect of Re- tempering. Lime with Cemeot. Moitar Impervious to 
Water. Effect of Freezfng. 

Abt. 3. Coscretb lOO 

Hortar. Aggregate. Proportions : theoiy, determination, data 
forMttmateB Mixing. Laying. Strength. Coat. 

Abt. 8. Abtificial Stone 1136 

Portland. McMuriiie. Prear. Raosome. Borel. 

CHAPTER V. QUARRTINO. 
Methods of Quarrying ; by band tools ; by explosives, — the drills, 
the explosives ; by cbonneliDg and wedging. 116 

CHAPTEB VI. STONE CUTTINa. 

Abt. 1. Tools. 19S 

Eighteen hand tools illustrated and described. Machine tools de- 
scribed. 

Akt. i. Hbtbods of FoBifTNa thb Biirfaces 139 

Four metliods Illustrated and described, 

Abt. 8. Methods of Fiitibhino the Surfaces. 181 

Eight methods Illustrated and described. 

CHAPTER VII. STONE MASONBY. 
Deflultions: parts of tbewall, kinds of masonry. Ashlar Masonry: 
dressing, bond, backing, pointing, mortar required, when employed, 
specifications. Squnred-stone Masonry : description, mortar required. 
speciflcatjous. Hubble Ma.4onry : description, mortar required, when 
employed, apeciflcalions. Slope-wall Masoniy. Slone Paving. Rip- 



ovGoQi^lc 



? CONTENTS. 



np. SireDgth of Stooe MaBoury : examples, safe preesure. Ueas- 
uremeot of masour;. Cost: quarrjliig, dressing, price of stone; 
examples— U. 8. public buildlogs, rellroada. tmmela, bridge piers, 
arch culyerlB ; summary. 1 

CHAPTER Vin. BRICK MASONRY. 
Horlar. Bond. CompressiTe Strength : results of eiperimeniB, 
aafe pressure. TrHnaverse Strength : strain on liatcl. Ueasurement 
of Brick-work. Data for Estimates : brick, labor, mortar requii^. 
Cost. Speclfloitlons : for btiUdinifH, sewera, arches. Brick e*. Stone 
MasoDty. Brick Maaonry Impervious to Water. EfQoreBcence. . . 1 



PABT III. FOUNDATIONS. 
CHAPTER IX. INTRODUCTORY. 
DxmnTioirs, add Plah or Propobrd Dibcdbsion. . . . . li 
CHAPTER S. ORDINARY FOUNDATIONS. 

Odtlinx of Comtmtb. li 

Abt. I. Thb Boil 11 

Examioation of the Bite. Bearing power of Soils : rock, clay, sand, 
seml-liquld aoilii ; summnry. Methods of ImproTlng Bearing Power: 
increasing depth, drainage, springs, consolidating the Boll, sand piles, 
layen of sand. 

Abt. 2. DsmeinMO thk Poormos. V 

Load to be Supported. Area Required, Center of Pressure and 
Center of Base. Independent Piets. Effect of Wind. Offsets for 
Masonry Footings. Timber.F90tbigi. Sleel-r»11 Footings. Inverted 

AbT. 8. PHEFABIVa THE Bkd. 3 

On Rock. On Firm Earth. In Wet Qronnd : coffer-dam, con- 
crete, grillage. 

CHAPTER KI. PILE FOUNDATIONS. 

Dbtinitiorb. ... 9 

Abt. 1. DKSCRiPTtuiia. AHD Methods or Dritthb Piles. .... 3 
Deecriplion : Iron piles ; screw piles ; disk piles ; sheet piles ; bear 
ing piles,— speciHcations, caps and aboes, splicing. Pile Driving Ma- 
chines: drop iiaramer, — friction clutch, nipper; sleam-hainmer, drop- 
hammer w, steam-hammer ; gunpowder pile-drivers ; driving with 
dynamilc; driving with water Jet; jetM. hammer. Cost of Piles. 
Cost of Pile Driving: railroad construction, bridge construction, 
bridge repairs, foundations, harbor and river work. 



ovGoQi^lc 



TABLE OF C0HTEW18. 



Abt. 3. Bearido PowBtt o? Ptleb. 388 

Hetbods of DelerminiDg Supporting Power. Batlonal Formula. 
CompuisoD of Empiricul Formulas: BeAufoy'i, Nyslrom's, Moaon'S, 
Sander's, McAlpine'a, Traulwlne's. tbe Author's. SuppottlDg Power 
Determined by Experimeot ; examples, factor of eattty ; Bupporting 
power or screw and disk plles^ 

Abt. 8. Arrahoehbnt or thb Foundation. 350 

position of Piles. Bawing-off. Fialshing Foundation: piles and 
grillage, piles and concrete, lateral yielding. Cuablng's Pile Founda- 
aon. 

CHAPTER XII. FOUNDATIONS UNDER WATER. 
Difficulties to be OrBRCoxE. Oihtjnb or Contsntb. . . 367 

Abt. 1. The Coffbr-Dak Process 368 

Construction of tbe Dam. Leakage, pumps. Preparing the 
Foundation 

Art. S. The Crib abd Opeit Caibbok {"bocess 266 

Deflaitlona. PriDciple, Consiructlon of tbe Caisson. Construc- 
tion of the Crib. Excavating tne Site. 

Akt. 8, Drbdgino throooh Wells. 871 

Principle. Excavator, Noted Examples: Poughkeepale, Atcba- 
falaya,aud Hawhesbury bridges; brick cylinders. Frictional Reust- 
ance. Coat. • 

Abt. 4. Pnedhatic Frocebs STiS- 

Vacuum Process. Plenum Process. History. Pneumatic Hies, 
bearing power. Pneumalic CsissouH : the caisson, Ibe crib, thecoSer- 
dam, machinery, air-lock. Excavators; sand lift, mud-pump, water 
column, blasting. Rate of Sinking. Quidlng tbe Caisson. Noted 
Examples: Havre deOrace, Blair, St Louis, Brooklyn. Forth Bridges. 
Pbysiologicsl Effects of Compressed Air, Examples of Cost : at 
Havre de Grace, Blair, and Brooklyn, and in Europe. 

Art. 5. The FRBBziNa Process. 807 

PriDcipIe. History, Details of Process. Examples. Advantages. 

Art. 6. Compabisom of Methods. 809 



PART IV. MASONKY STKUCTURE8. 

CHAPTER Xm. MASONRY DAMS. 

Classification op Dams. 811 

Abt. 1. Stabilitv of Gbavitv Dams 3ia 

Principles. Stability against Sliding; destroying forces, resisting 
forces, co-etlicient of friction, condilion of equilibrium, factor of 
safety. Stability against Overturning : by moments. -^overturning mo- 



ovGoQi^lc 



TABLB OF COyrSKTS. 



ment, realattng moment, condltloa for equilibrium, factor of safety ; 
by resolution of forces. Btablllt; Bgalust Crushing : method of flnd* 
log maximum pressure, tenalou od masonry, limiting pressure. 

Abt. 3. Oirn.niB8 of the Desioii 834 

Width on Top. The Profile : theory, examples. The Plan ; 
straight crest e«. straight toe ; gravity t». arch dams ; curved gravity 
dams. Quality of Hasotuy. BlbIiogt«phy. 

Abt. 8. RocK-E^LL Dams. .^ 8Sl 

Wood. EartlL Bock-fill aad utasoniy dams compared. 

CHAPTER XIV. RETAIHISQ WALLS. 

Definltloni. Methods of Failure. Difficulties. 888 

Abt. 1. Theoketical PoBuuiiAS 84C 

The Three Assumptloiis. Theories: Coulomb's, Weyraach's, 

AsT. 3. EupiucAi, Rules US 

Bnglish Rules. American Rules. Details of Construction : quality 
of masonry, drainage, land ties, relieving archee. 

CHAPTER XV. BRIDGE ABUTMENTS. 

DUcussloD of General Forms. Quality of Masonry. Foundation. 

Wing Abutment, —design, and table of contents of various sizes. 

U -Abutment,— design and table of contents of varlotu sizes. T-Abut. 

ment, — design and table of contents of various sizes 808 

CHAPTER IVI. BRIDGE PIERS. 
Sblectioh or Sits and ABRAnaBMEirT or Sfaits. ... 866 

Art. 1. Thbobt of Stability 887 

Methods of fUlure. Stability against Sliding : effect of wind, cur- 
rent, ice; rcslstiug forces. Btabll it j against Overturning: by mo- 
ments ; by resolution of forces. Stability against Crushing. Example 
of method of computing stability. 

Abt. S. DsrAiLe or Corbtbuctioh 877 

Dimensions : on lop, at bottom. Batter. Cross Section. Spcclflca- 
tlons. Examples : (^ro. Grand Forks, Blair, Henderson, St. Croix 
River ; iron tubular ; wooden barrel. Tables of Contents of different 
styles and sizes of bridge piers. Specifications. 

CHAPTER XVn. CULVERTa 

Abt. 1. Water Wat Re4UIEEd 891 

The Factors. The Formulas : Meyer's, Talbot's. Pmctioal method 
of finding area of water way. 

Abt. 2. Box Ain> Pipe Culvxbts. 896 

Sfens Son Culvert : foundntlou, end walls, cover, spec! B cations. 



ovGoQi^lc 



TABLE 07 CONTENTS. 



ExuDples: Siuidard, West Shore, Cuudlon. Table ot Conteols and 
cosl of the TsriouB Btyle» bdcI sixea 8M 

VilTified Pipe Oalwrlt : CooBtructloa. Example. I^ble of Cod- 
tents 407 

Iron Pipe OuUierlt : ConBCriictlon. Blze and Weight of Pipe. Ex- 
unplw ; A., T. Ss B. F.. and C, B. Sc Q. slaodards. Tahle of Quan- 
tity of Materials Required 413 

Timber CtUf>erl : C, M. & St. P. standard box cnlverta. C, U. & 

Q. standard barrel culvert. 417 

Art. 8. Abcb CTn.TKRT 419 

Qeoenl Form : qjlay of wfng walls, joluing wings and body, seg- 
mental M, semi-circular. Examples : diagrams illuniatlng details, and 
also tables giving dimenalona, and contents, and cost, of all dzes of 
each of the standard forms of the Illinois Central. C. E. & N., A., T. 
&S.-F. (both semi-circular and segmental), and a standard form. 
Specifications. 

CHAPTER XVra. MASONRY ARCHES, 
Deflnitions: parts and kinds of arches; line of resistance. . . . . 440 

Art. 1. Thbori of tkb HASotniT Arch 444 

External Forces. Methods of Failure. Criteria ot Safety: sliding, 
rotation, crushlug, — unit, pressure, open joints. Location of Line of 
Resistance : hypothesis of least prossure ; hypothesis of least crown 
thrust, joint of rupture ; Winkler's hypothesis ; Navier's principle. 
Rational Theory of the Arch: symmetrical load,— two methods; 
unsymmetrical load ; criterion for line of resistance. Scheffler's 
Theory : two examples ; erroneous application ; reliability of. Ran- 
kine's Theory : curvature of linear aich, method of testing stability, 
reiiabillty. Other Theories. Theory of the Elastic Arch. StablUty 
ot Abutments and Piers. 

Art. 2. Rules Drrivkd frou Practick 4M 

Empirical Formulas : thickness of the arch at the crown,— Ameri' 
can, French. English practice; thicknessatthespringing,— American, 
French, English practice ; dimensions of abutments. Dimensions of 
Actual Arches and Abutments. Illustrations of Arches. Minor De- 
tails : backing, spandrel fliling. drainage. Brick Arches ; bond ; ex- 
amples,— tunnel, Philadelphia sewers, Washington sew era. Specifica- 
tions: stone arches, brick arches. 

Art. 6. Arch Ckntrrs. SU| 

Load to be supported, method of computing. Outline forms of 
Centera : solid rib, built rib, braced wooden rib, trussed frame. Ex- 
amples: centers for Vosburg tunnel, stone bridges, and Cabin John 
Arch. Striking Centers: method, time. 



ovGoQi^lc 



TABLE OF COHTZNIB, 



APPENDIX I. SPECIFICATIONS FOB MASONBY. 



Ofiiieral R&IIroad IKuoaty. , . 
HiMOUTj of Railroad Building*. . 
Architectural Haaonry 



APPENDIX II. SFPPLEMENTABY NOTES. 

Labor Bequired !d QuanrlDg. 544 

Coat of Cutting Qranite. MS 

Coat of Laying Cut Stone. C45 

Coat of Breaking Stone for Concrete 64$ 

Coit of Imboddlng Large Stones in Concrete. M7 

Cnublng SlrenglL of Sewer Pipe. 647 

Holding Power of Dilft-botta. 547 



ovGoQi^lc 



jvGooi^le- 



MASONBT C0NSTRU0TI02SI. 



DJTEODTJCnON. 

Ukdeb this geaeni head vill be diBoaBsed the snbjecta nlating 
to the use of atone and brick as employed by the en^neer or archi- 
tect in the conetraction of boildings, retaining walls, bridge piers, 
cnlverts, arches, etc., indnding the fonndations for the samo, 
For conTenience, the snhject will he divided as follows : 

Part I. Description and Characteristics of the Material*. 

Put IL Methods of Preparing and XTsing the Materiaia. 

Part III. Foundations. 

ftrt TV. MaeoQr; Structaree. 



ovGoQi^lc 



■* nu flnt ooat of muoniy Bhould be Its oaljr cceL ThonriiMipentiactuic* 
decay and drift nway, though emiMukmeDts abould crumble uid wvib ouL 
uuaamjr ahoold itiuia u one great mow ol loUd rock, firm and enduring.* 



ovGoQi^lc 



PART I. 

THE MATERIAIA 



NATURAL STONE. 

Art. 1. BBQUiBrrBS fob Good Buildino Stonb. 

1. The qnalities ThJoh are most important in stone uaed lor 
Donstrnction are cheapness, durability, strength, and beanty. 

2. CHlAmsB. The primary factor which detflrminee the Talne 
of a stone for strnctand purposes is its cheapness. The items which 
contribute to the cheapness of a stone are abundance, proximity of 
quarries to place of use, facility of transportation, and the ease with, 
which it is quarried and worked. 

The wide distribution and the great rariety of good building 
etone in this country are snch that suitable stone should everywhere 
be cheap. That such is not the case is probably due either to a 
lack of the developme'it of home resources or to a lack of con fidence 
in home products. The several State and Govenunent geological 
surreys have done much to increase oar knowledge of the building 
stones of this country. 

The lack of confidence in home resources has very frequently 
caused stones of demonstrated good quality to be carried far and 
wide, and frequently to be laid down upon the outcropping ledges 
of material in every way their equal. The first stone house erected 
in San Francisco, for example, was built of Btone bronght from 
China ; and at the present day the granites mostly employed there 
are bronght from New England or from Scotland, Yet then are 
no stones in onr country more to be recommended than the Califor- 
nia granites. Some of the prominent public and private buildings 
in Oinoiunati are oonetructed of stone that was carried by water and 



ov($OOi^lc 



i HATUBAI. BTOTTB. [CHAP. I. 

railway a distance of aboat 1500 miles. Withia 150 miles of Cin- 
cinnati, in the snb-carboniferouB limestone district of Kentncky, 
there are very extensive deposits of dolomitic limestone that aSord 
a beaatifnl building stone, which can be quarried at no more ex< 
pense than that of the granite of Maine. Moreover, this dolomitfi 
is easily carved, and requires not more than one third the labor to 
give it a surface that is needed by granite. Experience has shown 
that the endurance of this stone under the influence of weather is 
very great ; yet because it has lacked authoritative indorsement 
there has been little market for it, and lack of confidence in it bat 
led to the transportation half-way across the continent of a stent 
little, if any, superior to it. 

Development of local resources follows in the wake of good in- 
formation concerning them, for the lack of confidence in home prod- 
ucts can not be attributed to prejudice. 

The facility with which a stone may be quarried and worked is 
an element affecting cheapness. To be cheaply worked, a stone 
must uot only be as soft as durability will allow, but it should have, 
no flaws, knots, or b&rd crystals. 

3. I>tnABI£ITT. Next in importance after cheapness is dura- 
bility. Bock is supposed to be the typo of all that is unchangeable 
and lasting ; but the truth is that, unless a stone is suited to the 
conditions in which it is placed, there are few substances more liable 
to decay and utter failure. The durability of stone is a subject 
upon which there is very little reliable knowledge. The question 
of endurance under the action of weather and other forces can not 
be readily determined. The external aspect of th§ stone may fail 
to give any clue to it ; nor can all the testa we yet know determine 
to a certainty, in the laboratory, just how a given rock will with- 
stand the effect of onr variable climate and the gases of our cities. 
If our land were what is known as a rainless country, and if the 
temperature were uniform throughout the year, the selection of a 
durable building stone would be much simplified. The cities o' 
northern Europe are full of failures in the stones of importan 
structures. The most costly building erected in modem times, per- 
haps the most costly edifice reared since the Great Pyramid, — tht 
Parliament House in London, — was built of a stone taken on tfat 
recommendation of a committee representing the best scientific and 
tochDical skill of Great Britain, The stone selected woa submitted 



ovGoQi^lc 



ABT. 2.J TESTS UP BUILDmrS STONES, 5 

torarioQs teste, but the corroding inflnence of a Ijondonatmc^phere 
was overlooked. The great strnctare was built, and now it seems 
questionable whether it can be made to eodure oe long as a timber 
building would stand, so great is the effect of the gases of the 
atmosphere upon the stone. This is only one of the numerous in- 
stances that might be cited in which a neglect to consider the 
climatic conditions of a particular locality in selecting a building 
material has proved disastrous, 

"The great difference which may exist in the durability of stones- 
ol the same kind, presenting little difference in appearance, is 
strikingly exemplified at Oxford, England, where Christ Church 
Cathedral, built in the twelfth or thirteenth century of oolite from 
a quarry about fifteen miles away, is in good preservation, while 
many colleges only two or three centuries old, built also of oolite 
from a quarry in the neighborhood of Oxford, are rapidly cmmbling 
to pieces. " * 

4. Stbevoth. The strength of stone is in some instances a 
cardinal quality, aa when it is to form piers or columns to support 
great weights, or capstones that span considerable intervals. It is 
also an indispensable attribute of stone that is to be exposed to 
mechanical violence or unusual wear, as in steps, lintels, door- jambs, 
e*c. 

5. Beautt. This element is of more importance to the archi- 
tect thau to , the engineer; and yet the latter can not afford to 
neglect entirely the element of beauty in the design of his most 
utilitarian structures. The stone should have a durable and pleas- 
ing color. 

Art, 3. Tests of the Quautt of Bitildikg Stones. 
9. As a general mie, the densest, hardest, and most uniform 
stone will most nearly meet the preceding requisites for a good 
building stone. The fitness of stone for structural purposes can bt 
determined approximately by examining a fresh fracture. It should 
be bright, clean, and sharp, without loose grains, and free from any 
dull, earthy appearance. The stone should contain no " drys," i.e., 
seams containing material not tlioroughly cemented together, nor 
"crow-foots," i.e., veins containing dark-colored, uncemented 
material. 

* BauUne's CItQ Engineering, p. 862. 



ovGoQi^lc 



6 NATURAL STONE. [OHAP. I. 

The more form&t tests employed to determine the qaalitiee of a 
building stone are: (1) veight or density, (S) hardness and tough- 
neaa, (3) strength, (4) darability. 

1. Wmght of Stone. 

7. Weight or dennty is an important property, since npon it 
depends to a large extent the strength and dnrabilit; of the stone. 

If it is desired to find the exact weight per cabic foot of a given 
stone, it is generally easier to find its specific gravity first, and then 
mnltiply by 62.4, — weight, in pounds, of a cabio foot of water. 
Tjiis method obriates, on the one hand, the expense of dressing a 
sample to regular dimensions, or, on the other, hand, the in- 
aocnracy of meosaring a roagh, irregalar piece. Kotice, howcTer, 
that this method determines the weight of a onbic foot of the solid 
stone, which will be more than the weight of a cubic foot of the 
material as used for stractaral parposes. In finding the specific 
gravity there is some difficnity in getting the correct displacement 
of porona stones, — and all stones are more or less poruns. There 
are varione methods of overcoming this difficulty, which give 
slightly diSerent resnlta. The following method, recommended by 
General Gillmore, is most frequently nsed: 

All loose grains and sharp comers haying been removed from 
the sample and its weight taken, it is immersed in water and 
weighed there after all babbling has ceased. It is then taken ont 
of tiie water, and, after being compressed lightly in bibalons paper 
to absorb the water on its surface, is weighed again. The specific 
gravity is found by dividing the weight of the dry stone by the 
difference between the weight of the saturated stone in air and in 
water. Or expressing this in s formnla, 

W 
Specific gravity = ^^ _^ ^ , 

in which W^ representfl the weight of dry stone in air, W, the 
weight of saturated stone in air, If, the weight of stone immersed 
in water. 

The following table containa the weight of the atones most fre- 
qnently met with. 



ovGoQi^lc 



iBT. 2.] 



TZaXS OF BriLOING BTOKEB. 

TABLE 1. 
Wbioht of BmLDisG Btokbb. 





PODMM FEB CUHIO 


Foot. 




lUn. 


Hu. 


Mod. 




161 
14« 
167 

1B7 
160 


178 
174 
180 
161 
176 










170 






174 





2. HaBDHEBB AHD TorOHNESS. 

8. The apparent hardness of a etone dependB upon (1) the 
. hardness of its component minerals and (2) their state of aggrega- 
tion. The hardness of the component minerals is determined by 
the resistance they offer to being scratched ; and Tariss from that 
of talc which can easily be scratched with the thnmb-nail, to that 
: of qnartz which scratches glass. Bat however hard tiie mineral 
constituents of a stotie are, the apparent hardness of the stone itself 
depends npon the state of aggregation of the particles. Many 
rocks composed of hard materials work readily, because their grains 
are loosely coherent; while others composed of softer materials are 
qnite toagh and difiBcalt to work, owing to the tenacity with which 
the particles adhere to each other. Obviously a stone in which the 
grains adhere closely and strongly one to another will be stronger 
and more dnrable than one which is loose textured and friable. 

The toughness of a stone depends upon the force with which the 
particles of the component minerals are held together. 
' Both hardness and toughness should exist in a stone used for 
fltoops, pavements, road-metal, the facing of piers, etc. Ko experi- 
ments have been made in this country to test the resisting power of 
atone when exposed to the different kinds of service. A table of the 
resistance of stones to abrasion is often quoted,* but as it contains 
only foreign stones, which are described by local names, it is not of 
much value. 

* For enunple, Habui's Olvtl Engineering, p. IS. 



ovGoQi^lc 



8 KATUR1.L 8T0NB. [CHAP. I. 

3. STHSNarH. 

TTnder thia head will be included (1) ornihing or compreaaifd 
Btreogth, (2) transYerse strength, (3) elasticity. Usually, when 
simply the strength la referred to, the crashing strength is intended. 

9. CBTTSHnro Stbevoth. The crushing strength of a stone is 
tested by applying measured force to cubes nntil they are cmahed. 
The resnlts for the crushing strength vary greatly with the details 
of the experiments. Several points, which should not be neglected 
either in planning a series of experiments or in using the results 
obtained by experiment, will be taken up separately, although they 
are not entirely independent. 

10. Form of Test Specimen. Experiments show that all brittle 
materials when subjected to a oompresBive load fail by shearing on 
certain definite angles. For brick or stone, the plane of rupture 
makes an angle of about 60° with the direction of the compressing 
force. For this reason, the theoretically best form of test specimen 
would be a prism having a height of about one and a halt times the 
least lateral dimension. The result is not materially different if 
the height is three or four times the least lateral dimenBion. But 
if the test specimen is broader than high, the material is not free 
to fail along the above plane of rupture, and conseqaently the 
strength per unit of bed-area is greater than when the height is 
greater than the breadth. 

However, notwithstanding the fact that theoretically the teat 
specimen should be higher than broad, it is quite the universal 
custom t« determine the crushing strength of stone by testing 
cubes. 

11. Blie of the Gabe. Although the cube is the form of test 
specimen generally adopted, there is not equal unanimity as to the 
size of the cube; but it is oonclnsiTely proven that the strength per 
square inch of bed-area is independent of the size of the cube, and 
therefore the size of the test specimen is immaterial. 

General Gillmore, in 1875, made two sets of experiments which 
seems to prove that the relation between the crushing stongth and 
the size of the cube can be expressed by the formaU 

ID which y is the total crushing pressure in pounds per square inch 




ovGoQi^lc 



ABT. 2.] STBBNQTH OB BniLDINQ BTONES. ft 

of bed-oreft, a is the crashlDg prearare of a l-inch cnbe of the Bome- 
material, and x is the length Id inches of an edge of the cnbe under 
trial. For tvo sampleB of Berea (Ohio) aandatone, a was 7000 and 
flSOO IbB., respectively.* 

Besnlta by other obaervers with better machines, particularly by 
General Gillmore f with the large and accurate testing-machine at 
Watertown (Masa.) Arsenal, X uniformly show this sappoaed law to 
be without any foundation. Unfortunately the above relation 
between strength and bed-area ie frequently quoted, and has found 
a wide acceptance among engineers and architects. 

Two inches is the moat common size of the cube for oompression 
tests. 

12. Cushions. Homogeneous stones in small cubes appear in 
all cases to break as shown in Fig. 1. The 
forms of the fragments a and b are, approxi- 
mately, either conical or pyramidal. The 
more or less disk-shaped pieces e and d are 
detached from the sides of the cnbe with a 
kind of explosion. In the angles « and /, the 
stone ia generally found crushed and ground 
into powder. This general form of breakage 
occurs also in non-homogeneons stones when ?">■ i- 

crushed on their beds, but in this case the modification which the- 
grain of the stone produces mnst be taken into account. 

The nature of the material in contact with the stone while- 
nnder pressure is a matter of great moment. If the materials which 
press upon the top and bottom of the specimen are soft and yielding 
and press out sidewise, they introduce horizontal forces which 
materially diminish the apparent crnshing strength of the stone. 
If the pressing surfaces are hard and unyielding, the resistance of 
these surfaces adds considerable to the apparent strength. 

■ Report on Btrengtb of Batlding Stona, Appendix, Baport of Obltt ot EngiDeeta- 
of U. 8. A. (or 187t 

tNotos OD the ComprsBBlTe BeBlstanoe at Ftaestons, Brlok Plan, HydiauUa 
Caments, H ort&rs, and Oonaret«, Q. A. OlUmore. John Wile; A Sons, New York, 

ises. 

{Report OD the " Teste of Metals," etc., tor tha year ending JaoeSO, ISSi, pp. 
Ue, 166, IST, 197, 312, 918, 316; the aama being Sen. Ex. Doo. No. 85, 49th Cong., 
Xet SesBloD. For a dlseosBlon at Uiese data by the antboT, see ^igtauriag JfrnM, 
VoLzlz.pp. Sll-SU. 




ovGoQi^lc 



10 NATURAL STOKE. [CHAP. L 

Formerly steel, wood, lead, and leather were mnch nsed oa 
pressing snifaces. Under certain limitations, the relative crashing 
strengths of atones with these different pressing sarfacee are 100, 
89, 65, and 62 respectively.* 

Testa of the strength of blocks of stone are nsefal only tn com- 
paring different stones, and give no idea of the strei^gth of strnct- 
nres bnilt of snch stone (see § 246) or of the crnabing strength of 
etone in large masses in its native bed (see § 273). 

Then, since it is not possible to have the stone under the same 
conditions while being tested that it is in the actual atroctnre, it is 
best to test tbe stone under conditions that can be accnrately 
deacribed and readily duplicated. Therefore it ia rapidly coming to 
be tbe custom to test^ the stone between metal pres^ng surfaces. 
Under these conditions the strength of tbe specimen will vary greatly 
with tbe degree of smootfaneBS of its bed-anid!aceB. Hence, to obtain 
definite and precise resnlts, these enrfaces abootd be robbed or 
gronnd perfectly smooth; bat as this is tedions and expensive, it is 
qnite common to rednce tbe bed-snrfaces to planes by plastering 
them with a thin coat of plaster of Paris. With the stronger 
etoDea, specimens with plastered bed will show less strength than 
those having rubbed beds, and this difference will vary also with 
the length of time tbe plaster is allowed to harden. With a 
stone having a atrengtb of 5,000 to 6,000 pounds per square inch, 
allowing the plaster to attain its maximnm strength, this differ- 
ence varied from 5 to 30 per cent., the mean for ten trials being 
almost 10 per cent, of the strength of the specimen with rubbed 
beds. 

13. Sreuing the Cube. It is well known that even large 
atones can be broken by striking a number of comparatively light 
blows along an; particular line ; in which case tbe force of the blows 
gradu^Iy weakens tbe cohesion of tbe particles. This principle finds 
application in the preparation of teat specimens of atone. If the 
specimen ia dressed by hand, the coDcnssion of the tool greatly 
affects ita internal conditions, particnlarly with test specimens of 
small dimensions. With 3-incb onbea, t^e tool-dreseed specimen 
uaually shows only abont 60 per cent, of the strength of the sawed 



ovGoQi^lc 



»■] 



STRENGTH OF UUILDINQ STONES. 



11 



.sample. The sawed Bample most nearly represente the conditions 
•of ftctaal practice. 

Unfortuaatelj, experimenters seldom state whether the 
^specimens were tool-dressed or sawed. The disintegrating 
'effect of the tool in dressiog ia greater with small than with 
large specimens. This may acconnt for the difference in 
:strength of different sizes of test Bpeoimen as seems to be 
jhown by some experiments. 

AU stones are strongest when laid on their natural bed, 
i.«., when the pressure is perpendicular to the stratification ; • 
■and with sedimentary rocka there is a very great difference 
in the two positions. Hence, in preparing the test specimen 
the natural bed should be marked, and the cube should be 
-tested upon its native bed. 

14. Data en Cmihlng Strength. The strength of the principal 
classes of building stone in use in the United States is about as 
iollowa : 

TABLE L 



CBUBKiMa Stekhoth 


or UuBsa o> 


Stoitb. 








Knna or Snn. 


Pounds per Bqiuralncb. 


TootpwSq 


ivafooL 




Ko. 


1U< 


MIo.. 


Mmx. 


Trap Bocki of N. J 


20,000 
18,000 
8,000 
7,000 
5,000 


£4,000 
31.000 
30,000 
30.000 
15.000 


1,440 
860 
580 
500 
860 


1,780 
1,010 
1,440 








1,060 





16. Cruihing Strength of Slabe. Only a few experiments have 
been made to determine the crushing strength of slabs of stone. 
The strength per square inch of bed-snr^e was considerably 
greater than that for cubes ; but a study of the results of all of the 
reliable experiments • fails to discover any simple relation between 



" tor ISU—aeo. Ex. Doo. No. 35, 49th 



ovGoQi^lc 



12 NATDEAL STOITB. [CHAP. L 

the oriisliing etrength of cubes aad dabs. It is probable that the 
effect of the precming sarface is 80 great as to completely maek the 
TariatioQ due to height of Bpecimeu. More experiments on this 
flnbject are very much needed. 

16. TSAXSVIBSE Strxvoth. When stones are used tor lintels, 
etc, their transverse strength becomes important. The ability of a 
stone to resist as a beam depends upon its tensile strength, since 
that is always mnch less than its compressive strength. A knowl- 
edge of the relative tensile and compreaaive strength of stones is 
valnable in interpreting the effect of different pressing surfaces in 
compreasiye tests, and also in determining the thickness required 
for lintels, sidewalks, cover-stones for box culverts, thickness o£ 
footing courses, etc. 

Owing to the smalt cross section of the specimen employed in 
determining the transverse strength of stones, — usually a bar 1 inch 
square, — the manner of dressing the sample affects the apparent 
transverse strength to a greater degree than the compressive strength' 
(see § 13); and it is even more unfortunate, since the strength of 
the stone as used in actual practice is nearly proportional to the 
strength of sawed samples. 

The following formulas are useful in computing the breaking 
load of a slab of stone. Let W represent the concentrated center 
load plus half of the weight of the beam itself, in pounds ; and let 
/>, d, and I represent the breadth, depth, and length, in inches, 
respectively. Let R = the modulus of rupture, in lbs. per sq. in. ;, 
let C =■ the weight, in pounds, required to break a bar 1 inch 
square and 1 foot long between bearings ; and let L = the lengthi 
of the beam in feet. Then 

The equivalent unifomdy distributed weight is equal to twice the- 
concentrated center load. 

Table 3 on the following page gives the values of R, the mod- 
ulus of rupture, and of C, the co-efficient of transverse strength, 
required in the above formulas. 

Example, — To illustrate the method of using the above formulas, 
assume that it is desired to know the breaking load for a limestone 
Blab 3 inches thick, 4 feet wide, and 6 feet long. Then d = 48 ;. 



ovGoQi^lc 



ABT. 2.] 



emXSQ^ OF BUILDINO STONES. 



TABLE 8. 
TbjUibtbbsb Strxkoth ot Stohx, Brick, and Uobtas. 



HooDura or Buftcbb. 



Blue^tone flagging 

Granite 

Limestone 

" oolitic, from Ind., sawed. 
Jfarble 



Slate 

Brick (§69). 

Concrete— see g Iff^. 

Mortar, neat Portland. 1 jeaj old. 
Uortar, 1 part Portland cement, 1 

part sand. 1 jear old 

Hortar, 1 part Portland cement, 3 

parts sand, 1 ^ear old 

Uortar, neat Rosendale, 1 year 
Hortar, 1 part Rosendale cement, 

1 part sand, 1 year old 

Uortar, 1 part Rosendale cement, 

3 parts sand, 1 year old ... . 



860 


3,700 


900 


1,800 


140 


1,500 


2,190 


2,888 


144 


2.160 


676 


1.260 


1,800 


5.400 


M9 


800 




1.168* 




»46» 




e88» 


416 


600 


848 


586 


888 


406 



(( = 3 ; / = 73 ; R = 1500 lbs.,— the "averse" value from the 
tahle; — and C = 83. Substituting these values, we have 

W=ll^R = L|i^X_9i500 = 6000 poundfl; 

or, using the other form, 

W= ^C = ^^^3 = 5976 pounds, 

vrhich agrees with the preceding except for omitted decimals. 
Hence the breaking toad for average quality of limestone is 6000 
ponnds concentrated along a line half-waj between the ends ; the 
QnLformly distributed load is twice this, or 13,000 pounds. The 



D.qitizeabyG00l^lc 



14 HATi/RAL 8T0NB. [CHAP I, 

question of what margin ahoald ba allowed for safety is one that caix 
not be determined in the abstract ; it depends upon the accnracy 
with which the maximum load is estimated, upon the manner the 
load is applied — whether with shock or not, — upon the care with 
which the stone was selected, etc This subject will be discussed 
farther in connection with the use of the data of the above table in 
subsequent parts of this volume. 

17, Elasticitt. But very few experiments have been made to 
determine the co-efficient of elasticity, the elastic limit, and the 
" set " of stone. Data on these points would be valuable in deter- 
mining the effect of combining masonry and metal, of joining 
different kinds of masonry, or of joining new masonry to old ; in 
calculating the effect of loading a masonry arch ; in proport)'>ning 
abutments and piers of railroad bridges subject to shock, etc 
The following is all the data that can be found: 
TABLE 4. 
Co-KmcisNT 07 Blabticitt of Stonb, Brick, asd Uobtab. 



Haverstrew Freestone * 

Portland &toue (oClite limeBtoDe)! 

Martlet 

Portland GrantteJ. 

Blatef 

Graflon Limestone} 

Richmond Qranitet 

Brick, medium— mesD of IS experiments* 

Louisville Cement Mortar, 4 months old : t 

Neat cemenl 

1 part cement, 1 part sand. 

I part cement, 3 parts sand 

Ulster Co. (N. Y.) Cement Uortar, S3 monthe 
old:" 

8 parts cement, 8 porta sand 

1 part cement, 8 parts sand 

hvtland Cement Mortar, 32 months old* 



950,000 
1,680,000 
8,600,000 
S,G00,O00 
7,000,000 
8,000,000 
18,000,000 
8,600,000 

800,000 

600.000 

1,800,00P 



eu.oop 

686.000 



1, as quoted hj Stonej. 



ovGoQi^lc 



ART. i.} BTBBNOTH OF BUILDINQ 8T01TBS. 16 

18. BiBLiaeEAPHIOAI. A large nnmber of testa bave been 
applied to the boilding stones of tbe United States. For the 
results and detuls of some of the more important of these tests 
see: Report on Strength of Sailding Stoae, den. Q. A. Gillmore, 
Appen. II, Beport of Chief of Engineers, U. S. A., for 1875; 
Tenth Ceosna of the XS. 8., Vol. X, Beport on the Qnarr; 
IndnBtry, pp. 330-35; the several annnal reports of tests made 
with the V. S. Ooverament testing machine at the Watertown 
(Mass.) Arsenal, pablished by the U. S. War Department under 
the title Report on Tests of Metals and Other Materials; 
Transactions of the American Society of Civil Engineers, Vol. II, 
pp. 146-Sl and pp. 187-9^; Jourital of the Association of En- 
gineering Societies, Vol. V, pp. 176-79, Vol. IX, pp. 33-43; 
Engineering Jfews, Vol. XXXI, p. 135 (Fe£. 16, 1884); and the 
reports of the various State Geological Sarveys, and the com- 
missioners of the varions State capitols and of other pnblio 
buildings. 

By vay of comparison the following reports of tests of building 
stones of Great Britain may be interesting: Prooeedinga of the 
Inatitnte of Civil Engineers, Vol. CVII (1891-98), pp. 341-69; 
abstract of the above, Engineerinff Mws, Vol. XXVIII, pp. 279-82 
(Sept. 22, 1892). 

In consnlting the above references or in nsing the results, the 
details of the manner of making of the experiments shonld be 
kept clearly in mind, particalarly the method of pieparing and. 
bedding the specimen, 

4. DUKABILTTT. 

19. *' Although the art of bnilding has been practiced from the 
earliest times, and constant demands have been made in every age 
for the means of determining the best matenals, yet the process of 
ascertaining the durability of stone appears to have received but 
little definite scientific attention, and the processes usually employed 
for solving this question are still in a very unsatisfactory state. 
Hardly any department of technical science is so much neglected as 
that which embraces the study of the nature of stone, and all the 
raried resources of litbology in chemical, microscopical, and physical 
methods of investigation, woDderfuUy developed within the last 



ovGoQi^lc 



t6 NATURAL STONE. [CHAP. I 

quarter century, have never yet been properly applied to the aeleo- 
tion and protection of atone used for building purposeB." * 

Examples of the rapid decay of building atones have already been 
referred to, and numerouB others oonld be cited, in vhich a stone 
vbich it was snppoeed would last forever has abready begun to 
decay. In every way, the question of durability is of more interest 
to the architect than to the engineer ; although it is of enough 
importance to the latter to warrant a brief discussion here. 

20. Sebtaiictite Aoehtb. The destructive i^ents may be clas- 
sified as mechanical, chemical, and organic. The last are onim- 
portont, and will not be considered here. 

' 81. Mechanical A^ots. For our climate the mechanical agents 
are the most efficient. These are frost, wind, rain, fire, pressure, 
and friction. 

The action of frost is usually one of the main causes of rapid 
decay. Two elements are involved, — the friability of the material 
«nd its power of absorbing moisture. In addition to the alter- 
nate freezing and thawing, the constant variations of temperature 
from day to day, and even from hour to hour, give rise to molecular 
motions which affect the durability of stone as a building material. 
This effect is greatest in isolated columns, — as monuments, bridge 
piers, etc. 

The effect of rain depends upon the solvent action of the gases 
which it contains, and upon its mechanical eSect in the wear of 
pattering drops and streams triclding down the face of the wall. 

A gentle breeze dries out the moisture of a building stone and 
tends 'to preserve it; but a violent wind wears it away by dashing 
sand grains, street dust, ice particles, etc., against its face. The 
extreme of such action is illustrated by the vast erosion of the sand- 
stone in the plateaus of Colorado, Arizona, etc., into tabular miias, 
isolated pillars, and groteequely-shaped hills, by the erosive force of 
sand grains borne by the winds. The effect is similar to that of the 
sand blast as used in various processes of manufacture. A violent 
wind also forces the rain-water, with all the corrosive acids it con- 
tains, into the pores of stones, and carries off the loosened grains, 
thus keeping a fresh surface of the stone exposed. Again, the 
swaying of tall edifices by the wind must cause a continual motion, 

■ Tamil Ceanu o( the U. S., Tot. X, Bepcet on tbe Qomtt lD^utEr,Tb Ml 



ovGoQi^lc 



ART. 3.] DUaABlLITT OF BUILDINO STOSES. 17 

not only in tlie joints between tlie blocks, bat among the grains of 
the stoneB themaelrea. Many of these have a certain degree o( 
fiexibility, it is true; and yet the play of the grains mnat gradually 
increase, and a tendency to disintegration result. 

Experience in great fires in the cities shows that there is no 
stone which can withstand the fierce heat o£ a maes of burning 
buildiogs. Sandstones seem to be the least aSected by great heat, 
and granite most. 

Frictiou aBects sidewalks, pavements, etc., and has already 
been referred tr> (§ 8). It may also aSect bridge piers, sea-walls, 
docks, etc. 

The efiect of pressure in destroying stone is one of the Ic^t 
importance, provided the load to be borne does not too nearly eqnnl 
the crushing strength. The pressure to which stone is subjected 
does not generally exceed one tenth of the ultimate strength as 
determined by methods already described. 

22. Chemical Agents. The principal ones are acids. Every 
constituent of stone, except quartz, is subject to attack by acids; 
and the carbonates, which enter as chief constituents or as cement- 
ing materials, yield very readily to such action. Oxygen and am- 
monia by their chemical action tend to destroy stones. In cities or 
manufacturing districts sulphur acids and carbonic acid hare a 
very marked effect. These all result from the combustion of gas, 
coal, etc., and some are also the residuary gases of many kinds of 
manufactories. The nitric acid in tbe rain and the atmosphere 
exerts a perceptible infinence in destroying building stone. 

23. BsBiBTHro Abektb. The durability of a building stone de- 
pends upon three conditions; viz., the chemical and mineralogical 
nature of its constituents, its physical structure, and the character 
and position of its exposed surfaces. 

24. Chemical Composition. The chemical composition of the 
principal constituent mineral and of the cementing material has an 
important effect upon the durability of a etone. 

A siliceous stone, other things being equal, is more durable than 
a limestone; but the durability of the former plainly depends upon 
the state of aggregation of the individual grains and their cement- 
ing bond, ae well as on the chemical relation of the silica to the 
other chemical ingredients. A dolomitic limestone is more durable 
thkn a pure limestone. 



ovGoQi^lc 



18 KATCBAL STONE. [CHAP. T. 

A stone that absorbs moisture abundantly and rapidly is apt to 
be iajttred by alternate freezing and thawing; hence clayey conetit- 
iients are injnrions. An argillaceous stone is generally compact, 
and often baa no pores visible to the eye; yet such will disintegrate 
rapidly either by freezing and thawing, or by corrosive vapors. 

The presence of calcium carbonate, as in some forms of marble 
and in earthy limestones, renders a building material liable to rapid 
attack by acid vapors. In some sandstones the cementing material 
is the hydrated form of ferric oxide, which is solnble and easily 
removed. Sandstones in which the cementing material is siliceous 
are likely to be the most durable, although they are not so easily 
worked as the former, A stone that hae a high per cent, of alumina 
(if it be also non-cry stalUne), or of organic matter, or of protoxide 
of iron, will usually disintegrate rapidly. Such stones are gen- 
erally of a bluish color. 

26. Seasoning. The thorough drying of a atone before, and the 
preservation of this drynesa after, its insertion in masonry are com- 
monly recognized as important factors of its durability; bnt the 
exact nature of the process of seasoning, and of the composition 
of the quarry-sap removed by thorough drying, have never been 
determined. The quarry water may contain little else than ordinary 
well-water, or may bo a solution more or less nearly saturated, at the 
ordinary temperature, with carbonate of calcium, silica, double salts 
of calcium and magnesium, etc. In the latter case, hardening re- 
sults from the drying, and an exact knowledge of its nature might 
throw important light on the best means for the artificial preserva- 
tion of stone. Again, water may exist in large quantity, in chemical 
combination, in the silicates {e.g., chlorite, kaolin, etc.), or in the 
hydrated iron oxides which constitute the cement of a building 
stone. 

26. Physical StrnotiiTe. The physical properties which con- 
tribute to durability are hardness, toughness, homogeneity, con- 
tiguity of the grains, and the structure — whether crystalline or 
amorphous. 

Although hardness (resistance to crushing) is often regarded as 
■the moat important element, yet resistance to weathering does not 
necessarily depend upon hardness alone, bnt upon hardness and the 
non-absorbent properties of the store. A hard material of close 
and firm texture is, however, in those qualities at least, especially 



ovGoQi^lc 



AllT. 2.J BURaBILITT: op BUILDIKG S10KE8. 19 

fitted to resist friction, as in stoops, pavements, iind road metal, and 
the wear of rain-drope, dripping rain-water, the blows of the waves, 
etc 

Porosity is an objectionable element. An excessive porosity in- 
creases the layer of decomposition which is caused by the acids of 
the atmosphere and of the rain, and also deepens the penetration of 
frost and promotes its work of disintegration. 

If the constituents of a rock differ greatly in hardnese, texture^ 
solubility, porosity, etc., the weathering is unequal, the surface ia 
roughened, and the sensibility of the stone to the action of frost is. 
increased. 

The principle which obtains in applying an artificial cement, 
snch aaglue, in the thinnest film in order to secure the greatest 
binding force finds its analogy in the building stones. The thinner 
the films of the natural cement and the closer the grains of the pre- 
dominant minerals, the stronger and more durable the stone. One- 
source of weakness in the .famous brown-stone of Hew York City- 
lies in the separation of the rounded grains of quartz and feldspar 
by a superabundance of ocherous cement. Of course the further 
separation prodaced by fissure, looseness of lamination, empty 
cavities and geodea, 'and excess of mica tends to deteriorate still 
further a weak building stone. 

Experience has generally shown that a crystalline structure re- 
sists atmospheric attack better than an amorphous one, This prin- 
ciple has been abundantly illustrated in the buildings of New York 
City. The same fact is generally true with the sedimentary rocks 
also, a crystalline limestone or good marble I'esisting erosion better 
than earthy limestone. A stone that is compactly and finely granular 
will exfoliate more easily by freezing and thawing than one that is 
coarse-grained. A stone that is laminar in structure absorbs mois- 
ture unequally and will be seriously affected by unequal expansion 
and contraction, — especially by freezing and thawing. Such a stone 
will gradually separate into sheets. A stone that has a granular 
texture, as contrasted with one that is crystalline or fibrous, will 
crumble sooner by frost and by chemical agents, because of the 
easy dislodgment of the individual grains. 

The condition of the surface, whether rough or polished, in- 
fluences the durability, — the smoother surface being the better. 



ovGoQi^lc 



20 NiXUBAL 6T0KB. [CHAP. I. 

The stone is more durable if the exposed Burfoce is vertical than if 
mclmed. The lamination of the stone should be horizontal. 

27. HZTEODS OF TESTISO StTBABILITT. It hus long been recog- 
nized that there are two ways in which we can 'form a' Judgment of 
the durability of a building stone, and these may be distinguished 
as natural and artificial. 

28. Natural Hethoda. These must always take the precedence 
whenever they can be used, becai^e they iuTolve (1) the exaot 
ftgenciea concerned in the atmospheric attack npon stone, and (S) 
Jong periods of time &r beyond the reach of artificial experiment. 

One method is to visit the quarry and observe whether the ledges 
that have been exposed to the weather are deejdy corroded, or 
whether these old surfaces are still fresh. In applying this test, 
consideration must be given to the modifying effect of geological 
phenomena. It has been pointed out that "the length of time the 
ledges hare been exposed, and the changes of actions to which 
they may have been subjected during long geological periods, are 
unknown; and since different quarries may not have been exposed 
to the same action, they do not always afford definite data for re- 
liable comparative estimates of durability,' except where different 
specimens occur in the same quarry." 

North of the glacial limit, all the products of decomposition 
have been planed away and deposited as drift-formation over the 
length and breadth of the land. The rocks are therefore, in gen- 
eral, quite fresh in appearance, and possess only a slight depth of 
cap or worthless rock. The same classes of rock, however, in the 
South are covered with rotten products from long ages of atmos- 
pheric action. 

A study of the surfaces of old buildings, bridge piers, monu- 
ments, tombetonee, etc., which have been exposed to atmospheric 
influences for years, is one of the beet sources of reliable information 
concerning the durability of stone. A durable stone will retain the 
tool-marks made in working it, and preserve its edges and corners 
sharp and true. 

29. Artificial Hethods of Testing IhirabiUty. The older but 
less satisfactory methods are: determining (1) the reeistance to 
crushing, (2) the absorptive ]>ower, (3) the resistance to the expan- 
iiion of frost, by saturating the stone with some solution which will 
crystallize in the pores of the stone and produce an effect similar to 
frost, (4) the aolntolity in acids, and (5) microscopical examination. 



ovGoQi^lc 



AKT. 2.J DUKABILITY OF BLILDISQ STONEB. Sil 

SO. Absorptive Power. The ratio of absorption dependslargely 
OB the density, — a dense stone absorbing less water than a lighter, 
more poroae one. Compactness is therefore a matter of impor- 
tance, especially in cold climates; for if the water in a stone is onco 
allowed to freeze, it destroys the surface, and the stone veryspeedUy 
crnmblee away. Other things being eqnal, the less the absorption 
the better the stone. 

To determine the absorptive power, dry the specimen and weigh 
it carefully; then soak it in water for 34 hours, and weigh again. 
The increase in weight will be the amonnt of absorption. Table 4 
shows the weight of water absorbed by the stone as compared with 
the weight of the dry stone — that is, if 300 units of dry stone weigh 
301 units after immersion, the absorption is 1 in 300, and is recorded 

68 1-300. 

Dr. Hiram A. Cntting, State Greologist of Vermont, determined 
the absorptive power * by placing the specimens in water under th» 
receiver of an air-pnmp, and found the ratio of absorption a littl» 
larger than is given in the following table. It is believed that the 
results given below more nearly represent the conditions of actual 
practice. The values in the " Mai." column are the means of two- 
or three of the largest results, and those in the "Min." column of 
two or three of the smallest. The value in the last column is tha 
mean for 20 or more specimens. 

TABLE C. 
Abbobftivk Power of Stone, Brick, akd IfosTAs. 



^^^^ 


RiTio or AoKoamo*. 


Uu. 


um. 


Avenge 




1-150 
1-150 
1-30 
1-15 
1-4 

1-a 





1-500 
1-840 
1-50 
1-10 


1-768 

1-^oa 

1-88 
1-U ' 
1-10 
1-4 















31. Sffoet of Frost. To determine the probable effect of froet 
npon s stone, carefully wash, dry, and weigh samples, and then wet 
■ Van Nostnuid'* Englo'g Ha(., voL zzlv. pp. 401-0G> 



ovGoQi^lc 



93 ' NATURAL 8T0NB, [CHAP. I. 

them and expose to alternate freezing and. thawing, after which 
wash, dry, and weigh again. The Ices in weight meaanres the rela- 
tive durability. 

A quicker w&y of accompli ehing essentially the aame result is to 
heat the specimens to 500° or 600° F., and plunge them, while liot, 
. into cold water. The following comparatiTO resnlts were obtained 
bj the latter method : * 

RateUte R>Uo ot LoM, 

White brick 1 

Red brick 3 

BrowD-stoue (gandsioue [rom Conn.) 5 

Kova ScDtU s&ndaione 14 

32. Brard'i Tert. Bmrd's method of deteruiining the effect of 
frost is much used, although it does not exactly conform to the con- 
-ditions met with in nature. It consiBts in weighing carefully some 
«mall pieces of the stone, which are then boiled in a concentrated 
■solation of sulphate of soda and afterwards hung up for a few days 
in the open air. The salt crystallizes in the pores of the stone, 
-expands, and produces an effect somewhat similar to frost, as it 
causes small pieces to separate in the form of dust. The specimens 
are again weighed, aud those which suffer the smallest loss of weight 
are the best. The teat is often repeated several times. It will be 
'Soeu that this method depends upon the assumption that the action 
■of the salt in crystallizing is similar to that of water in freezing. 
This is not entirely correct, since it substitutes chemical and 
mechanical action for merely mechanical, to disintegrate the stone, 
thus giving the specimen a worse character than it really deserves. 
The following results were obtained by this method: f 

fietoUve Ratio ot LOM. 

Hard brick 1 

Light dove-colored ssadstone from Seneca, OliJo. ... 2 

Coarse-graiaed BandMone from Nova BcotU S 

Coane-graiaed sandsUme from Little Falls, N. J S 

Coarse dolomite marble from Pleasantville, N. T. . . . 7 

, Coaise-gTHined sandstone from Conn. IS 

Boft brick 16 

. Fine-gratned BudMone from Conn. 19 

• Tenth Censng of thaU. 8., ToL x., Repeat on the Qnurr Indnstrr, P- 3S1 For 
3 table showing eesenUalljr the same remilta, aoe Tan Noettand's Eni^'g Mag., toU 
xlT. p. 537. 

t Tenth CeniBB, voL X., Report of the Qoairr JnAtMzj, p. SSG. 



ovGoQi^lc 



A[{T. 2.] DURABILITT OF sniLDINQ BT0NB8. 33 

S3. Effect of the Atmoaphere. To determine the effect of the 
atmosphere of a Urge city, where coal ia used for fuel, soak clean 
«inall pieces of the stone for several days in water which contains one 
per cent, of sulphuric and hydrochloric acids, agitating frequently. 
If the stone contains any earthy matter likely to be dissolved by the 
gases of the atmosphere, the water will be more or less cloady or 
mnddy. The following results were obtained by this method: * 

BslatlTS Ratio of LOM. 

White brick 1 

Bed brick S 

NovaBcolta atone ft 

Brown-sloae 80 

34. Microsoopical Examinatioo. It is now held that the best 
method of determining the probable durability of a building stone 
is to study its surface, or thin, transpai-ent slices, under a micro- 
scope. This method of study in recent years has been most fruit- 
ful in developing interesting and vainable knowledge of a scientifio 
and truly practical character. An examination of a section by means 
of the microscope will show, not merely the various substances which 
compose it, bnt also the method according to which they are 
arranged and by which they are attached to one another. For 
example, "pyrites ia considered to be the enemy of the quarryman 
and constructor, since it decomposes with ease, bnd stains and dis* 
colors the rock. Pyrites in sharp, well-defined crystals sometimes 
decomposes with great difficulty. If a crystal or grain of pyrites is 
embedded in soft, porous, light-colored sandstones, like those which 
come from Ohio, its presence will with certainty soon demonstrate 
itself by the black spot which will form about it in the porons 
stone, and which will permanently disfigure and mar its beauty. 
If the same grain of pyrites is situated in a very hard, compact, noU' 
absorbent stone, the constituent minerals of which are not rifted or 
cracked, this grain of pyrites may decompose and the products be 
washed away, leaving the stone untarnished." 

36. KETEODB of Pbesebtiho. Yitruvins, the Boman architect, 
two thousand years ^o recommended that stone should be quarned 
in summer when driest, and that it should be seasoned by being 
allowed to lie two years before being used, so as to allow the nataial 

* Tenth Ccmsni, toL z^ fieport on the Qtuur; ludaaOT, p. 381 



ovGoQi^lc 



24 NATURAL BTONE. [CHAF. 1. 

Bap to evaponite. It is a notable fact, tbat in the erection of St 
Fanl's Cathedral in London, England, Sir Christopher Wren re- 
quired that the stone, after being quarried, should be exp(»ed for 
three years on the sea-beach, before its introduction into th& 
building. 

The surfaces of buildings are often covered vith a coating of 
paintj coal-tar, oil, panifiBne, soap and alum, rosin, etc., to preeerTo, 
them. 

Another method of treatment consists in bathing the st^ine ia 
euccessive solutions, the chemical actions bringing about the forma- 
tion of JDBoluble silicates in the pores of the stone. For example, if 
a stone front is first washed with an alkaline fluid to remove dirt,. 
and this followed by a succession of baths of silicate of soda or 
potash, and the surface is then bathed in a solution of chloride of 
lime, an insoluble lime silicate is formed. The soluble salt is then 
washed away, and the insoluble silicate forms a durable cement and 
checks disintegration. If lime-water is substituted for chloride ot 
lime, there is no soluble chloride to wash away. 

There are a great many applications that have been used for the 
prevention of the decay of building stones, as paint, oil, coal-tar, 
bees-wax, rosin, paraffine, etc., and numerous chemical preparations 
similar to that mentioned in the paragraph just above ; but all are 
expensive, and none have proved fairly satisfactory.* 

It has already been stated that, in order to resist the eSects of 
both pressure and weathering, a stone should be placed on its nat- 
□ral bed. This simple precaution adds considerably to the dara~ 
bility of any stone. 



Abt. 3. Classificatioit akd Deschiption of Bcildinq Stones. 

36. CUBsmoAtloV. Building stones are variously classified 
according to geological position, physical structure, oud chemical 
composition. 

37. OeaJ<^cal Classification. The geological position of rocks 
iias but little connection with their properties aa building materials. 
As a general rule, the more ancient rocks are the stronger and the 

* For on elaborate and valuable article by Prof. Eggleston on the causes ot il«cap 
and the metbods of preserrtng building aloDes, see Trans. Am, 8oc. of C. S., voL 
XT. pp' UT-TM ; and for a disciiaBloa on the same, see Baine volume, pp. TD5-1S. 



ovGoQi^lc 



ART. 3.] DBSCBIPTIOH OF BUILDINQ STONES. 35 

more darable ; but to this there are many ezoeptions. Accordjt^ 
to the Qsoal geological claaaification, rocks are divided into igneons, 
metamorphic, and aedimentary. Qreenstone, basalt, and lava are 
eiampleB of igneoae rocks ; granite; marblo, and slate, of meta- 
morphic ; and sandBtone, limeBtoae, and clay, of sedimentary. Al- 
though clay can hardly be classed with building stones, it is not 
entirely oat of place in this connection, since it is employed in 
making bricks and cement, which are important elements of 
masonry. 

38. Fhyaioal Claasifloation. With respect to the atmctnral 
character of large masses, rocks are divided into stratified and aa- 
Btratifled. 

la their more iniaute structure the unstratified rocks present, 
for the most part, an aggregate of crystalline grains, firmly adhering 
together. Granite, trap, basalt, and lava are examples of this class. 

In the more minute structure of stratified rocks, the following 
Tarieties are distinguished : 1. Compact crystalline structure ; ac- 
companied by great strength and durability, as in quartz-rock and 
marble. 2. Slaty structure, easily split into thin layers ; accom- 
panied by both extremes of strength and durability, clay-elate and 
hornblende-slate being the strongest and most durable. 3. The 
granular cry</oiiine,. structure, in which crystalline grains either 
adhere firmly together, as in gneiss, or are cemented into one masi 
by some other material, ae in sandstone ; accompanied by all degree* 
of compactness, porosity, strength, and durability, the lowest ex- 
treme being sand. 4. The compact granular structure, in which 
the grains are too small to be visible to the unaided eye, as in blue 
limestone ; accompanied by considerable strength and durability. 
fi- Porous, granular structure, in which the grains are not crystal* 
line, and are often, if not always, minute shells cemented together; 
Accompanied by a low degree of strength and durability. 6. The 
conglomerate structure, where fragmente of one material are embed- 
ded in a mass of another, as graywacke; accompanied by all degrees 
of strength and durability. 

A study of the fractured surface of a stone is one means of 
determining its structural character. The even fracture, when the 
surfaces of division are pianos in definite positions, is characteristic 
of a cryetaUine structure. The uneven fracture, when the broken 
nirface presents sharp projections, is characteristic of a granular 



ovGoQi^lc 



CO NATUEAL STOKE. [CHAP. I. 

BtixLctare. The slaty fracture gives an even aar&oe for planes ol 
divisioQ parallel to the lamination, and uneven for other directions 
of division. The conchoidal fracture preeeitts smooth concave and 
'wnvex surfaces, and is characteristic of a hard and compact atrnct- 
are. The earthy fracture leaves a roagh, dull snrbce, and indi- 
cates Boftress and brittleaesa. 

' 39. Chemical Clauifloation. Stones are divided into three 
classes with respect to their chemical composition, each distin- 
guished by the earth which forms its chief constituent , viz., sili- 
ceous stones, argillaceous stones, and calcareous Etoues. 

Siliceous Stones ore those in which silica is the characteristic earthy 
constituent. With a few exceptions their structure is crystalline- 
granular, and the crystalline grains contained in them are hard and 
durable ; hence weakness and decay iu them generally arise from 
the decomposition or disintegration of some softer and more perish- 
able material, by which the grains are cemented together, or, when 
they are porous, by the freezing of water in their pores. The prm- 
«ipal siliceous stones are granite, syenite, gneiBS, mica-slate, green- 
stone, baealt, trap, porpliyry, quartz-rook, hornbleude-alate, ana 
sandstone. 

Argillaceous or Clayey Stones are those in which alnmina, although 
it may not always be the most abundant constituent, exiets in suf- 
ficient quantity to give the stone its characteristic properties. The 
principal kiuds are slate and graywacke-slatsk 

Calcareous Stones are those in which carbonate of lime pre- 
dominates. They eflorveace with the dilute mineral acids, which 
oombine with the lime and set free carbonic acid gas. Sulphurio 
acid forms an insoluble compound with the lime. Nitric and mu- 
riatic acids form compounds with it, which are soluble iu water. 
By the action of intense heat the carbonic acid is expelled in gas- 
eous form, and the lime is left in its caustic or alkaline state, when 
it is called quicklime. Some calcareous stones consist of pure car- 
bonate of lime ; in others it is mixed with sand, clay, and oxide 
of iron, or combined with carbonate of magnesia. The durability 
of calcareous stones depends upon their compactness, those which 
are porous being disintegrated by the freezing of water, and by the 
chemical action of an acid atmosphere. They are, for the most 
part, easily wrought. The principal calcareous stones are nurble, 



ovGoQi^lc 



AKT. 3,] DESCaiPTIOS OF BUILDIKO STONES. 37 

«ompact limeetone, granul&r limestone (the calcareous stone of tht 
geological classification), fiad magnesian limeBtone or dolomite. 

40. SlHBIPTIOV OF BuiLDDie BTOITZB. A few of the mora 
prominent classes of building stones will now be briefly deecribed. 

41. Trap. Although trap is the strongest oi builiiing material^ 
and exceodingly durable, it is little used, owing to the great diffi- 
culty with which it is quarried and wrought. It ia an exceedingly 
tough rock, and, being generally without cleav^e oi- bedding, is 
especially intractable under the hammer or chisel. It is, however, 
fometimes used with excellent effect in cyclopean architecture, the 
blocks of various shapes and sizes being fitted together with no 
effort to form regular courses. The " Palisades" (the bluff skirting 
the western shore of the Hudson River, opposite and above New 
York) are composed of trap-rock, — much used for road-metal, street 
pavements, and railroad ballast. 

42. Granite. Granite is the strongest and most durable of all 
ihe stones in commpn use. It generally breaks with regularity, 
and may be quarried in simple shapes with facility ; but it ia ex- 
tremely hard and tough, and therefore can only be wrought into 
elaborate forme with a great expenditore of labor. For this reason 
the use of granite is somewhat limited. Its strength and durability 
commend it, however, for foundations, docks, piers, eta, and for 
massive buildings ; and for these purposes it is in nse the world 
over. 

The larger portion of our granites are some shade of gray in 
color, though pink and red varieties are not uncommon, and black 
Tarieties occasionally occur. They vary in texture from very fine 
«nd hom<^;eneous to coarsely porphyritio rocks, in which the indi- 
vidual grains are an inch or more in length. Excellent granites are 
ionnd in New England, throughout the Alleghany belt, in the 
Rocky Mountains, and in the Sierra Nevada. Very large granite 
, ijuarrioe exist at Vinalhaven, Maine ; Gloucester and Quincy, Ma^ 
sachoeette; and at Concord, New Hampshire. These quarries fur- 
nish nearly all the granite nsed in this country. An excellent 
granite, which ie largely nsed at Chicago and in the Northwest, is 
found at St. Olond, Minnesota. 

At the Vinalhaven quarry a single block 300 feet long, 20 feet 
vide^ and 6 to 10 feet thick was blasted out, being afterwards broken 
tip. Until recently the largest single block ever quarried and 



ovGoQi^lc 



ZB HATCRAL 8T0SB, [CQAP. tt 

dressed in tbiB conntrj was that used for the General Wool Mcnn- 
ment, now in Troy, New York, which measured, when completed,. 
60 feet in he^ht bj 5^ feet square at the baae, being only 9 feet 
shorter than the Egyptian Obelisk now in Central Park, New York> 
la 1887 the Bodwell Qranite Company took ont from ita qDarrieB in: 
Maine a granite shaft 115 feet long, 10 feet square at the base, and 
weighing 850 tons. It is claimed that this is the largest single 
quarried stone on record. 

43. Marbles. In common langnage, any limestone which will take' 
a good polish is called a marble ; but the name is properly applied 
only to limestones which have been exposed to metamorphic action,, 
and have thereby been rendered more crystalline in texture, and 
have had their color more or less modified or totally removed. 
Marbles exhibit great diversity of color and texture. They are- 
pure white, mottled white, gray, blue, black, red, yellow, or mot- 
tled with various mixtures of these colors. Marble is confessedly 
the moat beautiful of all building materials, but is chiefly employed 
for interior decorations. 

44, Iiimeitoues. Limestones are composed chiefly or Li^^ly of 
carbonate of lime. There are many varieties of limestone, which 
difFer in color, composition, and value for engineering and building 
purposes, owing to the differences in the character of the deposita 
and chemical combinations entering into them^ "If the rock is 
compact, fine-grained, and has been deposited by chemical agencies, 
we have a variety of limestone known as travertine. If it contains 
much sand, and baa a more or less concfaoidal fracture, we have a 
siliceous limestone. If the silica ia very fine-grained, it is hom- 
stona If the silica is distributed in nodules or flakes, either in 
seams or throughout the mass, it is cherty limestone; if it contains 
sihca and clay in about equal proportions, hydraulic limestone ; if 
clay alone is the principal impurity, argillaceous limestone ; if iron 
is the principal impurity, ferruginous limestone ; if iron and clay 
exceed the lime, ironstone. If the ironstone is decomposed, and 
the iron hydrntcd, it is rottcnstone; if carbonate of magnesia forms 
one third or less, magnesian limestone ; if carbonate of magnesia 
forms more than one third, dolomitic limestone." 

The lighter-colored and fine-grained limestones, when sawed and 
used as ashlars, are deservedly esteemed as among our best building 
materials. The; are, however, less easily and accurately worked 



ovGoQi^lc 



ART. 3.] DE8CEIPTI0N OF BL'ILllINO STONES. 29 

under tiie chisel than BandBtones, and for this reason and their 
greater rarity are &a less generally nsed. The gray limestones, like 
that of Lockport, New York, when hammer-dressed, have the ap- 
pearance of light granite, and, mnce they are easily wrought, they 
are advantageously used for trimmings in buildings of brick. 

Some of the softer limestones possess qualities which Bpecially 
commend them for bailding material?. For example, the cream- 
colored limestone of the Paris basin {calcaire grassier) is so soft that 
it may be dressed with great facility, and yet hardens on exposure, 
and is a durable stone. Walls laid up of this material are frequently 
planed down to a common surface, and elaborately ornamented at 
«mall expense. The Topeka stone, found and now largely nsed in 
Kansas, has the same qualities. It may be sawed' out in blocks 
Almost as easily as wood, and yet is handsome and durable when 
placed in position. The Bermuda stone and eoquina are treated in 
the same way. 

Large quantities of limestones and dolomites are quarried in 
nearly all of the Western States. These are mostly of a d^II grayish 
color, and their uses are chiefly local. The light-colored oolitio 
limestone of Bedford, Indiana, is, however, an exception to this 
rule. Not only are the lasting qualities fair and the color pleasing, 
but its fine even grain and softness render it admirably adapted for 
carved work. It has been very widely used within the last few 
years. This stone is often found in layers 20 and 30 feet thick, and 
is much used for bridge piers and other massive work. There ai-e 
noted limestone quarriesat Dayton and Sandusky, Ohio; at Bedford, 
Ellettsville, and Salem, Indiana; at Joliot, Lemont, Orafton, and 
Chester, Illinois; and at Cottonwood, Kansas. 

46. Sandstoaes. " Sandstones vary much in color and fitness for 
architectural purposes, but they include some of the most beautiful, 
durable, rnd highly valued materials used in construction. What- 
ever their differences, they have this in common, that they are 
chiefly composed of sand — that is, grains of quartz — to a greater or 
less degree cemented and consolidated. They also frequently con- 
tain other ingredients, as lime, iron, alumina, manganese, etc, by 
which the color and texture are modified. Where a sandstone is 
composed exclusively of grains of quartz, without foreign matter, it 
may be snow-white in color. Examples of this variety are known 
in miiny localities. They are rarely nsed for building, though capa* 



ovGoQi^lc 



30 NATURAL 8T0HB. [OHAP. I. 

ble ot feeing employed for that purpose with excellent effect. They 
bnye been more generally valued as famishing material fortheman- 
niacture of glass. The color of sandstones is frequently bright and 
handsome, and couBtitntes one of the many qualities which have 
rendered them so popular. It is UBually caused by iron; when gray, 
blue, or green, by the protoxide, as carbonate or silicate ; when 
brown, by the hydrated oxide ; when red, by the anhydrous oxide. 
The purple sandstones usually derive tbia shade of color from s 
amall quantity of manganese. 

" The texture of sandstones varies with the coarseness of the 
sand of which they are composed, and the degree to which it is con- 
solidated. Usually the material which unites the grains of sand 
is silica; and this is the best of all cements. This silica has been 
deposited from solntion, and sometimes fills &I1 the interstices be> 
tween the grains. If the process of consolidation has been carried 
fill enough, or the quartz grains have been cemented by fusion, the 
sandstone is converted into quartzite, — one of the strongest and most 
dnrable of rocks, but, in the ratio of its compactness, difBcult to- 
work. Lime and iron often act as cements in sandstones, but both- 
are more soluble and less strong than silica. Hence the finest and 
most indestructible sandstones are such as consist exclusively of 
~ grains of quartz united by siliceous cement. In some sandstones 
part of the grains are fragments of feldspar, and these, being liable 
to decomposition, are elements of weakness in the stone. The very 
fine-grained sandstones often contain a lat^e amount of clay, and 
thus, though very haiidsome, are generally less strong than those- 
which are more purely siliceous. 

" The durability of sandstones varies with both their physical 
and chemical composition. When nearly pure silica and well ce- 
mented, sandstones are as resistant to weather as granite, and very 
much less affected by the action of fire. Taken as a whole, they 
may be regarded as among the most dnrable of building materials. 
When first taken from the quarry, and saturated with quarry water 
(a weak solution of silica), they are frequently very soft, but on ex- 
posure become much harder by the precipitation of the soluble silica 
contained in them. 

48. "Since they form an important part of all the groups of 
sedimentary rocks, sandstones are abundant in nearly all countries; 
EUid as they are quarried with great ease, and are wrought with th» 



ovGoQi^lc 



ART. 3.] DBSCRIPTIOK 07 BUILDING 810X18. 3t 

hammer and ofaiael vith much greater facility than limeatonesr 
grauitee, and moat other kinds of rocks, these qualities, joined to 
their rarioas and pleasant colors and their durability, hare mada 
them the most popular and nsefnl of building stonea. In the 
United States, we hare a rer; large number of sandBtonea which are 
"jxtensively used for building purposes. 

" Among these may be mentioned the Dorchesier stove of New 
Brnnswiok, and Brown-stone of Connecticut and New Jersey. 
These have been much used in the boiIdingB of the Atlantic cities. 
The latter has been very popular, but experience has shown it to b& 
Berioualy lackiug in durability. 

" Among the sandstones most frequently employed in the build- 
ing of the interior are : — 

1. " The Ohio stone, derived from the Berea grit, a member of 
the Lower Carboniferous series in Northern Ohio. The principal 
quarries are located at Amherst and Berea. The stone from Am- 
herst IB generally light drab in color, Tory homogeneous in testure„ 
and composed of nearly pure silica. It is very resistant to fire and 
weathering, and is, on the whole, one of the best and handsomest 
building stones known. The Berea stone is lighter in color than 
tiie Amherst, but sometimes contains sulphide of iron, and is then 
liable to stain and decompose. 

2. '■' The Waverly sandstone, also derived from the Lower Car- 
tioniferona seriea, comes from Southern Ohio. This is a fine- 
grained hom<^neous stone of a light-drab or dove color, works with 
facility, and is very handsome and durable. It forms the material 
of which many of the finest buildings of Cincinnati are constructed, 
and is, justly, highly esteemed there and elsewhere. 

3. " The Lake Superior sandstone is a dark, piirpliah-brown 
stone of the Potsdam ago, quarried at Bass Island, Marquette, etc 
This is rather a coarse stone, of medium strength, but homogeneous 
and durable, and one much used in the Lake cities. 

4. "TheBt. ffeneftefti s/one is a fine-grained sandstone of a del- 
icate drab Or straw color, very homogeneous in tone and texture. 
It is quarried at St. Genevieve, Missouri, and is one of the hand- 
somest of all our sandstones. 

6. " The Medina sandstone, v\ao)y forms the base of the Upper 
Silurian seriefi in Western New York, furnishes a remarkably strong 



ovGoQi^lc 



32 katural stoke. [chap, l 

And dnrable stone, mncii used forpaTement and corbing in the 
Luke cities. 

6. " The coal-measures of Pennsylvania, Ohio, and other Western 
States sapply exceUent Bandstonee foT building purposes at a large 
number of localities. These vary in color from white to dark red 
or purple, though generally gray or drab. While strong and 
durable, they are mostly coarser and less handsome than the sand- 
stones which have been ennmerated above. This is the scarce from 
which are derived the sandstones used in purely engineering stmct- 
nres. " • 

47. Other Hames. There is a great variety of names of more 
or less local application, derived from the appearance of the stone, 
the use to which it is put, etc., which it would be impossible to 
classify. The same stone often passes under entirely different 
names in difEerent localities; and stones entirely different in their 
«ssential characteristics often pass under the same name. 

48. LocATloir of ftlTAKBlzs. For information coaceming the 
location of quarries, character of product, etc., see: Tenth Censas 
«f the U. S., Vol. X, Beport on Quarry ludnstry, pp. 107-363; 
Report of Smithsonian Institution, 1885-86, Part II, pp.*357-^488; 
Merrill's Stones for Building and Decoration, pp. 45-312 — substan- 
tially the same as the preceding — and the reports of the various 
State geological surveys. 

49. Cost. See §§ 226-38. 

• Frirf. 1. B. KewbaiTT. 



ovGoQi^lc 



CHAPTER IL 
BRICK. 

61. Bbice is made by snbmittiiig clay, which has been prepared 
properly and moulded into shape, to a temperature which oonrerts 
it into a semi-vitriSed mass. 

Common brick is a moat valuable substitute for stone. Its 
-comparative cheapness, the ease with which it is transported and 
handled, and the facility with which it ia worked into structures of 
«ny desired form, are its valuable characteristics. It is, when prop- 
erly made, nearly aa strong as the best building stone. It is but 
slightly affected by change of temperature or of humidity; and ia 
also lighter than stone. 

Notwithstanding the good qnalities which recommend brick as 
a substitute for stone, it is very little used in engineering structures. 
It is employed in the construction of sewers and bridge piers, and 
for the lining of tunnels. Brick could many times be profitably 
sabstituted for iron, stone, or timber in engineering Btructnres. 
This is especially true since recent improvements in the process of 
manufacture have decreased the cost while they have increased the 
'quality and the uniformity of the product. The advantages of 
employing brick-work inst(«d of stone masonry will be discussed in 
■connection with brick masonry in Chapter VIII. Probably one 
thing which has prevented the more general use of brick in engi- 
neering is the variable quality of the product and the trouble of 
proper inspection. 

S2. Fbocssb of XAvrrTAcnmE. The Clay. The quality of the 
brick depends primarily upon the kind of clay. Common clays, of 
which the common brick is made, consist principally of silicate of 
alumina; bat they also usually contain lime, mf^nesia, and oxide 
of iron. The latter ingredient is useful, improving the product by 
giving it hardness and strength; hence the red brick of the Eastern 
States 18 often of better quality than the white and yellow brick 
jnade in the West. Silicate of lime renders the clay too fusible,' 



ovGoQi^lc 



34 BEICK. [CHAP. II. 

and causes the bricks to soften and to become distorted in tbe pro- 
cees of buming. Carbonate of lime is certain to decompose iii 
bnrning, and tbe caustic lime left behind absorbs moistare, prevents 
the adherence of the mortar, and promotes disintegration. 

Uncombined silica, if not in excess, is beneficial, as it preserres 
the form of the briclc at high temperatures. In excess it destroys 
the cohesion, and renders the bricks brittle and weak. Twenty-fira 
per cent, of silica is a good proportion. 

63. Moulding. Id the old process the cla; is tempered with 
water and mixed to a plastic state in a pit with a tempering wheel, 
or in a primitive pug-mill; and tlaen the soft, plastic clay is pressed 
into the moulda by band. This- method is so slow and laborioaa 
that it has beeu almost entirely displaced by more economical and 
expeditious ones in which the work ia done wholly by machinery. 
There is- a great variety of machines for preparing and moulding 
the clay, which, however, may be grouped into three classes, accord- 
ing to the condition of the clay when moulded: (1) soft-mud 
machines, for which the clay is reduced to a soft mud by adding 
about one quarter of its volume of water; (S) stiff-mud machines, 
for which the clay <■; reduced to a stiff mud; and (8) dry-clay 
machines, with which the dry, or nearly dry, clay is forced into the 
moulds by a heavy pressure without having been reduced to a plastic 
mass. These machines may also be divided into two classes, aocord* 
ing to tlie method of filling the moulds: (1) Those in which a con- 

. ttnuons stream of clay is forced from the pug-mill through a die 
and is afterwards cut up into bricks; and {'i) those in which the 
clay is forced into moulds moving under the nozzle of the pug-mill. 

64. Burning. The time of burning varies with the character of 
/the clay, the form and size of kiln, and the kind of fuel. With the 

older processes of burning, the brick, when dry enough, is built up 
in sections — by brick -makers called " arches, " — which are usually 
about 5 bricks {3i feet) wide, 30 to 40 bricks (30 to 30 feet) deep, 
and from 35 to 50 courses high. Each section or "arch" has an 
opening— called an "eye" — at the bottom in the center of its width, 
which runs entirely through the kiln, and in which the fuel used in 
burning is placed. After the bricks are thus stacked up, tbe entire 
pile is enclosed with a wall of green brick, and the joints between 
the casing bricks are carefnlly stopped with mud. Bumiug, includ- 
ing drying, occupies from 6 to 15 days. The brick ia first subjected 



D.qitizeabyG00l^lc 



CLASSIFICATION OF COUUON BBICE. &^ 

to a moderate heat, and when all moistnre has been expelled, the 
heat i8 increased slowly until the "arch-brick," i. e., those next to 
the "eye," attain a white heat. This temperature is kept up until 
the burning is complete. Finally, all oponinga are closed, and tha 
mass slowly cools. 

With the more modeiii processes of burning, the principal yards 
have permanent kilns. These are usually either a rectangular space 
eurronnded, except for very wide doors at the ends, by permanent 
brick walls having fire-boxes on the outside; or the kiln may bo 
entirely enclosed — above as well as on the sides — with brick masonry. 
The latter are usaally circular, and are sometimes made in com- 
partments, each of which has a separate entrance and independent 
connection with the chimney. The latter may be built within the 
kiln or entirely outside, but a downward draught is invariably 
secured. The fuel, nsnally fine coal, is placed near the top of the 
kiln, and the down draught causes a free circulation of the fiame 
and heated gases about the material being burned. While some 
compartments are being fired others are being filled, and still 
others are being emptied. 

55. FiBE Buck. Fire bricks are osed whenever very higb 
temperatures are to be resisted. They are made either of a veiy 
nearly pure clay, or of a mixture of pure clay and clean sand, or, in 
rare casee, of nearly pure silica cemented with a small propoition 
, of clay. The presence of oxide of iron is very 'injurious, and, as a 
rnle, the presence of G per cent, justifies the rejection of the brick. 
In specifications it should generally be stipulated that fire brick 
should contain less than 6 per cent, of oxide of iron, and less tliao 
an aggregate of 3 per cent, of combined lime, soda, potash, and 
ma^csia. The sulphide of iron — pyrites — is even worse in its 
effect on fire brick than the substances first named. 

When intended to resist only extremely high heat, silica should 
be in excess; and if to be exposed to the action of metallic oxides, 
which would tend to unite with silica, alumina should be in excess. 

Good fire brick should be uniform in size, regular in shape, 
homogeneous in textnre and composition, eaeily cat, strong, and 
infasible. 

66. CLAMincATioH 07 COHOir Bbiok. Bricks are classified 
according to (1) the way in which they are moulded; (H) their 
position In the kiln while being burned; and (3) their form or use. 



ovGoQi^lc 



96 BBicE. [chap. n. 

■I- The method of moulding givee rise to the following terms: 

8o/t-mud Brick. One moulded from claj which has been reduced 
to & soft mnd by adding water. It may be either hand-monlded or 
machine-monl ded. 

Sliff-mud Brick, One moulded from clay in the condition of 
stift mud. It is always machine-moulded. 

Pressed Brick. One moulded from dry or semi-dry clay. 

Re-pressed Brick, A soft-mud brick which, after being par- 
tially dried, has been subjected to an enormous pressure. It is 
ako called, but less appropriately, pressed brick. The object of 
the re-preaeing is to render the form more regular and to increase 
the strength and density. 

Slop Brick. In moulding brick by hand, the monlds are some- 
times dipped into water just before being filled with clay, to pre- 
Tent the mnd from sticking to them, Brick moulded by this 
process is known as slop brick. It is deficient in color, and has a 
■comparatively smooth surface, with rounded edges and comers. 
*rhiB kind of brick is now seldom made. 

Snnded Brick. Ordinarily, in making soft-mud brick, sand is 
sprinkled into the moulds to prevent the clay from sticking ; the 
brick is then called sanded brick. The sand on the surface is of no 
serious advantage or disadvantage. In hand -mo aiding, when sand 
is used for this purpose, it is certain to become mixed with the clay 
and occur in streaks in the finished hrick, which is very undesira- 
ble ; and owing to details of the process, which it is here unneces- 
jwry to explain, every third brick is especially bad. 

Machine-made Brick. Brick is frequently described as "ma- 
chine-made;" but this is very indefinite, since all grad^ and kinds 
are made by machinei^. 

a* When brick was generally burned in the old-style np-draught 
kiln, the classification according to position was important ; but 
with the new styles of kilna and improved methods of burning, the 
quality is bo nearly uniform throughout the kiln, that tiie classifica- 
tion is less important. Three grades of brick are taken from the 
-©Id-stylo kiln: 

A ri'h or Clinker Bricks. Those which form the tops and sides of 
the arches in which the fire is built. Being over-hnmed and par- 
iially vitrified, they ta% hard, brittle, and vre^k. 



ovGoQi^lc 



BEQCISITES FOU OOOD BRIOE. 37 

Bodi/, Cherry, or Bard Bricks. Those taken from the interior' 
of the pile. The best bricks in the kiln. 

Salmon, Pale, or Soft Bricks. Those which form the exterior of 
the mass. Being underburned, the; are too soft for ordinary work, 
nnlesB it be for filling. The terms salmon and pale refer to the 
color of the brick, and hence are not applicable to a brick made of 
a clay that does not bum red. Although nearly all brick clajs burn- 
red, yet the localities where the contrary is true are suEBciently 
nnmeroiiB to make it desirable to use a diSerent term in dcsignating- 
the quality. There is, necessarily, no relation between color, and 
strength and density. Briok-makers naturally iiave a prejudice 
against the term soft brick, which doubtteas explains the nearly 
uniTersal prevalence of the lesa appropriate term — salmon. 

3. The form or use of toiokB glres rise to fha fblloving daadfr 
cation: — 

Compass Brick. Those having one edge ehorter than the other. 
Used in lining shafts, etc. 

Feather-edge Brick. Those of which one edge is tMnner thtut 
the other. Used in arches ; and more properly, but leas frequently, 
called vouSEoif brick. 

Face Brick. Those which, owing to tiniformity of size and 
color, are snitable for the face of the wall of buildings. Sometimet 
face bricks are simply the bent ordinary brick ; bat generally the 
term is applied only to re-preased or pressed brick made specially tor 
this purpose. They are a little larger than ordinary bricks (g 62), 

Sewer Brick. Ordinary hard brick, smooth, and r^alar in 
form. 

Paving Brick. Very hard, ordinary brick. A Titrifled ola} 
block, very much larger than ordinary brick, is sometimes used fot 
paving, and is called a paving hriok, but mora often, and more 
properly, a brick paving-block. 

57. ExqinilTU TOB OooD BXIOX. 1. A good brick ehonld have 
plane faces, parallel aides, and ahnrp edges and angles. 2. It should 
be of fine, compact, nniform texture ; shoald be quite hard; and 
should give a clear ringing sound when stmck a sharp blow, 3, It 
shoald not absorb more than one tenth of ite weight of water. 4. 
Its specific gravity should be 2 or more. 5. The crnshing strength 
tft half brick, when ground fiat and pressed botweec thick metal 



ovGoQi^lc 



38 BKiCK. [chap. n. 

plfttes, shoold be at least 7,000 poanda per square inch. 6. Its mod* 
aim of rupture shotild be at least 1,000 pounds per square inch. 

1. lu regularity of form re-pressed brick racks first, dry-cla; 
briclE next, then stifi-mud brick, and soft-mud brick last. Begu- 
Urit; of form depends largely upon the method of burning. 

2. The compactness and uniformity of testare, which greatly 
inSnence the durability of brick, depend mainly upon the method 
of moulding. As a general rule, hand-moulded bricks are best in 
iMb respect, since the clay in them is more uniformly tempered b&- 
fore being moulded ; bnt this advantage is partially neutralized by 
the presence of sand seams (§ 56). Machine-moulded soft-mud 
bricks rank next in compactness and uniformity of testnre. Then 
oome machine -moulded stiff-mud bricks, which Tary greatly in 
durability with the kind of machine used in their maaufactnre. 
By some of the machines, the brick is moulded in layers (parallel to 
Any face, according to the kind of machine), which are not thor- 
oughly cemented, and which separate under the action of the frost. 
Jn compactness, the dry -clay "brick comes last. However, the rela- 
tire value of the products made by the different processes varies 
-with the nature of the clay used. 

3. The absorptive power is one of the most important elements, 
since it greatly affects the durability of the brick, particularly its 
redfitance to the effect of frost (see §§ 31 and 33). Very soft, un- 
der-burned brick will absorb from 25 to 33 per cent, of their weight 
«f water. IVeak, light-red ones, such as are frequently used in fill- 
ing in the interior of walls, will absorb about 20 to 25 per cent. ; 
while the best brick wilt absorb only 4 or 5 per cent. A brick may 
be called good which will absorb not more than 10 per cent See 
Table 9 (page 45). 

4. The specific gravity of a brick does not indicate its quality^ 
and depends mainly upon the amount of burning and the kind of 
fnel employed. Over-barned arch lirinks, being both smaller and 
heavier than the better body bricks, have a considerably greater 
Bpecifio gravity, although inferior in quality. 

6. The crashing strength is not a certain index of the value of a 
brick, although it is always one of the 'tems determined in testing 
brick — if a testiog-machine is at hand. For any kind of service, 
the durability of a brick is of gre»tcr importance than its ability 
to resist crushing, — the latter is only remotely connected with doi^ 



ovGoo^^lc 



ABSORBING POWER. 89 . 

Iiility. Teets of the crashing etreagth of iudividnal bricks are dw> 
fnl only in comparing different kinds of brick, and give do idea o( 
the strength of walls built of such bricks (see § 246). Furthermore, 
the crushing strength can not be determined accurately, since it 
Taries greatlv with the size of the specimen and with the details of 
the experiments (see § 60). 

6. Owing to both the nature of the quality tested and the facility 
with which sncb a test can be made, the determination of the 
transTerse strength is one of the best means of judging of the 
tioality of a brick. The transverse strength depends mainly npon , 
tiie toughness of the brick, — a quality of prime importance in bricks 
nsed for paving, and also a quality greatly affecting the resistance to 
frost. 

68. ABBOiaiKQ POWZB. The less the amount of water absorbed 
1^ a brick the greater, in all probability, will be its durability. 
The amount of water absorbed is, then, an important consideration 
in determining the quality of a brick. There are different metboda 
in nse for determining the amount of water taken up by a brick, 
and these lead to slightly different results. Some experimenters dry 
the bricks in a hot-air chamber, while some dry them simply by ex- 
posing them in a dry room; some experimenters immerse the bricks 
in water in the open air, while others immerse them uuder the re- 
oeiver of an air-pump; some immerse whole brick, and some use 
small pieces; and, again, some dry the surface with bibulous paper, 
while others allow the surface to dry by evaporation. Air-drying 
most nearly represents the conditions of actual exposure in ma- 
sonry strnctures, since water not expelled in that way is in such a 
condition as not to do any harm by freezing. Immersion in the 
open air more nearly represents actual practice than immersion in 
a vacuum. The conditions of actual practice arc beet represented 
by testing whole brick, since some kinds have a more or less im- 
penriouB skin. Drying the surface by evaporation is more accurate 
than drying it with paper; however, neither process is susceptible 
of mathematical accuracy. 

The absorbing power given in Table 9, page 45, was determined 
by (I) drying whole brick in a steam-heated room for three weeks, 
(2) weighing and (3) immersing them in water for forty-four 
hours; and then (4) drying for four hours — until all the water on 
the surface was evaporated, — and, finally, (5) again weighing them. 



ovGoQi^lc 



.40 BBiOE. [chap. n:. 

The results in the table represent the mean of sereral obseiratioiis. 
If the brick had been kiln-dried, or weighed before the surface 
water waa entirely remoTed, the apparent absorption woold hava 
been greater. 

Comparing the absorbing power of brick as given in the table 
on page 45 with that of stone on page 20, we see the absorbing^ 
power of the best briok is abont equal to that of average lime- 
stone and sandstone, and mach greater than marble and granite. 
For a method of rendering brick non-abeorbent, see g§ 368-64. 

69. TkUTBTIBIE BtbUQIX. The experiments necessary to 
determine the transverse strength of brick are easily made (§ 16)> 
give definite results, and furnish valuable information coucerain^ 
the practical value of the brick; hence this test is one of the best la 
use. 

Table 6 gives the results of ezperiments made by the anther on 

Illinois brick. The averages represent the results of from six to fifteea 

TABLE 6. 

Tkaksvebbb SntEROTH or Iixciou Bbick. 

(Somiiurind from Table •, pass «.) 



u^' 


.„„..„. 


HomiLoa or RDmu Dt 

Laa.FBBS«.lF.> 


TaBaaSracnani.* 




M^ 


Mi„. 


ATtrace. 


Haz. 


Mta. 


A«r. 


' 


Soft-clHV. hand-moulded, 
-best 50)1 in kllii 

Soft-clay , machine- mould- 
ed, -bert OMinkilD.... 

Stifl-clay, mftchine-mould- 
ed,-be»t SOK in kiln. . . . 


2.388 

3.8B4 

1,475 

406 

4,848 


846 

1.136 

764 

160 

S.a86 


1,409 

i.Tia 

1,114 

SS6 

8.817 


134 

143 

83 
37 
341 


47 

68 

43 

8 

134 


78 
OS 















experiments on brick from three localities. The "Max." and 
"Min." columns contain the average of the two highest and the 
two lowest results respectively. 

The results in Table 7 were obtained under the direction of the 
Ohief Engineer of the Lehigh Valley R. B. Each result represents 

* For deaniUoD, aee f 16, 



ovGoQi^lc 



OBUSHINO STBENGTH. 41. 

the mean of seven to nine experiments on bricka from different: 
locatitiea. The results in Table 6 are oonuderabl; greater thao 

TABLE 7. 
TBAnerKSSB Btsesoth 





HoDDLUB or Bomru if 
Lbb. PM> 64- I>. 


Co-stfiohmt of Tuhb- 




«»i. 


HiD. 


Arencs. 


Ifu. 


Uln. 


Athmb- 




M4 
645 
444 


1,048 
710 
004 
269 


1.8I» 
800 
697 
878 


86 
30 


68 
8S 
38 
10 








Medium 

Soft 


8» 







thoBe in Table 7, the difference being dne probably more to recent 
improTements in the manofactnre of brick and to the method of 
selection than to locality. The brick from irhich the reenlta in 
Table 6 were derived were obtained from mannfactarerswell known 
for the high quality of their products. 

ea CBTnmvo STBrarOTH. It has already been explained (g§ 7 
to 14) that the results for the crushing strength of stone vary 
greatly with the details of the experiments; but this difference is 
even greater in the case of brick than in that of stone. In testing 
stone the nniform practice is to test onbes (§ 10) whoso taoee are 
carefolly dressed to parallel planes. In testing brick there is no 
settled custom. (1) Some experimenters test half brick while others 
test whole ones; (3) some grind the preesed surfaces accurately to 
planes, and some level up the surfaces by putting on a thin coat of 
plasterof P&nB, while others leave them in the rough; and (3) some 
test the brick set on end, some on the side, and others laid flat- 
wise. 

1. From a series of experiments* on soft brick, the author con- 
cludes that the crushing strength per square inch of a quarter of a 
brick is about, half that of a whole one; and that a half brick is 
about two thirds, and three quarters of a brick about fiee sixths, as 
strong per square inch as a whole one ; or, in other words, the 
strength of a quarter, a half, abd three quarters of a brick, and a 

* JSyiRMriiy JAiM, ToL icxLp. 88. 



ovGoQi^lc 



42 BRICK. [chap. II. 

whole one, are to each other as 3, 4, 5, and 6 respectively. The 
reason for thie difference is apparent if a whole brick be conceived 
«B being made up of a number of cnbea placed side by aide, in whiuh 
case it is clear that the interior cubes will be strotiger than the 
exterior ones because of the side support derived from the latter. 
For experiments showing the marked effect of this lateral support, 
see § 373. The quarter brick and the half brick have lees of this 
lateral support than the whole one, and hence have correspondingly 
less crushing strength. 

2. The strength of the specimen will vary greatly with the degree 
of smoothness of its bed-sur&ces. To determine the difference 
between reducing the pressed surfaces to a plane and leaving them 
in the rough, the author selected six bricks of regular form and 
apparently of the same strength, and tested three in the rough and 
the other three after having reduced the pressed surfaces to planes 
by laying on a coating of plaster of Paris, which, after drying, was 
ground off to a plane. The amount of plaster remaining on the 
aurfaces was just sufficient to fill up the depressions. Both sets 
were tested in a hydraulic press between cast-iron, parallel (self- 
adjusting), pressing surfaces. The average strength of those that 
were plastered was 2.06 times the strength of those that were not 
plastered. This difference will vary with the relative strength of 
the brick and the plaster. The average strength of the bricks whose 
surfaces were plastered was 9,170 pounds per square inch, which is 
more than that of the plaster used; and therefore it is highly 
probable that if the surfaces had been reduced to planes by grind- 
ing, the difference in strength would have been still greater. See 
also the last paragraph of § 12. 

3. As before stated, some experimenters test brick flatwise, some 
edgewise, and some endwise. Since bricks are generally employed 
in such s position that the pressure is on the broadest face, it seems 
a little more satisfactory to lay the brick flatwise while testing it; 
but since the only object in determining the crushing strength of 
brick is to ascertain the relative strength of different bricks, — the 
crushing strength of the brick is only remotely connected with the 
crushing strength of the brick-masonry (§ 246), — the position of the 
brick while being tested is not a matter of vital importance. Doubt- 
less the principal reason for testing them on end or edgewise is to 
bring them within the capacity of the testing-machine. Howevei^ 



ovGoQi^lc 



OaUSHINO 8TBENQTH. 43 

tbere is one good reason against testing brick flatwise; viz., all 
homogeneous grannlar bodies fail under oompreesion by shearing 
along planes at about 45° with the pressed surfaces, and hence if 
ihe height is not snfficient to allow the shearing etreasea to act 
freely, an abnormal strength is developed. See also § 10. 

The relative strength of brick tested m the three positions — flat- 
wise, edgewise, and endwise — ^varies somewhat with the details of 
the ezperimeats; bnt it ie reasonably well settled that the strength 
-of homogeneous brick flatwise between steel or cast-iron pressing 
snrtaces is one and a half to two times as much as when the brick is 
-tested on end. A few experiments by theauthor* seem to indicate 
that the strength edgewise is a little more than a mean between the 
atrength flatwise and endwise. If the bnck is laminated (see para- 
graph 3, § 57), the relative strength for the three positions — flat- 
wise, edgewise, and endwise— will vary greatly with the direction of 
the grain. 

61. Comparatively few experiments have been made to deter- 
mine the strength of brick, and they are far from satisfactory, since 
the manner of making the experiment is seldom recorded. The 
differences in the details of the experiments, together with the 
differences in the quality of the bricks themselves, are suf&cient to 
«anse a wide variation in the results obtained by different observers. 
The following data are given for reference and comparisons. 

The results in Table 8 (page 44) were made with the 17. S. 
testing-machine at the Watertown (Mass.) ArsenaLf In each 
experiment the pressed sur&ces were " carefully ground flat and 
«et in a thin facing of plaster of Paris, and then tested between steel 
pressing sarfaces." 

The experiments given in Table 9 (page 45) were made by the 
anther, on Illinois brick. The bricks were crushed between self- 
adjusting cast-iron pressing surfaces. Although No, II shows an 
average absorption, a moderate transverse strength, and a high crush- 
ing strength, this particular brand of brick disintegrated rapidly by 
the frost. This is characteristic of this class of brick, and is caused 
by the clay's being forced into the moulds or through the die in such a 
wayas to leave the brick in laminee, not well cemented together. A 
critical examination of the brick with the unaided eye gave no indi- 



ovGoQi^lc 



[chap ir 



1 



Hi 


llig ii 


ill 


s ill ii 


Hill 






II 


S 1 
J: I I ! I I : ; : ! ! ! ^: : "s ^>'3 

•t It 


II 




1 

s 

1 


1 
1 

•E 




|Jl 


■■ \ 


: i-s 


: :'S 


: ■! 


: 





jvGooi^le 



fm |s!|M;BS|isil' 

tutw 



i 



» 



ill 



ill 



3NB 



i|?i?^ 



sjiiimiyiisi 



|:- 1 s:-5 ;;: s ss ; 



ill 
III! 

I! 

1 ! 



.11 

If I. 



jvGooi^le 



46 BBICE. [CSAP. n. 

cation of a lamiD&ted Btructure, and yet compressing the brick ia 
two positions — sidewisu and edgewise — never failed to reveal auch 
Btructure. The crushing strength in the table was obtained whett 
the preBsure was applied to the edges of the laminie. la experi- 
ments Nos. 13, 13, and 1 4 the pressed surfaces were so nearly mathe- 
matical planes that possibly these bricks stood more than they would 
haTe done if their beds had been plastered. The strength of No. 15 
was beyond the capacity of the machine; a whole brick, on end, stood 
11,083 ibs. persq. in. without any cracks or snapping sounds — which 
usually occur at about half of the ultimate strength. 

Rankine says thivt " strong red brick, when set on end, should 
require at least 1,100 Iba. per aq. in. to crush them; weak red ones, 
550 to 800 lbs. persq. in.; and fire bricks, 1,700 lbs. per. sq. in."* 

Experiments on the brick in general use in Berlin gave for 
" ordinary" brick, on edge, a atrengtii of 2,930 lbs. per eq. in.; and 
for " selected" brick, 3,C70 lbs. per sq. in.f 

The brick used in the New York reservoir, when laid flat and 
packed with sand, showed an average strength, for fonr specimens, 
of 2,770 lbs. per sq. in.; and two samples tested between wood 
averaged 2,660 lbs. per sq. In.t Prof. Pikeg tested half brick flat- 
wise between sheets of pasteboard with the following results: St. 
Louis brick, 6,417 lbs. per sq. in. (the average of six trials); and 
pressed brick, 2,519 lbs, per sq. in. (the averse of thirteen sam- 
ples from ten localities). 

62. SIZE AST) Weight. In England the legal standard size for 
brick is 8J X 4f X 2J inches. In Scotland the average size is 
about 9^ X 4i X H inches; in Germany, 9J X 4} x 2| inches; in 
Austria, llj X 5J- X 2jE inches; in Cuba, 11 X 5^ X 2^ inches; and 
in South America, 12J X 6J X 2i inches. 

In the United States there is no legal standard, and the dimen- 
sions vary with the maker. In the Eastern States 8i X 4 X 2i 
inches is a common size for brick, of which 23 make a cubic foot; 
but in the West the dimensions are usually a little smaller. The 
National Brick-makers' Association in 1887 and the National 
> Civil En^ne«ring, pp. 866 and T69. 

f Van Nostrand's Engineering Magnzlnu, vol, xxzlv. p. 340. From ftbetjscta of 
the Inst, of C. E. 

X Jonr. Frank, Inst,, vol. Ixv. p. 333; aUo TninB. Am. Soo. of C. B., vol. U. pp. 
195-86. 

i Jour. Assoc. Engineering Boo., vol. iv. pp. 368-67. 



ovGoQi^lc 



SIZE, WKIGHT, ASD COST, 4? 

Traders and Builders' Aasociation ia 1889 adopted 8i X 4 X 3^ 
inches as the standard size for common brick, and 8| X 4^ X 2^ 
for face brick. The price should vary with the size. If, reckoned 
according to cnbic contents, brick 8x4x3 inches is worth tlO 
per thoosand, brick 8i X 4^ x %i is worth tl3.33 per thousand, 
and 8^ X 4^ X 3^ is worth llo per thousand. Further, where brick 
is laid by the thousand, small bricks are doubly expensive. Since 
bricks shrink in burning, in proporiiion to the temperature to which 
they are exposed, the amount differing with the different kinds of 
clays, it is impossible to have the size exactly uniform. Be-pressed 
and machine-monlded bricks are more nearly aniform in size than 
hand- moulded. 

The size of brick and the thickness of the mortar joint should 
be such that brick may be laid fiat, edgewise, or set vertically, and 
still fit exactly These proportions are seldom realized. 

Re-pressed brick weighs about 150 lbs. per en. ft. ; common 
hard brick, 125 ; inferior, soft brick, 100. Common bricks will 
average about H lbs. each. 

63. Con. Brick is sold by the thousand. At Chicago, in 1687, 
the " best Bewer" brick cost $9 ; common brick, from W to 17. 



ovGoQi^lc 



CHAPTEB IIL 
UME AND CEMENT. 

64, Cl^ABSIilCATIOK. Considered as materialB for nae in th* 
^bnilder's art, the prodacts of calcination of limestone are classified 
as common lime, hydraulic lime, and hydraulic cement. If the 
limestone is nearly pure carbonate of lime, the produot la common 
lime, which will slake npon the addition of water, and mortar made 
'of it will harden by absorbing carbonic acid from the air, but will 
not barden under water. If the limestone contains more impari- 
ties, the prodnct is hydraulic lime, which will slake npon the addi- 
'4ion of water, and mortar made of it will harden either in air or 
under water by the chemical action between the hydranlic lime and 
4he water used in making the mortar. If the limestone contains 
■still more impurities, the product is hydraulic cement, which will 
not slake upon the addition of water but most be reduced to a paste 
by grinding, and which will set either in air or under water by the 
-chemical action between the cement and the water used in making 
the mortar. ' Common lime is BometimoB called air-lime, because a 
paste or mortar made from it requires exposure to the air to enable 
it to " set," or harden. The hydraulic limes and cements are also 
-called water-limes and water-cements, from their property of 
hardening under water. 

Common lime is nsed in making the mortar for most architect- 
nral masonry, and until recently it was generally employed in 
-engineering masonry; but the opinion is rapidly gaining ground 
■that only cement mortar should be employed in engineering struct- 
ures requiring great strength or being subject to shock. On most 
first-clues railroads hydraulic cement mortar is used in all masonry 
fltrnctures. This change in practice is largely due to the better 
appreciation of the superiority of hydranlic cement as a building 
material. Although it has been manufactured for about fifty 
years, the amount nsed was comparatiTely limited until withiu 
'recent years. At present large quantities are imported from 



ovGoQi^lc 



AST. 1.] COUHON LIUE. 49 

Sarope, and very much more is made in thia coantrj. Hydranlio 
lime is neither manafaotared nor nsed id this coantrj. 

The following discuBsiOD coDcerning common sad hydiaallo 
limes is given as preliminary to the stndy of hydraulic cemente 
rather than becanse of the importance of these materials in engineer- 
ing constrncti>n 

Abt. 1. COUUOK LiHK. 

65. DEflOXHTioir. The limestones which famish the common 
lime are seldom, if erer, pure; bat osnally contain, besides the car- 
bonate of lime, from 3 to 10 per cent, of imparities, — snch as silica, 
alumina, magnesia, oxide of maagaaese, and traces of the alkalies. 
Lime — varioasly designated as common lime, qaickltme, or osnstia 
lime — is a protoxide of calcium, and is prodnced when marble, or 
any other variety of pare or nearly pare carbonate of lime, is 
calcined with a heat of sufficient intensity and dnration to expel 
the carbonic acid. It has a specific gravi^ of ^.3, is amorphoos, 
highly canstic, has a great avidity for water, and when bronght into 
contact with it will rapidly absorb nearly a quarter of its weight of 
that anbstanoe. This absorption is accompanied and followed by a 
great elevation of tempeiatare, by the evolution of hot and slightly 
caustic vapor, by the bnrsting of the lime into pieces; and finally 
the lime ia reduced to a powder, the volnme of which is from two 
and a half to three and a half times the volame of the original lime 
— the increase of bulk being proportional to the purity of the lime- 
stone. In this condition the lime is said to be slaked, and is ready 
for nse in making mortar. 

The paste of common lime is unctuous and impalpable to sight 
-and touch ; hence these limes are sometimes called fat or rich limes, 
as distingaished from others known as poor or meager limes. These 
latter asnally contain more or less silica and a greater proportion of 
«ther impurities than the fat limes. In slaking they exhibit a more 
moderate elevation of temperature; evolve less vapor; are seldom 
reduced to an impalpable homogeneous powder; yield thin paste ; 
and expand less. They are less valuable for mortar than the fat 
limes,' but are extensively employed as fertilizers. When used for 
building pnrposee they should, if practicable, be reduced to powder 
by grinding, in order to remove all diuiger of sabseqaent slsJdng. 



ovGoQi^lc 



60 LIMB AND CEKENT. [CHAP. Ill: 

66. Tumo. Good lime may be known by the following 
oharacteristios: 1. Freedom from cinders and clinkers, with not 
more than 10 percent, of other impnrities, — as silica, ainmina, etc. 
2. Chiefly in hard lamps, with bnt little dust. 3. Slakes readily 
in water, forming a very fine smooth paste, withoat any residue. 
i. DissoWes in soft water, when this is added in sufficient qnanti- 
ties. These simple tests can be readily applied to any sample of 
lime. 

67. PSXBSETDTO. As lime abstracts water from the atmospher^- 
aud is thereby slaked, it soon cmmblea into a fine powder, losing- 
all those qn^tiee which render it of valae for mortar. On thi» 
account great care mast be taken that the lime to be need is freshly 
bamed, as may be known by its being in hard Inmpa rather than 
in powder. Lime is shipped either in bnlk or in casks. Tf in bulk, 
it is impossible to preserve it for any considerable time; if in casitSr 
it may be preserved for some time by storing in a dry place. 

Common lime, when mixed to a paste with water, may be kept 
tor ut indefinite time in that condition without deterioration, it 
protected from contact with the air so that it will not dry np. It 
is cnstomary to keep the lime paste in casks, or in the wide, shallow 
boxes in which it was staked, or heaped np on the gronnd, covered 
over with the sand to be subsequently incorporated with it in mak- 
ing mortar. It is conTenient for some parpoees to keep the slaked 
lime on hand in s state of powder, which may be done in caska 
under cover, or in bnlk in a room set apart for that purpose. The 
common Hmes contain imparities which prevent a thorongh, 
uniform, and prompt slaking of the entire mass, and hence the 
necessity of slaking some days before the lime is to be used, to 
avoid all danger to the masonry by subsequent enlargement of 
Tolnme and change of condition. 

A paste or mortar of common lime will not harden under water, 
nor in continuously damp places esclnded from contact with the 
air. It will slowly harden in the air, from the surface toward the 
interior, by desiccation and the gradnal absorption of carbonic-acid 
gas, by which process is formed a snbcarbonate with an excess of 
bydrated base. 

68. Con. Lime is sold by the barrel (about 230 pounds net), 
or by the bushel (75 pounds). At Chicago the average price, in 
1898, was from 55 to 60 centa per barrel. 



ovGoQi^lc 



XKt. 2.] ETDRAULIC LIKB. 



Abt. 2. Htdkaulic Like. 



69. SbsobiptioN- Hydraulic lime is like common lime in that 
it will Blake, and differs from it in that it will harden under water. 
Hydraulic lime may be either argillaceona or Biliceons. The former 
ia derived from limeatonea contaJning . from 10 to KO per cent, of 
day, homogeneoDBly mixed with oarbonate of lime as the principal 
In^^ient; the Utter from siliceous limeBtones containing from 12 
to 18 per cent, of silica. Small percentages of oxides of iron, car- 
bonates of magueeia, etc., are generally present. 

Daring the bnrning, the carbonic acid is expelled, and the silica 
and alamina entering into combination with a portion of the lime 
form both the aUioate and the alnminate of lime, learing in the 
bnrnt prodoot an exoesB of quick or oanstic lime, which indacea 
slaking, and b«comes hydrate of lime when broaght into contact 
with water. The prodnot owes its hydranlioity to the crystallizing 
energy of the alnminate and the silicate of lime. 

Hydraolio lime is slaked by sprinkling with just sufficient water 
to slake the free lime. The free lime has a greater aridity for the 
water than the bydranlio elements, and ooQseqnently the former 
absorbs the water, expands, and diaintegratea the whole mass while 
the hydranlic ingredients are not affected. Hydraulic lime ia 
Qsaally slaked, screened, and packed in socks or barrels before 
being sent to market. . It may be kept without injury in this form 
as long as it is protected from moisture and air. 

No hydrsolic lime is manufactured in the United States. It is 
manufactured in sereral localities in Europe, notably at TeiJ and 
Scillj, in France, from which places large quantities were formerly 
brought to this country. 

Abt. 3. Htdraulio GsicEin:. 

70. CLASsmcATioir. Hydraulic cement may be divided accord- 
ing to the method of manufacture into three classes, viz. : Portland 
cement, natural cement, and pozzaolana. The first two differ from 
the third in that the ingredients of which the first two are composed 
must be roasted before they acqaire the property of hardening under 
water, while the iugredients of the third need only to be pnlrerized 
and mixed with water to a paste. 



ovGoQi^lc 



03 LIMB AND CEMENT. [CHAV. III. 

71. Portland. Portland cement is prodaced by calcining a 
mixtore containing from 75 to 80 per cent, of carbonate of lime and 
30 to 33 per cent, of claj, at snch a high temperatare that the silica 
and alamina of the clay combines with the lime of the limestone. 
As the quantity of nncombined lime is not Bnf&oient to oaose the 
mass to slake to a powder apon the addition of vater, the cement 
mnst be redoced to powder by grinding. 

To aecore a complete chemical combination of the clay and the 
lime, it ia necessary that the raw materials shall be rednoed to a 
powder and be thoroughly mixed before baming, and also necessary 
that the calcination shall take place at a high temperature. These 
are the distingniahing characteristics of the mannfacture of Portland 



In a general way Portland cement differs from natural cement 
iby being heavier, slower setting,, and stronger. 

72. Portland cement derives its name from the resemblance 
which hardened mortar made of it bears to a stone found in the isle 
of Portland, oS the south coast of England. Portland cement was 
made first in England about 1843, and in America about 1874. 

Until recent years nearly all the Portland cement used in this 
country was imported, but at present (1898) about one fifth of the 
-consumption is of domestic manufacture. The beat American 
Portland ia better than the best imported, and is sold equally cheap. 
In 1896 Portland cement was made at twenty-six places in the 
TJnited States. Kaw material suitable for the manufacture of Port- 
land cement exists in great abundance in nature, and with proper 
«are a high-class Portland cement may be produced in almost any 
part of the country. 

In recent years the amotint of cement nsed in this country has 
greatly increased, but the proportion of Portland used has increased 
at a much more rapid rate. In 1887 only about one fifth was 
Portland, while in 1897 one third was Portland. 

73. 5'atnral Cement. Tfatural cement is produced by calcining 
at a comparatively low temperatare either a natural argillaceous 
limestone or a natural magnesian limestone without pulverization 
or the admixture of other materials. The stone ia qnarried, broken 
into pieces, and burned in a kiln. The burnt cement is then 
crushed into small fragments, ground, packed, and sent to market. 

In the process of manufacture natural cement is distiagnished 



ovGoQi^lc 



AKT. 3.] HYDRADLIC CBUENT. 53 

from Portland, in nsing a natnrsl instead of an artificial mistnre 
and in calcining at a lower temperature. As a prodnct, natural 
cement is distingaisbed from Portland in weighing less, being less 
strong, and as a rule setting more qaickly. 

In Enrope in making tbis class of cement argillaceona limeetono 
is generally need, and the product is called Boman cement. In the 
United States magnesian limestone is usually employed in making 
this cement; and formerly there was great diversity in the term 
used to designate the product, domestic, American, and natural 
being employed. In the early editions of this rolnme, the antbor 
called this class of cement Rosendale, from the place where it was 
first made in this conntry — Boseudale, Ulster Co., N. Y. The 
term natural is now quite generally used, and on the whole it seems 
the best. 

74. In 1896 natural cement was made in sizty-eight places in 
serenteen states in this conntry, and it may safely be assumed that 
there is no very large area in which a atone can nqt be fonnd from 
which some grade of natural cement can be made. 

Kearly one half of the natural cement made in this conntry 
comes from Ulster Co., N. Y., and nearly half of ihe remainder 
comes from near Louisville, Kentucky. 

76. PozznoLAlTA. Pozznolana is a term applied to a combina- 
tion of silica and alumina which, when mixed with common lime 
and made into mortar, has the property of hardening under water. 
There are several classes of materials possessing this property. 

Pozznolana proper is a material of volcanic origin, and is the 
first substance known to possess the peculiar property of hydrau- 
licity. The discovery was made at Pozzuoli, near the base of 
Monnt Vesuvius, — hence the name. Vitruvius and Pliny both 
mention that pozzuolana was extensively nsed by the Romans befor» 
their day; and Yitmvius gives a formula for its use in monolithic 
masonry, which with slight variations has been followed in Italy 
ever since. It is as follows: " 12 parts pozznolana, well pulverized; 
6 parts qnartzoee sand, well washed; and 9 parts rich lime, well 
slaked." 

Trass is a volcanic earth closely resembling pozzaolana, and is 
employed substantially in the same way. It is found on the Rhine 
between Mayeuce and Cologne, and in various localities in Holland. 

Argnes is a species of ocheroos sand containing so large a pro- 



ovGoQi^lc 



64 LIMB AND CKUBKT. [CHAP.III. ' 

portion of olay that it con be mixed into a pute with water withont 
the addition of lime, and used in that stats for common mortar. 
Mixed with rich lime it yields hydrealic mortar of considerable 
energy. 

Brick dnst mixed with common lime produces a feebly hydraulic 
mortur. 

76. Slag Cement. Slag cement is by far the most important of 
the pozznolana cements. It is the prodnct obtained by mixing 
powdered slaked lime and finely pnlrerized blast-f ornace slag. The 
amonnt of slag cement mannfactared is very small as compared 
with Portland or natural cemeot, and apparently mnch more is 
mauofactared in Earope than in America. Probably most of the 
so-called pozznolana cements are slag cements. It is claimed that 
«lag cement mortar will not stain the stone laid with it. 

77. WEiaHT. Cement ix generally sold by the barrel, althongh 
not necesBarily ih a barrel. Imported cement is always sold in 
barrels, but American cement is sold in barrels, or in bags, or leas 
frequently in bulk. 

Portland cemeat usually weighs 400 pounds per barrel gross, 
aod 370 to 380 pounds net. A bag of Portland usually weighs 95 
pounds, of which four are counted a barrel. 

Natural cement made in or near Rosendale, K. Y., weighs 318 
pounds per barrel gross, and 300 net. Cement made in Akron, 
N. Y., Milwaukee, Wis., Utioa, 111., LouisTille, Ky., weighs 285 
pounds per barrel gross, and 265 net. Cloth bags usually contain 
■one third, and paper bags one fourth of a barrel. 

Slag oement weighs from 335 to 350 pounds net per barrel. 

78. Cost, The price of hydraulic cement has decreased greatly 
in recent years, owing chieSy to the development of the cement 
industry in this country. At present the competition among 
-domestic manufacturers gOTems the price. In 1898 the prices in 
car-load lots were about as follows: 

Imported Portland cement at Atlantic ports 11.50 to t3 per 
barrel in wood, and at Chicago t2 to 13.50. American Portland at 
eastern mills is II. SO to 11.75 in wood, and in the Mississippi valley 
$1.75 to t%. The price in paper bags is about 10 cents per barrel 
less than in wood, and about 15 cents per barrel cheaper in cloth 
bags than in wood— provided the cloth bags are returned to the 
jnill, freight prepud. 



ovGoQi^lc. 



ABT. 4.] TESTS OF CEUEXT. 56 

Natural cement in the Rosendale (N. Y.) district costs f. o. b. 
mills 50 cents per barrel {300 poanda net) in bnlk, 60 cents in 
paper, and 70 cents in vood. The price at the western mills in 
recent years va& 50 cents per barrel (265 poands net) in cloth (the 
sacks to be returned, freight prepaid), 55 cents in paper, aud 60 
■cents in wood. 

Slag cement is made in this country only at Chicago, where it 
«ell8 at prices bnt little below those of similar grades of Portland 
■cements. The imported pozzaolana sells sabstantially the Bame as 
-similar grades of Portland. 

Art. 4, Tests of'Cement. 

79. The yalne of a cement Taries greatly with the chemical 
composition, the temperstnre of calcination, the fineness of grind- 
ing, etc. ; and a slight variation in any one of theee items may 
greatly affect the physical properties of the product. Unless the 
process of manafactnre is conducted with the ntmost care, two lots 
of cement of the same brand are liable to differ considerably in 
physical properties. Therefore the testing of cement to determine 
its fitness for the nse proposed is a matter of very great importance. 
The properties of a cement which are examined to determine its 
-constractive valne are: (1) color, (2) thoronghness of barning, (3) 
iiotiTity, (4) soundness, (5) finenesB, (6) stieagth. 

80. CoLOB. The color of the cement powder indicates but 
little, since it is cbie&y dne to oxides of iron and manganese, which 
in no way affect the cementitiona valne; bnt for any given brand, 
Tariations in shade may indicate differences in the character of the 
Tock or in the degree of barning. 

With Portland cement, gray or greenish gray is generally con- 
sidered beat; bluish gray indicates a probable excess of lime, and 
brown an excess of clay. An nndne proportion of under-burned 
material is generally indicated by a yellowish shade, with a znarked 
difference between the color of the hard-bnrned, ongronnd particles 
retained by a fine sieve and the finer cement which passes throngh 
the sieve. 

Natural oemeuts are usually brown, but vary from very light to 
very dark. 

Slag cement has a manve tint — a delicate lilac. 



ovGoQi^lc 



66 IIME AND CEMEST. [CHAP. III. 

SI. TEOKOUOrorEBB of BUBimtQ. The higher the temperatorO' 
of baming the greater the weight of the cliaker (the ungronnd 
cement). Tiro methodB have been employed in ntilizing this priu- 
ciple as a test of the thoronghnesB of barning, viz. : (1) determine 
the weight of a unit of volume of the groond cement, and (2) 
determine the specific gravity of the cement. 

83. Weight. For any particular cement the weight varies with 
the temperatare of burning, the degree of fineness in grinding, and 
the density of packing. Other things being the same, the harder- 
burned varieties are the heavier. The finer a cement is groand the 
more bnlky it becomes, and conseqnently the less it weighs. Kence 
light weight may be canaed by laudable fine grinding or by objec- 
tionable nnder-buming. 

The weight per nnit of volume ia usually determined by sifting: 
thecementintoameasnre, and striking the top level with a straight- 
edge. In careful work the height of fail and the size of the meas- 
nring vessel are specified. The weight per cabic foot is neither 
exactly constant, nor can it be determined precisely; and is of very 
little service in determining the valae of a cement. However, it is 
often specified as one of the requirements to be fulfilled. The fol* 
lowing valnes, determined by sifting the cement with a fall of thre» 
feet into a box having a capacity of one tenth of a cubic foot, may 
be taken as fair averages for ordinary cements. The difference in 
weight for any particnlar kind is mainly dne to s difference in fine- 
ness: 

Portland TO to 90 lbs. per cublo foot, or Oi to 112 lbs. per buKhel. 
Kalural CO to S6 lbs. per cubic foot, or 63 lo 70 lbs. per busheL 

Specifications for the reception of cement frequently specify the 
net weight per barrel; but this ia a specification for quantity and 
not quality. 

83. Specific Gravity. The determination of the ^ecific gravity 
of a cement is the only real test of the thoroughness of burning. 
The specific gravity is determined by immersing a known weight 
of the cement in a liquid which will not act upon it (usually turpen- 
tine or benzine), and obtaining the volume of the liquid displaced. 
The specific gravity is equal to the weight of the cement (in 
grammes) divided by the displaced volume (in cubic centimetres). 

A variety of forms of apparatus for use in making this test ar& 



ovGoQi^lc 



ART. 4.] TBSTS OP CEMEHT. 57 

npOQ the market, bnt as Beveral of the volameters in ordinary nse 
iti chemical and physical laboratories are snitable for this porpose, 
it is unneceaaary to describe any of them here. Ab a alight differ- 
ence in specific gravity is frequently accompanied by a considerable 
difference in the qnality of the cement, great care is necessary in 
making the test. It is necessary that all the air-bnbblea contained 
in the cement powder be eliminated, bo that the volume obtained 
be that of the cement particles only. The cement shoald be passed 
through a sieve, say No, 80, to eliminate the lamps. The tempera- 
ture of the liquid should not he above 60" Fahr., and should not 
change during the test. A change of 1° C in the turpentine 
between the readings of the volumeter will make a difference of 
0.08 in the resulting specific gravity. 

The specific gravity of Portland cement varies from 3.00 to 3.S5, 
usually betweeu 3.05 and 3.17. Natural cement varies from %,75 to 
3.05, and is asnally between 2.80 and 3.00. Slag cement has a 
specific gravity of 2.72 to 2.76. The specific gravity of cement 
decreases with age owing to the absorption of water and carbonic acid 
from the air. 

German authorities state that the specific gravity of fresh Fort- 
land cement is between 3.13 and 3.^5. English specifications re- 
quire 3.10 for fresh Fortland and 3.07 for cement 3 months old. 
By the specifications of the Canadian Society of Civil Engineers 
the minimum for fresh Portland is 3.09. Many specifications fix 
3.00 or 3.05 for the lower limit. 

84. AcnriTT. When cement powder is mixed with water to a 
plastic condition and allowed to stand, the cement chemically com- 
bines with the water and the entire mass gradually becomes firm 
and hard. This process of solidifying is called setting. Cements 
differ very widely in their rate and manner of selling. Some 
occnpy bat a few minutes in the operation, while others require 
several hours. Some begin to set comparatively early and take 
considerable time to complete the process, while others stand con- 
siderahle time without apparent change and then set very quickly. 

A knowledge of the activity of a cement is of importance both 
in testing and in using a cement, since its strength ia seriously 
impaired if the mortar is distnrbed after it has begun to set. 
Ordinarily the moderately slow-setting cements are preferable, since 
they need not be handled so rapidly and may be mixed in larger 



ovGoQi^lc 



^58 LIUE AND CEUENT. [CHAP. III. 

qnaa titles; bat in some cases it is necessary to nse a rapid-setting 

- cement, as for example vben an Inflow of water is to be prevented. 

To determine the rate of setting, points have been arbitrarily 

fixed where the set Is said to begin and to end. It is very difficnlt 

to determine these points with exactness, particalarly the latter; 

bnt an exact determioation is not necessary to jadge of the fitness 

«f a cement for a partlcnlar use. For this purpose it is ordinarily 

snfficient.to say that a mortar has begun to set when It has lost its 
..plasticity, i.e., when its form cannot be altered without prodacing 
■ a fracture-; and that it has set hard when It will resist a slight 

pressure of the thnmb-nail. Cements will increase iu hardness long 
'.after they can not be indented with the thnmb-nail. 

For an accurate determination of rate of set two standards are 

in nse, viz. : Oillmore's and the Qerman. 

S6. Gillmore'i Test. Mix the cement with water to a stifE 
aplastic mortar (see §§ 103-4), and make a cake or pat 2 or 3 inches 

in diameter and about ^ inch thick. The mortar is said to have - 

begun to set when it will just support a wire ^^-iuch in diameter 
"weighing J pound, and to have " set hard " when it will bear a jfj- 

inch wire weigliing 1 pound, A loaded wire used for this purpose 
iis. frequently called a Vicat needle, after Vicat, its inventor. The 

interval between the time of adding the water and the time when 
■the light wire is just supported is the time of beginning to set, and 

the interval between the time the light wire ia supported and the 
'time when the heavy one is just supported is the time of setting, 

86. German Teat.* " A slow-setting cement (one setting in not 
'less than two hoars) shall be mixed three minutes, and a qnick- 

setting cement (one setting in less than two hours) one minute, with 

water to a stiS paste. The consistency of the cement paste for this 

cake shall be such that, when wrought with a trowel on the plate, 
' the paste will only begin to ran towards the edge of the same after 

the paste haa been repeatedly jarred. As a rule, 27 to 30 per cent. 

of water will suffice to give the necessary consistency to a Portland 

cement paste, f 

" For the exact determination of the time of beginning to aet, 

and for determining the time of setting, a standard needle 300 

■Bpaclfleatlonsot the Fmsslan Hlnlatsr of Pnblto Works, lalj 38, IBST. 
t Apparentl; this mortar la more moist thui the "plaatlo mortar" ordlnaTQf 
. empLoyed ia tbls 0001X17 {■«« tt 103-1). 



ovGoQi^lc 



ABT. 4.] TESTS OF CEMENT. 59 

£raiiLmee (11 oz.) in weight and 1 squaro luillimetire (0.0006 square 
inch) in cross-section is used. A metal ring 4 centimetres (1.575 
inches) in height and 8 centimetres (3.15 inchee) clear diameter 
(inside) la placed on a glass plate, filled with cement paste of the 
above coDstatency, and brcoght under the needle.* The moment 
at which the needle is no longer capable of completely penetrating 
the cement cake is considered the beginning of the time of setting. 
The time elapaing between this and the moment when the standard 
needle no longer leaves an appreciable impreaeion on the hardened 
«ake is considered the time of setting." 

To facilitate the making of this teat, an apparatos ia provided 
which GonBista of a light rod freely sliding throngh an arm; and 
carrying in its lower end the penetrating needle. The amoaut of 
penetration is read by an index moving over a gradnated scale. 

87. Elements Affecting Bate of Set. The amount of water 
employed is important. For data as to the amoant of water to be 
oaed, see gg 103-1. The less the water, the more rapid the set. 

It is nsnally specified that the temperature of tbe water and air 
shall be from eo" to 65" F. The higher the temperatnre, the more 
rapid the set. To prevent the anrface of the teat specimen from 
^ hardening by drying, it ia specified that the pat shall be immersed 
in water at 60° to 66° F. The setting under water is much alower 
than in air even thongh the air be aatnrated with moiatnre and be 
at the aame temperatnre as the water, dne to the mechanical action 
of the water. 

Other things being the same, the finer the cement is ground the 
-quicker it sets. 

Cements nsually become slower setting with age, particularly if 
exposed to the air — Portlands nsnally but slightly. 

The standard testa for activity are usually made on neat cement 
on accoant of the interference of the sand grains with the descent 
of the needle. The rate of setting of neat mortar gives bat little 
Indication of what the action may be with aand. Sand increases 
the time of aetting — but very differently for different cements. 
With some cements a mortar composed of one part cement to three 
parts sand will require twice as long to set as a neat mortar, whits 
with other cements the time will be eight or ten times as long. 



ovGoQi^lc 



60 LIUB AND CEKBNT. [CHAP. III. 

Sulphate of lime (plaster of Paria) greatly influences the rate of 
Betting of PortJand cements. The addition of 1 or 2 per cent i» 
sufficient to change the time of setting from a few mianteG to aeveral 
hours. Cement vhich has been made slow-setting by the addition of 
sulphate of lime, nsaally becomes quick -setting again after exposure 
to the air; cement which has not had its time of setting changed 
by the addition of sulphate of lime, usually becomes slower setting 
with age and may finally lose the power of setting. Cement which 
has become slow-setting by the addition of sulphate of lime will 
become quick-setting if mixed with a solution of carbonate of 
soda. 

A weak solution of chloride of lime usually cansee the cement 
to set more slowly; while a strong solution usually accelerates the 
rate of setting. 

SB. Time of Bet. A few of the quickest natural cements when 
' tested neat with the minimum of water will begin to set in 5 to 10 
minutes, and set hard in 15 to 20 miautes; while the majority will 
begin to set in 30 to 30 mtDutes and will set hard in 40 to 60 
minutes; and a few of the slowest will not begin to set under 60 
minutes. 

The quickest of the Portlands will begin to set in 20 to 40 min- 
utes; but the majority will not begin to set nnder 2 or 3 hours, 
and will not set hard nnder 6 or 8 hours. The 18S7 standard 
Qerman specifications reject a Portland cement which begins to 
set in less than 30 minutes or which sets hard in less than 
3 hours. 

89. SOTnrsMEBB. Sonndneaa refers to me ability of a cement to 
retain its strength and form unimpaired for an indefinite period. 
SouuduesB is a most important element; since if a cement ultimately 
loses its streugth it is worthless, and if it Anally expands it becomes 
a destructive agent. A cement may be unsound because of the 
presence in it of some active elements which cause the mortar to 
expand or contract in setting, or the unsoundness may be dne to 
exterior agencies which act upon the ingredients of the cement. 
Most unsound cements fail by swelling and cracking under the 
action of expansires; but sometimes the mortar fails by a gradual 
softening of the mass without material change of form. The ex- 
pansive action is usually dne to free lime or free magnesia in the 
oement, but may be caused by sulphur compounds. The priuoipal 



ovGoQi^lc 



JlSI. 4.] TB3TS OF CBUEITT. . 61 

«xterior agencies acting npon a cement are air, sea-water, and 
■extremes of heat and cold. 

The presence of small quantities of free lime in the cement is a 
freqnent «aase of ansoandneas. The lime slakes, and caasee the 
mortar to svell and crack — and perhaps finall; disintegrate. The 
-degree of heat employed in the burning, and the fiaeness, modify 
the effect of the free lime. Lime bnrned at a high heat slakes 
more slowly than when barned at a low temperatare, and is there- 
fore more likely to be iajarions. Finely gronnd lime slakes more 
qaickly than coarsely gronnd, and hence with fine cement the lime 
may slake before the cement has set, and therefore do no harm. 
The lime in finely gronnd cements will air-slake sooner than that ia 
-coarsely gronnd. 

Free magnesia in cement acts very mnch like free lime. The 
action of the magnesia is mach slower than that of lime, and hence 
its presence is a more serions defect, since it is less likely to be 
detected before the cement is ased. The effect of magnesia in 
cement is not thoronghly understood, bat seems to vary with the 
composition of the cement, the degree of burning, and the amount 
of water used in mixing. It was formerly held that 1( or 2 per 
cent, of magnesia in Portland cement was dangerous; bnt it is now 
known that 5 per cent, is not injnrions, while 8 per cent, may pro- 
duce expansion. Since many of tbe natnral cements are made of 
magnesinm limestone, they contain mnch more magnesia than 
Portland cements; bnt chemists are not agreed as to the manner in 
which the different constituents are combined, and consequently are 
not agreed either as to the amount or effect of free m^nesia in such 
a cement. Fortunately, it is not necessary to resort to a chemical 
analysis to determine the amonnt of lime or magnesia present, for 
a cement which snccesaf ally stands the ordinary test for soundness 
-(§ 92) for 7, or at moat 28 days, may be used with reasonable con- 
'fidence. 

The effect of lime and magnesia seems to be more seriona 
in water than in air, and greater in sea-water' than in fresh 
water. 

90. The action of sulphur in a cement is extremely rariable, 
depending upon the state in which it may exist and upon the 
natnre of the cement. Sulphur may occur naturally in the cement 
or may be added in the form of sulphate of lime (plaster of Paris) 



ovGoQi^lc 



63 LIHK AKD CEMENT, [CHA.P. III^ 

to rotard the time of set (§ 87). TJnder certfun conditioiu the 
Bnlphnr ma; form snlphides, which on espoeare to the air oxidize 
and form aalphates and canae the mortar to decrease in strength. 
Many, if not all, of tbe slag cements contain an excess of salphides, 
and are therefore nnfit for oae in the air, particularly a very dry 
atmoephere, althongh under water they may give satisfactory reaalta 
and compare favorably with Forthind cement. 

91. Teita of BoundBeu. Sereral methods of testing sonndness 
have been recommended. Of those mentioned below, tbe first two 
are called cold tests, since the mortar is tested at ordinary tempera- 
tares; and the others accelerated or hot tests. 

92. Th« Fat Test. The ordinary method of testing sonndness- 
is to make small cakea or pats of neat mortw 3 or 4 inches in 
diameter, aboot half an inch thick and h&riog thin edges, upon a- 
sheet of glass, and examine from day to day, for 38 days (if 
possible), to see if they show any cracks or signs of distortion. 
The amount of water nsed in mixing (see § 104) within reason- 
able limits seems to have no material effect on the rosnlt. The 
German standard specifioations reqnire the cake to be kept 24 
honrs in a closed box or nnder a damp cloth, and then stored in 
water. The French, to make snre that the pats do not get dry 
before immeraion, recommend that the cakes be immersed immedi- 
ately after mixing without waiting for the mortar to set. Some- 
really sound natural cements will disintegrate if immersed before 
setting has begun. 

The first evidence of had quality is the loosening of the pat from 
the glass, which generally takes place, if at all, within one or two 
days. Good cement will remain firmly attached to the glass for two- 
weeks at least. The cracks due to expansion occur usually at the 
edges of the pat, and radiate from the center. These cracks should 
not be confnsed with irregular hair-like shrinkage cracks, which 
appear over the entire surface when the pats are made too wet and 
dry oat too much while setting. 

93. A cement high in snlphides, as for example one made of 
blast-furnace slag, will successfully pass the above, the usual, test 
for soundness; and still the mortar when exposed in the air will 
show a marked decrease in strength and perhaps finally dis- 
integrate. The presence of an excess of snlphides may be sns- 
peoted in any cement made- from blast-fnmace slag. A slag 



ovGoQi^lc 



AET. 4.] TEBT3 OF CBMBNT. . 63: 

cemeiit u indicated by a manve or delicate lilac tint of the dry- 
powder. 

Tlierefore, in making the pat teat, it is wise to expose a pat in 
the air as well as one under water. Any ealphidea in the cement. 
will be revealed by brown or yellowish blotches on the pat exposed 
in air, and also by a greenish color of the interior of the pat exposed 
nnder water. The pat in air is not as good a test of expansives as- 
the pat ander water, owing to a possible deficiency of water and to^ 
greater shrinkage cracks. 

If there are any considerable indications of anlphidee, before- 
accepting the cement a chemical analysis shonld be made to deter- 
' mine the snlphnr and the probable nltimato action of the cement. 
Any cement containing sulphides in appreciable quantities is at. 
least doubtfal and shonld probably be rejected. Slag cements- 
asQally contain 1 to 1.5 per cent, of snlphides. 

Another excellent method of examining for the presence of enl- 
phidea is, in making the test for tensile strength (gg 99-111^), ta 
store part of the briqnettes in air and part in water. Any material, 
difierence in strength between the two lots is sufficient ground for- 
rejecting the cement for nse in a dry place. Of course due con- 
aideratiou should be given to the possible effect of evaporation of: 
water from the briqnettea stored in air, 

64. Expansion Test, Various experimenters test the soundness 
of cement by measaring the expansion of a bar of cement mortar. 
The French Commission recommend the measnremeat of the expan- 
sion of a bar 33 inches long by ^ inch square, or the measurement 
of the increase of circumference of a cylinder. The German 
standard tests require the measarement of the increase in length of 
a prism i inches long by 3 inches square. The apparatus for 
m^ing these tests can be had in the market. The testa require 
very delicate manipulation to secure reliable reanlte. 

96. Accelerated Testt. The ordinary teats extending over a 
reaeonable period, sometimes fail to detect unsoundness; and many 
efforts have been made to utilize heat to accelerate the action, with 
a Tiew of determining from the effect of heat during a short tim^ 
what would be the action in a longer period under normal condi- 
tions. Some of these tests have been fairly succesafnl, but none 
have been extensively employed. It is difficult to interpret the 
teste, as the results vary with the per cent of lime, magnesia, snt- 



ovGoQi^lc 



«4 LIMB AKD CEIIEKT. [CHAP. III. 

phatfis, «tc., present, aod with their proportions relatire to eacb 
other and to the whole. There is a great diversity as to the value 
of accelerated tests. Many natural cemeots which go all to pieces 
in the accelerated tests, particularly the boiling test, still stand 
well in octnal service. This is a strong argnmeot against drawing 
adverse conclnsions from accelerated tests when applied to Portland 
cement. 

The warm-water lest, proposed by Mr. Faija,* a British 
Authority, is made with a covered vessel partly fall of water 
maintained at a temperatare of 100° to 116° F., in the upper 
part of which the pat tg placed until set. When the pat is 
set, it is placed in the water for 24 honrs. If the cement remains 
firmly attached to the glass and shows no cracks, it is very probably 
sound. 

The hot-water test, proposed by Mr. Maclay.f an American 
authority, is snbstantislly like Faija's test above, except that 
Maclay recommends 195° to 300° F. 

The hailing test, suggested by Professor Tetmajer, the Swiss 
authority, consists in placing the mortar in cold water immediately 
after mixing, then gradually raising the temperature to boiling 
after about an hoar, and boiling for three hours. The test 
apecimeu consists of s small ball of such a consistency that when 
flattened to half its diameter it neither cracks nor runs at the 
edges. 

The kiln tetts consist of exposing a small cake of cement 
mortar, after it has set, to a temperature of 110° to 120° G. 
{166° to 248° F.) in a drying oven until all tlie water is driven off. 
If no edge cracks appear, the cement is considered of constant 
Tolume. 

The flame test is mads by placing a ball of the cement paste, 
About 2 inches in diameter, on a wire gauge and applying the 
flame of a Sanson barner gradually nntil at the end of an hour the 
temperature is about 90° C. (194° F.). The heat is then in- 
creased until the lower part of the ball becomes red-hot. The 
appearance of oraoks probably indicates the presence of an expansive 
element. 



ovGoQi^lc 



ART. 4.J TESTS OF CEMBKT. 65 

The chloride-of-lime lest is to mix the paste for the cakes with 
■A solution of 40 grammee of calciom chloride per liter of vater, 
allow to Bet, immerBe in the same solntion for 34 honre, and then 
■«za[nine tor checking and softening. The chloride of lime accel- 
erates the hydration of the free lime. The chloride in the eolation 
Qsed in mixing causes the slaking before setting of only so mocb of 
the free lime as is not objectionable in the cement. The chloride 
of calcium has no effect upon free magnesia. 

96. FUTEHESB. The question of fineness is wholly a matter of 
eoonomy. Oement until ground is a mass of partially Titrified 
clinker, which is not affected by water, and which has no setting 
power. It is only after it is ground that the addition of water 
indnces crystallization. Gonseqaently the coarse particles in a 
cement have no setting power whatever, and may for practical 
purposes be considered as so mnch sand and essentially an adnl- 
(eraot. 

There is another reason why cement shonld be well ground. A 
mortar or concrete being composed of a certain quantity of inert 
material bound together by cement, it is evident that to secure a 
strong mortar or concrete it is essential that each piece of aggregate 
■ball be entirely surrounded by the cementing material, bo that no 
two piecee are in actual contact. Obviously, then, the finer a 
cement the greater surface will a given weight cover, and the more 
economy will there be in its use. 

Fine cement can be produced by the maanfacturers in three 
ways: 1, by supplying the mill with comparatively soft, under-burnt 
took, which is easily reduced to powder; 2, by more thorough 
grinding; or 3, by bolting through a sieve and returning the 
unground particles to the mill. The first process pmduces an in- 
ferior quality of cement, while the second and third add to the cost 
of manufacture. 

It ia possible to reduce a cement to an impalpable powder, but 
the proper degree of fineness is reached when it becomes cheaper to 
nse more cement in proportion to the aggregate tban to pay the 
extra coat of additional grinding. 

97. Heasoring Fineness. The degree of fineness is determined 
by weighing the per cent, which will not pass through sieves of a 
specified number of meshes per square inch. In the paat, three 
.sieves have been used for this purpose, viz., sieves having 60, 75, 



ovGoQi^lc 



66 LIME AMD CEMENT. [CHAP. III. 

and 100 meshes per linear inch, or 2,500, 5,625, and 10,000 meshes, 
per sqaare inch reapectirely. These sieves are nsnally referred to 
by the nnmber of meshes per linear inch, the first being known aa 
No. SO, the second as No. 75, and the third aa No. 100. In each 
case the diameter of the mesh is abont eqaal to that of the wire. 
The per cent, left on the coareer sieves has no special significance, 
and hence the use of more than one sieve has been a!mos1 aban- 
doned. More recently in this conntry a No. 120 sieve (14,400 
meahea per square inch) has been employed, and sometimes a 
No. 200. On the continent of Enrope the sieve generally used has 
70 mesbes per linear centimetre, corresponding to 175 meshes per 
linear inch (30,625 per square inch). 

9S. Data on Fi&eneu. Nearly all Portland cements are so 
ground as not to leave more than 20 per cent, on a No. 100 sieve, 
and many of them will not leave more than 10 per cent, ob & 
No. 100 sieve or more than 20 per cent, on a No. 200 sieve; and 
some mannfactarers claim less than 10 per cent, on a No. 200 sieve. 
As a rule, American Portlands are finer gronod than German, and 
German finer than English. 

Most of the natural cements are nsually gronnd so as to give 
not more than 20 per cent, on the No. 100 sieve, and many of them 
will not leave more than 10 per cent, on the No. 100 sieve, and a 
few vill leave only 10 per cent, on the No. 200 sieve. 

A common specification is that not more than 10 per cent, shall 
be left on a No. 50 sieve. Such a test simply prevents the adultera- 
tion of the cement with very coarse particles, but does not insure 
any considerable proportion of impalpable powder (approximately 
that which will pass a No. 200 sieve), which alone gives value to 
the cement,* 

Since the natural cement is not so hard burned as the Portland, 
there is more impalpable powder in proportion to the per cent, left 
on the teat sieve than with the Portland; and consequently a severe 
test for fineness is not as important for natural cement as for 

*Tbcre hae receutly been iatrodac^ an article called sand-oemeDt, which la 
made by mixing oement and silica aand and gricdlng the mixture. The grind' 
iDg ot Uie mixture greatly Increases the flneness ot the cement. A mixture at 
1 part Dement and 3 parts silica sand when reground will carry nearly as much 
sand as the original pure cement, which shows the strildng sSeot of tlie Teiy 
Aoe grinding ot the cement. 



ovGoQi^lc 



ART. 4.] TESTS OF CEMENT. 67 

Portland, Farther, since nataral cement is mnob clieaper than 
Portland, it is more economical to use more cement than to 
require extra flneneBS. Again, since nataral cement is weaker, 
it IB not ordinarily naed with as large a proportion of sand as 
Portland, and henoe fineness is not aa important with nataral as with 
Portland. 

For varions specifications for fineness, see Art. 5, pages 78d- 
78&, partionlarly Tables 10c and lOd, pages 78/, 78g. 

99. TeitsILB Stkeksth. The strength of .cement mortar is 
nsnally determined by submitting a specimen having a cross section 
of 1 square inch to a tensile stress. The reason for adopting tensile 
teste instead of compresBive is the greater ease of making the former 
and the less variation in the resnltB. Mortar is eight to ten times 
as strong in compression as in teosion. 

The accnrate determination of the tensile strength of cement is 
a mnch less simple process than at first appears. Many things, 
apparently of minor importance, exert snch a marked infiuence apon 
the results that it ia only by the greatest care that trustworthy testa 
can he made. The rariations in the resnlts of different experienced. 
operators working by the same method and upon the same material 
are frequently very large. In one particular test case,* the lowest 
of nine results was but 30 per cent, of the highest, the remainder 
being evenly distributed between the two extremes. Similar varia- 
tions are not at all unusual. The variation is chiefly due to differ- 
ences in making the test specimen. Unfortunately, there is at 
present no detailed standard method of procedure in making the 
tests, and conseqnently all that can be done is to observe with the 
most oonscientions care the rules that have been formnlated, and 
draw the specifications in accordance with the personal equation of 
the one to make the tests. 

100. Heat vs. Sand Tests. It is very common to test neat-cement 
mortar, but there are two serious objections to this practice. First, 
most neat cements decrease in tensQe strength after a time. This 
decrease seems to be due to a change in the molecular structure of 
the cement, the crystals growing larger with increase of age, thna 
producing a crowding which results in a decrease of the tensile 
strength. This decrease is most marked with high-grade Portlands 

'Engineering 2ltat, ToL zxxv. pp. IBIMil. 



ovGoQi^lc 



68 LIUB AND CEMENT. [CHAK III. 

vbich attain their strength rapidly, and naaally occurs between 
three months and a year. A second objection to neat teste is that 
coarselj-gronnd cements show greater strength than finely-groand 
cements, although the latter mixed with the osoal proportion of 
sand will give the greater strength. 

On the other hand, more skill is required to secare uniform 
resalts with sand than with neat cement. 

101. The Sand. The quality of the sand employed is of great 
importance, for sands looking alike and sifted through the same 
sieve gire reenlts varying 30 to 40 per cent. 

The standard sand employed in the official Oerman testa is a 
natnral quartz sand obtained at Freienwalde on the Oder, paasmg 
a sieve of 60 meshes per square centimetre (ZO per linear inch) and 
canght upon a sieve of 120 meshes per square centimetre (SS per 
linear inch). The standard "sand" recommended by the Com- 
mittee of the American Society of Civil Engineers is crushed quartz, 
used in tbe maanfacture of sand-paper, which passes a No. 20 sieve 
(wire Ko. 28 Stnba's gangs) and is caught on a Ko. HO sieve (wire 
ifo. 30 Stnbs's gauge), the grains being from 0.03 to 0.03 inch in 



The crushed quartz consists of sharp, glossy splinters, while the 
standard German sand is composed of nearly spherical grains having 
a rough surface like ground glass. The quartz contains about 50 
per cent, of voids, while the German standard sand oontMns only 
about 40 (see Table 10^, page 79t.) The crushed qaartz will give 
less strength than standard sand. Ordinarily common building sand 
will give a higher strength than standard sand, since nsually tbe 
former oonslsta of grains having a greater variety of sizes, and con- 
eeqnently there are fewer voids to be filled by the cement (see Table 
lO^r, page 79i.) 

102. The Amonnt of Water. Tbe amount of water necessary 
to make the strongest mortar varies with each cement. It is com- 
monly expressed in per cents, by weight, although in part at least 
it depends upon volume. The Tariation in the amonnt of water 
required depends upon the degree of fineness, the specific gravity, 
the weight per nnit of Tolume, and the chemical composition. If 
the cement ia coarsely ground, the voids are less, and consequently 
the volume of water reqnired is less. If the specific gravity of one 
.cement is greater than that of another, equal volumes of cement 



ovGoQi^lc 



ART. 4.] TESTS OF CKMBNT. 69 

vill reqaire different Tolnmesof water. The chemical composition 
has the greatest inSaeDce upon the amount of water neceBsary. 
Part of the water is req aired to combine chemically with the cement, 
and part acts physically in reducing the cement to a plastic mass; 
and the portion required for each of these effects differs with differ- 
ent cements. The dryness and porosity of the sand mb.y also 
appreciably affect the quantity of water required. The finer the 
sand, the greater the amount of water required. Again, the same 
consistency may be arrived at in two ways — by using a small quan- 
tity of water and working thoronghly, or by using a larger 
quantity and working less. (For instrnotions coQcerning mixing, 
see § 106). 

Attempts have been made to eetabliab a standard consistency, 
hut there is no constant relation between the consistency and the 
maximum strength. With one cement a particular consistenoy may 
give maximum strength, while with another cement a different con- 
sistency may be required to develop the greatest strength. The 
relationship between consistency and strength will vary also with 
the details of the experiment. In reporting the results of tests the 
quantity of water employed should be stated. 

There are two distinct standards of consistency for the mortar 
employed in testing cements, — the plastio and the dry. 

103. Plastic Mortar. This grade of mortar is that com- 
monly employed in the United States and England, and is fre- 
qnently naed in France.* Tbere are two methods of identifying 
this degree of consistency, viz. : the Tetmajer method and the 
Boulogne method. The Tetmajer method reqnires more water 
than the Boulogne method — for Portland this excess is abont 3 
per cent, of the weight of the cement, and for natural abont S 
per cent. 

The Tetmajer method is much used on the continent of Europe. 
It is as follows: The plasticity shall be such that a rod 0.1 of an 
inch in diameter and weighing 0.66 pounds will penetrate 1.35 
inches into a box 3 inches in diameter and 1.57 inches deep, filled 
with the mortar.f 

The Boulogne method is treqnently used in France. It is aa 

* Bee loot note, page 71. 

f For Ml Ulustratloa ot tbe apparatua, sea Trans, Amer. Soo. ot 0, E., toL z^ 
P.U. 



ovGoQi^lc 



70 LIME AND CEMENT. [CHAP. III. 

folIoTH: * " The qnantity of water ie aficertaiQed bj a prelimioarT' 
experiment. It is recommended to commence with a rather anmller 
qnantity of water than may be ultimately raqnired, and then to 
make freeh miiinga with a Blight additional qnantity of vat«r. 
The mortar is to be Tigoronsly worked for five minntes with a trowel 
on a marble dab to bring it to the required consiBtency, after which 
the foar following testa are to be applied to determine whether the 
proportion of water is correct: 1. The consistency of the mortar 
shonld not change if it be ganged for an additional period of three 
minntes after the initial five minates. Z. A small qaantity of the 
mortar dropped from the trowel upon the marble slab from a height 
of aboat 0.50 metres (20 inches) Bhoald leave the trowel clean, and 
retun its form approximately without cracking. 3. A small qnan- 
tity of the mortar worked gently in the hands should be easUy 
moulded into a ball, on the surface of which water ahoald appear. 
When this ball is dropped from a height of 0.50 metres (20 inches), 
it should retain a rounded shape without cracking. 4. If a slightly 
smaller quantity of water be nsed, the mortar should he ornmbly, 
Mid crack when dropped upon the slab. On the other hand, the 
addition of a further qnantity of water — 1 to 2 per cent, of the 
weight of the cement — would soften the mortar, rendering it more 
sticky, and preventing it from retaining its form when allowed to 
fall upon the slab." 

104. With any particular cement the exact amount of water to 
prodnce the above degree of plasticity can be determined only by 
trial, but as a rule the qnanti^ reqnJred by the Boulogne method 
will he about as follows: 

For neat cement: Portland, 23 to 36 percent.-, natural, from 
30 to 40, usually from 32 to 36 per cent. 

For 1 part cement to 1 part sand: Portland cement, 13 to 15 
per cent, of the total weight of cement and sand; natnral, 1? to 
20, usnally 18 to 19 per cent. 

For 1 part cement to 2 parts sand: Portland, 12 to 13 per cent. 
of the total weight of the sand and cement; natural, 12 to 16, 
DBoally 13 to 15 per cent. 

For 1 part cement to 3 parts sand: Portland, 11 to 12 per cent. 

■ From alMttacU of lust at C. K. 



ovGoQi^lc 



ART. 4.] TESTS OF CEHEtTT. 71 

at the total weight of the sand and cement; nataral, 1% to 13 pel 
«ent. 

106. Dry Mortar. This, grade at mortar is employed in the 
Oerman and French * governmental teste of tensile strength. The 
rales for the identification of this degree of consistency are not vei; 
Bpecific. " Dry mortars" are nsnally described as being" as damp 
as moist earth." 

The German government does not recognize tensile tests of neat 
cement mortar; bat for 1 to 3 sand mortars specifies that the weight 
of water oaed for Portland cement shall be eqnal to 10 per cent, of 
the total weight of the sand and cement. 

The French Commission gires a mlef for 1 to 2, 1 to 3, and 
1 to 6 mortars, with either Portland or nataral cement, whioh is 
eqsiraleDt to the following formnla: 

w = \WR + iS, 

ia whidk to = the weight, in grammes, of water required f<Nr 
1,000 grammes of the sand and cement; 
W = the weight, in grammes, of water required to re- 
dnce 1,000 grammes of neat cement to plastio 
mortar (see g 104) ; 
It = the ratio of the weight of the cement to the weight 
of the sand and cement. 
For a 1 to 3 mortar the preceding formnla gives 8.5 per cent., 
which seems to show that the French standard reqaires less water 
than the German. 

The cement laboratory of the city of Philadelphia employs the 
»boTe formula, but Dses 60 for the constant iDStead of 4S. For a 
1 to 3 mortar, the Philadelphia formula gives 10 per cent., which 
agrees with the German standard, 

106, Mixing the Hortar. The sand and cement shonld be 
thoroughly mixed dry, and the water required to rednce the mass 
to the proper consistency shonld be added all at once. The mixing 

• The Frenoli GommlBeloD TeoommeodB dry mortar tor tanslla taets oolj; uid 
Also Tooommends that, attor an Internatlooal agreemeDt to that elteot, plaatla 
mortars be eraplOTdd for all t«Bts to the exolimlOD of dry mortars. 

t Carter and Oleaeler'a CodcIhsIods adopted by the Frenah Conunlsalon la 
telerenoe to Teats ot Oementa, p. 21. 



D.qilizMb,G00>^le 



72 



LIVE AND CEMENT. 



[chap, riu 



shoald be prompt and thorough. The mass shonld not be dmply 
tnrned, but tbe mortar should be rubbed agunst the top of th& 
Blate or glass mixing-table vith a trowel, or in a mortar with a 
pestle. Insufficient working greatly decreases tbe strength of tbft 
mortar— frequently one half. The inexperienced operator is very 
liable to nse too mncfa water and too little labor. With a slow- 
Betting cement a kilogramme of the dry materials should be strongly 
and rapidly rubbed for not less than 5 minutes, when the consist- 
ency should be such that it will not be changed by an additional 
mixing for 3 minutes. 

Usually the mortar is mixed with a trowel on a atone slab; but 
when many batches are required, there is a decided advantage in 
mixing the mortar with a hoe in a short Y-shaped trough on th» 
floor. Various machines have been devised with wbich to mix the 
mortar. The jig mixer * is an apparatus in which the materials ar& 
placed in a covered cup, and shaken rapidly up and down. The 
Eaija mixer f consists of a cylindrical pan in which a mixer 
formed of four blades revolves. The- 

ir-^v — -:f- ( latter seems to give the better result^ 

j bat neither ue nsed to any considerable 
I extent. 

107. The Form of Briquette. The 
'f briquetterecommeaded by the Committee 
/ of the American Society of Civil En- 
• gineers, Pig. 2, is the form ordinarily 
used in this country and in England. 
The form generally employed in con- 
tinental Europe is somewhat similar to- 
the above, except that the section is 5- 
square centimetres (0.8 square inch) and 
the reduction to prodnce the minimum 
section is by very much more abrupt 
cnrres.J The latter form gives only 70 
to 80 per cent, ae much strength as the former, 

■ For llliiBtr&t«d deearlptlan, see Trans. Amer. Soc. ot O. E., vol. zit. p. SOO-1. 

t For BiiUsh torm, see Traas. Am. Soc. ot C. E., vol. xvll. p. 333 ; luid lor th» 
Amerloaii torm, see catalogue ot Blahl^ Bros. Teatlnt; Machine Co., Philsdelptila. 

t For an elaborate dlaansslon ol the beat torm of briquette, see Johnton'ft 
KateilaU ot Oonstmiition, p. tSS-SB. 




oiGoQi^lc 



ART. 4.] TESTS OF CEKEKT. 7* 

The moalds are made of brass aDd are single or multiple, the 
latter being preferred where a great nnmber of briqaettee is required, 
The moolds are in two parts, to facilitate lemoTol from the 
briqnette withoat breaking it. 

108. Konlding the Briquette. Id moalding the briquette there 
are two general methods employed, corresponding to the two stand- 
ard consiBtenciea of the mortar. 

109. Plaalic Mortar. The rales of this section (109) apply to 
hand-moulding. 

The Committee of the American Society of Civil Engineers* 
recommendations areas foUowe: " The moalds while being charged 
shonld be laid directly on glaas, alate, or eome non-absorbing- 
material. The mortar should be firmly pressed into the monlda. 
with a trowel, without ramming, and struck off level. The moald- 
ing mast be completed before incipient setting begins. As soon as. 
the briquettes are hard enoagh to bear it, they should be taken 
frem the moalds and kept covered with a damp cloth until they 
are immersed." 

The French Commieaion recommends the following method;*' 
" The moalds are placed npon a plate of marble or polished met^ 
which has been well cleaned and labbed with an oiled cloth. Six 
moulds are filled from each ganging if the cement be sluw-eetting, 
and foar if it be quick-aetting. Salficieat material is at once placed 
in each mould to more than fill it. The mortar is pressed into the 
moald with the fingers so as to leave no voids, and the side of the 
moald tapped several times with the trowel to assist in disengaging 
the babbles of air. The excess of mortar is then removed by slid- 
ing a knife-blade over the top of the mould so as to prodnce no 
compression npon the mortar. The briquettes are removed from 
the mould when sufficiently firm, and are allowed to remain tor 21 
bonrs npon the plato in a moist atmosphere, protected from currents 
of air or the direct rays of the sun, and at a nearly constant tem- 
perature of 15"* to 18° C. (59° to ei-i" F.)." 

110. Various machines have been devised for moulding bri- 
qnettee of plaatio mortar, but none are used to any considerablA 
extent, t 

* Cuter uid Oleoeler'B OoaaluBlons kdoptad b7 the Frenoh Oammlselon In 
refsreooe to Teats ol Oemeuts, p. 33. 

t Tor HI lUartmted deMrlptlonotBuMU'slevermMhliu^iM mane. Amer. Boo. 



ovGoQi^lc 



74 LIUE AND CEUBNT. [CHAP. III. 

In Canada, and to some extent io England, the briquettes are 
moalded bj applying a pressure of 20 ponnds per sqaare incb on tbe 
Bnrface of tbe briquette.* Some adyocate a presaare of 1,000 to 
1,500 ponnds npon the upper face of tbe briqaette.| 

111. Dry Mortar. The rules of tbia section (111) areforftond- 
moulding. 

Tbe German standard rules are: \ " On a meted or tbiok glass 
plate fire sheets of blotting-paper soaked in water are laid, and on 
these are placed five moulds vetted with water. 250 grammes 
(8.76 oz.) of cement and 750 grammes (26.25 oz.) of standard sand 
are weighed, and thoroughly mixed dry in a vessel. Then 100 
cnbic centimetres (100 grammes or 3.5 oz. ) of f rosb water are udded, 
and the whole mass thoroughly mixed for five minntee. With the 
mortar so obtained, the moulds are at once filled, with one filling, 
so high as to be rounded on top,. the mortar being well pressed in. 
By means of an iron trowel 5 to 8 oentimetree (1.96 inches to 3.14 
inches) wide, 35 centimetres (13.79 inches) long, luid weighing 
about 250 grammes (8.75 oz.), the projecting mortar is pounded, 
first gently and from the side, then harder into the moulds, until 
the mortar grows elastic and water flnahes to the snrface. A 
pounding of at least one minute is absolutely essential. Au addi- 
tional filling and pounding in of the mortar u not admissible, since 
the test pieces of the same cement should have the same densities 
at the Afferent testing stations. The mass projecting over the 
mould is now cut oS with a knife, and the surface smoothed. The 
mould is carefully taken off and the test piece placed in a box lined 
irith zinc, which is to be provided with a cover, to prevent a non- 
Duiform drying of the test pieces at different temperatures. 
Twenty-four hours after being made, the test pieces are placed 
ander water, and care must be taken that they remain under water 
during the whole period of hardening." 

The French Commission recommend tbe following for sand 



ot 0. R, vol. nrll. p. Ml ; ditto of JamleaoD'a lever macblDe, see The Transit 
(Iowa State University), December, 1B89, or Engineering NeiB», voL nv. p. 138, or 
Trans. Amer. Soo. of C. E., vol. uv. p. 30a, 

■Trans. Canadian 9oc. of O. E., vol, lx.p.U, " Floal Beport Ol ths Oonimlttaa OB 
* Standard Uetbod of Testing Cements." 

f SpaldlDR's Hydiaulio dement, p. 130. 

j Eagineerina Sewt. vuL xvL c. 916. 



ovGoQi^lc 



ART, 4.] TB8T8 OP CEHEHT. 75 

raortara: " Sufficient mortar is ganged at once to make six 
briquettes, reqairiog 350 grammea of cement and 750 grammes of 
normal sand. The moald is placed npon a metal plate, and upon 
top of it is fitted a guide having the same section as the mould and 
a height of 125 millimetres (5 inches). 180 grammes of the mortar 
are introdaced and roughly distriboted in the mould and guide vith 
a rod. By means of a metallic pestle weighing 1 kilogramme, and 
having a base of the form of the briquette but of slightly less 
dimensions, the mortar is pounded softly at first, then stronger and 
stronger until a little vater escapes under the bottom of the mould. 
The peetle and guide are thoD removed and the mortar cut off level 
with the top of the mould." 

Ilia. TheB&bm6 hammer apparatus is much used, particularly 
in Germany. It consiats of an arrangement by which the mortar 
is compacted in the mould by a anccession of blows of a hammer 
weighing Z kilogrammes (4.4 pounds) npon a plunger sliding in a 
guide placed npon top of the mould. The machine is arranged to 
lock after striking 150 blows. A high degree of density is thna 
produced, and more regular results are obtained than by hand. 
The apparatus is alow.* 

The Tetmajer apparatus f is simUar in character to the B&bm6 
hammer. " It consists of an iron rod carrying a weight upon its 
lower end, which is raised through a given height and dropped upon 
the mortar in the mould. The ram weighs 3 kilogrammes. This 
machine is used in the Zurich laboratory, and Prof. Tetmajer regu- 
lates the number of blows by requiring a certain amount of work 
to be done upon a unit volume of mortar, — 0.3 kilogrammetre of 
work per gramme of dry material of which the mortar ia composed. 
This apparatus is anbject to the same limitations in practice as the 
Bdhme hammer, in being very slow in use and somewhat expensive 
in first cost." 

llli. Storing the Briquettes. It is asual to store the briquettes 
nnder a damp cloth or in a moist chamber for 24 hoars, and then 
immerse in water at a temperature of 60° to 65° F. For one-day 
testa, the briquettes are removed from the moulds and immersed as 

* For an llluBtrated deaoHpUon, see En^ntermg Uriel, vol. rflL p. 300 ; Tnna, 
Amer. Boo. ot 0. E., vol. zxi. p. 34. 

-t French OommiMlon'e Beport, toL L p. 387. 



ovGoQi^lc 



76 LIME AND CEUEt^T. [CHAP, III. 

soon as they have begun to set. The volnma of the water shoald 
be at least toar times the volame of the immersed briquettes, and 
the vater sboald be reDewed every seven days. 

The briquettes shonld be labeled or nnmbered to preserve their 
identity. Keat-cement briqnettCB may be stamped with steel dies, 
as may also sand briqaettes, provided a thin layer of neat cement !» 
spread over one end in which to stamp the number. 

lUc. Agt vhea Tested. Since in many cases it is impracticable 
to extend the tests over a longer time, it has become customary to- 
break the briquettes at one and seven days. This practice, together 
with a demand for high tensile strength, has led mannfactnrers to 
increase the proportion of lime in their cements to the highest 
possible limit, which brings them near the danger-line of unsonnd- 
Dess. A high strength at 1 or 7 days is nsnally followed by a 
decrease in strength at 28 days. Steadily increasing strength at 
long periods ia better proof of good quality than high results during 
the first few days. The German standard test recognizes only 
breaks at 28 days. The French standard permits, for slow-setting 
cements, teste at 7 and 38 days, and 3 and 6 months, and 1, 2, 
etc., years; and for rapid-setting cements, from 3 to 24 hours for 
neat mortar and 24 hoars for sand mortars. In all cases the time 
is counted from the instant of adding the water when mixing the- 
briquette. The briquettes should be tested as soon as taken from 
the water. 

llld. The Testing Haohine. There are two types in common' " 
nse. In one the weight is applied by a stream of shot, which runs 
from a reservoir into a pail suspended at the end of the steelyard 
arm; when the briquette breaks the arm falls, automatically cutting 
off tbe flow of shot. In the other type, a heavy weight is slowly 
drawn along a graduated beam by a cord wound on a wheel turned 
by the operator. The first is made by Fairbanks Scale Co., and tb» 
second by Biehle Bros., and also by Tinins Olsen, both of Phila- 
delphia. 

Fig. 3 represents a cement-testing machine which can b& 
made by an ordinary mechanic at an expense of only a few 
dollars. Athongh it does not have the conveniences and is not 
as accurate as the more elaborate machines, it is valuable where 
tbe quantity of work will not warrant a more expensive one, and 
in many oases is amply sufficient. It was devised by F. W. Bmca 



ovGoQi^lc 



ABT. 4.] 



TSSTB OF CBUENT. 



77 



for nse at Fort MarioD, St. Angaatine, Fla., and reported to the 
Engineering News (vol. t. pp. 194-96) bj Lieatenant W. M. Black, 
U. S. A. 

The machine coDsiBta easentiallj of a connterpoiBed wooden lerer 
10 feet long, working on a horizontal pin between two broad 
nprights 20 inches from one end. Along the top of the long arm 
ranB a grooved wheel carrjing a weight. The diGtaitces from the 
f nlcmm in feet and inches are marked on the sarface of the lever. 
The clamp for holding the hriqaette for tensile teats ia suspended 
from the short arm, IS inches from the fnlcrnm. Pressure for 
shearing and compreBsive Btresses is commnnicated throngh a loose 
upright, set nnder the long arm at any desired distance (generally 
-fi or 12 inches) from the f alcmm. The lower clip for tensile strains 
1 to the bed-plate. On this plate the cnbe to be crashed 




r- B, 



rests between blocks of wood, and to it is fastened an npright with 
a square mortise at the proper height for blocks to be sheared. The 
rul on which the wheel runs is a piece of light T-iroo fastened on 
lop of the lever. The pin is iron and the pin-boles are reinforced by 
iron waahers. The clampe are wood, and are fastened by clevis 
joints to the lever arm and bed-plate respectively. When great 
etreeses are desired, extra weights are hung on the ead of the long 
arm. Pressores of 3,000 pounds have been developed with this 
machine. 

For detailed drawings of a more elaborate home-made cement- 
testing machine, see Proceedings Engineers* Clnb of Philadelphia, 
Tol. V. p. 194, or Engins^ing News, vol. it, p. 310. 

III0, The Clips, The most important part of the testing 
machine are the clips, by means of which the stress is applied 
io the briquette. 1. The form mnst be saoh as to grasp the 



ovGoQi^lc 




78 LIME AND CEUENT. [CHAP. HI. 

briqaeite on fonr Bjmmetrical sQrfaces. 3. The earfsce of con- 
tact mast be large enongh to prevent the 
0\ briquette from being crashed between the 

J points of contact, 3. The clip mast tnm 

vithont appreciable frictioQ vhen under 
^^ stresB. 4, The clip must not spread ap- 

I I preciably while anbjectod to the maximom 

load. 

The form of clip recommended by the 
Committee of the American Society of 
Civil Engineers is shown in Fig. i. This 
/-^ form does not offer snfficient bearing sar- 
face, and the briquette is frequently crashed 
at the point of contact. The difficulty ia 
remedied somewhat by the use of rubber* 
tipped clips. 

Whatever the form of the machine or 
clips, great care should be taken to center 

□ II — I the briquette in the machine. 
I 111/. The Speed., The rate at which 
I I the stress is applied makes a material 
''— ' difference in the strength. The following 

^°- *■ data are given by H. Faija,* an English 

authority, as showing the effect of a variation in the speed of 
applying the stress: 

lute. Teoaile 8tr«i«lli. 

100 pounds Id 130 seoonda 400 pouods. 

100 '■ " 60 " UB " 

100 " •■ 80 " 480 " 

100 " " 15 " 4M " 

100 " .. 1 .. 498 " 

The French and German standard specifications require 660 
pounds per minute. The American Society of Civil Engineers 
recommends 100 pounds per minute for strong mixtures, and half 
this speed for weak mixtures. The Canadian Society of Civil 
Eogineera recommends 200 pounds per minute. 

111;. Data on Tensile Strength. Owing to the great variation 

• Tisns. Amer. 8oo. of C. E., vol. rrli. p. 9S7, 



ovGoQi^lc 



ABT, 4.] TEST3 OF CEMENT. 78fl 

in the manner of making the tests, it is not possible to j^ire any very 
Talaable data on the strength that good cement shonld show. In 
188fi a Committee of the American Society of Civil Engineers 
recommended the values given in Table 10 below. At least the 
minimum valnes there given are required in ordinary specifications, 
and the maximum valnes are sometimes employed. Many of the 

TABLE la 
Tehtbilx Stbenoth of Cemknt Uobtabs. 





m POUKDB FU 84OIU ImjH, 




ForUand. 


Katutal. 


Clear Crment. 
1 day— 1 Lour, or until aet, in idr, tbe remainder 


Kin. 
IW 
SCO 
SCO 
UO 


Hu. 

140 
560 
700 

800 


60 
100 
800 

80 
BO 
200 


Max. 


1 week— 1 day In dr, the remainder of the time 




4 weeks— 1 day in air, the remainder of the time 




1 veai^l day in air, the rematoder of the lime 




1 Part Cmcbmt to 1 Part Sabd. 
1 week— 1 day in air, the remainder of the time 












1 year— 1 day in air, tbe remainder of tbe time 








1 Part Ckmbbt to 8 Pahta 8aot>. 
1 week— 1 day in air, tbe remainder of the time 


80 
100 
200 


135 
300 
850 




4 weeks— 1 day In air, the remainder of the time 






1 vear— 1 day in air, tbe remainder of tbe time 










' 



tetter cements commonly give resalts above the moximam values 
in the table. Natural cement, neat plastic mortar, will generally 
■how 50 to 75 ponnds per aqnare inch in 7 days, and 100 to SOO in 



ovGoQi^lc 



"78J LIMB AND CEMENT. [cHAP. IIL. 

28 days. Good Portland cement, neat plastic mortar, vill abov 
100 to 200 pounds per sqaare inch in one day, 100 to 600 in 
7 days, and 600 to 800 in 28 days. With 3 parts sand, Portland 
cement, plastic mortar, will give at least 100 ponnds per sqiiare 
inch in 7 days, and 200 in 28 days. Of coarse the strength varies 
greatly with the method ol testing. In consulting aathoritiee on 
this subject, it shoald be borne in mind that the strength of cement, 
particularly Portland, has greatly increased in the past 10 years. 
The specifications should be drawn to correspond with the personal 
eqnation of the one who is to test the cement. 

For various specifications for tensile strength, see Art. 5, pages 
78C-78A, particnlarly Tables 10c and lOd, pages 78/, 78j. 

For additional data on the strength of mortars composed of 
different proportions of cement and sand, see Fig. 5, page 91. 

lllh. EftUATlHO THE Eebults. It not infrequently occurs that 
several samples of cement are submitted, and it is required to 
determine which is the most economical. One may be high-priced . 
and have great strength; another may show great strength oeat 



TABLE 10a. 
BsnATiTB EOoHOHT OP Ckmbmts Tbstbd Neat at T Dats. 





.m™™. 


t™. ».„„.. 


0...^ 


BlUTITI 




















1 


1 


1 


1 


1 

i 

I 


j 


ill. 


1 


A 


M.O 


98.1 


638 


81. B 


$a.80 


100.0 


79.96 




B 


88.0 


95. 9 


771 


100.0 


3.B4 


98.8 


94.26 







88.7 


96.6 


477 


61.9 


2.40 


96.8 


57.28 




D 


91.8 


100.0 


891 


60.7 


2.46 


98.8 


47.66 




B 


81.6 


88.8 


860 


86.6 


a.47 


»8.1 


70.79 





and be coarsely ground. If tbe cement is tested neat, then 
strength, Gnenesa, and cost should be considered ; but if the cement 



ovGoQi^lc 



ART. S.] SPECIFICATIONS SOS, CEMENT. ISC 

IB tested with tbe proportiou of sand naoall; employed in pnustioe, 
then only strength and coet need to be considered. 

Table 10a (page 78b) shows the method of dedacing the relatirs 
cconom; when the cement ta tested neat; and Table 10^ shows the 

t 

' TABLE It*. 

RKL&TITB SoOHOlfT OF CBMBNTe TZSTKD WTtB SaKD AT 1 DATS. 





TDniLa 


v?r" 


CamtrtEaa. 


BaLtrm Eoohokt. 


OMum. 


Ponwtopor 


B/tMtn. 


Cart per 


IMatlTCL 


udBslSllM 
Co*. 


Buk. 


A 


168 


WA 


«3.80 


100.0 


SS.40 




B 


176 


100.0 


3.84 


08.3 


98.80 







166 


M.8 


2.40 


95.8 


».S8 




D 


180 


76.7 


2.4D 


9S.8 


71.94 




E 


m 


76.7 


3.47 


98.1 


71.40 





method when tbe cement is tested with sand. The data ore from 
. -actaal practice, and the cements are the same in both tables. 
Resalts similar to the above coald be deduced for any other 
age; the circomstanoes under which the cement is to be asad 
shonld detonnine the age for which tbe comparison should be 
made. 

The ahoTe method of equating the resnlts gives the advantage 
^ a cetnent which g^ns its strength rapidly and which is liable to 
be nnsonnd. Therefore this method shonld be used with discretion, 
particularly with short- time tests. 

Art. 5. Specifications foh Ceuent. 

lilt. Oement is eo variable in quality and intrinsic valne that 
no considerable quantity should be accepted without testing it to 
see that it conforms to a specified standard. A careful study of 
Art. 4, preceding, will enable any one to prepare such specifications 
as will suit the special requirements, and also give the iostmctions 



ovGoQi^lc 



?8<j LIUB AKD CEMENT. [CHAP. III. 

neceaaary for applying the tests. A few specifications vill be given, 
to setre as gnidee in preparing others. 

SPBOIEIOATIOITS FOB QUAUTT. 

.lliy. OSBKAK POBTLAVD. The following are the most im- 
portant paragraphs from the standard specifications of the German 
gOTernment aa given in the official circnlar issned hy the Jhtiniater- 
of Pnblic Worka of Prnssia nnder date of July 38, 1887 :* 

" Timt of aetting. Accordfog to Oke purpoe for which It la Intended, 
quick or ilow-MttlDg Portland cement may be demanded. Blow-aettlng cements- 
are tboee that wt in about tvo horns." The teat U made at deacrlbed \\k 
%». 

" Cantlant^ ^ FtrfuoM. The Tolume of Portland cement should remain 
conatanL The declsiTe test of this should be that a cake of cement, made on a. 
glass piste, protected from sudden drying and placed under water after 24 
hours, should show, ereo after long submersion, no signs of crumbling or of 
cracking at the edges." For method of making the test, see § 93. 

" Finentu </ Grinding. Portlaud cement must be ground so fine that no 
more than 10 per cent, of a sample shall be left on a BJeve of 900 meshes per 
M^uare centimetre (6,800 per square Inch). The thickness of the wins of the 
sieve to be one-half the width of the meshes." Notice that a store having 90(^ 
neahes per square centimetre (5.BO0 per sq. la.) Is the staadard, although 
devea of 5,000 meshes per square centimetre (89,000 per sq. in.) are frequently 
used. 

" Ttet of Btrmgth. The binding strength of Portland cement is to b» 
determined hy testing a mixture of cement and sand. The test is (o l>e con. 
ductad for tensile and compreBalve strength according to a uniform method, 
and is to be performed upon test specimens of like form, like cross section, and 
with like apparatus. It Is recommended, besides, to determine the strength or 
neat cement. The tests for tension are to be made upon briquettes of S sq. 
cm. (0.78 sq. in.) cron section at the place of rupture, the Icfits f or compressioo. 
upon cubes of SO sq. cm. (7.8 sq. In,) areft." 

" TtntUt and CompreMtw Strength. Slow-setting Portland cement, when, 
mixed with standard sand in the proportion of I part of cement to S of sand, 
by weight, 28 days after l>elng mixed — one day In air and 37 In water — must 
possess a tecdle strength of not less than 16 kllog. per square centimetre (S!2& 
lbs. per square inch), nod s minimum compreeaUe strength of 160 kllog. per 
square centimetre (S,250 lbs. persq. In.). Qnick-setting cements generally show 
a lower strength after 28 days than that given shove. The Ume of setting^ 
must, therefore, be given when stating figures relative to strength." The test is, 
made as described In the second paragraph of §111 or the first paragraph oC 
gill*. 

■ Translation from Trans. Amer. Soo. ot 0. E., voL xa. pp. 1A-3L 



ovGoQi^lc 



ART. 5.J SPECIPICATIOSS FOK OEMSNT. 78« 

111k. EXQUBH PoRTtAlTD. Ill Great Britain tbere are no 
official apecifications, but the following proposed * by Mr. Henry 
Faija are much need: 

" PitmiMi to be Buch tbBt tbe cement will all pau through a siere haTlDg 
430 boliM (30*) to the square Inch, and leave onlj 10 per cent residue when 
rifted through a rieva hnviug 2,500 holes (50') to the sqUKre lach. 

" Rcpanthn or OorUraetion, A pat made and submlLted to mcdat heat and 
warm water at a temperature of 100* to 1 15* F., ihall show do lign of expaa- 
■lon or contraction (blowing) In twenty-Four boun, 

" T»7itiU Stnngih. Brlquettei of iloiB-tetting Portlnud, which have been 
gauged, treated, and teated fu the prescribed rononer, to cany aa average ten- 
■ile strain, without fracture, of at least 17S lbs. per sq. in. at the expiration of 
8 dajs from gauging; and those teated at the uplraiiou of 7 days, to show ut 
Increaae of at least 50 per cent, over the strength of those at S days, but to 
carry a mlallnum of SOO lbs. per sq. In. 

"For guicktetting Portland, at leut 1 TO lbs. p<>rsq. In. at 8 days, and an 
increase at 7 day b of 20 to 30 per cent., but a minimum of 400 lbs. per sq. In. 
Very high tensile strengths at early dales generally Indicate a cement verginff 
on an unsound one." 

1111. Fkehoh FOBTlahs. The following are the requirement* 
of the SerTicee Maritimes dee Ponts et OhanSB6eB,f and are fre- 
qnentl; employed in France: 

" J)muttg. A liter measure Is looeely filled with cement, t^reviously 
acreened through a sieve of 180 mesbe* to the linear Inch, and weighed. Thl» 
test la used for comparison of different lots of the same cement, the weight o( 
1 liter of which must exceed a certain figure determined for the cement la 
question. No general requirement as to density Is made." 

" Chemteal Cmnpotilioa. Cement containing more than 1 per cent, of sul- 
pharic anhydride (=1.7 per cent, sulphate of lime) Is rejected, while that con- 
taining more than 4 per cent, of oxide of Iron Is declared suspicious. Cement 
containing lets than 44 parts of silica and alumina to 100 ot lime Is also con- 
sidered auspicious." 

3Vm« tf Setting. The test for time ot setting Is made as described In g 8S 
(page 58). " Cement whlcli begins lo set In leas tUan SO minutes or seta com- 
' pleiely In less than 8 hours is refused." 

" Otnutaney of Vvlwne. Pats on glass are Immersed lo f#a-ua(«- kept 
at a temperature of GB* to 6G* F., and examined for cracking or change of 
form." 

" TttuUe BtrenglA. The amount of water to be employed is determined aa 

• Trans. Amer. 8oc. ot C. E., toL zvU. p. DSS ; vol. xuc. (1S93) p. 60. 
fOandlOt's "OlmentsandChaiixHydraulloB," Paris, 1891, pp.l5IMl. 



ovGoQi^lc 



78/ 



LIMB AND CEHKirr. 



LcHAP. in. 



In third parsgTuph of g lOS (page 09). The briquettea are moulded as d«- 
•cribed in ia the third pangiaph of g 109, page 78. 

" For aeat cemeDt, the teuile strength at T days must be at least 30 kllog. 
per sq. cm. (2S<1 lbs. per sq. in,); at 28 dajs, 95 kilog. per «q. cm. (4ffT Iba. 
per sq. in.); at 12 weeks, 46 kllog. per sq. cm. (SS9 lbs. per sq. in.). The 
tenaile strength at 28 days must exceed that at 7 dajs bj at least S Ulog. per 
sq. cm. (71 lbs. per sq. ia.). The tensile alreiigth at 12 weeks must be greater 
than that at 28 dajv uoleas the latter shall be at least ESC kilog. per sq. cm. 
<781 Iba. per «q. In.). 

"For 8 ports crushed quartz to 1 part cement, with 13 per cent, water 
(moulded as described in the third paragraph of g 111], the tensile strength 
must be at T days at least 8 kilog. per sq. cm. (Ill lbs. per sq. In.); at 28 days 
«t least IS kllog. per sq. cm. (318 lbs. per sq. in.): and at 12 weeks, IS kilog. 
persq. cm. (258 lbs. persq. in.). The Strength at 12 weeks must la all cases 
be greater than that at 38 days." 

111m. American Practice. Tables lOc and lOd give the ayerage 
reqairemeats for fineneae and tensile strength of Portland and 
natnr&l cements, for rarions classes of work in the United States. 
These Talaee may be regarded as repreaeotatlre of the average 
American practice : 



TABLE 10c. 



AhKRIOAN RDQOIRRtmHTS B 



88XJ. S. A. Engineers.. 
10 Cities 

ORailways. 

6 Bridges 

8 Aqueducts 

81 SpectflcatloDs 



Hast Cement. 



A«e when Tsnad, Dart. 



ovGoQi^lc 



ART. 5.] 



BPECIPICATION'8 FOE CEMENT. 



78y 



24 U. S. A. EDg1n««r8 , . 
lOCHIes 

4IUnwft7S 

2Bddgea. 

8 Aquedueto. 

01 SpeciflcatloDi. 



Ace vlHii Txtad. Dkti. 



178 



n 



II In. f BILASELFEU ; Hatvsai ASB f OETLAITO. The follow- 
ing IB an abstract of the Bpecificationa used in 189? by tbe Depart- 
ment of Public Works of the City of Philadelphia. Thes* 
Bpecificationa are inserted aa ahowing the extreme of American 
practice in the high degree of fineneas and great strength required. 
Compare these resolta with those in Tables lOc and lOf^. Tho 
Philadelphia Bpecificationa are not included in theae tablea. 

NATDKAL CEMENT. 

" i^eeifie Oravily. Tbe Bpeclflc graTitf shall not be less than 2.7. 

" Fin^nau. The residue shall Dot leave more Ibsn 2 per cent on a No. GO 
•teTe, nor IS od a No. 100 sieve, nor S5 on a No. 200 steve, the slevm harln; 
2,400, 10,300, ftod 85,700 meshes per square Inch and the diameter of tbe wire 
being 0.00W, 0.004S, sod 0.0020 of ui loch respectlvel;. 

" CoTutanry of Voivme. Pals ol neat cement one half toch thick vitb 
thin edges, Immersed In water after haid set, shall show no slgos of checUn; 
or dMntegratlon. 

" Time <^ Batting. It shall begin to set in not less tban 10 minntea, and set 
- baid In less than 30 minutes. 

" Tatuila BtrtngtA. The tensile Urength of dry mortar [see last paragraplk 
of % lOS] shall not be lees than in tbe accompaDying table : 



ovGoQi^lc 



LIHE AlTD OEUEMT. 



[chap. in. 





PODHItt PU S4CAU IKCH. 




Sew. 


*Qa2^ 




100 
200 
800 








38 d»y> (1 day In »!r, 37 days in water) 


aoo 



POBTLAND CEMENT. 

■' Bpeeiflc OravH]/. The specific gravity Bhall not be le« thui 8.0. 

" FineMU. The residue sball not be more than 1 per cent, on a No. BOsieTC, 
10 on a No. 100 deve. aod 80 on a No. 200 sieve. 

" Conttaaei/ of Volume. Same as for natural cement above. 

" Tinu of Betting. The cement shall not develop initial Kt in leu than 80 
■nloutes. 

" TeniHe StrtngVi. The tensile strength of dry mortar [see last paragragh 
■ «t g 105] shall not be less than in the BccranpanylDg table : 





Poems piB SqDABK Ihcb. 




Nest. 


IQuwts. 




175 
800 
000 








2S day* (1 day In air, 37 days in walor) 


MO 



SPECIFIOATIOHS FOB DELIYEBT AND STOBAOE. 

lllo. The preceding Bpecifications preecribe the quality of the 
cement; and the following refer to the quantity, the sampling, and 
the storage. The tests onder the former are made in the laboratory, 
those under the latter on the work. 

Paekage. The oeoient shall be delivered in strong barrels* lined with 

* It Is enstomary to specify that the eement, partJeaUrly Portland, shall be 
delivered tn barrals. The onl; reason tor shipping In banela Is that the oement I> 
abetter protected from the weather In barrels than In bage. The ai^uments in 



ovGoQi^lc 



,1^T. 5.] SPECIFICATIONS FOR CEHENT. 78f 

'paper so a* to be reAsoiubly protected from tbe air and dampneu. Each 
packi^ ■ball be labeled wltb tbe brand, tbe numufacturer'a luune, and the 
.groM weight. 

Weight The net welgbt of a bunel of cemeat shall be uodarstood to be 
876 pnuQda of Ponlaod, and 800 pounds of Eastern natural or 36S pounds of 
Wenera nutural [see § 77, page 04]; and bags shall coutaiu an aliquot part of 
« barrel. A variation of 2 per ceuL is allowable in the weight of Individual 
parkageg. Any brokea barrel or torn bag may be rejected or accepted at half 
Its original weight, — at the option of the Inspector. 

Tine of Dttivery. Tbe Inspection and leMs will occupy at leaat ten * days, 
«nd the Contractor ahall submit tbe cement for sampling at least ten* day* 
before desiring to use iL Tbe Inspeclor shall be promptly notified upon the 
receipt of each shipment. 

Sampling. The cement from which to test the quality ahall be selected by 
tnklug, from the Inierior of each of slif well-distributed barrels or bag! In 
'S.ich car-load, auffldent cement to make from five to ten briquettes. These 
■Isf portions, after being thrown together and thoroughly , mixed, will be 
•ssumad to represent ttie average of the whole car-load. 

Storage. All cement when delivered ahall be fully protected from tbe 
weather ; nnd shall not be placed upon the ground without proper blocking 
under it. Accepted cement may be re-inspected at any time ; and if found 
U> be damaged, it shall be rejected. Any cement damaged by water to sucb 
«D extent as to show upon the outside of the barrel will be rejected. 

Ijitptelioa Mark*. Cement which baa been accepted may be so labeled by 
the Inspector ; and the Contractor shall preserve these labels from deface- 
ment and ahall prevent their imitation. Rejected cement shall be so marked ; 
and tbe Contractor ahall promptly remove such cement. 

Baai0for Bt^eeting. Bach shipment of cement shall be tested for quandtj 
•ud quality. If the average weight of the barrels or bags tested is less than 
the weight specified, a corresponding deduction shall be made in the price ; 
and If ten per cent, falls lo conform to the requiraments for quality, the entfra 

favor of ahlppinit In bags are ; 1. The coat Is lees, sinoe the cost of the barrel is 
«)lmlDated. 3. The oement 1b more eaaiiy handled, since the weight of a unit Is 
leee. 3, la oloth bags the oement Improvee by seasoning, i.e.. the contact with 
the air hydrates any free lime due to Improper chemical aomblnatlon or Imperfeet 
ealoInatloD. i. The praetloe of shipping In barrels Is only a survival from the time 
when the best oement was of European manufactore, which of neoesslty was 
shipped In barrels because of the exoeealve moisture In the holds of vassala, S. In 
Eumpe Portland cement Is usually abippod In cotton-dnek bags. 

* For Important work this time Is nsnally made thirty days. 

t This is the nnmber epeolfied by the Pennsylvania Railroad, a rood noted for 
earetul and thorough work. It Is frequently speolQed that ten samples shall be 
taken; and In Important worklwhere a single barrel of poor oement may materially 
.■Seat the strength of tlia work, it Is sometlmea speolQed that eaoh and every 
4)arral shall be tested. 



ovGoQi^lc 



78; LIME AlfD CEUENT. [CHAP. III. 

shipment mftj be rejected. Tbe fkilnre of a (hlpment to meet the speclflca- 
tioDS for quality ma; prohibit further um of that brand Od the work. 

Barrels coDtalnlDg a Urge proportion of lumps Bhall be rejected. 

Spinal to Tett. Tbe Eagioeer rewrrea tbe right to refuK to teat tny^ 
bnnd which In his judgment la unsuitable for the work,* A barrel or bag 
which b not plainly labeled with the brand and maker's name shall not b« 
tested, and shall be immediately removed. 



* This provision Is somettmss Inserted to avoid the trouble and delay of teatlos 
aaj brand which the Engineer Is reasonably oertaln Is nnflt for the work owing Uk 
Its genend lepatation tor poor quality or lack of uniformity. 



ovGoQi^lc 



CHAPTEB in* 

BAND, QRAY£L, AND BROKEN STONE. 

112. Sand is need ia making mortar; and gravel, or sand and 
broken stone, io making concrete. The qualities of the sand and 
brokea Btone have an important effect upon the strength and cost 
of the mortar and the concrete. The effect of the variation in these 
materials is generally overlooked, even though the cement is subject 
to rigid specifications. 

AnT. 1. Sand. 

lis. Sand is mixed vith lime or cement to rednce the cost of 
the mortar; and is added to lime also to prevent the cracking which 
would occur if lime were used alone. Any material may be used to 
dilute the mortar, provided it has no effect upon the durability of 
the cementing material and is not itself liable to decay. Pulverized 
Btone, powdered brick, slag, or co&l cinders may be used ; but 
natural sand is by far the most common, although fine crushed 
stone, or "stone screenings," are sometimes employed and are in 
some respects better than natural sand. 

In testing cement a standard natural sand or crushed quartz ia 
employed; but in the execution of actual work usually local natural 
Band must be employed for economic reasons. Before commencing 
any considerable work, eH available natnral sands and possible sub- 
stitnteB should be examined to determine their values for use in 
mortar. 

114. SxannTBS fob Good Sard. To be sniteble for use in 
morter, the sand shonld be sharp, clean, and coarse ; and the gnuns 
should be composed of durable minerals, and the size of the grains 
should be such as to give a minimum of voids, i.t., luterstieeB 
between the grains. 

The usual specifications are simply : " The sand shall be sharp, 
clean, and coarse." 

114a. Snrabilitj. As a rule ocean and lake sands are more 



ovGoQi^lc 



"Vgfr SAJITD. [chap. IIIO. 

durable than glacial sands. The latter are rock meal gronnd in the 
.geological mill, and nsnally consist of silica with a considerable ad- 
mixture of mica, hornblende, feldspar, carbonate of lime, etc. The 
silica is hard and durable ; but the mica, hornblende, feldspar, and 
carbonate of lime are Boft and friable, and are easily decomposed 
bj the gases of the atmosphere and the acids'of rain-water. The 
lake and ocean sands are older geologically ; and therefore ore 
usually nearly pure quartz, since the action of the elements has 
eliminated the softer and more easily decomposed constituents. 
Some ocean sands are nearly pure carbonate of lime, which is soft 
and friable, and are therefore entirely unfit for nse in mortar. 
These are known as calcareous sands. 

The glacial sands frequently contain so large a proportion of 
Boft and easily decomposed constituents as to render them unfit for 
use in exposed work, as for example in cement sidewalks. Instead 
■of constructing exposed work with poor drift sand, it is better either 
to ship natural silica sand a considerable distance or to secure 
'Crushed quartz. Gmshed granite is frequently used instead of sand 
in cement sidewalk construction; but granit« frequently contains 
mica, hornblende, and feldspar which render it nnanitable for this 
kind of work. 

However, as a rule the physical condition of the sand is of more 
importance than its chemical composition. 

114d. Sharpness. Sharp sand, i.e., sand with angnlar grains, 
is preferred to that with rounded grains because (1) the angular 
grains are roagher and therefore the cement will adhere better; and 
(3) the angnlar grains offer greater resistance to moving one on the 
•other under compression. On the other hand, the sharper the sand 
the greater the proportion of the iuterstices between the grains 
(compare line 4 of Table 10^;, page 79t, with the preceding lines of 
the table); and oonseqnently the greater the amount of cement 
required to produce a given strength or density. But a high 
degree of sharpness is more important than a small per cent, of 
yoids. 

The sharpneBB of sand can be determined approximately by 
rubbing a few grains in the hand, or by cmahing it near the ear 
and noting it a grating sound is produced; but an examination 
through a small lens is better. Sharp sand is often difficult to 
obt«n, and the requirement that " the sand shall be sharp" is 
practically a dead letter in most specifications. 



ovGoQi^lc 



AKI. 1.] CLSANH£B6. 79* 

114c. OlMumvij. Cleao eaad a neceesary for the atroogert 
mortar, sioce an eorelop of loam or orgftoic matter about the sand 
gruDB will prevent the adherence of the cement. The cleanneBa of 
«and may be judged by preaaing it together in the hand while it ii 
'damp; if the sand sticks together when the pressare is removed, it 
U entirely unfit for mortar parposes. The cleanness may also be 
tested by robbing a little of the dry sand in the palm of the hand; 
if the huid is nearly or quite clean after throwing the sand oat, it 
IB probably clean enough for mortar. The cleanness of the sand 
may be tested qnantitatirely by agitating a qnantlty of sand with 
water in a graduated glass flask; after allowing the mixture to 
settle, the amonnt of precipitate and of sabd may be re^ from the 
padnation. Care shonld be taken that the precipitate has folly 
settled, dnce it will condense considerably after its upper surface is 
clearly marked. 

Sand is sometimes washed. This may be done by placing it on 
« wire screen and playing upon it with a hose; or by placing it in 
An inclined revolving screen and drenching with water. When 
only comparatively small qnantities of clean sand are reqnired, it 
«au be washed by shoveliog into the npper end of an inclined 
: Y-shaped trough and playing upon it with a hose, the clay and 
lighter organic matter floating away and leaving the clean sand in 
the lower portion of the trough, from which it can be drawn off by 
removing for a short time plugs in the sides of the trough. Sand 
■can be washed fairly cleaa by this method at an expense of about 
10 cents per cubic yard exolasive of the cost of the water. For a 
sketch and deacription of an elaborate machine for washing sand 
by paddles revolving in a box, see Engineering Newt, vol. ili. 
IMge 111 (Feb. 16, 1899). By this method the cost of thoroughly 
washing dirty sand is about 15 cents per cubic yard. 

Although it is cnstomary to require that only dean sand shall 
1m used in making mortar, a small quantity of very finely powdered 
-clay will not materially decrease the strength of the mortar. In 
some instances clay to the amount of 10 per cent, of the sand seems 
not to decrease the strength of the mortar.* Mortar containing 
^nsiderable clay is much more dense, plastic, and water-tight; and 
it occasionally convenient for plastering surfaces and stopping leaky 
joints. Such mortar is not affected by the presence of water. 



ovGoQi^lc 



"iQd SAVD. [OHAP. ino. 

In cDgiaeeriag literatare but lew definite specifications for the 
cleanneBB of sand can be foand, a diligent search revealing only tho 
foUowing: For bridge vork on the New York Central and Uadsoa 
Biver R. B., the specifications required that the sand shall be bo 
clean as not to soil white paperirhea rnbbed on it. For the retain- 
ing walls on the Chicago Sanitarj Canal, the snapended matter 
when shaken with water was limited to 0.5 per cent. For the dam. 
on the Monongahela Hirer, built under the direction of the 
U. S. A. engineera, the snBpeoded matter was limited to 1 per 
cent. For the dam at Portage, N. Y., bnilt by the Stat« Engineer, 
the " aggregate of the impurities " was limited to 5 to 8 per cent. 
The contamination permissible ia any particular case depends upon 
the cleannees of the sand aTsilable and upon the difficulty of 
obtaining perfectly clean sand. Sand employed in masoniy con- 
Btmction frequently contains 5, and sometimes 10, per cent, of 
suspended matter. 

Hid. Fineness. Coarse sand is preferable to fine, since (1) the 
former has leas sarface to be covered and hence requires leas 
cement; and (2) coarse sand requires loss labor to fill the interstices 
with the cement. The sand should be screeued to remove the 
pebbles, the fiueaeas of the screen depending upon the kind of work 
in which the mortar is to be used. The coaner the sand the 
better, even if it may properly be designated fine gravel, provided 
the diameter of the largest pebble is not too nearly equal to tho 
thickness of the mortar joint. 

Table IO0 gives the results of a series of esperiments to deter- 
mine the effect of the size of grains of sand upon the teoBile 
strength of cement mortar. The briquettes were all made at the 
same time by the same person from the same cement and sand, the 
only difference being in the fineneaa of the sand. The table clearly 
shows that coarse sand is better than fine. Notice that the results 
in line 4 of the table are lai^r than those in line 3. This is 
probably due to the tact that the sand for line 4 has a greater range 
of sisee and consequently fewer voids. If this explanation ia true, 
then since the sand in each line of the lower half of the table has 
greater variety of sizes than those ia the upper half, the ooarse sand 
is relatively better than appears from Table IO0. 

Table 10/ shows the fineness of nstnral sands employed in 

actual oonstruotioD ; and as the sands were to alt appearances of 

- the same character, this table also shows at least approximately tho 



ovGoQi^lc 



or FnfRKBsa ot Sans upok thx Tekulb Stbehoth or 1 : 2 
Ckukt Hortab. 







TsmiLB 


Si»inrrH 


a ttmna na mtoi 


LUniCB, 


Bd. 


Bull UDOBT awmMK 






^""" 










TDftylu 


IMo. 


SUoa. 


a Km. 


11 Km. 




No. 4 Mid No. 8 


248 


442 


SS9 


.70 


0«0 




■■ 8 •' " W 


m 


846 


478 


613 


673 




" 18 *' " » 


186 


aeo 


818 


897 


89a 




■' 80 " " SO 


SU 


381 


em 


403 


440 




■' 80 " " 80 


149 


306 


388 


376 


818 




" 60 *• " 76 


isa 


S14 


360 


276 


806 




'■ 76 " " 100 


98 


IDS 


311 


306 


368 




Puaing No. 100 


98 


1S5 


101 


330 


371 





ravumm. 


3 


r- 




Pw Otnt. bj weight, oaii^t 


on One Mo. 




Pn 







a 


M 

ai 


10 

16 


ao 


«0 

9 


TB 


iw 


100. 


h 


1 


30 


11 


8 


2 


700 


3 





39 


20 


18 


10 


13 


6 




1 


447 







33 


31 


11 


17 


30 


8 




1 


870 







18 


15 


10 


19 


88 


6 




1 


841 







9 


10 


6 


11 


45 


16 




1 


883 







18 


16 


7 


8 


88 


16 




1 


809 
















1 


6 


60 


38 


3 


248 

























6 


M 


SCO 













8- 


8 


16 


4S 


80 


6 


189 



jvGooi^le 



79/ BAKD. [chap. Ilia. 

«fFect of fineneas npoD tensile strength. This table agrees irith 
the preceding in showing that the coarser sand makes the stronger 
mortar. This oonol[ision is perfectly general. 

If the Toids are filled vith cement, uniform coarse grains giT» 
greater strength than coarse and fine mixed ; or, in other words, 
for rich mortar coarse grains are more important thau small voids. 
Bnt if the voids are not filled, then coarse and fine sand mixed give 
greater strength than uniform coarse grains] or, in other words, 
for lean mortar a small proportion of voids is more important than 
cnarse grains.* 

As a rule, the sand ordinarily employed in making cement 
mortar is much too fine to give maiimnm Btrength or to permit 
the use of a minimum amonnt of cement. For example, the sanda 
in lines 13 and 14 of Table lOjr (page 7di) are mnch used in actual 
work, and have approximately the same degree of fineness as th& 
sands in the last three lines of Table 10/, which give a moch weaker 
mortar than the preceding sands of Table 10/. 

1140. Specifications seldom contain any nnmerioal reqairement 
for the finonees of the sand. The two following ore all that can 
be fonnd. For the retuning-wall maeoory on the Chicago Sanitary 
Canal the requirements were that not more than SO per cen/ ihalt 
pass a No. 50 sieve, uid not more than 12 per cent, shall pass a 
No. 80 sieve. For the Portage Dam on the Oenesee Biver, bnili 
by the New York State Engineer, the apecificatious were that at 
least 75 per cent, should pass a No. 30 sieve and be canght on a * 
No. 40. 

The fineness of the sand employed in severd noted works is as 
follows, the larger figures being the nnmber of the sieve, and the 
smaller figures preceding the nnmber of a sieve being the per cent, 
retained by that sieve, and the small number after the last sieve 
number being the per cent, passing that sieve: Poe Lock, 
St. Mary's Fall Canal, * 20 "30 •"40"; concrete for pavement 
fonudations in the City of Washington, D. C, • 3 ' 6 ' 8 ■* 10 * 20 •• 
40'60»80'; Genesee (N. Y.) Storage Dam, •20'30''60* 
100 •; Rough Biver (Ky.) Improvement, " 30 " 30 " fiO "; St. Begis 
sand, Soulanges Canal, Canada, '*%0"30''fiO"; Grtiud Coteau 



•Beport of Chief of BD^neera, U. B. A., ISSS, p. 3863, or Jour. Wsat. Boo. of 
Eiigrs., vol.11, p. B19; and Beport of Opentloiu of tbe EngUieeriitg Department ol 
the DiMilot of Oolumbla, 1S9S. p. lU. 



ovGoQi^lc 



AKi. 1.] VOIDS. 79J;- 

■and,* Sonlanges Guial, Canada, " 20 " 30 ** fiO **. Tables 10/ and- 
IC^ abow the fineoees of a number of nataral sands employed in 
actual work. 

114/*. Voids. The smaUer the proportioD of Toids, i.e., the 
interstices between the grains of the sand, the less the amount of 
oement reqnired, and oonsequentl; the more economical the saud. 

The proportion of Toids ma; be determined by filling a Tossel. 
irith sand and then determiniog the amount of water that can ba 
pnt into the reesel with the sand. This qoantity of water divided 
by the amoont of water alone which the Tesael will oontain is thd, 
proportion of Toids in the sand. The qoantiUea of water as above- 
may be determined by Tolnmes or by weight. The proportion at 
voids may be determined for the sand loose or rammed, ttie latter- 
being the more appropriate, since the mortar is either oompreeaed 
or rammed when naed. In either case it is more accurate to drop- 
the sand throngh the water than to pour the water upon the aand» 
nnoe with the lattw method it is difficult to eliminate the air- 
bobbles, — ^particularly it the sand be first rammed. If the sand iai 
dirty and tbe water is ponred upon it, there is liability of the clay'a; 
being washed down and puddling a stratum which will prevent th» 
water penetrating to the bottom. If the bubbles are not excluded^ 
or if the water does not penetrate to the bottom, the reenlt obtained 
is less tiian the true proportion of voids. Again, if the sand ia 
dropped throngh a considerable depth of water, there is liability 
that the sand may become separated into strata having a single Biz» 
of grains in each, in which odse the voids will be greater than if 
the Beven^ sizes were thoroughly mixed. 

The per cent, of voids Tariee with the moisture of the sand. 
A small per cent, ot moisture has a surprisiag effect npon the- 
volume and consequently npon the per oent. of voids. For 
example, fine sand containing 2 per .cent, of moisture uniformly 
distributed has nearly 20 per cent, greater volume than the Bam» 
sand when perfectly dry. This effect of moisture increases with 
the fineness of the sand and decreases with tbe amount of water 
present. 

114^. Table lOf?, page 79t, shows the voids of a number of both 
artifici^ and natural sands. An examination of the table showa 

■A 1 to 9 mort&r wttb UiU aond mta only 79 per oenL as strong ae tba pre- 
McUugi and with a 1 to S moTtarotil7Tl p«r eent.—Truia.OaD. Sod. ol 0. K, vol. is. 



ovGoQi^lc 



79A BAKD. [chap. Ilia. 

that the voids of natnnl sand vhen rammed •my bom 30 to 37 
per cent. Sands Nob. 10, 11, and 12 are rtry good; bnt Nob. 13 
and 14 are very poor. All five are freqaentlj employed in actoal 
work. Compare the flneneos of these aanda irith those in Table 
10/, page 79e. 

114A. The following obserrattons may be nsefnl in inveeti^ting 
the Telative merits of different aands : 

The prc^rtion of Toids is independent of the size of the 
grains, bnt depends npon the gradation of the sizes; and vuriep 
with the form of the grains and the roaghneflB of the snrfaoe. A 
mass <A perfectly smooth spheres of uniform size wonld have the 
flame proportion of voids, whether the spheres be large or small. 
A mass ot perfectly smooth spheres packed as closely as possible 
wonld have 26 per cent, of voids; bnt if the spheres are packed aa 
loosely as possible the voids woald be 48 per cent. A promiscnons 
mass of bird-flhot has about 36 per cent, of voida. The difference 
between this and the theoretical minimum per cent, for perfectly 
emooth spheres is due to the variation in size, to roaghnesa of the 
flurface, and to not securing in all parts of the mass the arrangement 
of the shot necessary for minimum voids. German standard aand 
has gnuns nearly spherical and nearly uniform in size, having 
slightly rough surface, and has 41 per cent, voids loose (see line 3, 
Table 10^). The difference in the per cent, of voids between this 
sand and a mass of spheres uniform in size and perfectly spherical 
is due to irregularitieB in form and to roughness of surface of the 
sand grains, hnd to not secnring the arrangement of the grains 
necessary for minimum voids. Crushed stone retained between the 
same sieves as German standard sand has 65 per cent, of voids (see 
lines 1 to 3 of Table lOg), the excess of this over German standard 
aand being due to the rough surfaces and sharp corners preventing 
the grains from fitting closely together. 

If the mass consists of a mixture ot two sizes such that the 
smaller grmns can occupy the voids between the larger, then the 
proportion of voids may be very much smaller than with a single 
size of grains. For this reason a mixture of two grades of sand of 
widely different sizes has a smaller per cent, of voids than any one 
size alone, — compare lines 1 to 9 with the remainder of Table 10^. 

The best sand is that which has graios of several sizes such that 
the smaller grains fit into the voids of the larger, the proportion of 
any particnlar size being only sufficient to fill the voids between 



ovGoQi^lc 



AKT. 1.] 



I 

Z i ^ 
II 



mi 






SSo53StiS&S8i:S8S8£S8 



SSKSSSSSSSSS : 



ss^sessssssK 



S3S7«S3S««3« 



ooosgaB-sB-ssg 



ssss^g^^'ssi 



°Si=3"K:8sS 



,41=. 
lull'"" 
iipi 

_|Sj.|. .... 



ovGoQi^lc 



79;' SAND. [chap. iiio. 

the grains of the next larger aize. If the grains are spherical and 
the diameter of the smaller ia about one fifth of the diameter of the 
larger, the smaller grains irill jnst fit into the interstice between 
the larger ones. The smaller the voids the greater the economy, 
and the denser and stronger the mortar. 

The finer the sand the more nearly nniform the size of the 
grains, and consequently tha greater the proportion of voids. Also 
the finer the saud the less sharp it is, and consequently the greater 
the surface to be covered. Therefore a coarse sand is to be preferred 
to a very fine one — see Tables lOe and 10/, page 79e. Further, the 
advantage of coaree sand over fine increases as the proportion of 
cement decreases, since vrith the smaller proportions of cement the 
voids are not filled. 

114t. Concluiion. An examination of the preceding data shows 
, that very fine sand makes a much weaker mortar than coarse sand, 
and also that natural sands vary considerably in the proportion of 
voids and consequently differ in the amount of cement required to- 
prodnce any particular strength. Therefore before adopting a sand 
for a work of any considerable magnitude, all available sands should 
be carefully examined with reference to (1) their effects npon the 
strength of the mortar, (2) their per cent, of voids or the amoani 
of cement required with each, and (3) their cost. If mortar of any 
particular strength is desired, the proportion of cement shoald be- 
adJQsted according to the fineness and voids of the best available 
sand. 

114y. StoITE SoSEEinveB. The finer particles screened out of 
cmshed stone are sometimes used instead of sand. For ths 
physical characteristics of stone screenings see Nob. 16 and 17, 
P^e 79t. 

Experiments show that sandstone screenings give a slightly 
stronger mortar than natural sand, probably because of the greater 
sharpness of the grains. Crushed limestone usually makes a con- 
siderably stronger mortar, in both tension and compression, than 
natural sand, and this difference seems to increase with the age of 
the mortar.* Part of the greater strength is unquestionably due 
to the greater sharpness of the limestone screenings, and the part 

■ ADDual Report of Obiet of EoglneerB, U. S. A., 18S3, Port 3, p. 3015 ; do. 1S94, 
I>ut 4, p. 2331; do. 1B9S, Fart i, p. 29G8; Jour. Weat. Soo. ot Eugra., voL ii, pp. 394 
audUO. 



ovGoQi^lc 



AET. 2.] QRATEL. 79ife 

that increaaee with the age of the mortar seems to be due to some 
chemical action between the limestone and the cement. 

114it. Cost and Weight of Band. The price of reasonably good 
sand varies from 40 cents to tl.60 per yard, according to locality. 

Sand is Bometimea sold by the ton. It weighs, when dry, from 
80 to 115 poands per cnbic foot fsee Table lO^r, page 79t), or aboat 
1 to 1^ tons per cabio yard, ' 

Abt. 2. Gravel and Bbosen Stoke. 

116. The term gravel is sometimes used aa meaning a miztnre 
of coarse pebbles and sand, and sometimes as meaning pebbles with-' 
ont sand. In this volume, gravel will be understood as a miztare 
of coarse pebbles and sand. 

llfia. Gravel and broken stone are mixed with cement mortar 
t« make an artificial stone called concrete (Art. 2, Chap. IV), The 
quality of the concrete varies greatly with the condition of the 
gravel or broken stone, but unfortanately too little attention ig 
given to the character of this component. 

115£. GkAvel. To be suitable for use in making concrete, 
gravel sboald be clean, and it should be composed of durable 
minerals, and the size of the pebbles and grains should be sach as 
to give minimum voids. 

The investigation of the suitability of gravel for use in concrete 
is essentially the same as that of sand, which has been fully con- 
sidered in the preceding article. 

The physical characteristics of pebbles and gravel are given near 
the foot of Table lOA, page 80. Judging from the little data that 
can be found in engineering literature and from all the information 
gathered by an extensive correspondence, gravels No. 16 and Xo. 
17 of the table are representative of the gravels employed in actual 
work. 

Concerning No. 18 notice that 65 per cent, passed a Ko. 5 
screen; and therefore this mixture could more properly be called 
gravelly sand. If one fifth of the material passing the No. 5 sieve 
be omitted, the voids of the remainder will be only 15 per cent, 
when rammed ; in other words, if one-tenth of this gravel were 
sifted on a No. 5 sieve and that portion retained on the sieve were 
mixed with the remainder of the original, the voids would be 
reduced to 15 per cent., which would improve the qu^ity of the 



ovGoQi^lc 



791 BKOKEH STONB. [OHAP. IlIO. 

gravel for making concrete. This is a Talnable hint as to the pos- 
sible advantage of sifting «T«n a portion of the gravel. 

115c. BsoKZV Stohx. Any bard and durable stone ia saitable 
for use in making concrete. It is usual to specify that the stone 
for concrete shall be broken to pass, every way, through a S-incb 
ring, although it is sometimes broken to pass a 1-incb ring. The 
stone should be broken small enough to be conveniently bandied 
and easily incorporated with the mortar. The finer the stone is 
broken the greater its cost, and the greater the surface to be coated; 
and consequently the greater the amount of cement required. 
Approximately cubical pieces are preferable to long, thin, sj^intery 
fragments, since the latter are liable to break under pressure or 
while being rammed into place, and thus leave two uncemented 
surfaces. 

116d. Toidi. The proportion of voids, i.e., interstices between 
the fragments, may be determined in either of two ways as follows : 

1. The voids may be found by filling a vessel with the aggregate, 
and then pouring in water until the vessel ia full. The amount of 
water required to fill the voids divided by the amount of water 
alone tbe vessel will contain la the proportion of voids in the 
aggregate. The amount of water in each case may be determined 
by weight or by volume. 

For Borne preoantlons applicable in this case, particularly in 
determining the voids of broken stone contuning considerable fine 
material, see g 114/. If the material is porous, it is beat to wet it, 
80 as to determine tbe voids eitfirior to the fragments. The water 
absorbed by tbe material should not be included in the voids, since 
when the concrete is mixed the E^gregate is usually dampened, 
particularly if it is porous. Of course in wetting the aggregate 
before determining the voids no loose water should remain in the 
pile. The voids may be determined for the material either loose ' 
or compacted. The proportion of the voids ia found to determine 
the amount of mortar required to fill the voids of the concrete in 
place; and therefore it is better to determine the voids in the com- 
pacted mass, since the concrete is usually rammed when laid. The 
•compacting may be done by shaking or by ramming, the latter 
.heing the better since it more nearly agrees with the conditions 
Tinder which the concrete is used, and further since in compacting 
by shaking tbe smaller pieces work to tbe bottom and the latter to 
the top, which separation increases the voids. 



ovGoQi^lc 



ART. 2.] VOIDS. 79to 

This method usually gives results slightly too Bmall, owing to 
the difficulty of excliidiDg all the air-bubbles. However, a high 
degree of accuracy can not be expected, since the material is neither 
uniform in composition nor uniformly mixed. 

3. To find the voids determine the specific gravity of a frag* 
ment of the material (§ 7) , and from that the weight of a unit of 
Tolume of the solid; and also weigh a unit of volume of the aggre- 
gate. The difference between these weights divided by the firat 
gives the proportion of voids. 

il5e. Table lOA, page 80, shows the per cent, of voids ia 
various grades of broken stones used in making concrete. 

The per cent, of voids in broken stone varies with the hardness 
of the stone, the form of the fragmente, and the relative propor- 
tions of the several sizes present. The last is the most important. 
If broken stone passing a 3j-inch ring and not a |-inch screen 
be separated into three sizes, any one size will give from 52 to 54 
per cent, of voids loose, white equal parts of any two of the three 
sizes will give 48 to 50 per cent., and a mixture in which the 
volume of the smallest size is equal to the sum of the other two 
gives a trifle less than 48 per cent. Hotico, however, that un- 
screened crushed stone has only 32 to 35 per cent, voids — see lines 
7 and 11 of Table lOA. This is a very excellent reason for not 
screening the broken stone to be used in making concrete. 

A mass of pebbles baa only about three fourths as many voids 
as a mass of broken etone having pieces retained between the same 
screens. Notice, however, that gravel, i.e. pebbles and sand, has a 
leas proportion of voids than pebbles alone. 

lis/. Coat and Weight. The cost of breaking stone for con- 
crete varies from fiO to 75 cents per cubic yard according to kind 
of stone and size of plant.* The original cost of the stone and 
transportation expenses are too variable to attempt to generalize. 
Ordinarily the cost of broken stone is not more than tl.50 to $2.00 
per cubic yard f. o. b. cars at destination. 

The weight of broken stone varies from 85 to 120 lbs. per cubio 
foot (see Table lOh, page 80) ; or about 3200 to 3200 pounds per 
oabic yard. 

* ^r addlUonal dsta, see Bapplemeiital Notes, No. 5, p. (MS. 



jvGooi^le 



BBOKBV 8TOVB. 



[chap, mo, 











SSg SSSS S|£3 SSSS 3S3 




li 


1 


ESS SXS3 SSS= S3SS SgS 




k 


A 


«3S ass $SiiS 3S3E SSS j 


1 


1 


SSS SS3S SS3S «33S S38 ! 


i 


' 


t 








" i 


\\° \\ 




-" 




1 

s 

1 

1 

I. 


e 








M 


\\- \\ 




"-S 


> 








» : 


i- ;: 




ae» 


s 








- i 


;- ;; 


: : »2e- 


s 






=o- ;= 


22 :o 


= o 8Sg 


'A 


' 


ioo o|ss :| 


=2 ;s 


Issss 


r 
1 

1 


llpl 


^ 


»as |=>sa ;= 


» :S »S*3 »t-«> j 


5. 


iSS = :=S 1 


;ss|= 


:S SSH 


i. 


oSo : : 


o o 


;=-o ; 


;» o»o 


1 




i 


1 _ 


= = J- 




'" 1 




Is 


-«« »«..- -.SS 2S35 SK2 



jvGooi^le 



FART II. 

METHODS OF PREPARING AND USING THE 
MATERIALS. 



CHAPTER IV. 

HORTAB, CONCRETE, AND ARTIFICIAL STONE. 

Abt. 1. MOBTAB. 

116. Mortar is a miztare of the paete of cement or lime vith 
sand. Id common mortar, the cementing sabfitanee is ordinary 
lime; in hydraulic mortar, it is hydraalio cement. 

117. COKKOV Like Kobtar. Mortar made of the' paate of 
«ommon or fat lime ia extensirely used on account of (1) its intrin- 
sic oheapnesB, (2) ita great economic advantage oving to ite great 
increase of volume in slaking, and (3) the simplicity attending the 
mixing of the mortar. On account of the augmentation of volnme, 
the paste of fat lime shrinks in hardening, to snch an extent that 
it can not be employed as mortar without a large doee of sand. 

Ab a paste of common lime sets or hardens very slowly, even In 
the open air, nnlees it he sabdivided into small particles or thin 
films, it is important that the volnme of lime paste in common 
mortar shonld he bat slightly in excess of what is sufficient to coat 
all the grains of sand and to fill the voids between them. If this 
limit be exceeded, the strength of the mortar will be impaired. 
With most sands the proper proportions will be from S.5 to 3 
ToInmeB of sand to 1 volume*<of lime paate. Generally, if either 
less or more sand than this be used, the mortar will be injured, — 
in the former case from excess of time paste, and in the latter from 

81 



ovGoQi^lc 



83 UOKTAR, [chap. IT. 

porosity. Notice that the yolnme of the Tesulting mortar is abonfe 
equal to the volame of the sand alone. 

118. The ordinary method of slaking lime coosistB in placing 
the lamps in a layer 6 or 8 inches deep in either a vater-tight box, 
or a basin formed in the sand to be nsed in mixing the mortar, and 
ponring npon the Inmps a qnantity of water %i to 3 times the 
Tolnme of the lime. 

This process is liable to great abnse at the hnnds of the work> 
men. They are apt either to nse too mach water, which reduces 
the slaked lime to a semi-flnid condition and thereby injares itg 
binding qaalitiee; or, not having nsed enongh water in the first 
place, to seek to remedy the error by adding more after the slaking 
has well progressed and a portion of the lime is already rednced to 
powder, thns snddeuly depressing the teraperatnre and chilling the 
lime, which renders it granalar and lumpy. It is also very im- 
portant that the lime should not be stirred while slaking. The 
essential point ia to secnre the redaction of all the lumps. Cover- 
ing the bed of lime with a tarpanlin or with a layer of aand retains 
the heat and accelerates the slaking. All the lime neceasary for 
any required qnantity of mortar should be sl^ed at least one day 
before it is incorporated with the sand. 

After the lime is slaked the sand is spread evenly over the paste^ 
and the ingredients are thoroughly mixed with a shovel or hoe, & 
little water being added occasionally if the mortar is too stiff. 

119. Mortar composed of common lime and sand is not St for 
thick walls, becaase it depends npon the slow action of the atmos- 
phere for hardening It; and, being exclnded from the air by tho 
sorronnding masonry, the mortar in the interior of the mass 
hardens only after the lapse of years, or perhaps never." The 
mortar of cement, if of good quality, seta immediately; and, as far 
aa is known, continues forever to harden without contact with the 
(ur. Cement mortar is the only material whose strength increases 
with age. Owing to its not setting when excluded from the air, 
common lime mortar shonld never be used for masonry coostrnctioQ 
under water, or in soil that is constantly wet; and, owing to its 
weakness, it is nnsnitable for structures requiring great strength, or 

• Lime mortar taken (rom the walls olaocieiit buildings has been touod to bo- 
onlf GO to 80 per aent. saturated with carbonio acid otter nearly 3,000 years ol ex- 
poanre. Lime mortar 3.000 jaars old has been found la aubterranean vaalta, In. 
«iaaUy the ooodlUou, exoopt tor a tbln eruat on top, at freshly mixed mortar. 



ovGoQi^lc 



ART. 1.] UBTEODS OF PROPOBTIONIKO. S3- 

subject to shock. Its use in engineeriDg masonry has been aban- 
doned on all flrst-class railroads. Cement iB bo cheap that it coald 
pro&tablj be Bab8titnt«d tor lime in the mortar for ordinary 
masonry. 

130. Htdbaitlic Like Hoktab. With mortars of hydraalio 
lime the volume of sand should not be less than 1.8 times that of 
the lime paste, in order to secure the best results regardless of cost. 
The usual proportions are, hovever, for ordinary work, the same 
as in common mortars, care being taken to incorporate sufficient 
paste to coat all the grains of sand and to fill up the voids between 
them. 

. 121. Htssaitlic Cekeh Hoeiab. Hydraulic cement mortar 
hardens simultaneonsly and uniformly throughout the maas, and if 
itbft cement is good continues to gain in hardness with age, — th& 
' slov-setting cements for a longer time than the quick-setting. For 
the best results the cement paste should be jnst saSlcieut to coat 
. the grains and fill the voids of the sand. More cement than this 
adds to the cost and weakens the mortar (see § 100). If the amoant 
of cement is not sufficient to coat all the grains and fill the voids, 
the mortar will be weak and porous, and hence will not be darable. 
A dense, impervious mortar is particularly desirable for masonry 
exposed to sea-water, to exclude the water from the interior of the 
mass and prevent its chemical as well as physical action apon the 
cement. 

182. Xethoda of Proportioning. In laboratory work the propor- 
tions of the cement and sand are uniformly determined by weigh- 
ing; but there is no uniform practice of measuring the proportions 
on the work. One of the three following methods is generally 
employed. 

1. By Weight. The most accurate bnt least oommoa method 
is to weigh the ingredients for each batch. This method is incon- 
venient in practice, and adds somewhat to the cost of the work; 
and therefore occasionally the weight of a unit of volume of the 
sand and of the cement is determined, and the relative volumes of 
the ingredients are fixed accordingly, the actual proportioning being 
done by volumes. Cement is bought and sold by weight, and 
hence it is very appropriate to proportion the mcrtar by weight. 

2. Packed Cement and Loose Sand. A commercial barrel of 
cement is mixed with one or more barrels of loose sand, i.e., the 
proportioning is done by mixing one volnme ot packed cement with. 



ovGoQi^lc 



S4 MOBTAtt. [chap. it. 

one or more Tolames of loose Band. This method ie freqnentlj' 
oaed. As far as the cement is concerned, it is as accurate as tbe 
first, since the veight and volnme of a barrel of cement ma; readil; 
be known when only whole barrels are used, — as is nsnally tbe case. 
Hven though the cement is received in bags, the barrel of packed 
cement ia still a ooDTenient nnit, for an integral number of bags, 
usnally three or four, are eqnal in weight to a barrel. As far as 
the sand is concerned this method is not as accurate aa the first. 
TFhe weight of the sand is affected by the amount of moisture 
present; but a small amount of moisture sfEects the Tolnme in a 
greater proportion than the weight. For example, the addition of 
3 per cent, of water (by weight) thoroaghly mixed with dry sand 
increases tbe Tolume of the sand nearly SO per cent.* Therefore if 
the mortar is proportioned by Tolumes, damp sand will give a richer 
mortar than dry sand. The effect of moisture on the volume ia 
greater the finer the sand, and decreases as the amount of moistare 
inoreasee. Measuring the sand by volumes is inaccurate also owing 
to the packing of the sand. 

Except for the inaccuraci^ in measuring the sand, tbia method 
gives practically the same results for Portland as the first method, 
since ordinarily a unit of volume of packed cement and of sand 
weighs substantially the same; viz., 100 pounds per cubic foot. 
Since natural cement when packed in barrels nsn^ly weighs about 
75 pounds per cubic foot, a mortar of 1 part natural cement to 
1 part sand by weight is equivalent to IJ parts cement to 1 part 
«and by Tolumea of packed cement and loose sand. 

3. Loose Cement and Loose Sand. A volume of host cement is 
mixed with one or more volumes of loose sand. The actual propor- 
tioning is usually done by emptying a bag or fractional part of a 
barrel of cement into a wheelbarrow, and filling one or more wheel- 
barrows equally full of sand. As far as the sand is concerned, this 
method is as inaccurate as the second ; and it is also subject to great 
Tariationa owing to differences in specific gravity, fineness and 
packing of the cement. Even though inaccurate, it is very fre- 
quently employed. It is the most convenient method when the 
cement is shipped in bulk, — which is only rarely. 

Occasionally the actual proportioning is done by throwing into 

• Foret, Ohiet o( Laboratorr Ponts et CbausB^eB, In Engineering Nttot, vol. 
xxTll. p. 810. For BimlUr data B«e Beport at Chief ol Engineers, U. 8. A., 1S9B, 
P.298S. 



ovGoQi^lc 



ABT. 1.] mSINQ THZ UOBTA.B. 85 

the mortor-box one shoveUtil of cement to one or more ahoveUnla 
of aand. This is rery crnde, and should never be permitted. 

Since a commercial barrel of Portland nill make 1.1 to 1.4 
barrels if measured loose, a mortar composed of 1 part Portland 
cement to 1 part sand, by weight, is equivalent to 0,7 to 0.8 parts 
«ement to 1 part sand by Tolnmes of loose cement and loose sand; 
and a mortar composed of 1 part nataral cement to 1 part aand, 
by weight, is equivalent to O.SO to 0.75 parts cement to 1 part of 
Band by volames of loose cement and loose sand. 

122a. For a tabular statement incidentally showing the relatire 
omonntB of cement required by the three methods of proportioning, 
see Table 11, page 88. 

138. Proportions in Praotiee. The proportions commonly used 
in practice are: for Portland cement, 1 TOlame of cement to 2 or 3 
Tolnmes of sand; and for natural cement, 1 volume of cement to 1 
or 2 Tolnmes of sand. The specifications are usually defective in 
not defining vhich method is to be employed in proportioning. 
This is a matter of great importance. Compared with the second 
method of proportioning in 1 122, the third requires for Portland 
only 0.7 to 0.8 as much cement, and for natural cement only 0.1 
to 0.5 as much. 

134. *^^^Tlg the Hortar. When the mortar is required in small 
quantities, as for use in ordinary masonry, it is mixed as follows: 
Aboat half the sand to be nsed in a batch of mortar is spread evenly 
over the. bed of the mortar-box, then the dry cement is spread 
evenly over tbe sand, and finally the remainder of the aand is 
spread on top. The sand and cement are then mixed with a hoe 
or by turning and Te-taining with a shovel. Tbe misiug.can he 
done more economically with a shovel than with a hoe; but the 
eftectiveneas of the shovel varies greatly with the manner of using 
it. It is not sufficient to simply turn tbe mass; but the sand and 
cement should be allowed to run off from tbe shovel in such a 
manner as to thoroughly mix them. Owing to the difficulty of 
getting laborers to do this, the hoe is sometimes prescribed. If 
ekiUfuUy done, twice taming with a shovel will thoroughly mix 
the dry ingredients; although four turnings are sometimes specified, 
and occasionally as high as six (see g 260). It is very important 
that the sand and cement be thoroughly mixed. When thoroughly 
mixed it will have a uniform color. 

Tbe dampness of tbe sand is a matter of some importance. It 



ovGoQi^lc 



¥6 itosTAS. [chap. IV. 

the sand U yarj damp vhen it is mixed with the cement, anfficieat. 
moistare may be given oS to cause the cement to set partially^ 
which may materially decrease its strength. This ie particnlarly 
noticeable with qnick-settiog cementB. 

The dry mixture is then shoveled to one end of the box, and 
water is poured into the other. The sand and cement are then 
drawn down with a hoe, small qaantities at a time, and mixed with 
water nntil enough has been added to make a stiff paste. The 
mortar shonld be vigoroasly worked to insnre a uniform product. 
When the mortar is of the proper plasticity the hoe should be clean 
when drawn out of it, or at most hut very little mortar shonld stick 
to the hoe. 

Cements vary greatly in their capacity for water (see § 104), the- 
naturals requiring more than the Portlands, and the fresh-ground 
more than the stale. An excess of water is better than a deficiency, 
particularly with a quick-setting cement, as its capacity for com- 
bining with water is very great ; and farther an excess la better tlian 
a deficiency, owing to the possibility of the water evaporating 
before it has oombined with the cement. On the other hand, an 
excess of water makes a porous and weak mortar. If the mortar is 
stiff, the brick or stone should be dampened before laying; else tho 
brick will absorb the water from the mortar before it can set, and 
thus destroy the adherence of the mortar. In hot dry weather, the 
mortar in the box and also in the wall should be shielded from the 
direct rays of the sun. 

When mortar is required in considerable quantities, as in making 
concrete, it is usually mixed by machinery (see § 156n). 

125. Qboitt. This is merely a thin or liquid mortar of lime or 
cement. The interior of a w^l is sometimes laid up dry, and the 
grout, which is poured on top of the wall, is expected to find ita 
way downwards and fill all voids, thus making a solid mass of the 
wall. Qront should never be used when it can he avoided. It 
made thin, it is poroos and weak; and if made thick, it fills only 
the upper portions of the wall. To get the greatest strength, the 
mortar should have only enongh water to make a stiff paste — the 
less water the better. 

126. Data fos Estihates. The following will be found use- 
ful in estimating the amounts of the different ingredients necessary 
to produce any required quantity of mortar: 

Lima weighs about 330 pounds per barrel. One barrel of lima 



ovGoQi^lc 



ART. 1,] DATA FOR ESTIMATES. S7 

-vill make abont %i barrels (0.3 ca. yd.) of stiff lime paste. Ooe 
barrel of lime paste and three barrels of sand will make abont 
three barrels (0.4 ca. yd.) of good lime mortar. One barrel of 
nnalaked lime will make aboat 6.75 barrels (0.95 en. yd.) of 1 
to 3 mortar. 

Portland cement weighs 370 to 380 pounds per barrel net (see 
§ 77, page 54). The capacity of a Portland cement barrel Taries 
from 3.20 to 3.75 ca. ft., the aven^ being 3.49* or practically 
3.50 en. ft. A barrel of Portland will make from 1.1 to 1.4 bar- 
rels if measnred loose. A cubic foot of packed Portland cement 
(105 ponnds) and aboat 0.33 en. ft. of water will make 1 en. ft. of 
■Stiff paste; and a cubic foot of loose cement (gently shaken down 
bat not compressed) will make abont 0.8 ca. ft. of stiff paste.' 

Natural cement weighs from 865 to 300 pounds per barrel net 
(see § 77, page 54). The capacity of a natural cement barrel varies 
from 3.37 to 3.80 ca. ft., the average being 3.52,* or practically 
3.50 en. ft. A barrel of natural cement will make from 1.33 to 
1.50 barrels if measnred loose. Volame for volnme, natural 
cement will make abont the same amonnt of paBt« as Portland; or 
a cable foot of packed natnral cement (75 pounds) sod abont 0.45 
CO. ft. of water will make 1 en. ft. of stiff paste, and a cable foot 
of loose cement (gently shaken down, but not compressed) will 
maKe abont 0.8 ou. ft. of stiff paste. 

128. Quantities for a Yard of Mortar, Table 11, page 88, 
shows the approximate quantities of cement and sand required for 
■a cubio yard of mortar by the three methods of proportioning 
described in § 122. The table is based upon actual tests made by 
mixing 3^ cubio feet of the several mortars; f but at best such date 
-can be only approximate, since so much depends upon the specifio 
gravity, fineness, compactness, etc., of the cement; upon the fine- 
ness, hamidity, sharpness, compactness, etc., of the sand; and 
npon the amoont of water used in mixing. The sand employed in 
-deducing Table 11 contained 37 per cent, of voids when measured 
loose; and the plasticity of the mortar was each that moiatnre 
flashed to the sarface when the mortar was stmok with the baok 
-of the shovel Qsed in mixing. 

The Tolnme of the resulting mortar is always less than the sum 
* Tbb Tbchkooiufh, UnlTeratt; of lUlnoia, No. II, p. 101. 
t Bf L. a Bablii, Aulstuit U. S. Englneer—Me Baport ol Ohlsf of 

>D.aA., ia»t.p.383e. 



ovGoQi^lc 






[chap. it. 



3 S S S S 



jvGooi^le 



Data tor bsthiates. 



TABLE 13. 
Amovst or HoRTAB beiiiiikbd fob a Cdbio Tabd ot Habohbt. 



^o.■ 


DBKnoFTloii or HuontT. 


Hoaru. 


Mla. 


Hu. 






0.08 
0.00 
0.10 
0.25 
0.85 

0.S8 
O.SO 
0.13 
0.20 












Brickwork,— UADcUrd rize(g 266) uid i" jolata 

rtof" .... 

rtoj"" .... 

CoDcrete-sM Tablea ISd and 18«. pagM llSf, IISA. 


0.1» 
0.8S 
0.40 


10 


large stoaes, rough bammsr dreaaed 

8quared.alone masonry,— 18" couraea and I" jdots. . . . 
ir ■• " ■• " .... 


0.80 
O.IS 
0.26 



of the TolnmeB of the cement and sand, or of the paste snd sand, 
becaase part of the paste enters the Toids of the sand; but the 
Tolnme of the mortar is always greater than the snm of the Tolames 
of the paste and the solids in the sand, becaase of imperfect mixing 
and also becanse the paste coats the grains of sand and thereby 
increases their size and consequently the Tolnme of the interBtices 
betveen them. This increase in Tolnme varies with the dampness 
and compactness of the mortar. For example, the volume of a 
rather dry mortar with cement paste equal to the voids, when 
compacted enough to exclnde great voids, was 136 per cent, of the 
BQm of the volumes of the paste and solids of the sand ; and the 
same mortar when rammed had a volume of 102 to 104 per cent. 
If the paste is more than equal to the voids, the per cent, of in- 
crease is less; and if the paste is not equal to the voids, the per cent, 
of increase is more. The excess of the volume of the mortar over 
that of the sand increases with the fineness of the sand and with 
the amount of mortar used in mixing. 

129. Hortar for a Yard of Xaionry. Table 13, page 89, gives 



ovGoQi^lc 



■flO UOUTAB. [CHAP. IV. 

data ooDcemiag the amoatit of mortar required per cubic jard for 
the different classcB of masoniy, extracted from sncceeding pages 
of this Tolame; and are collected here for greater cooTenienoe in 
making estimates. 

130. STBEiroTE OF KoBiAK. The Btrength of mortar ia 
dependent upon the strength of the cementing material, upon the 
strength of the sand, and npon the adhesion of the former to the 
latter. The kind and amount of strength required of mortar 
depends npon the kind and purpose of the maeonry. If the blocks 
are large aad well dressed, and if the masonry is subject to com- 
pression only, the mortar needs only hardness or the property of 
resisting compression ; hard sharp grains of sand with comparatively 
little cementing material woald satisfy this requirement fairly well. 
If the blocks are small and irregular, the mortar should have the 
capacity of adhering to the surfaces of the stones or bricks, so as to 
prevent their displacemeDt ; in this case a mortar ricti in a good 
cementing material should be used. If the masonry is liable to be 
enbject to lateral or oblique forces, the mortar should possess both 
adhesion and cohesion. 

131. Tensile Strength. Fig. 5 shovs the effect of time upon 
the strength of various mortars. The diagram represents the 
average results of a great nnmber of experiments made in connec- 
tioQ with actual practice. Eesnlts which were nniformly extremely 
high or low as compared with other experiments were exclnded on 
the assnmption that the difference was dae to the method of monld- 
ing and testing. Since the individual v&lnes plotted were them- 
selves means, there were no very erratic results, and consequently 
the lines are quite reliable. There were fewer experiments for the 
larger proportions of sand to cement, and hence the curves are less 
accurate the larger the proportion of sand. 

The line for the strength of lime mortar probably represents the 
nuudmnm value that can be obtained by exposing the mortar freely 
to the air in small briquettes. This line is not well determined. 

Unusually hard-burned Portland cements when tested neat 
will show a greater strength than that given in the diagrams. 
Very fine cement when mixed with sand will show greater strength 
ihan that given by Fig. 5. Ag^n, the diagram shows neat cement, 
both Portland and natural, stronger than any proportion of sand, 
while frequently neat cement mortar is not as strong as a mortar 
■composed of one part sand and one part cement — particularly at 



ovGoQi^lc 



•. 1.] 



TENSILE 8TRENOTH. 



91 



"the greater agee. However, notwithetaading these exceptions, it 
is b^ieved that the resnlts represent fair average practice. The 
proportions of sand to cement were determined by weight. 

132. The reenlta in Fig. 5 are tabulated in another form in 
, to show the effect of varying the proportions of the sand 




«nd cement, and also to show the relative strength of natural and 
Portland cement mortars at different ages. The curves of Fig, 6 
are especially nsefnl in discussing the question of the relative 
eoonomy of Portland and natural cement (g 136). Por example, 
assnme that we desire to know the strength of a 1 to 2 natursj 
cement mortar a year old, and also the proportions ot a Portland 
cement mortar of equal strength. At the bottom of the lover 



ovGoQi^lc 



92 MORTAB. [OHAP. IT. ■ 

right-haud dugrun of Fig. 6 find the proportion of sand in thd 
mortar, which in this case is 2 ; follow th« correeponding line np 
ootil it iDtflreects the " natural " line. The elevation of this in- 
tersection above the biee, as read from the figure at the side of 
the diagram, b the strength of the specified mixture, which in this 





~ 






II 1 1 1 












iTHr:^ 










^'i:"" n 


J. 












" I 




\ 












"^ 


y 












s 




s 










K 




S 










^m 






^'? 










s 






4:, 






C 




\' 




T- 




^ 






a-. . 


-L. 




"^ 


'§ 






4s B r , 


5 








1 1 1 1 


(P 








/^oi'/*nbr|- 


,t:~ 


\ 








\ 








r— 1 


" 1 


?" 
















\ 










■fc*" 






s , 








»0 


^^ 












*■" 




^ 




^i. 






Pr.' 






^ 








lS"° 






^- ■/ 






'v. 










N^ 




"^ 


no 








\ 






( 


- 


1 


-- 















MM 






- 




Itknth 


- sao 


































\ 


























I 




\ 


^ 








\ 






'^'f 






aa» 




\i 








, 
















~ 








- — 








J 1 




4 














^^' 






\ 








r~f 


1, ■. ' 


• tfCi» 








1 Year 




\ 
























\. 










\ 




s 


5 














c 














'I 


\ 










54 






s 
















Sj 




























xo 




— 








— 


- 



Parts dand to I Part Cement by ]A/eight 



oaae ia aboat 350 ponnds per sqnare inch. The second part of the- 
problem then is to detennine the proportions of a Portland cement 
mortar which will have a strength of 250 ponnda per square inch. 
Find the 250 point on the Bcale at 'the side of the diagram, and 
imagine a horizontal line paaaing tbroogh this point and intersect* 



ovGoQi^lc 



ART, 1.] COHPRESSIVE 8TKEK0TB. 93 

ing the " Portland " line; from tbia point of inteTsection draw a 
Tertical line to the base of the diagram, and thia point of intersec- 
tion gives the required number of volames of sand to one Tolnme of 
cement, which ii^ this case is 5.5. Therefore a 1 to 2 natural 
mortsr a year old has a strength of 250 pounds per square inch, and 
is then eqniralent to a 1 to 5.5 Portland mortar, 

183. Compreuive Strenifth. But few experiments have been 
made npon the compreBsive strength of mortar. An examination 
of the results of about sixty experiments made with the Watertown 
testing-machine seems to show that the compressive strength of 
mortar, as determined by testing cubes, is from 8 to 10 times the 
tenule strength of the same mortar at the same age. This ratio 
increases with the age of the mortar and with the proportion of 
•and. The standard German specifications require that the com- 
pressive strength of cement mortar shall be at least 10 times the 
tensile strength. 

Data determined by submitting ou.fes of mortar to a compreesive 
stress are of little or no value as showing Xbe strength of mortar 
when employed in thin layers, as in the joints of masonry. The 
strength per anit of bed area increases rapidly as the thickness of 
tiie test specimen decreases, but no experiments have ever been 
made to determine the law of this increase for mortar, 

134. Adhesive Strength. Unfortunately very few experimeDta 
have been made on the adhesive strength of mortars, i.e., the 
power with which mortars stick to brick, stone, etc. It is com- 
monly assnmed that, after the lapse of a moderate time, the 
adhesive and cohesive strengths of cement mortars are about eqnal, 
and that in old work the former esoeeds the latter. Modem 
experiments, however, fail to establish the truth of this assumption, 
and indicate rather that the adhesion of mortar to brick or stone ia 
much less, during the first few months, than its tensile strength; 
and also that the relation between the adhesive strength and 
cohesive strength (the resistance of the mortar to pulling asunder) 
is very obscure. The adhesion of mortars to brick or stone varies 
greatly with the different varieties of these materials, and particu- 
larly with their porosity. The adhesion also varies with the quality 
of the cement, the character, grain, and quantity of the sand, the 
amoant of water used in tempering, the amount of moisture in the 
stone or brick, and the age of the mortar. Some cements which 
exhibit high tensile strength give low values for adhesion; and ooa- 



ovGoQi^lc 



[CUAP, IV. 





TABLE 18. 


— Adhbbivb Btkknotb 


OF 


HOBTABB 




i\ 






iierigo Id bHtTS ilreBEtli In 




s s 

a 1 




fi 


ill 


P 


AOthOrttf. 


i 




■fi ;:;:;; 




b«i«™b_..im 


i 




i 




- " :::-.:- 


! 


If, 


S 


s 


;r 


^'?*™-^ 


Ii 












&• 


i;: 


IS 


£ 


l^(. WarnD..'n 
















IS 


w 
la 


f 


ax8 


tt 




o^B. aiiiiiior..-a 


s 








" ■■ ::::- 
















3 
a. 




































kSfcr::iS! 




tu 














a 


"mi 


*L» 


K 


"iK«|Dr.Bitlims..::ia» 


















S 
^ 




Robenion.'i.'hBSS 
l.l.a>.nn'.'.'.'.lfm 


































il 






:::::: 






•85 








a 




<9 

40 


w 

M 

70 
IT 
» 





i 


ssr'.-?.".is 




' po'ni™i.. ..."... .."... 




1 


M.llet IM 












- 
















- 






atjt.Wue brick 












_ 






mnbtm red 












• " 




H 








" " 



















I OouwpanjiclM Ui wmant air Md oiit bcron taUn(. 1 



D.qilizMb,G00>^le 



ART. 1.] COST OP UORTAR. 95 

Tersolj, cements which are apparently poor when tested for cobeeiOQ, 
show excellent adbeeive qaalities. 

The table * on the preceding page givea all the reliable data 
known. A comparieon of the table with the diagram on page 92 
ehows that the adhesion of a mortar ia far leas than its tendle 
strength at the same age, bnt faila to ahow any definite relation 
between the two. In the experiments by Dr. B5hme at the Boyal 
Testing Laboratory at Berlin the mortar waa made with standard 
qaartz sand, and the tensile strength of the mortar and its adheaion 
to common brick were determined separately. By comparing the 
tensile and adhesive strengths at the same ages, it was fonnd that 
the former was always abont ten times greater than the latter when 
the mortar consisted of one part of cement to three or fonr parts 
of sand, and from six to eight times greater when the cement waa 
nsed neat or with one or two parts of sand. In the experiments 
made by Prof. Warren, of Sydney UniTereity, New Sonth Wales, 
tbe tensile strength of neat Portland cement mortar waa three and 
a half times ite adhesion to brick. The result of twelve thousand 
epecial testa by Mr. Mann was that pnre Portland cement of 425 
pounds tensile and 5,640 pounds compressive strength pe'rsqnara 
inch has bnt 60 to 80 pounds adhesion to limestone, and that the 
ratio of tenaila to adhesive strength varies from 5 te 1 to 9 to 1. 

135. Cost or Uobtas. Knowing the price of the materials it 
is very easy, by the nse of Table 11, page 88, to compute the cost 
of the ingredients required for a cubic yard of mortar. The 
expense for labor is quite variable, depending upon tbe distance the 
materials must be moved, the qnantity mixed at a time, etc. As a 
rough approximation it may be assumed that a common laborer can 
n^ix 3 jarda f er day, at a cost of, say, 50 cents per cubic yard. If 
the mixing is done by machinery, the cost may be as low as 25 
cents per cabic yard. The cost of a cubic yard of mortar composed 
of 1 part Portland cement and 2 parts sand, both by weight ia 
then ahont as follows : 

Cemeat 3.80 bbls. (see page 88) @ S8.00 = $8.40 

Sand 0.78CU. yd. (seepage 88j igi .60 = .89 

LatMr, handling msterial« and mlziog i day® (l.fK) = .60 

Talalco*tofi eubie yard of mortar = 19.29 



ovGoQi^lc 



96 HOBIAB. [CHA.F. 17. 

136. Hatnral vs. Fortland Cement HorUr. It is sometimes a 
qaestion whether Portland or nataral cement Bhoald be osed. If 
a qnick-setttng cement ig reqaired, then natnral cement is to be 
preferred, since as a rale the natural c«menta are quicker setting, 
although there are many and marked exceptions to this role. Other 
things being the same, a slow-setting cement is preferable, since 



■ k' 
\^ 
t' 

fltrftt SatHt ra / Atrr Cement by Weight 
Flo. Ta,— Cost of OENEtrr Mobtib. 

it is not so liable to set before reaching its place in the wall. This 
is an important item, since with a quick-setting cement any slight 
delay may necessitate the throwing away of a boifal of mortar 
or the remoTal of a atone to scrape oat the partially-set mortar. 

Generally, however, this qaestion shoald be decided npon 
economical gronnds, which makes it a question of relative strength 
and relative prices. The tensile strength of natnral and Portland 
cement mortars is shown in Fig. 6, page 93, The cost of mortar 
of Tarions proportions of sand may be computed as in the preceding 
flection; but as the cost of labor is uncertain and ig sabstantially 
the same for both kinds of mortar, it is sufficient to deal with the 
cost of the materials only. Assuming Portland cement to cost 13 
per barrel, natural (1 per barrel, and sand 50 cents per cubic yard, 
and using Table II, page 86, the cost of the materials in a onbiA 
yard of mortar is as in Fig. 7a. 







































































s, 












t 




\, 












\ 






<"■ 










^> 






S; 


V 










\ 






N 












"-.^ 










— 








i'i^ 

























— 


















J 








- 


t 








jvGoogle I 



'. 1.] HATCRAL VS. PORTLAND CEMENT UORTAS. 



I- 



J 



Cost of tlortar in DoUara per Cubic Yard 
Wia. 75.— RcATm Ecohoht or Natdbu, ahq Pdbtumd CramrrH. 



1 








/ 


\ 










1 






/ 


\ 










i 




1 






— 


!Sn; 


^ 






/ 












\ 




%50 








V 


^ 


k 






1 












^ 


N 






1 












-A 




\ 



fitrta Sanel to I fbrf Cement by Vlikiqht 
Fta. Tc— EcoiroMio Pkopobtiom of Sasd. 



ovGoQi^lc 



' 98 M0RT4E. [CHAP, IT- 

By plotting the strength of Portland and tiatnral cement mortar 
6 months old and the coet of a yard of mortar as given in Fig. loy 
Fig. 7b is obtained, which ehovs the relation hetween the strength 
at 6 months and the coat of the mortar made of the two kinds of 
cement. Kotice that for any tensile strength under abont 370 
ponnde per square inch, either natnral or Portland cement may be 
need, bnt that the former is the cheaper. In other words, Fig. 7^' 
shows that if a strength of abont 370 poands per sqnare inch at 
6 months is sufficient, nataral cement is the cheaper. Nearly all 
carefnlly condncted testa of the strength of cement mortar 6 
months old or over give a aimilar resnlt, except that the above- 
limit ifl nsnally between 300 and 350 pounds. A considerable 
change in prices does not materially alter the resnit, and hence the 
conclnsion may be drawn that if a strength of 300 to 350 pounds- 
per square inch at 6 months is sufficient, nataral cement is more 
economical than Portland. Incidentally Fig. 7c, page 97, shows 
the same relation. However, in this connection it should not be 
forgotten that other considerations than strength and coat may 
govern the choice of a cement; for example, aniformity of prodnct^ 
rapidity of set, and sonndnesa are of equal or greater importance 
than strength and cost. 

Mortar made of two brands of Portland or nataral cement will 
differ considerably in economic values, and hence to be of the 
highest value the above comparison should be made between the 
most economical Portland and the most economical natnral cement, 
as determined by the method described in § lllA. 

Short-time tests do not warrant any general conclnsion as to the 
relative economy of nataral and Portland cements, since the 
strength at short times varies greatly with the activity of the 
cement. For example, the two apper diagrams of Fig. 6, page 
93, when plotted as in Fig. 7& show Portland to be the more 
economical, while other similar experiments show nataral cement to 
be the more economical. 

137. Economic Proportion of Sand. Fig. 7c, page 97, ehowa 
the ratio of strength to cost for different proportions of sand, for 
both Portland and nataral cement; in other words. Fig. 7c shows. 
the tensile strength in poands per square inch for each dollar of the 
cost of a cubic yard of mortar. For example, if a nataral cement 
mortar at 6 months has a tensile strength of 280 pannda per squarfr 
inch, and costs t3.95 per yard, the strength per dollar is: 380 •{- 



ovGoQi^lc 



ART. 1.] EFFECT OP RE-TEMPEBIHO. ^9 

2,95 = 94.9 pounds per square inch. In this way Fig. 7c waa 
oonBtnicted, aaing the cost of mortar as given in Fig. la and the 
strength as determined by L. C. Sabin in connection with the con- 
strnotiAn of the Poe lock on the St. Mary's Falls Canal.* Accord- 
ing to this diagram the most economic mortar, either natural or 
Portland, conaists of 3 parts sand to 1 part cement, by weight. 

A atndy of the resnlta of other eiperimenta shows that the above 
coQclnsiona are not general. The maximom ratio aa above is 
different for different ages for the same cement, and. at the same 
age is different for different cements. The above ratio varies (1) 
with the activity of the cement, which determines the strength 
neat at different ages; (3) with the fineness, which determines the 
sand-carrying power of the cement; (3) with the fineness of the 
sand, which determines the surface to be covered by the cement; 
and (4) with the cost of the cement and the sand. If the strength 
of any particular cement with the various proportions of sand ia 
known for a particalar age, and the price of the cement and sand 
also is known, the most economic propertion of sand can be com- 
puted as above. To determine the most eoonomio mortar, the 
most economic cement should be selected as described in § lllA, 
and then be mixed with the most economical proportion of sand as 
,,- above. 

Strictly, the maximum ratio of strength to cost determined as 
above ia not necessarily the most economical mortar. The work in 
band may not require a mortar as strong as that giving the maxi- 
mum patio of strength to cost, in which case a mortar having a 
smaller proportion of cement may be used; and similarly, if the 
work requires a mortar stronger than that giving the maximum 
ratio of strength and cost, then a mortar must be used which con- 
tains a greater proportion of cement. 

1S8. Effect O? Hl-TBUMEnro. Frequently, in practice, 
cement mortar which has taken an initial set, is re-mixed and used. 
Masons generally claim that re-tempering, i.e., adding water and 
re-mixing, is beneficial; while engineers and architects usually 
specify that mortar which has taken an initial set shall not be need. 

Be-tempering makes the mortar slightly less "short" or 
"brash," that is, a little more plastic and easy to handle. Be- 
tempering also increases the time of set, the increase being very 

• Baport of Chlet of Eagluaere, U. S. A., 1B93, page 3019, Table L 



ovGoQi^lc 



1° !>-''. 

^Ts ...... ^f.' fj^ ^- ■ 

100 KOBTAR. [chap. IV. 

different for different cements. Bat on the other hand, re-temper- 
ing Ttsnally weakens & cement mortar, A qnick-settiog natDral 
cement Bometimee loses 30 or 40 per cent, of ita strength b; being 
re-tempered after standing 20 minntes, and 70 or 80 per cent, by 
being re-tempered after standing 1 hour. With slow-setting 
cements, particnlarly Portlands, the loss hj re-tempering immedi- 
ately after initial set (§ 84) is not material. A mortar which has 
been insnfiiciently worked is sometimes made appreciably stronger 
by re-tempering, the additional labor in re-mixing more than com- 
pensating for the loss caused by breaking the set. 

The loBsof strength by re-tempering is greater for qnick-setting 
than for slow-setting cement, and greater for neat than for sand 
mortar, and greater with fine sand than with coarse. The loss 
increases with the amonnt of set. If mortar is to stand a consider- 
able time, the injaiy will be less if it is re-tempered several times 

' daring the interval than if it is allowed to stand nndistnrbed to the 
end of the time and is then re-mized. Re-tempered mortar shrinks 
more in setting than ordinaiy mortar. This fact sometimes 
accounts for the cracks which frequently appear npon a troweled 
surface. 

The only safe mle for practical work is to require the mortar 
•to be thoroughly mixed, and then not permit any to be used which 
lias taken an initial set (§ 84). This rule should be more 
atrennoDsly innsted upon with natural than with Portland cements, 

-and more with qnick-setting than with slow-setting varieties. 

139. Lna WITH Cekeft. Cement mortar before it begins to 
aet has no cohesive or adhesive properties, and is what the maeoo 

-calls " poor," "short," "brash "; and conseqnently is difficnlt to 
use. It will not stick to the edge of the brick or stone already laid 
sufficiently to give mortar with which to strike the joint. The 
addition of a small per cent, of lime paste makes the mortar ' ' fat '* 
or " rich," and more pleasant to work. The substitution of 10 to 
20 per cent, of lime paste for an equal volume of the cement paste 
does not materially decrease the strength of the mortar, and 
frequently the addition of this amount of lime slightly increases its 
strength. In all cases the substitution of 10 to 20 per cent, of lime 
decreases the cost more rapidly than the strength, and hence is 

' economical; but the snbstitation of more than about 20 per cent, 
decreases the strength more rapidly than the cost, and hence is not 

• Monomioal. The economy of using lime with cement is, of coutse, 



ovGoQi^lc 



ART. 1.] HORTAB lUPERTIODS TO WATER. * 101 

greater with Portlaad thao with n&tural cement owing to the 
greater cost of the former. 

If the mortar is porone, i.e., if the voids of the sand are not 
filled with cement, the addition of lime will make the mortar more 
dense and plastic, and will also increase its strength and cost. The 
increase in strength is not proportional to the increaae in cost, hat 
the increased plasticity and density jastify the increased cost — the 
former adds to the ease of using the mortar, and the latter to its 
dnrability. 

The addition of lime does not materially aSect the time of set, 
«nd nsoally slightly increases it. 

It has long been an American practice to reinforce lime mortar 
by the addition of hydraulic cement. The mortar for the 
"ordinary brickwork" of the United States pnblic hnildings is 
composed of " one fourth cement, one half sand, and one fourth 
lime." The cement adds somewhat to the strength of the mortar, 
bnt not proportionally to the increase in the cost of the mortar. 

140. XOSTAS iKFEBTloirs TO Wateb. Nearly every failnre 
«f masonry is due to the dis integration of the mortar in the ontside 
of the joints. Ordinary mortar — either lime or cement — absorbs 
water freely, common lime mortar absorbing from 50 to 60 per cent. 
of its own weight, and the beet Portland cement mortar from 10 
to 20 per cent. ; and conseqnently the; disintegrate under the 
action of frost. Mortar may be made practically non-absorbent 
by the addition of alnm and potash soap. One per cent., by 
weight, of powdered alnm is added to the dry cement and sand, and 
thoroQghly mixed; and about one per cent, of any potash soap 
{ordinary soft^eoap made from wood ashes is very good) is dissolved 
in the water nsed in making the mortar. The alum and soap com- 
bine, and form oompoonds of alnmina and the fatty acids, which 
are insoluble in water. These compounds are not acted npon by 
the carbonic acid of the air, and add considerably to the early 
Btrengtb of ihe mortar, and somewhat to its nltimate strength. 

With lime mortar, the alnm and soap has a slight disadvantage 
in that the oompounds which render the mortar impervious to water 
also prevent the air from coming in contact with the lime, and 
conseqnently prevent the setting of the mortar. On the other 
han^, the alum and soap compounds add considerably to both the 
early and the nltimate strength of the mortar. 

This method of rendering mortar impervious is an application 



ovGoQi^lc 



103 MORTAR, [nHAP. IT. 

of the principle of Sylvesier's method of repelling moietute from 
external walls by applying alam and soap washes alternately on the 
outside of the wall (see § 363). The same principle is applied in 
McMnrtrie's artificial stone (see g 162). The alam and soap are 
easily need, and do not add greatly to the cost of the mortar. The 
mixture conld be advantageonsly used in plastering, and in both 
cement and lime mortars of ontside walls or masonry in damp 
places. It has been very Bnccessfnlly used in the plastering of 
cellar and basement walls. It should be employed in all mortar 
need for pointing (§ 204). 

The addition of a small amount of very finely powdered clay 
(§ 114c) decreases the permeability of mortar; but since clay absorh» 
and parts with water with the changing seasons, the use of clay ia 
not efficient in preventing disintegration by freezing and thawing. 

141. Fbeeziho or Uobtab. The freezing of mortar before it 
has set has two effects: (1) the low temperature retards the setting 
and hardening action; and (3) the expansive force of the freezing 
water tends to destroy the cohesive strength of the mortar. 

142. Effect on Lime Mortar. The freezing of lime mortar 
retards the evaporation of the water, and consequently delays the 
combination of the lime with the carbonic gas of the atmosphere. 
The expansive action of the freezing water is not very serious upon 
lime mortar, since it hardens so slowly. Consequently lime mortar 
is not seriously injured by freezing, provided it remains frozen 
nntil fully set. Alternate freezing and thawing somewhat damages 
its adhesive and cohesive strength. However, even if the strength 
of the mortar were not materially affected by freezing and thawing, 
it is not permissible to lay masonry during freezing weather; for 
example, if the mortar in a thin wall freezes before setting and 
afterwards thaws on one side only, the wall may settle injariously. 

When masonry is to be laid in lime mortar during freezing 
weather, frequently the mortar is mixed with a minimum of water 
and then thinned to the proper consistency by adding hot water 
just before using. This is nndesirable practice (see § 118). When 
the very best results are sought, the brick or stone should be 
warmed — enough to thaw off any ice upon the surface is sufficient 
' — before being laid. They may be warmed either by laying them 
on a tomace, or by suspending them over a slow fire, or by w-^ting 
with hot water, or by blowing steam throagh a hose against them. 

143. Effect OS Cement Hortar. Owing to the retardation of the 



ovGoQi^lc 



ABT. I.] EFPECT OF FfiEEZlKQ. 103 

low temperatare, the settiDg and hardening may be so delayed that 
the water may be dried ont of the mortar and not leave enongh for 
the chemical action of hardening; and consequently the mortar will 
he weak and cmmhly. This wonld be anbatantially the same as 
aaing mortar with a dry porons brick. Whether the water evapo- 
rates to an iDJnrions extent or not depends npon the hamiditj of 
the air, the temperature of the mortar, the activity of the cement, 
«nd the extent of the exposed anrface of the mortar. The mortar 
in the interior of the wall is not likely to be injured by the loss of 
water while frozen; but the edges of the joints are often thus seri- 
•onsly iujured. In the latter case the damage may bo fnlly repaired 
by pointing the masonry (§ 304) after the mortar has fnlly set. 

On the other hand when the cement has partially set, if the 
expansive force of the freezing water is greater than the cohesive 
strength of the mortar, then the bond of the mortar is broken, and 
on thawing ont the mortar will crumble. Whether this action will 
take place or not will depend chiefly npon the Btreugth and activity 
of the cement, upon its hardness at the time of freezing, aud upon 
the amount of free water present. The relative effects of these 
several elements is not known certainly; but it has been proven 
conclusively that for the best results the following precautions 
shonld he observed: 1. Use a qnick-aetting cement. 3. Make the 
mortar richer than for ordinary temperatures. 3. Use the mini- 
mum quantity of water in mixing the mortar. 4. Prevent freezing 
as long as possible. 

There are various ways of preventing freezing: 1. Cover the 
masonry with tarpaulin, straw, manure, etc. 2. Warm the stone 
and the togredients of the mortar. Heating the ingredients is not 
of much advantage, particularly with Portland cement. 3. Instead 
of trying to maintain a temperature above the freezing point of 
fresh water, add salt to the water to prevent its freezing. The 
usual mle for adding salt is: "Dissolve 1 pound of salt in 18 
gallons of water when the temperature is at 33° Fahr., and add 
3 ounces of salt for every 3° of lower temperature." The above 
rule gives a alight excess of salt. The following rnle is soientifically 
correct and easier remembered: " Add one per cent, of salt for each 
Fahrenheit degree below freezing." Apparently salt slightly 
decreases the strength of cement mortar setting in air, and slightly 
increases the strength when setting in water.* 
' • Beport of Clael of Engliieera, U. B. A., 189S, pp. 2963-71, 301S. ' 



ovGoQi^lc 



104 HOBTAB. [chap. IV. 

Alternate freezing txti thawing is more damaging than contin- 
QOBB freezing, since with the former the bond may be repeatedly 
broken; and the damage dne to sncceaBive disturbance increases 
with the nnmber. 

144. Practice has abown that Portland cement mortar of th» 
nsnal proportions laid in the ordinary way is not materially injured 
by alternate freezing or thawing, or by a temperature of 10° to 
15" F. below freezing, except perhaps at the exposed edges of thfr 
joints. Under the same conditions natural cement mortar is liable 
to be materially damaged. 

By the use of salt, even in less proportions than specified above^ 
or by warming the materials, masonry may be safely laid with. 
Portland at a temperature of 0° P. ; and the same may nanally b& 
done with natural cement, -although it will ordinarily be necessary 
to re-point the masonry in the spring. Warming the materials is 
not as efTective as using salt. 

14fi. CHAHeE 07 VoLVU IK SsTmre. The Committee of th» 
American Society of Civil Engineers draw the following oonoln- 
sions:* 1. Cement mortars hardeniug in air diminish in linear 
dimensions, at least to the end of twelve weeks, and in most oase» 
progresdvely. 3. Cement mortars hardening in water increase ia 
like manner, but to a leas degree. 3. The contractions and expan- 
«ions are greatest in neat cement mortars. 4. The quick-setting 
cements show greater expansions and contractions than the slow- 
setting cements. 5. The changes are less in mortars containing- 
sand. 6. The changes are less in water than in air. 7. The con- 
traction at the end of twelve weeks is as follows: for neat cement 
mortar, 0.14 to 0.33 per cent. ; for a mortar composed of 1 part 
cement and 1 part sand, 0.08 to 0.17 per cent. 8. The expansion, 
at the end of twelve weeks is as follows: for neat cement, 0.04 to> 
0.35 per cent. ; for 1 part cement and 1 part sand, 0.0 to 0.08 per 
cent. 9. The contraction or expansion is essentially the same m 
all directions. 

146. Elabticitt, Cohprebbioit, ash Set op Hobiak. For 
data on elasticity see page 14. The evidence is so conflicting that 
it is impossible to determine the coefficient of compression and of 

• Bee the " Report ot Progreas at the Oommittee on the OompraBalve Strei^th of 
Cemente and the Oompreasion ot Hortare aod Settlement of Maaourj," in the 
Tr&nsaoUoiiB ot that Sooiety, vol. xvU. pp. 3IS-3T ; also s similar report In vol. zri 
pp. 717-33. 



ovGoQi^lc 



ART. 1.] ELASTICITY, COUPRGBSIOST, AKD SET OF VOBTAB. 105> 

Bet of mortar, even tipproziniat«ly. For mach valoable data on. 
this and related sabjecta, see tho " Report of Progress of the Com- 
mittee od the GompressiTe Strength of CemeDta and the Compres- 
sionof Mortars and Settlement of Masonry," in the Transactiods of 
the American Society of Civil Engineers, vol. xti. pp. ?17-33„ 
Tol. xvii. pp. 213-17, and also vol. zriii. pp. 2S4r-80. The aeverai 
annnal reports of tests made with thB United States GoTemment 
teating-nuichiue at Watertown contain valuable data — particnlarljr 
the report for 1884, pp. 69-247 — bearing indirectly npon this and 
related Babjecte; bat since some of the details of the eiperimenta 
are wanting, and since the fnndamental principles are not well 
enough anderatood to carry ont intelligentlj a series of experimentB, 
it is impossible to draw any v^aable ooBolosionB from the data. 



ovGoQi^lc 



[chap. it. ' 



ABT. 2. CONCBETB. 

147. Concrete consists of mortar in which is embedded small 
pieces of some hard material. The mortar is often referred to oa 
the matrix; and the embedded fragments, as the aggregate. Ood* 
«rete is a epeoies of artificial stone. It is sometimee called b4toQ, 
the French eqaivaleht of concrete. 

" Concrete is admirably adapted to a Tariety of moat importast 
Qses. For fonndatioos in damp and yielding soils and for snbter- 
ranean and submarine masonry, under almost every combination of 
circnmstancee likely to be met with in practice, it is superior to 
brick masonry in strength, hardneBs, and dnrability; is more 
economical; and in some cases is a safe snbstitnte for the best 
natural stone, while it is almost always preferable to the poorer 
varieties. For submarine masonry, concrete poBsesaeB the advan- 
tage that it can be laid, under certain precautions, withont ezhaast- 
ing the water and without the use of a diving-bell or submarine 
armor. On account of its continuity and its impermeability to 
water, it is an excellent material to form a substratum in soils 
infested with springa; for sewers and conduits; for basement and 
sustaining walls; for columns, piers, and abutments; for the 
hearting and backing of walls faced with bricks, rubble, or ashlar 
work; for pavements in areas, basements, sidewalks, and cellars; 
for the walls and floors of cisterns, vaalts, etc. Groined and 
vaulted arches, and even entire bridges, dweiUng-hoases, and fac- 
tories, in single monolithic masses, with saitable ornamentatioD, 
have been constructed of this material alone." 

Concrete ia rapidly coming into use in all .kinds of engineering 
constructions. It enables the engineer to build his su pc I'S t rue t lire 
on a monolith as long, aa wide, and as deep iis he may think best, 
which can not fail in parts, but, if rightly proportioned, must go 
«11 together — if it fulls at all. 

148. The Xobtab. The matrix may be either lime or cement 
mortar, but is usually the latter. The term concrete is almost 
universally nnderstood to be cement mortar with pebbles or broken 
stone embedded in it. Lime mortar is wholly unfit for use in large 
masses of concrete since it does not set when excluded from the air 
(see § 119). 



ovGoQi^lc 



^RT. 3.] THE A08RE0ATE. 107 

The cement mortar may be made as already described in Art. 1 
preceding. 

149. Thx Aooee&atz. The aggregate may cooBist of small 
pieces of any hard material, as pebbles, broken stone, broken brick, 
shells, slag, coke, etc. It is added to the mortar to rednoe the 
cost, and withio limits also adds to the strength of the coDcrete. 
Ordinarily either broken stone or gravel is nsed. Coke or blast- 
farnace slag is nsed when a light and not strong concrete is desired, 
se for the fonndation of a pavement on a bridge or for the fioon 
of a tall building. Of course a soft porons aggregate makes a weak 
concrete. 

What«Ter the aggregate it should be free from dust, loam, or 
any weak material. The pieces shonld be of graduated sizes, so 
that the smaller shall fit into the voids between the larger. When 
this condition is satisfied less cement will be required and conae- 
<]Qently the cost of the concrete will be lees, and at the same time 
ita strength will he greater. Other things being eqnal, the rongfaer 
the snrfaces of the fragments the better the cement adheres, and 
consequently the stronger the concrete. 

160. It is sometimes specified that the broken stone to he nsed 
in making concrete shall be screened to practically an nniform size; 
bnt this is nnwise for three reasons; viz.: 1. With graded sizes the 
smaller pieces fit into the spaces between the larger, and conse- 
quently less mortar is required to fill the spaces between the 
fragments of the stone. Therefore the unscreened broken stone is 
more economical than screened broken stone. 3. A concrete con- 
tuning the smaller fragments of broken stone is stronger than 
thongh they were replaced with cement and sand. Experiments 
shoir that sandstone screenings give a considerably stronger mortar 
than natural sand of eqnal fineness, and that limestone screenings 
make stronger mortar than sandstone screenings, the latter giving 
from 10 to 50 per cent, stronger mortar than natural sand.* 
Hence, reasoning by analogy, we may oonclade that including the 
finer particles of broken stone will make a stronger concrete than 
replacing them with mortar made of natural sand. Farther, 
experiments show that a concrete containing a considerable propor- 

•ADUiua Beport ot Ohlel of En^naera, U. S. A., 189S, Fart 3, p. SOIB; do. IBM, 
Part i,p.a8U; do. JBSB, Put 4, p. 3959 ; Jour. West. Boo. ol Sng^, vol. U. pp. 3U 
and 400. 



ovGoQi^lc 



108 CONCEBTE. [CHAP. IT. 

tioa of brokeo atone is itronger than the mortar ^one (see the second 
and third paragrapha of g 153). Since the mortar alone ia -weaker 
than the concrete, the leas the proportion of mortar the strontrer the 
concrete, provided the roids of the aggregate are filled; and ther^ 
fore concrete made of broken stone of graded aizes is atronger than 
that made of practically one size of broken stone. 3. A aingle size 
of broken atone has a greater tendency to form archea while being 
rammed into place, than atone of graded aizea. 

Therefore concrete mode with acreened atone ia more ezpenaive 
and more liable to arch in being tamped into place, and is leas denae 
and weaker than concrete made with anacreened stone. 

In short, screening the atone to nearly one size is not only a 
needless expense, bnt ia also a positive detriment. 

The dnat ahonld be removed, aince it baa no atrengtb of itself 
and adda greatly to the surface to be coated, and also prevents the 
oontact of the cement and the body of the broken atone. Particles 
of the size of aand grains may be allowed to remain if not too fine 
nor in excess. The small particles of broken atone shoald be 
removed if to do so redacea the proportion of voida (g§ 115d, 115e). 

161. Oravel n. Broken Stone. Often there is debate as to the 
relative merits of gravel and broken atone oa the aggregate for con- 
crete ; bnt when compared npon the same basis there is no room for 
doubt. 

In the preceding section it was shown that finely crashed stone 
gave greater tensile and compressive strengths than equal propor^ 
tiona of sand; and hence reaaoning by analogy, the conclaaioa is 
that concrete composed of broken atone ia atronger than that con- 
taining an equal proportion of gravel. Thia element of strength is 
dne to the fact that the cement adheres more closely to the rongh 
snrfaces of the angalar fr^menta of broken stone than to the 
smooth surface of the rounded pebbles. 

Again, part of the resiataDce of concrete to crushing ia dne to 
the frictional resistance of one piece of ^gregate to moving on 
another; and consequently for this reason broken stone ia better 
than gisvel. It is well known that broken atone makes better 
macadam than gravel, aince the rounded pebbles are more easily 
displaced than the angular fragments of broken stone. Concrete 
differs from macadam only in the use of a better binding material : 
and the greater the frictional resistance between the particles the 
stronger the mass or the less the cement required. 



ovGoQi^lc 



IRT. 2.] THBOBY 0» THE PE0P0BTI0H8. 109 

A Beriea of experiments* m&de bj the City of Washington, 
D. C, to determine the relative valne of broken atone and gravel 
for concrete, vhich are enmrnarized in § 157^, P&ge 112r, gives the 
following reaolts: 



Stbbnoth of QiuTaL Cohcbbtb ih txbkb 


or BitoEBM Stohb Cow 








Fonuini CnONT 


10 days. 88 per cent. 


76 per cent. 


45 ■' 78 " '■ 


91 " ■■ 


8momh». 96 " " 


119 " " 


6 " 48 " '■ 


78 '■ ■' 


1 yew. 88 " " 


108 ■• " 


JfM» 68 " " 


98 " ■' 



Each reealt is the mean for two 1-foot cnbes, except tbat the valnes 
for a year are the means for five cnbes. Kotice that the gravel 
coocrete is relatively weaker for the earlier ages, owing probably to 
the greater internal friction of the broken-stone concrete. 

In a BericB of forty-eight French expert m eels, f the crashing 
strength of gravel concrete with Portland cement is only 79 per 
cent. OB great as that of brokea-stone concrete. The grarel had 40 
per cent, of voids, while the broken stone had 47 per cent., which 
favored the gravel concrete (see § 154). The resalta of these test» 
are shown graphically in Fig. 8, page 112a. 

153. Since frequently gravel is cheaper than broken stone, » 
mixture of broken stone and gravel may make s more efficient con- 
crete than either alone, {. e., may give greater strength for the same 
coat, or give leas cost for the same strength. 

153. Thzobt of tee Fsopoetiokb. The voids in the aggre- 
gate should always be filled with mortar. If there is not enough 
mortar to fill the voids, tbe concrete will be weak and porona. On. 
the other hand, more mortar than enongh to fill the voids of the 
aggregate increases the cost of the concrete and also decreases its- 
strength. The' decrease in strength dne to an excess of mortar is 
nsnally greater than wonld be prodnced by aubstitnting the same 
amount of aggregate, since ordinarily the sand and the aggregate 
have approximately the same per cent, of voids, while the sand has 
the greater, and also the smoother, surface. 



ovGoQi^lc 



110 CONCKBTB. [chap. IT. 

A correctly proportioned concrete is always stronger than the 
mortar alone. For example. Table 13a * shows thajt a concrete 
containing a oonaiderable proportion of pebbles is stronger than the 
mortar alone — compare lines 2, 5, and 8 with the preceding line of 
each gronp, reepectirely. The reeults are for grarel concrete, and 
they wonld be more striking tor broken-stone concrete, since the 
cement adheres better to broken stone than to either sand or 
grarel. 

TABLE 18a. 
KxLATivs Stbrksth or Hortab aso Qrayml Coxcbxtc 







Tbibothb 


■ niUTKKJ). 


Bar. No. 


Froport.0.-. 


c™b^«™.gth. 






Cnoaot. 


Bud. 


PebblM. 


tbkt of tlie MorMr. 




1 


2 





8 1S8 


100 per cent. 








8 


3783 


136 " " 








9 


a 414 


136 ■' " 




I 


8 





1406 


ITO POT oanl. 








B 


1661 


114 " " 








S.6 


1084 


10» " " 




1 


4 





1068 " 


100 percent 








S 


1391 


131 " " 








8.5 


laai 


lU *■ " 



The average strength of twenty-fonr cnbes ranging from 6 to 
16 inches on a side, made nnder the direction of Gen. Q. A. Oill- 
more,t and composed of 1 volame of cement, 3 volumes of sand, 
and 6 volumes of broken stone, was 15 per cent, more than that of 
corresponding cnbes made of the mortar alone. In another series 
of the same experimeQtB,J the average strength of eight cubes of 

* Dr. B. Dyeberhoff, a 0«rmau aathorlty, as qooted In " Der Portland Cemaut 
nnd seina Anvrendaagen Im Bauwasaa." p. 90. 

f IToteB on the OompresalTe BeslBtanoe ol Preeatonee, Brlok Flars, Hydnulio 
Cement, Kortar, and Ooneretee, Q. A. Glllmore. John Wiley A Bona, New Tork, 
188S, pp. lST-10 and 14S-U. 

} Ibid., pp. 141-43. 



ovGoQi^lc 



THEOBT OF THB PH0P0BTI0H8. 



Ill 



AST. 2.] 

concrete oompoeed of 1 pftrt cement, 1^ parta aand, uid 6 pHti 
broken stone was 9fi per cent, of that of corresponding cnbes of the 
mortar slone, which is interesting ss showing that a lean concrete 
is nearly as strong as a veiy rich mortar, 

A correctly proportioned concrete is also stronger than either a 
richer or a leaner mixture — see Table 13t, page 113/. 

164. For the strongeet and densest concrete, the voids of the 
aggregate shoald be filled with a rich strong mortar; bnt if a 
oheaper concrete ia desired, fill the roida of the aggregate with sand 
and odd as mach cement as the cost will justify. In other words, 
to make a cheap concrete, nse as lean a mortar as the cironmstanoea 
warrant, but nse enough of it to fill the Toida of the oggr^ate. 
Sand is so cheap that there is no appreciable saTing by omitting it; 
and the use of it makes the concrete more dense. 

The strength cf a concrete varies nearly with the amount and 
strengUi of the cement used, provided the mortar is not more than 
enoagh to fill the voids. Table 13d shows the strength of con- 
crete in terms of the cement employed. The data from which 
TABLE IW. 

BZIATION BBTWEEN THE CbUSHINO STXBKQTE OF CONCBRTK ABD THB 

Phopoetion of Ckmkmt. 
Hoitai Equal to the Tolda Id ilie Aggregate. 



R- 


VolunH Loom. 


C-JS""' 


iSK'SSSS. 


OOMDt. 


Sud. 


keCaal. 


RolMiTe. 


AcnuU. 


TheoretlMl. 


RelMln. 








0.00 


1.00 


4.467 


6,000 


1.00 






8 


0.88 


0.67 


8.781 


8,800 


0.68 








0.8S 


O.SO 


2.M8 


8,800 


0.80 








o.ao 


0.40 


a,oio 


8,000 


0.40 








0.17 


0.S8 


1,796 


1.600 


0.88 








0.14 


0.28 


1,885 


1.400 


0.88 



this table was made are the same as those summarized in Table 
13/, page llZq. The actnal crashing strengths were plotted, and it 
was fonnd that they conld be reasonably well represented by a right 
line paning through the origin of co-ordinates. The values for this 
average line are shown in next to the last oolnmn of Table IRA. 



ovGoQi^lc 



l\Sa CONCRETE, [chap. IT. 

Theae ezperimeDte seem to prove that the strength of concrete 
Taries as tbe qnaotit; ot cement, provided the voids are filled with 
mortar. The same conclneioa is proved by the data snnunarized 




Portland Cemeni -Barrels per Cubic Yard 

no. 8.— KiLmoR nrmni tbi BTRuraTD or the Cohckiti ihd tbi Ahddht of CncrK 

in Fig. 8, The diagram presents the reaalta of forty-eight experi- 
ments oa 4-inch cabes.* Each point represents two experiments, 
the age of the mortar in one being 7 days and in the other 28 days. 
The points with one circle aronnd them represent the strength of 
broken-stone concrete, and tbe points with two circles gravel con- 
crete. Both the sand and the gravel employed in these experiments 
were very coarse, and conseqaently the amount of cement per cabic 
yard is nnnsnally great. 

155. When mortar is mixed with broken stone, the film ot 
mortar sarronnding each fragment increases tbe volume of the 
resnlting concrete. Table 13c, page ll%b, gives the reanlt of fifteen 
experiments to determine this increase in volume. The mortar was 

* Oandlot'a CemeDts ot Cbaux HydrauUques, pp. 340-U. 



ovGoQi^lc 



AET, 2.] THBOBT OP THB PBOPORTIOHS. 112S 

moderately diy, and the conciete was qaite dry, tnoistnre flushing 
to the surface only after vigorous tamping. The broken stone vu 
No. 10 of Table lOh, page 80, and contained 28 per cent, of roidf 
when rammed. 

Line 4 of Table 13c shows that if the mortar is eqnal to the 
voids, the rolnme of the rammed concrete is 7^ per cent, more than 
the Tolnme of the rammed broken stone alone. Possibly part of 
TABLE IBe. 

Ikcbkabk of Voluue by Uixiho Hortab witk Bbokkb Stone. 



E«r.No. 


Tolums of Hortar in 
t*rm« or the Voids tn 
the Broken Stone.- 


Volume ot Rain Died ■ 


TolcOinlheRMiiined 




70 per cent 


105.0 per cent 


16.8 percent 




aO ■' ■■ 


106.5 " " 


13.3 ■• " 




90 " " 


108. S " " 


9.5 " " 




100 " " 


107.5 '■ " 


7.0 " " 




110 " " 


109.0 •■ •■ 


4.9 " " 




130 " " 


UO.S '• •• 


8.8 " " 




180 " " 


iia.6 " •■ 


1.2 '• '• 


8 


140 '■ " 


114.0 '■ ■' 


0.0 " •■ 



) of Tolnme was due to imperfect mixing, altboagh it 
was believed that the mass was perfectly mixed. The table also 
shows that the voids in this concrete are eqnal to 7 per cent, of its 
volume; in other words, even though the volume of the mortar is 
equal to the volume of the voids, the voids are not filled. Appar- 
ently the voids can be entirely filled with this grade of mortar only 
when the mortar is abont 40 per cent, in excess of the voids. 

The increase in volume in Table ISc may be regarded as the 
maximum, since the mortar was qaite dry and the stone unscreened. 
With moderately wet mortar and the same atone, the increase in 
Tolnme was only abont half that in the table; and with moist 
mortar and atone ranging between 3 inches and 1 inch, there was 
no appreciable increase of volume. With pebbles the increase is 
only about two thirds that with broken stone of the same size. 
With fine gravel (So. 18, page 80) the per cent, of increase was 
considerably greater than in Table 13c; with mortar equal to 150 
per cent, of the voids, it was possible to fill oaly about 5 to 7 per 



ovGoQi^lc 



1130 COHCBSTB. [OHA.P. IT. 

cent, of the Toida. The mortar used in Table 13c vaa 1 Tolnme of 
cement to 2 Tolamea of sand, both measured loose; bat with richer 
mortan the inoraaee in volume was a little less, and with leaner 
mortars a little more. Theee differencee are so sniall that they may 
be disregraded. 

Notice that the voids io Table 13c are for the vet concrete. 
When the concrete has dried oat the voids will be more; since 
ordinarily all the water employed in making the concrete does not 
enter into chemical combination with the cement, and coneeqaently 
when the concrete dries ont the space occupied by the free water ia 
empty. 

156. Kethods of Determining the Proportions. There are two 
methods of fixing the proportions for a concrete; viz.: 1. adjiut 
the proportions so that tbe Toids of the aggregate shall be filled 
with mortar, and the voids of the sand with cement paste; or, %, 
fix tbe proportions witbont reference to the voids in the materials. 
These two methods will be considered in order. 

156a. With Reference to Ike Voidt. To find the correct prc- 
portions for a concrete, first determine the per cent, of voids in the 
rammed aggregate (§ llbd). Next determine the voids in the 
sand. Then use that proportion of cement which will fill the voids 
of the sand with cement paste (see g 126). The amount of mortar 
to be used depends upon the per cent, of voids in the aggregate and 
tbe density desired iu the concrete (see Table 13c, page 113d). 

The det^ls of the method of determining the amount of mortar 
and of cement will be illustrated by the following example. Assume 
the aggregate to be broken stone, unscreened except to remove the 
dust, containing 38 per cent, of voids when rammed (see No. 10, 
Table lOA, page 80). Also assome that a concrete tA maximum 
density ia desired ; and that therefore the mortar should be equal 
to about 140 per coat, of tbe voids (see Table 13c, page llZb). The 
aggregate compaote 5 per cent, in ramming (No. 10, Table lOA), 
and therefore a yard of loose material will equal 0.95 of a yard 
rammed. Adding mortar eqnal to 140 per cent, of tbe voids 
increases the volume to about 114 percent. (Table 13c); and there- 
fore adding the mortar will increase the volume of the ramn^ed 
aggregate to 0.95 X 1.14 = 1.08 cu. yd., which is the volume of 
concrete produced by a yard of loose aggregate. To produce a yard 
of concrete will therefore require 1 -f- 1.08 = 0.93 on. yd. of loose 
broken stone. Since the mortar is to be eqnal to 140 per oent. of 



ovGoQi^lc 



ABT. 2.] HETHODS OF DETZBUININO 'THE PBOPOBTIOIT&. HZd 

the voids, a yard of ooaorete will reqaire 1.40 x 0.38 = 0.S9 co. 
yd. of mortar. ABsnine the rammed eand to contain 37 per oent. 
of Toida (see Table lOjr, page 79i). Therefore to fill the voids of 
the sand vith cement paate will require 3? per cent, as much 
packed cement m loose aand; or in other words, the proportions of 
the mortar sbonld be abont 1 votome paclced cement to 2^ Tolumee 
loose sand. Interpolating from Table 11, page 88, ve aee that to 
produce a yard of this mortar will reqaire abont 2.40 bbl. of Port- 
land cement and 0.79 en, yd. of sand. Gonsequentlj a yard of the 
concrete will require 0.S9 X 3.40 = 0.94 bbl. of Portluid cement, 
and 0.39 X 0.79 = 0.31 en. yd. of aand. The quantities for a 
eobic yard of the rammed concrete are: 0.94 bbl. of packed Fort- 
land cement, 0.31 co. yd. of loose sand, and 0.93 co. yd, of loose 
brokea stone; and since 1 bbl. = 0.13 en. yd., the proportions 
are: 1 volnme ot packed Portland cement, S^ Tolnmea of looee 
sand, and 7^ volnmes of loose broken stone. 

156d. Without Raference to the Voids. tJsaally the proportions 
of a concrete are fixed withoat any reference to the method to be 
employed in measuring the cement, and also without reference to 
the voids in the sand and in the aggregate. The proportiona are 
nanally atated in Tolomee, that of the cement being the unit. For 
example, a concrete is described as being 1 part cement, 2 parts 
eand, and 4 parts broken stone. 

This method is inexact, in the first place, since it does not state 
the degree of compactueas of the cement. If the unit of cement is 
a commercial barrel of packed cement, the resulting concrete will 
be much richer than if the cement were measared loose (see § 126). 
In the second place, this method, in name and naoatly in fact, 
takes no account of the proportion of voids in either the sand or 
the aggregate. If the stone is screened to practically one size, it 
may have 4fi to $0 per cent, of voids when rammed; but if it is 
unscreened except to remove the dust, it may have only 30 per 
cent, of voids (see Table lOh, page 80). 

Ifi6c. To explain the method of testing whether or not the voids 
are filled in a concrete described in the above form, take the com- 
mon proportions: I volume cement, 2 volomes sand, and 4 volumes 
broken stone. If the cement is measured by volumes loose, as is 
usually the case, 1 volume of dry cement will make abont 0.8 of a 
volame of paste. If the sand is the best, it wHl probably have 
about 30 per cent, of voids when rammed (see Table lO^r, page 79t) ; 



ovGoQi^lc 



llZe COITCRETE. [chap. IT. 

and hflDce the 2 Tolames of BBnd Trill contain about 0.6 of a volame 
of Toids. Tbe cement is then S6 per cent, more than eaongb to 
fill the Yoida of the sand. Tbe cement and sand \Tben rammed 
will make 3 -|- (0.8 — 0.6) = 2.2 + volnmea of mortar.* If the 
broken atone is nnscreeiied, it will probably have aboat 30 per cenL 
Toide when rammed (aee Table lOA, page 80); and hence the 4 
Tolnmee of stone will contain 1.2 Tolnmes of voids. The excess of 
mortar is then 2.2 — 1.2 = 1.0 nnits, or 83 per cent, more than 
enongh to fill the voids of the broken stone. Tbe mortar and the 
broken stone will make 4 + (2.2 — 1.2) =: 5.0 + Tolnmes of 
rammed concrete. f 

For the materials assamed, the preceding proportions are very 
uneconomical, since there is 25 per cent, more cement than the 
voids in the sand and 83 per cent, more mortar than the voids in 
the broken stone. The possible saving in cement may be compated 
as follows : 25 per cent, of the cement coald be omitted in making 
the mortar. The mortar woald then be 3 volnmes, of which 0.8 
of a volume, or 40 per cent., is in excess of the voids in the aggre- 
gate. The omission of this snrplas mortar is equivalent to omitting 
0.40 X 0.75 = 30 per cent, of the original cement. The total 
snrplas of cement is then 25 -)- 30 = 55 per cent. It the above 
proportions were intended to give a conoreta of mazimnm density, 
then the mortar employed shoald be about 40 per cent, in excess 
of the void (§ 155). In this case, the surplus mortar wonld be 
(2.0 — 1.40 X 1.2) = 0.32 volumes, or 16 per cent, of the total 
mortar; and the snrpli^s cement in this mortar wonld be (0.75x0.16) 
= 12 per cent. Therefore the total snrplas cement is 25 4- 12 = 
37 per cent. Even in this case the proportions are uneconomical. 

166d. The above example shows bow extravagant the above 
proportions are with tbe beet grades of sand and broken stone. If 
the sand has 37^ per cent, of voids and the broken stone 40 per 
cent., then with the preceding proportions there will be practically 
no surplus cement, and there will be an excess of mortar of aboat 
25 per cent. In other words, with coarse sand and screened stone, 
tbe voids of the sand will be filled with cement paste, and the voids 

• The mortar whan rammed will make from 3 to 4 per cent, mora volume ttian 
the Bum ol tbe sand and the exoasa ot the pftate (aae the laat paragraph ol % 138, 
page 87). 

t The Tolume ot tbe oonuFets will be ellghtl; more than 6.0 units, slnoe some 
sand will remain twtweeii the tragmohts of stone, and thereby Inoreaaa the Tolams 
(see Table 13o, page 113b.) 



ovGoQi^lc 



ABT. 2,] DA.TA. FOB ESTIUATES. 113/ 

of the broken stone will be filled with mortar. However, it ia 
«X06edi&glf uneconomical to nse a very poroQB aggregate and 
attempt to make a very dense concrete. 

The above comparisoDs show how anscientific it is to proportion 
concrete regardless of the condition of the materials to be need. 

166s. Occasionally specifications state the qnality of the mortar 
to be nsed, and require the mortar and the aggregate to be so pro- 
portioned that the mortar shall at least be eqnal to the voids in the 
aggregate. Under this method of procedure, to gnard against lack 
of aniformity in the aggregate, imperfect mixing, and insafQcient 
tamping, it is cnatomary to require more than enough mortar to 
fill the voids, this excess varying from to 50 percent., but usnally 
being from 15 to 25. Apparently 15 per cent, ii frequently nsed 
in Grermany.* 

Xotice that this method is an approxiinatJon to that discussed 
In % 156(1 preceding. 

166/. Data for Eitimatei. Table 13d and Table 13e, pages 
112jr and 112A, give the quantities of cement, sand, and broken stone 
required to make a cubic yard of concrete, for the two methods of 
proportioning described in § 156a and § 156J, respectively. Each 
table gives the quantities for unscreened and also for screened 
broken stone; and Table IZd gives also the quantities of cement 
and gravel required for a cubic yard of concrete. 

The barrel of cement in both tables is the commercial barrel (^ 
packed cement. 

166g. Table ISd is recommended for general use. The first line 
gives a concrete of the maximum density and maximum strength, 
i.e., the quantity of mortar is snflScient to fill the voids (see g 155); 
and the successive lines give concretes of decreasing density and 
strength. The third and subsequent lines give concretes containing 
mortar equal to the voids, the mortar in the third line being 1 tp 3j 
in the fourth 1 to i, etc. 

The qnantilies were computed aa described in g 156a, and were 
afterwards checked by making 6-inch cabes of concrete. While 
the results are only approximate for any particular case, it is 
believed that they represent average conditions with reasonable 
accuracy. 

The quantities in the table are for stone nniform in qnality, and 

* Der FortUnd Oement uod selno Annendnngea im BauireHeD, pp. IM uid US. 



jvGooi^le 



n2g 



[CHAP. IT. 



a ^ 
s ": i 



3 •< 3 

! 1 







s 


s 


S 


s 


d d 




i 


* 


s 

e 


E 


d 


9 3 

d d 




t 


s 


s 


S 


d 


d d 




i 


2 


9 




3 
d 


d d 




i 
i 




s 


s 


s 

d 


S 3 

d d 




a 


£ 


^ 


s 


S 3 




^ 


£ a 




d 


d d 




i 


C 


s s 



jvGooi^le 



ART. 3.] 



IHSBBDIBNTS FOR A TABD 07 CONCBSTB. 



1 

s 

J. 


1 

■ 


1 


'N S 8 


s s s 


s a s 


8» d - 


O o ^ 


« d - 


1 


t* S S 


i S ft 


Si* 






« o o 


1 


iS E S 


$ C 8 


s s s 




o o e 


« a o 


1 


i* ^ S 


sec 


B 8 g 




M » o 


O O O 


I 


1 


»E S S 


S X s 


s s s 






« o o 


1 


^ s s 


■s s s 


$ S £ 


lio d o 


* o o 




1 


3* S S 


S 3 A 


S S S 




o o » 


o o o 


1 

1 

i 

s 
i. 


1 


1 


ix S 3 


S 5 S 


S s 3 


io = =i 


e o o 


o o o 


1 


^ s s 


9 3 S 


S 9 B 




o o o 


d o « 


1 


S» s s 


g g s 


8 S 3 




o o e 


ode 


i 


3« = S 


S S R 


8 S 8 




o o o 


* o o 


1 


i 


is s s 


ess 


s s s 


j=! = » 




o o o 


1 


Is s s 


3 s ai 


8 S S 


go o o 


o » « 


a a ts 


1 


5-2 8 S 


£ ? S 


ass 




» « o 


o o e 


1 


1 




« t- IB 


"■ " s 


1 


" 


- 


' 


i 


- 


- 


- 




1^ 








f » » 



jvGooi^le 



iiai coucsBTB. [chap. IV, 

for concrete thoroaghlj and Tigoronaly r&mmed; aad if it is desired 
certainl; to secare the deaseet coacrebe, it might he wise to iDcreaae 
somewhat the cement and sand given in the first line of Table 13d, 
The per cent, of increase should vaiy with the circnniBtances of the 
case iu hand (see § 156e). 

The proportions of the concretes can be determined b; remem- 
bering that a barrel of cement is eqnaltoO.lScn. yd. For example, 
for ansoreened broken stone and Portland cement, the 0.94 bbl. of 
cement iseqaal to 0.1^ ca. jd.; and the proportions are: 1 volame 
of packed cement, %.5 volames of loose sand, and 7.5 volames of 
loose nnscreened broken stone. It it be assamed that a barrel of 
packed cement will make 1.25 barrels when measured loose (see 
g 126), the above proportions become: 1 volame loose cement, 
2,0 Tolnmes loose sand, and 6.0 rolnmes loose nnsoreened broken 
stone. 

1S6A. Table 13e is given for use in determining the ingredients 
required for a concrete designed in the ordinary way — see § 156d. 
The quantities were computed sabHtantially as illustrated in 156<;. 
This table is not as accnrate as Table 13d, and besides many of the 
proportions are uneconomical (see the second paragraph of § 156c), 

156i. Froportione from Practice. While a statement of the 
proportions ased in practice may be of interest, it can not be of any 
great value since it is impracticable, if not impossible, to describe 
fully the circumstances and limitations under which the work was 
done. Farther the specifications and records from which such data 
must be drawn are frequently very indefinite. It is believed that 
the following examples are as accnrate as it is possible or practicable 
to make them, and also that they are representative of the best 
American practice. 

For foundations for pavements: 1 volume otnaiureU cement, 
2 volumes of sand, and i or 5, and occasionally 6, volnmes of broken 
stone; or 1 volume of Portland cement, 3 volumes of sand, and 6 
or 7 volumes of broken stone. Occasionally gravel is specified, and 
more rarely gravel and broken stone mixed. 

For foundations and minor railroad work: 1 volume of natnral 
cement, 2 volnmes of sand, and 2 to 6, nsnally i or 5, parts of 
broken stone. See also pages 532 and 535. 

For important bridge and tnnnel work: 1 part of Portland 
cement, 3 parts of sand, and 4 or 5 parta of broken stone. 



ovGoQi^lc 



AKT. 3.] PEOPOEnONS FHOM FRACTICB. 112/ 

For ateel'grill^e foundationB : 1 part Portland cement, 1 part 
aand, and 2 parts broken stoae. 

For the Melan steel and coocrete constmction the usual pro- 
portions are: 1 Tolnme of Portland cement, 2j Tolnmes of aand, 
fi Tolnmea of broken stone. 

For the retaining valU o;i the Chicago Sanitary Canal : 1 part 
natural cement, 1^ parta sand, and 4 parts unscreened limestone. 

For the dams, locks, etc., on the Illinois and HissiBsippi Canal: 
1 volume of loose Portland cement, 8 Tolnmes of gravel and broken 
■tone; or 1 volame loose natural cement and 5 volnmes gravel and 
broken stone. 

For the Poe Lock of the St. Mary's Fall Cuial: 1 part nataral 
cement, 1^ parts of sand, and 4 parts of sandstone broken to pass 
a Si'inch ring and not a f-inch screen. The broken stone had 46 
per cent, voidaloose and 38 when rammed. 

In harbor improTements the proportions of concrete range from 
the richest (used to resist the violent action' of iravea and ice) to 
the very lowest (used for filling in cribwork). At Buffalo, N. Y., 
an extensive breakwater built in 1890 by the TI. 8. A. engineers, 
consistbd of concrete blocks on the faces and a backing of coucrete 
deposited in place. Portland was used for the blocks and natural 
for the backing, the proportions being: 1 volume cement, 3 aand, 
and 8^ of broken stone and pebbles mixed in equal parts. 

For the concrete blocks used in constructing the Mississippi 
Jetties the proportions were: 1 part Portland cement, 1 part sand, 
1 part gravel, and 5 parts broken stone. 

For incidental information concerning proportions need in prac- 
tice, see Cost of Concrete, § 158a, page llSi;. 

156y. Watxs Requiees. There is a considerable diversity of 
opinion among engineers as to tlie amount of water to be used in 
makiog concrete. According to one extreme, the amount of water 
should be snch that the concrete will quake when tamped ; or in 
other words, it should have the consistency of liver or jelly. 
Aocording to the other extreme, the concrete sbonld be mixed so 
dry that when thoroughly tamped moisture jnst finshes to the sur- 
face. The advocates of wet mixture claim that it makes the 
stronger and more dense concrete; while the advocates of dry mix- 
ture claim the qtposite. The difference in practice is not as great 
aa in theory; the apparent difference is chiefly due to difference* 
in condition. 



ovGoQi^lc 



llSJt CONCEKTB. [chap. IV. 

It is nnqnefltionably trn« that diy miitnrea of neat oement, and 
also of cement and sand, are stronger than wet miztDies, prorided 
the amonat of water is enfficient for the crystallization of the 
cement. It is also certainly true that in even a dry mortar or con- 
crete, the water is considerably in excess of that necessary for the 
crystallization of the cement, this excess increasing with the amonnt 
of sand and aggregate. Of conrse the excess water is an element 
of weakness. Bat the amonnt of water to be nsed in making con- 
crete is nsnally a qnestion of expediency and cost, and not a qnee- 
tion of the greatest attainable strength regardless of expense. 

1. Dry mistares set more qnickly and gain strength more 
rapidly than wet ones; and therefore if quick set and early strength 
are desired, dry concrete shoald be preferred. 2. Wet concrete 
contains a great number of invisible pores, while dry concrete is 
liable to contaio a considerable nnmber of visible voids; and for' 
this reason there is liability that wet concrete will be pronounced 
the more dense, even though both have the same density. 3. Wet 
concrete Is more easily mixed ; and therefore if the concrete is mixed 
by band and the supervision is insufficient or the labor is careless, 
or if the machine by which it is mixed is inefficient, wet mixtures 
are to be preferred. 4. Wet mixtures can be compacted into place 
with less eSort than dry; but on the other hand tbe excess of water 
makes the mass more porous than though tbe concrete had been 
mixed dry and thoroughly compacted by ramming. Dry con- 
crete must be compacted by ramming, or it will be weak and 
porous; therefore if the concrete can not be rammed into place, it 
should be mixed wet and then the weight of the stones will bury 
themselves in the mortar, and the mortar will flow into the voids. 
5. A rich concrete can be compacted much easier than a lean one, 
owing to the lubricating effect of the mortar; vid hence rich con- 
cretes can be mixed dryer than lean ones. The quaking of concrete 
frequently is due more to an excess of mortar than to an excess of 
water. 6. Lean concretes should be mixed dry, since if wet the 
cement will find its way to the bottom of the layer and destroy the 
uniformity of the mixture. 7. Machine-made concrete may be 
mixed dryer than hand*made, owing to the more thorough incor- 
poration of the ingredients. 8. Gravel oonorete can be more easily 
compacted than broken stone, and hence may be mixed dryer. 
Cement and sand alone is more easily compacted than when mixed 
with ooaiur material, particularly broken stone; and therefore 



ovGoQi^lc 



ABT. 2.] WATKB KBQUIBED. IIZI 

mortar to be depoaited in moaa should be mixed diyer than concrete. 
9. In mixing dry b; band there is a tendency for the cement to ball 
np, or form nodnles of neat cement, vhile in mixing wet this does 
not occur. 10. If wet oonorete is deposited in a wood form, there 
is liability of the water exuding between the planking and carrying 
away part of the cement and thns weakening the face — which should 
be the strongest part of the mass. 

The conclnsion is that sometimes wet concrete must be used 
regardless of any qnestion of strength and cost; while with thorongh 
mixiog and vigoroas ramming, dry concrete is strongest bnt also 
most ezpeosiTe to mix and lay. 

IMk. The following experiments are the only ones of any im- 
portance made to determine the relative strength of wet and di; 
concretes. The mean crashing strength of fonr handred and 
ninety-six 1-foot cnbes * made with mortar as " dry as damp earth '* 
was 11 per cent, stronger than cnbes made with mortar of the 
"ordinary consistency nsed by the average mason," and 13 per 
cent, stronger than cnbes that " qnaked like liver nnder moderate 
ramming." The oabes were made of five brands of Portland 
cement, with broken stone and five proportions of sand varying 
from 1 to 1 to 1 to 6; half the oabes had a little more mortar than 
enoDgh to fill the voids, while the other half had only about 80 per 
cent, as mncb mortar as voids. One quarter of the cnbes were 
stored in water, one quarter in a cellar, one quarter under a wet 
cloth, and one quarter in the open air; and all were broken when 
approximately Z years old. The diSerenoe in the amonnt of mortar 
made no appreciable diSerenoe in the strength. 

The mean of twelve cubes of dry concrete was 61 per cent, 
stronger than corresponding cubes of quaking concrete, f 

A few minor experiments have been found confirming the above, 
and none have been found that contradict them. 

1661. I'he amount of water required to prodnoe any puiicnlar 
plasticity varies so greatly with the proportions of the ingredients, 
the kind and fineness of the cement, the dampness of the sand, the 
kind of aggregate, etc., that it is scarcely possible to give any valu- 
able general data. The water varies from 10 to 40 pounds per cubic 
foot of concrete. The only general rule that can be given is that for 

• Geo. W. Batter, 1b Beport oC the Mew York BUto Engtneer lor 1807, pp. ttS-IM, 
pKrtlcalarir Table ^, page 89S. 
t Feret, Xnginmring Uttot, voL zxrll, p. Sll, 



ovGoQi^lc 



llZm. OOUCSFtTE. [chap. IV. 

dry concrate the aggregate ahonld be wet but hare no free water la 
the heap; and that the mortar shoold be damp enongh to show 
water only when it is thorongfaly rammed, or so that water will 
flush to the surface when it is tightly squeezed for a coDsiderahle 
time in the hand. 

In the experiments referred to in the first paragraph of the 
preceding section, the average quantity of water for the different 
grades of dry mortar was 19. S Iba. per en. ft., and for the plastic 
21.1, and for the wet HZ. 5, the sand being reasonably dry. 

166m. Xiznre.— The value of the concrete depends greatly 
upon the thoronghsess of the mixing. Every grain of sand and 
every frt^^ent of a^regate should have cement adhering to every 
point of its surface. Thorough mixing should cause the cement 
not only to adhere to all the surfaces, hut should force it into 
intimate contact at every point. It is possible to increase the 
strength of really good concrete 100 per cent, by prolonged tritura- 
tion and rubbing together of its constituents. The longer and more 
thorough the mixing the better, provided the time does not trench 
upon the time of set or the working does not break and pulverize 
the angles of the stone. Uniformity of the mixtare ia as important 
as intimacy of contact between the ingredients. Of coarse tborongh- 
ness of mixing adds to the cost, and it may be wiser to nae more 
cement, or more concrete, and less labor. 

Concrete may he mixed by hand or by machinery. The latter 
is the better; since the work is more quickly and more thoroughly 
done, and since ordinarily the ingredients are brought into more 
intimate contact. Machine mixing is frequently specified. If any 
considerable quantity is required, machine mixing is the cheaper, 
ordinarily costing oi^y about half as much ae hand mixing. 

Ifi6n. Hand Kxing. The sand and aggregate are usually 
measured in wheelbarrows, the quantity being adjusted for a bag 
or barrel of cement. The dry cement and sand are mixed as 
described in the first paragraph of g 124 (page 85), which see. 
The proper quantity of water is then added, preferably with a 
spray to secure greater uniformity and prevent the washing away 
of the cement. The mass should he agmn turned until it is of 
uniform consistency. The broken stone, having previously been 
sprinkled but having no free water in the heap, is then added. The 
whole is then tamed antil every fragment is oorered with cement. 
Specifioationa nsoally require ooncret* to be turned at least foar 



ovGoQi^lc 



AHT. 2.] HIXIMG. UZn 

times, and freqnently six. The concrete appe&rs wetter eoob time 
it is tnrned, and should appear too dry antil the very lost. 

If gravel is used iostead of broken stone, the mixing is done as 
described for cement and sand. 

Ififto. Haohine Kizin^. A variety of concrete-mixing maobines 
are in nae. Some forms are intermittent and some continuous in 
their Action. Some of the latter automatically measnre the in- 
gredients, A simple variety of the former oonsists of a cubical box 
revolved slowly about a diagonal axis. The dry materials are 
inserted through a door, and the water is admitted through the 
axis daring the process of mixing, Sight or ten revolutions are 
sometimes specified; but eighteen or twenty are more frequently 
specified and give a mnch better concrete. Sometimes an inclined 
cylinder or long box revolving abont the long axis is employed. 
Another form consiata of a vertical box having a series of inclined 
shelves projecting alternately from opposite sides, the materials 
being thrown in at the top and becoming mixed by falling succes- 
sively from the inclined shelves. A modification of thta form sub- 
stitutes rods for the shelves, the mixing being accomplisbed by the 
ingredients in their descent striking the rods. Still another type 
form consists of a spiral conveyor or a bladed screw-shaft revolving 
in a trough in which the materials are thrown. All of these forms, 
and also modifications of them, are to be had on the market.. 

166p. Latdto. Aftermixing, the concrete is conveyed in wheel- 
barrows or in buckets swung from a crane, deposited in layers 6 to 
8 inches thick, and compacted by ramming. In dumping, the mass 
sbonld not be allowed to fall from any considerable height, as doing 
80 separates the ingredients. If in handling, the larger fragments 
become separated, they should be returned and be worked into the 
mass with the edge' of a shovel. 

The rammer aaually employed consists of a block of iron having 
a face 6 to 8 inches square and weighing anything up to 30 or 40 
pounds. The face of the rammer is sometimes corrugated, to keep 
the surface of the layer rough and thus aSord a better bond with 
the next, and also to transfer the compacting effect of the blow to 
the bottom of the layer. The tamping should be vigorous enoogh 
to thoroaghly compact the mass; bnt too severe or too long-con- 
tinned ponnding injarea the strength of the concrete by forcing the 
broken stone to the bottom of the layer, or by disturbing the 
incipient set of the cement. 



ovGoQi^lc 



112o COKCKETE. [CHAP. IT, 

When one layer is laid on another already portiallf eet, the 
entire sarface of the latter should be thoroughly wet; bnt water 
Bhonld not stand in paddlee. In case the first layer is fully set, it 
is wise to sweep the sarface with neat oement paste to make sure 
that the two layers adhere firmly. If the sand or gravel contains 
any appreciable clay and the concrete is mixed vet, clay is liable to 
be fiashed to the snrface and prevent the adherence of the next 
layer; therefore under these conditions particnlar care shoald be 
given to secure a good union between the layers. After the con- 
crete is in place it should be protected from the snn, and not be 
disturbed by walking apon it until fnlly set: this limit shonld be 
at least 13 hours and is frequently specified as 4 or 5 days. 

169q. Bepoiiting Concrete under Water. In laying ooocrete 
nnder water, an essential requisite is that the materials shall not 
fall from any height, but be deposited in the allotted place in a 
compact mass; otherwise the cement will be separated from the 
other ingredients and tbe strength of the work be serioasly im- 
paired. If the concrete is allowed to fall through the water, its 
ingredients will be deposited in a series, the heaviest — the atone — 
at the bottom and the lightest — the oement — at the top, a fall of 
even a few feet causing an appreciable separation. Of coarse con- 
crete should not be used in running water, as the cement would be 
washed out. 

A common method of depoeiting concrete under water is to 
place it in a V-shaped box of wood or plate-iron, which is lowered 
to the bottem by a crane. The box is so constructed that, on 
reaching the bottem, a pin may be drawn out by a string reaching 
to the surface, thus permitting one of the sloping sides to swing 
open and allowing the concrete to fall out. The box is then raised 
to be refilled. It usually has a lid. Concrete nnder water should 
not be rammed; bnt, if necessary, may be leveled by a rake or 
other suitable tool immediately after being deposited. 

A long box or tnbe, called a trhnie, is also sometimes used. It 
oonsists of a tnbe open at top and bottom, boilt in detachable sec- 
tions BO that.the length may be adjusted to the depth of water. 
The tnbe is suspended from a crane, or movable frame running on a 
track, by which it is moved aboat as the work progresses. The 
upper end is hopper-shaped, and is kept above the water; the lower 
end leats against the bottom. Tbe trSmie is filled in the beginning 
by placing the lower end in a box with a mnvuble bottom, filling 



ovGoQi^lc 



ABT. 2 ] BTBENGTH, 112^ 

the tabe, levering all to the bottom, and then detaching the 
bottom of the box. The tabe is kept fall of boncrete, as the mass 
issoee from the bottom more is thrown in at the top. 

Concrete has also been BnGcessfnll; deposited ander water hj 
enoloeing it in paper bags, and levering or sliding them down a 
chate into place. The bags get wet and the pressiin of the con- 
crete soon bursts them, thns alloviag the concrete to nnite into a 
solid mass. Concrete is also sometimes deposited nnder vater b; 
enclosing it in open-cloth bags, the cement oozing throngh the 
meshes snflQcientl; to nnite the whole into a single mass. 

When concrete is deposited in vater, a pnlpy gelatinous flnid i» 
washed from the cement and rises to the surface. This oaosee tho 
water to assume a milky hue; hence the term Untance, which 
French engineers apply to this substance. It is more abundant ia 
salt water than in fresh water. It sets totj slovlj, and sometimes 
scarcely at all, and its interposition between the layers of concrete 
forma strata of separation. The proportion of laitance is greatly 
diminished by using large immerBing boxes, or a tr6m!e, or paper 
or cloth bags. 

167. Stbevoth. The strength of concrete depends upon the 
kind and amoant of cement, and upon the kind, size, and strength 
of the ballast. Mortar adheres to broken stone better than to 
pebbles, and therefore concrete containing the former is stronger 
. than that containing the latter (see g 151). If the sizes of the indi- 
Tidoal pieces of the ballast are so adjusted that the smaller fit into 
the interstices of the larger, successively, then the cementing 
material will act to the best advantage and consequently the con- 
crete wUl be stronger. Bamming the concrete after it is in place 
brings the pieces of aggregate into closer contact, and conseqneutly 
makes it stronger. The strength of concrete also depends somewhat 
cpon the strength of the ballast, but chiefly upon the adhesion o£ 
the cement to the ballast. 

There are comparatively few experiments upon the strength) 
of concrete in which the data was complete enough to make the- 
reaalts of any considerable value. 

167a. Compressive Strength. In a series of experiments made 
by Geo. W. Bafter* to determine the crashing strength of concrete, 
three varieties of Portland cement were used, all of which wer& 

* Baport ot tlie Kew ToA state EnglnMr. IBST, pp. STS-tfO. 



ovGoQi^lc 



1127 



[chap. it. 



eqnal to the maximnm both neat and with sand in Table 10, pag« 
78a. The sand was pore, clean, sharp silica, containing 33 per 
cent, of voids. The aggregate was sandstone broken to pass a 
S-inch ring, having 37 per cent, voids when tamped. In half the 
blocks the mortar was a little more than enongh to fill the Toida; 
and in the other half the mortar was equal to about 80 per cent, of 
the voids. The mortar was mixed as " irj as damp earth." 

The test specimens were 1-foot oabee, and were stored nnder 
water for fonr months and then bnried in sand. The age when 
tested ranged from 550 to 650 days, the average being abont 600. 
The cnbes were crnsbed on the U. S. Watertown Arsenal testing- 
machine. The means are shown in Table 13/. The individoal 
molts agreed well among tbemflelves. 
TABLE 18/. 
Orubbirc Strehotb op PoB-n.uiD ConcKEXX. 
Toida of broken (tone praclfcally filled wlih morui^-we the lest. 

Age wben tsated eOO dajs. 



Bw.Ho. 




or HOBTAR. 


Nn. 


Cmmanm 


«.»™. 


Cement. 


Sud. 


''■?^S' 


)l».per«i.lD. 


ton, per 1. ft. 






1 




4.467 


833 






3 




8,781 


«ia 






8 




2,558 


181 






4 




3,018 


HI 






B 




1.706 


129 






A 




1.865 


«8 



The cubes snmmarized in Table 13/ were stored nnder water. 
Companion blocks stored in a cool cellar gave 83 per cent, as much 
strength; those fally exposed to the weather, 81 per cent.; and 
those covered with barlap and wetted several timee a daj for abont 
three months and afterwards exposed to the weather, 80 per cent. 

The cubes of Table 13/ were mixed as " dry as damp earth." 
Companion blocks of which the mortar was mixed to the " ordinary 
consistency nsed by the average mason," gave 90 per cent, as mnoh 
strength ; and those mixed to " qnake like liver nnder moderate 
ramming," 88 per cent. 



ovGoQi^lc 



ART. 2.] COUFBESSITE STRKKOTH. 112r 

167b. Table 13^ ebovs the results of a aeries of experimenti 
made by A. W. Dow, Inspector of Asphalt and Cement, Washing* 
ton, D. C* 

TABLE 1^. 
CKUSRins Strbhoth or Coscretr ik Poukim fbb BquAUR Ihoh. 





''™!r?^.'?.s:sssr' 


,..S;s.„. 


AOB or Cum wan Bbour. 


^; 


Hortu-. 


AatrvpiH. la 
Sliax from 
«J*" lo A". 


± 


Fn 

Cent. 

mi«d 

wUh 
Mortar 


DV* 


4t 

DTI 


.U 


Ho*. 


Tmt, 




Cemwt 


8ud. 


Broken 


QnTsl. 


1 




a 

3 

a 
a 
a 

3 

a 

3 
3 

a 
a 
a 


e 
e" 

6t 

s 

4 

6 
«• 

et 

8 
4 


e 

8 
3 
Port 

e 

8 
3 


iimlCe 
4B.S 
49.7 
89,6 
29.8 
8ft.5 
87.8 

u<iCe 
4S.8 
4B.7 
S9.B 
39.8 
80. B 
87.8 


nent. 
. 88.9 
88,9 
96.a 
120.1 
107.0 

100. e 

meot. 
68.0 
88.0 
.06.2 
130.1 
107.0 
100.8 


328 


689 


876 
096 


T06 


OlS 










4 

B 


87 
108 


421 
804 


861 
693 


844 
683 


788 
841 


7 


908 


1,700 


3.360 
1,680 

3.680 


2,610 

i.sai 

1.84(1 
8.O70 


8,060 


D 
10 
11 
13 


«04 
960 


1.680 
1.860 


3.700 

a.8ao 

3.760 
3.840 



tTbrMtourttwordliuuTitoiM, 01 



I fourtb gnnollihlc 



The Btrength of the cement is shown in Table 13A. Notioe that 
the Portland cement did not gain strength proportionally as fast u 
the natural cement; for example, the Portland mortar in line 4 is 
two aod two-thirds times as strong as the preceding natnral-oement 
mortar, while that in line 13 is not qaite as strong as the natnral- 
oement mortar immediately preceding. 



ovGoQi^lc 



[chap, it. 



TxaanM STBaHOTH 



TABLE 18%. 

or Cbhkht dbed ni Taslb 18;. 



B*v.No. 


iMWEnlMTBD. 


PunaSTunuw 






''°*^«»T. 


VMbXM. 


POHlMd. 




ld.y. 

Ijow. 




M 

180 


441 

889 








IS8 








827 










414 










486 




la 









Tbo fineneaa of the Band WM asfoUows:* 'S" 6'-* S" 10"20" 
40' 60» 80" 100" and contained 44.1 per cent, of Toids. With 
the nataral cement the vater naed was 0.317 on. ft. (20 Iba.) per 
en. ft. of rammed concrete, and with Portland cement 0.^ on. ft. 
(12 lbs.) — in both cases inclndtng the moigtare in the aand.f 

The broken stone vaa gneiss broken to pass a S^-inch ring, none 
pasaing a No. 10 sieve, the voids for each partionlar concrete being 
as stated in Table 13^. The gravel was clean qnartz passing a 
l^-inoh ring and only 3 per cent, passing a No. 10 ring, and had 
39 per cent, of voids. The per cent, of voids in the aggregate filled 
with mortar is stated in Table 13^. Each resnlt in the table is the 
mean of two cabea, except those for one year, which are the mean 
of five. Owing to the friction of the press with which the tests 
were made, the reenlts are 3 to 8 per cent, too high. 

*167c. Table 13i shows the relative strength of rich and lean 

*For azpl&nftUoa oE the nomanolature, see the seooDd paragraph off Hit. 
t The Hud oontklned iA per cent, of water, whlah inareased the volama of Uie 
Mnd ud made Uie mortar eUghtlT rloher than aa stated. 



ovGoQi^lc 



r.2.] 



COKPBBBSITE STfiENOTH. 



TABLS IK 
Rblatitb Btbckoth op Rich and Lsak Cohoketm. 







o™.„b™™. 


^No. 


OWMnt. 


Bud. 




ODdWeak. 


VourW««h«L 




8W110. 


^r 


B«UlJ*& 


^■r 


BelMln. 








Portbmd Mod-oement 








1 


H 


S 


413 


0.77 


400 


0.68 








4 


449 


0.S8 


«79 


0.03 








** 


SS6 


1.00 


741 


1.00 




1 


s 


4 


81S 


0.«1 


441 


0.60 











87S 


0.68 


477 


0.64 








6 . 


031 


1.00 


<89 


1.00 




1 


8 


S 


144 


0.6* 


274 


0.85 








e 


110 


0.C3 


182 


0.07 








7 


210 


1.00 


823 


1.00 








EngUib Portluid cement 






10 


1 


9 


S 


494 


0.60 


S66 


0.81 


11 






8 


Sll 


0.76 


065 


0.80 


12 






4 


810 


1.00 


818 


0.88 


18 






S 


S81 


0.71 


680 


0.07 


U 






« 


600 


0.61 


8S8 


1.00 


» 


1 


8 


8 


888 




.300 


0.68 








4 
B 

e 

Gwnuin 






868 
S8« 
86? 






















PortUnd cenwnt 
























B 
8 






708 
738 





















jvGooi^le 



112h concrete. [chap, it, 

concretes.* The water waa equal to 20 per cent, of the weight of 
the cement and the sand. The teat specimens for the Portland 
saud-cement were 9 inches square and 13 inches high, and for the 
remainder 12-inch cubes. All were crashed between sheets of rubber 
(see § 12, page 9). Each Talae in the table is the resnlt for a single 
cube. Table 13i is Talnable chiefly as shoving the relative strength 
of rich and lean concretes. The table shows that a moderately lean 
concrete ia stronger than a very rich one, which is in accordance 
with the conclnsion from Table 13a, page 110, that a concrete is 
stronger than the mortar alone. Table 13i also shows that the 
strength of the concrete increases with the richness of the mortar, 
which agrees with Table 13d, page 111, and Fig. 8, page 112a. 

167d. For data on the crnshiDg strength of grairil concrete, see 
Table 13a, page 110. 

For data on the cmahing strength of gravel and broken-stone 
oomoretes approximately 17 days old, see Fig. 8, pape 1 I2a. 

Ibte. The strength of concrete made of coke does not increase 
with age owing to the soft and friable nature of the aggregate. 
Apparently the maximum strength* of 1 volame loose cement, 
3 Tolnmes sand, and 6 volumes crushed coke is abont 600 to 700 lbs. 
persq. in. with Portland, and about 300 to 360 with natural cement. 

107/. Transverse Strength. Table 13^, pagellSv, is a summary 
of 191 tests OQ concrete bars 30 inches long and 1 inches sqaare.f 
The cement stood 49? lbs. per sq. in. neat at 7 days, and 309 lbs. 
with 3 parts sand at i weeks. In most of the bars the mortar waa 
made of pulverized sandstODC, altboagh in some cases river and pit 
sands were used. The aggregate was generally broken sandstone, 
but gravel and broken whtnatoue were also used. " In each case 
the voids io the ' sand ' were filled with cement, and those in the 
aggregate with mortar." 

The results are tabulated in the order of the ratio of the cement 
to the total sand and aggregate. Kotioe that the resnlts in the last 
line are proportionally higher than those in the remainder of the 
table. This difference is probably due to the fact that the apeci- 
mena for the first four liuea were made with natural sand and 
atone, while in those for the last line only crashed sandstone was 
nsed for both the sand and the aggregate. 

* W. B. Andaraaii, Btndent Can. 8oo. C. E., In Truis. Cu. Boo. 0. B.. voL xUL, 
Portl. 

t A. F. Bnioa, In Proa, ot Inst, ot O. E. (LoDdon), toL oxlll, pp. Sll-M. 



ovGoQi^lc 



ABT. %.] TBAKSTBBSB STRENOTH. 



TiBLB 18^. 
lIoDin.TJ8 or RuFTUSK ov PoBTLAND Co\cRKTB Barb, Pomme pkk 







A.. ,.w.,„ .„».«. 


r- 


0^. 


s... 


^ 


■ 


' 


• 


a 


,. 


a 


" 






2 




03 


145 


SIS 


266 


801 


303 


820 






84 




87 


m 


166 


104 


268 


280 


200 






S 






8S 


13D 


176 


101 


3U 


214 






3* 






81 


ISO 


IS6 


m 


100 


210 






8 




87 


118 


1S4 


187 


sie 


248 


2S3 



157^. In connection with the constniction of the Poe Lock of 
the St. Mary's Falls Canal * a series of one hundred and forty-Mf en 
oonorete beams 10 inches square were tested. The experiments 
were very carefully conducted, bnt there were bo many rariables 
that it is impossible to draw any general conclnsions therefrom. 
The beams made with Portland cement were tested when about 19 
months old and those with natural cement when about 13 months. 

157A. Weisht of Cohohztx. The weight of concrete varies 
with the materiulH and the proportions, and with the amount of 
Tsmming. Tho weight varies from 130 to 160 Ibe. per ou. ft., bnt 
is nsuaily from 140 to 160. The differeoce in weight of the con- 
crete due to the aggregate and to the ramming is greater than that 
dae to the difference in weight between Portland and natural 
cement. The maximum difference between Portland and natnral 
concrete, due to the greater weight of Portland cement, is 4 or 
5 lbs. per cubic foot. Concrete made of blaat-fnmaoe slag weighs 
from 110 to 130 lbs. per cubic foot; and that made of coke from 
80 to 90 lbs. per en. ft. 

168a. Cost or Cohcbste. The cost of concrete varies greatly 
with the materials, the proportions, the cost of material and labor, 
etc. 

The following is the analyeis of the compoeition and cost of the 
concrete employed for the foundations of tbe sea-wall at Lovell's 
Island, Boston Harbor: f 

* Report ol Ohlst or Engineen, V. B. A., 1896. Part 4, pp. 393a-SI. 

t Oompiled trora (MUmore's Limea, HydmaUa CemeiiU uid Hort&re, p. 347. 



ovGoQi^lc 



2tt ooncebh. , [oHAP. it. 

Cement, O.SS bbl 0.13 cu. yd. ®91M = 913a 

Band 0.35 cu. jd. ® TO 17 

Gimvel O.WciLyd. ® 27 U 



Total 



Labor, maklog mortar O.OSdajr* ® 1 30 = 08 

L*bor, mmklngcoDcraW. 0.11 dmyl ® IM 18 

Labor, trangporllug conorelA 0.08 dajl @ 1 90 OS 

Labor, packing oodgtMa 0.08 daya % 120 01 

mallabor 0.28 daya 88 

Toola, Implemanta, etc 11 

3t>lal cm( 1 «u. yd.cfeonortU.in place fS 11 

Th« pTDportiona for this concrete were 1 cement, 3 nnd, and 
i gnvel. It was niiuinalljr cheap, owing parti; to th^ use of 
pebbles inetead of broken stone. If the latter had been naed, it 
would hare ooet probably 4 to 6 times as mnoh as the gravel. The 
amonat of labor required was also nnnsnall; amall, this item idone 
being frequently 6 to 8 times as much as in this case. 

The following ia the aQalysis * of the coat of nearly 10,000 yards 
of concrete as laid for the fonndstions of a blast-fnraaoe plant near 
Troy, N. Y., in 1886. The conditions were unusually farorabla 
for oheap work. The concrete consisted of 1 Tolnme of packed 
cement to 7 of sand, gravel, and broken stone. 

Oemeat, 1.28 bbl O.I8cu.7d. ® |1 00 = <! 28 

Sand 0.10 " ® 80= 08 

Gravel O.M " « 80= 11 

BnAeaatone 0.74 " @ 141= 104 

Total mat«rUiU 1.88 " = (3 41 

Labor, handltng cement 0.03 dajr 

" unloading stone 0.14 " 

" mixing 0.86 " 

" auperlotendence 0.01 " 

Totallabor 1.02 " = I 0» 

roMeiMto/aouNifanliifwnergCa, injiau =$8 59 



* Tnna. Am. Boo. ol 0. E., toL xr. p. 8TS. 



® 


100 = 


« 


100 = 


« 


100 = 


® 


»01 = 



jvGooi^le 



ART. 2.] . COBT OF COXCBETE. 112x 

Th« followiog IB the cost of the concrete used in the constrno- 
tion of HiUnd Avenue reservoir, Pittsburg, Penn.* The stone wm 
hroken so as to pass through a ii^-incfa ring. The mortar was 
1 part Roaendale natural cement to 2 parta Band. The concrete 
Tras 1 part mortar to 2^ of atone. The concrete wae mixed by hand. 
Common laborers received tl.26 per day, and foremen t2.60. Tbs 
contract price, was t6.00 per yard. 

Quairytog alone ^OiS 

TraDsponlng stone SO 

Breaking stoae SO 

Cement® 9L8S perbbl 1 80 

Ssod, cost of digging 10 

Water 00 

lAbor, mlKlugand laylog 10 

IncideutaU 00 

Total eott ptr eiibie pard, inplac* (4 00 

The following is the cost of concrete in the foundations of an 
electric power-house at Pittsburg, Pa., in ISM.f The proportions 
were 1 volume of pocked cement, 3 volumes of sand, and 5 volumes 
of broken atone. The cost of labor was abnormally high. The day 
was ten working hours. 

Portland cement l.SSbbl O.lTcu.yd. ®t3.(IO 98.88 

8«Dd 0.00 " ® 1.80 0.«S 

Broken atone 0.90 " ® 1,30 1.13 

Lftbor O.ftldsr ® 1.70 1.09 

SuperlDteodence 0.07 " (Q 8.00 0.21 

Total cMl per ev.j/d., inplaea 10 00 

The following is the cost of constructing the concrete letaia- 
ing wall on the Chicago Sanitary Canal.]; Tlie average height of 
the wall was 10 ft. iu Sec. 11, and 33 ft. in Sec. 15. The thickness 
on top was 6 ft., and at the bottom it waa equal to half the height* 
The atone was taken from the adjacent canal excavstiou. The body 
of the wall was made with natural cement, bat the coping and 
facing, each 3 inches thick, were made with Portland cement. 

* Emlle Low In SnginMrijig NeiB», vol. illl. p. fl, S3. 

t E. T. Ohibu in The Polyleehaie, Benaselaer FolTteobnla Inatltate, toL, vlL 
p.l«E. 

X lonr. WMt. Soo. ta Bng^, tOL lU. pp. 1810.83. 



ovGoQi^lc 



112y CONCBBTB, [chap. IT. 

The proportions wen 1 Tolnme of cement, H Tolnmes of sand, and 
4 Tolnmee of nnacreened limestone. The oost of plant employed in 
Sec. 11 was t9,600, and in Sec. 15 was 125,420. The contract 
price for the ooncrete in Sec 14 was 12.74, and in Sec. 15 (3.40 
per OB. jA. 

bbor.gmeral 90.078 * (0.083 

onthewaU 108 -IIU 

mlzliig concrete 131 .2S0 

pUdng Hud remorlDg forma ISO .143 

tnnqxtrtlDg matori&li 143 .081 

quaiTTlaf stooe. SOS .370 

cnuhing none. 078 .138 

7b<al fir tabor $0.I»7B $1,074 

MalerUl, cement, natural ®90.6SperbU. 0.868 .930 

■■ PorlUiidig)»3.35 " " .800 .180 

aand ® fl.8apeTcn.7d. .460 .476 

ntal for maUrialt fl.6SS |1.086 

Uftcblnery, cost of opentiDg 407 .06? 

Total oott parol, yd. (8.010 (8.337 

For additional data concerning the cost of ooncrete, see 
§g 233-34, page 157. 

166b. The following items relate only to the labor of making 
ooncrete. 

Table 13i gives the details cf the cost per cnbic yard of the 
labor reqaired in mixing and laying coacrete for the Buffalo, ^. Y., 
breakwater, constmcted in 1867-89. The data were commnaicated 
by Capt. F. A. Mahan, Corpa of Engineers, U. S. A., who had 
charge of the work. The total amount of concrete laid was 14,587 
on. yds. The conditions nnder which the work was done yaried 
considerably from year to year.* 

Table 13m gives the details of the labor required in mixing and 
laying concrete in the oonstrnotion of the Boyd*s Corner dam.f 

■ Tbe work U folly deeoribed In Beport or Oblel ol EnglDeere, IT. a A., tor 1B90, 
pp. 3808-36. 

t From KD MoODDt of the ooiuitruotloii o( the Boyd's Oomer dam on the Croton 
Blver near New York CUty, by J. Jtmet B. Oroee, In Traaa. Am. Boo. ot O. E., toL 

iu.p.8ea. 



ovGoQi^lc 



ABT. S.] COST OF CONCBETE. 

TABLE ISk. 
Ck>BT OF Hixnro akd Li.Tnia Corobsth. 



TramporliDg cement from alore-houw. . 

UsBSuring cement 

Hiilog cement paste 

Meuurtng aand and pebbles 

Ueuuring broken stone 

Mixing concrete 

TruiBportlng concrete 

Spreading &nd Tsraming concrete 

Placing forms 

Building temponuT rallwaj 

Tofal lidor per eu. yd 



. tl.7»0 $3.098 11.538 



TABLE 13m. 
Labob Rbqtjired in Mixraa juis LiTma Cosobstr. 





hjjnn PBO Cowo Ti«D. 




Hew ToA Stonce BeBarrolr. 


Bt. Louis BeMTTolr. 




SSil^"J 


.£f2l,'?i. 








% 


IP 


111 


All WOTk on lerel 


HSS^"^^;;;;;; 


[..» 


;..» 


O.ISR 
OUG 

if 


1 


i 


can 


O.OST 


om 










fo.wi 

O.OtG 
0.U& 


0.(r» 


'i:i 


^s 


(.». 





















jvGooi^le 



llSo CONOBBTB. [chap. IT, 

I681;. The coat of mixing and laying 6 inclies of concrete for a 
pavement foandation is about 7 cents per acj. yd., for 1 part cement, 
2 partB sand, and i parte broken stone, turned six times — ezclasiTe 
of casting into place. With gravel inBtead of broken stone, the cost 
is about 6 cents per sq. yd. ; and with fonr turnings instead of six, 
the cost is about half a cent less than the prices above. 

1BB(J. Economio Concrete. The relative economy of natural and 
Portland cement mortars can be investigated as explained in 
§§ 136, 137. 

The relative strengths of gravel and of broken>stone concretes 
are stated in the last two paragraphs of g 151. The relative 
economy of concrete made with broken stone and gravel will vary 
with the cost of each; bnt aa a role, when gravel costs less than 80 
per cent, of that of broken stone, gravel is more economical. 

The strengths of both broken-stone and gravel concretes are 
given in Table 13y, page I13r, for both natural and Portland 
cements at diSereut ages. A study of these results shows that the 
relative strength of natural and Portland concrete is different at 
different ages. For example, taking averages for 10 days, the 
Portland concrete was & times as strong as the natural concrete; 
while at a year the Portland concrete was only 3 times as strong as 
the natural concrete. At 45 days and also at 6 mouths, the Port- 
land concrete was i times stronger than the natural concrete; and 
at 3 months 5 times as strong. Taking averages for like dates 
and compositions, the Portland cement concrete was 3.7 times 
as strong as natural cement concrete. With the data in Table 13d 
or 13e, puge llSjr or 112/<i, it is easy to compute the cost of each 
kind of concrete, ii the cost of a cubic yard of Portland cement 
coniTete J" more llian 3,7 times that of a cubic yard of natural 
cement concrete, then the latter is on the average the more econom- 
ical; but if Llie Portland cement concre^ costs less than 3.7 times 
thut of the natural cement concrete, then the former is on the 
average the more economical. Of course the relative cost will vary 
with the condition of the cement market and with the locality. 

198s. The following example, from actual practice, illnstratee 
the possibilities in the way of combinations between Portland and 
natural cements, and gravel and broken stone. The specifications 
called for a concrete composed of 1 volume of natural cement, 
2 volnmes of sand, and i volumes of screened broken stone. The 
contractor found that at current prices a concrete composed of 
1 volume of Portland cement and 9 Totames of gravel would cost 



ovGoQi^lc 



AKT. 3.] ARTIFICIAL STONE. 113ft 

aixmt the eame as the concrete specified. A teet of the strength of 
the two oonontes showed that at a week the Fortlaad-graTel con- 
crete was 1.6S timet as strong as the ostoral cement and broken- 
stone oonorete ; and at a month 1.S9 times as strong. Therefore th« 
Fortland-grarel concrete was the more economical, and was nied. 

Abt. 3. Artificial Stoke. 

169. Several kinds of artificial atone have come into nse within 
the last tventj-fire years tor architeGtDi:&I and artistic purposes, and 
ior the pavements of cellars, for footpaths, areas, and other locali- 
ties not subjected to the tread of heavy animals. They are all a 
-combination of hydraolio cement and sand, pebbles, etc. Some of 
them possess very connderable merit, and are of valne in districtfi 
"Where dnrable and cheap bailding-stone is not supplied by nature. 

The strength and hardness of all varieties of artificial stone vary 
directly with the ultimate strength and hardness attainable by the 
hydraulic ingredients employed. An obvious means of improving 
the quality of the stone, therefore, is the employment of the highest 
grades of cement. 

160. K^TOl-CoiOBET. As made by its inventor, Goignet, of 
Paris, its asosl ingredients are : Portland cement, silioeons hydraulic 
lime (like that obtuned at Toil, France), and dean sand, mixed 
tc^ether with a little fresh water. The proportions are varied con- 
siderably for diflerent kinds of work. The dry ingredients are first 
thoroughly mixed by hand, and again in a mill after moistening 
them very slightly with clean water. Moulds are then filled with 
the mixture, which is compacted by ramming. The pecnliaritiea 
of this stone result from (1) the small quantity of water Qsed in its 
manufacture, (2) a jadicloas choice of the qufdities and proportions 
-of the ingredients, and (3) the thoroughness with which the mixing 
^is done. It Is nothing more tiian hydraulic concrete, from which 
the coarse fragments have been omitted, and upon which have been 
conferred all the advantages to be derived from their thorongh 
manipulation. It is used in France to a considerable extent in 
-constructing the walls of honses, and in repairing masonry, — as 
bridge piers, culverts, etc. 

In this country a mixture of either natural or Portland cement 
and sand is frequently, but improperly, called B6ton-Coignet. 

161. FOETLAKS BtOBI. .This is a mixture of Portland cement 
jmd sand, or sand and gravel, compacted into form by tamping. 



ovGoQi^lc 



114 ARTIFICIAL 8T0NB. [OHAP. IV. 

When properly made it posaesseB the eaeential reqaiBitea of atreDgtb 
and hardness in a degree proportionate to the ralae of the cement 
employed. The proportions of 1 measnre of dry cement to 2 or 3}- 
measnres of sand will answer for most purposes. The manipulation 
should be prolonged and thorough to insure the production of a. 
homogeneous atone. It is much used for flagging, for which pur- 
pose the surface layer, to the thickness of about half an inch, may 
advantageously bs composed of 1 measure of cement to 1^ or 1} of 
sand. ' 

162. HcHtTBTBiE fiTOiTE. This stone consiata eesentially of the 
Portland stone described above, in the pores of which are formed 
compounds of alumina with the fatty acids by the double decom- 
position of alum and a potash soap (see § 140, page 101). Thefl& 
compounds are insoluble in water, are not acted upon by the car- 
bonic acid of the air, and add conaiderably to the early strength of 
the stone and somewhat to its ultimate strength. 

The peculiar merit of this stone is that its power of absorbing 
water la decreased by the ube of the alum and the soap. All mot- 
tars and most of the artificial stones absorb water freely,— porous 
mortar from 50 to 60 per cent, of its own weight and the best Port- 
land from 10 to 20 per cent, — and consequently they disintegrate 
rapidly under the action of frost. The absorbed water also disBolves 
the salts of magnesia, lime, soda, and potash (of all of which thero 
is always more or less in cement), and on evaporating leaves a white 
efflorescence on the surface, which injures the appearance of the 
wall. For these reasons the ordinary artificial stones are in dis- 
repute for architectural purposes. The absorptive power of the Mc- 
Murtrie stone la about twice that of granite, about equal to that of 
• the best limestones, and about one tenth or less of that of the best 
sandstones. It has been used in Washington, D. C, to a limited 
extent, the window trimmings of the National Museum and also the 
fronts of a few stores and dwellings being of this stone. It appears 
to have given entire satisfaction. 

163. Fbeab Stohe. Thisiscomposedof siliceous sand and good 
Portland cement, to which gum shellac is added. The composition 
used by the inventor was 1 measure of cement and 2J measures of 
sand moistened with an alkaline solution of shellac of sufficient 
strength to furnish an ounce of the shellac to a cubic foot of stone. 
The shellac adds to the early strength of the stone ; but it is not 
certain that it adds to the nltimate strength, nor is it certain that 



ovGoQi^lc 



AST. 3.] SOKSL 8T0NB. 116 

the sbeUac may not decay and ultimately prove an element of 



When mixed, it is rammed into wooden monlds, and after setting 
is laid away to eeaeon, — which reqaires several months for best 
resalts. It was mnch nsed in architectural work in the West a few 
years ago, bat did not gire satiBtaction. 

164. BUraoME 8to>x. This is made by forming in the in- 
terstices of sand, gravel, or any pnlverized stone a hard and 
insolnble cementing substance, by the natural decomposition of 
two chemical componnde in solution. Sand and the silicate of 
Boda are mixed in the proportion of a gallon of the latter to a 
bnshel of the former and rammed into moulds, or it may be 
rolled into elabe for footpaths, etc. At this stage of the process 
the blocks or slabs may be easily cut into any desired form. They 
are then immereed, under pressure, in a hot eolution of chloride of 
calcium, after which they are thoroughly drenched with cold water 
— for a longer or shorter period, according to their sizo— to wash 
out the chloride of sodium formed during the operation. In 
England grindstones are frequently made by this proceaE. 

165. SOESL BtOHB. Some years ago, K. Sorel, a French 6hemist> 
discovered that the oxychloride of magnednm possessed hydraulic 
energy in a remarkable degree. This cement is the basis of the 
Sorel stone. It is formed by adding a solution of chloride of mag- 
nesium, of the proper strength and in the proper proportions, to 
the oxide of magneaium. The strength of this stone, as well as its 
hardness, exceeds that of any other artificial stone yet produced, 
and may, when desirable, be made equal to that of the natural 
stone which furnishes the powder or sand used in its fabricatioa- 
The process is patented, and is used mainly in making emery-wheels. 
By incorporating large pebbles and cobble-stones in the mixture 
the stone can be .made quite cheaply, and is therefore suitftble fot 
fonndatioos and plain masslTe walls. 



ovGoQi^lc 



OHAPTEE V. 
QUARRYING, 

166. This is so lai^ a. sabject that it caiiQot be more than en- 
tered upon here ; for greater detail, see treatises on Qnarrying, Eock- 
bloBting, and Tunneling, 

167. Soubcsb of BuILDDTO Btohes. The bowlders, which are 
scattered promiscuonsly over the surface of the ground and also 
frequently buried in it, furnish an eicellent building stone for massive 
structures where strength is essential. They are usually of tough 
granite or of a slaty strnctnre, and are difficult to work. Sometimes 
they have a cleavage plane or rift, along which they may be split. 
They can be broken into irregular pieces by building a fire about 
them, and drenching them while hot with water, or they may be 
broken by explosives. 

Of oourse by far the greater quantity of stone is taken directiy 
from quarries. All building-stone deposits have usually a certain 
amount of covering, consisting either of a portion of the same de- 
posit, which has been disintegrated by atmospheric influences, or of a 
later deposit. This covering is called the "cap-rock" or "strip- 
ping," In opening the quarry, the solid portions of cap-rock are 
liroken op by blasting, and the whole is cart«d out of the way. After 
asufficientspaceisstripped, the next Btep necessary, when the quarry 
rock does not stand out in clifFa, is to excavate a narrow space on 
one side for a quarry face, either by blasting or by some of the 
methods to be described presently. 

168. Kethods of flUABBTino. After a considerable area has 
thus been laid bare, the stone is quarried in one of three ways, 

169. J. By Hand Tools. When the stone is thin-bedded, it may 
be quarried by hand-tools alone. The principal tools are pick, crow- 
bar, drill, hammer, wedge, and plug and fenthers. The layers are 
forced apart by the crow-bar or wedges. The flat pieces are broken 

jip with the hammer or by drilling holes for the plug and feathers. 

lis 



ovGoQi^lc 



> QUARBYINO BY BXPL08IVI8. IIT 

The plug ii a narrow wedge with pluie tocee; the feathers are 
wedges ^t od one side and roanded on the other (Fig. S5, page 128). 
When a plug is placed between two feathers, the three will slip into 
a cylindrical hole ; if the plug is then driven, it exerts a great force. 
If these plugs and feathers are placed a few inches apart in a row, 
and all driven at the same time, the stone will be cracked along the 
line of the holes, even though it be comparatively thick. 

The drill used to cut the holes for the plug and feathers is a bar 
of steel furnished with a wide edge sharpened to a blunt angle and 
hardened. It is operated by one man, who holds the drill with one 
hand and drives it vrith a hammer in the other, rotating the drill 
between blows. The holes are usually from | to } of an inch in 



Sandstonee and lunestones occnrring in layers thin enough to 
be quarried as above are usually of inferior quality, suitable only 
for slope walls, paving, riprap, concrete, etc. 

170. //. By Exploalvea. Generally, the cheapest method of 
qoarrying small blocks is by the use of ezplosiveB. However, ex- 
plosives are used mainly for detaching large blocks, which are after- 
wards worked up by means of wedges. In this method of quarry- 
ing, drill-boles are put down to the depth to which the rock is to. 
be split, and the requisite amount of powder or other explosive put 
in, covered with sand, and fired by a fuse. Sometimes numerous 
charges in a line of drill-holes are fired simultaneously by means of 
electricity. 

Quick-acting explosives, like dynamite, have a tendency to shatter 
the stone and break it in many directions, the texture being affected 
by the sudden explosion in the same manner as by the blow of a 
Irammer. Coarse gunpowder is generally preferred for quarrying 
stone. Light charges of powder lightly covered with sand are better 
than heavy charges tightly tamped ; * and experience goes to show 
that better work is done by repeated light bUiste in the same hole, 
than by a single heavy blast. By means of light charges often re- 
peated, a mass of rock may be detached without being broken up, 
which would be badly shattered by a single charge strong anough to 
detach It. 

In each locality the atrncture of the rock must be carefully 

* For an utlcle ihowlog tbat »n aiz-apaoe Bboold be left betneea Ui* explodr* 
and tbe tanqilDg, we Siigineering jVhut, vol. xviii. p. S83. 



ovGoQi^lc 



118 QUABBTUra. [chap. V, 

studied with a view to take advantage of the cleavage plaiies and 
Datura) joints. For quarrying each class of rocks there is a charac- 
teristic method employed, which is, however, varied in detail in 
different quarries. The minor details of qnarry methods are as 
various as the differences existing in the textures, structaree, 
and modes of occurrence of the rocks quarried. Much depends 
upon how the hlast is made. The direction ia which the rock is 
most liahle to break depends upon the structure of the rock and 
the shape of the drill-hole. Even such an apparently unimportant 
matter as the form of the bottom of the drill-hole into which the 
explosive is put has a very marked effect. If bored with a hand- 
drill, the hole is generally triangular at the bottom, and a blast in 
snch a hole will break the rock in three directions. In some qnar* 
Ties the lines of fracture are made to follow predetermined directions 
by putting the charge of powder into canistors of special forms.* 

171. Drills. The holes are bored by jumpers, chnrn-drills, or 
machine-drUIs. The first is a drill similar to the one need for drill- 
ing holes for plugs and feathers (§ 169), except that it is larger and 
longer. It is usually held by one man, who rotates it between the 
alternating blows from hammers in the hands of two other men. 
Chum-drills are long, heavy drills, usually 6 to 8 feet in length. 
They are raised by the workmen, let fall, caught on the rebound, 
raised and rotated a little, and then dropped again, thus cutting 
a hole without being driven by the hammer. They are more eco- 
nomical than jumpers, especially for deep holes, as they cut faster 
and make larger holes than hand-drills. 

172. Machine rock-drills bore much more rapidly than hand- 
drills, and also more economically, provided the work is of sufBcient 
magnitude to justify the preliminary outlay. They drill in any 
direction, and can often be used in boring holes so located that they 
could not be bored by hand. They are worked either by steam 
directly, or by air compressed by steam or water-power and stored 
in a tank called a receiver and thence led to the drills through iron 
pipes. 

A variety of rock-drilling machines has been invented, f but 
they can be grouped in two classes, viz., percussion-drills and rotat- 
ing drills. The method of action of the percussion-drill is the same 



ovGoQi^lc 



QD4HKYISU BY EXPLOSIVES. 119 

aa that of the churn-drill already described. The usual form ia 
that of a cylinder, in which a piaton ia moved by Btoam or com- 
pressed air, and the drill is attached to this pieton bo as to make a 
stroke with every complete movement of the piston. An aatomatio 
device cansea it to rotate slightly at each stroke. 

173. In the rotating drills, the drill -rod is a Jong tube, revolving 
about its axis. The end of the tnbe — hardened so as to form an 
annular cutting edge — is kept in contact with the rock, and by its 
rotation cuts in it a cylindrical hole, generally with a solid core in 
the center. The drill-rod is fed forward, or into the hole, as the 
■drilling proceeds. The dibria is removed from the hole by a con- 
stant stream of water which is forced to the bottom of the hole 
through the hollow drill-rod, and which carries the dibris up 
through the narrow apace between the outside of the drill-rod and 
the sides of the hole. 

The diamond drill is the only form of rotary rock-drill ezten- 
aively used in this conntry. The tube has a head at its lower end, 
in which are set a number of carbons or black diamonds. The 
diamonds usually project slightly beyond the circumference of the 
head, which is perforated to permit the ingress and egress of the 
"water used in removing the debris from the hole and at the same 
time prevent the head from binding in the hole. When it ia desir- 
able to know the precise nature and stratification of tho rock pene- 
trated, the catting points are bo arranged aa to cut an annular groove 
in the rock, leaving a solid coro, which is broken off and lifted out 
whenever the head is brought up. Where it is not desired to pre- 
serve the core intact, a solid boring-bit is used instead of the core- 
bit. They are made of any size np to 15 inchea in diameter. 

174. Explosives.* The principal explosivea are gunpowder, 
nitro-glycerine, and dynamite. Only a coarse-grained and cheap 
variety of the first is used in quarrying, the others being too sudden 
and too strong in their action. 

The pressure exerted by gunpowdpr when fired in a confined 
fipace depends upon the relative weight and quality of powder used, 
and upon the space occupied by the gases evolved. The absolute 
force of gunpowder, the force which it exerts when it exactly fills 
the space in which it ia confined, has never been satiBfactorily ascer- 

* * Taimollng,'' 



jvGooi^le 



120 QUA&BYINO. [chap. V. 

tallied. It has been vftrtoasly eetiinated &t from 15,000 to 1,500,000 
poandB per square inch. ExperimentB by Gen. Sodman ahow that 
for the powder used in gunnery the absolute force of explosion i& 
at least 300,000 pounds per sqiiare inch. " In ordinary quarrying^ 
a cubic yard of solid rock in place (or about 1.9 cubic yards piled, 
up after being quarried) requires from i to } pound of powder. 
In very refractory rock, lying badly for quarrying, a solid yard may 
Irequire from 1 to 2 poands. In some of the moat aucceEsful great- 
blasts for the Holyhead Breakwater, Wales, (where several thou- 
sands of poands of powder were exploded, usually by galvanism, at 
a single shot,) from 2 to 4 cubic yards (solid) were loosened per 
pound of powder ; but in many instances not more than 1 to I^ 
yards. Tnnnels and shafts require 2 to 6 pounds per solid yard,. 
usually 3 to 5 pounds. Soft, partially decomposed rock frequently 
requires more than harder ones." * 

The explosion of the powder splits and loosens a mass of rook 
whose. volume is approximately proportional to the cube of the Una 
of least resisloTicei^thai is, of the shortest distance from the charga 
to the surface of the rock, — and may be roughly estimated at ttoic& 
that cube ; but this proportion varies much in different cases. Thd 
ordinary mle for the weight of powder in small blasts is 

PowDBB, inpounds, = (Lute op Eesistance, in feet,)' -*- 32. 

Powder is sold in kegs of 25 lbs., costing about $2.00 to t2.2& 
per keg, exclusive of freight, — which is very high, owing to the lisk^ 

176. Most of the explosives which of late years have been tak- 
ing the place of gunpowder consist of a powdered substance, partly- 
saturated with nitro-glycerine — afuid produced by mixing glycerine 
with nitrio and sulphuric acids. , Nitro-glycerine, and the powders. 
containing it, are always exploded by means of sharp percussion, 
which is applied by means of a cap and fuse. The cap is a hollow 
copper cylinder, about i inch in diameter and an inch or two in 
length, containing a cement .composed of fulminate of mercnry and. 
some inert substance. The cap is called single-force, double-force, 
etc, according to the amount of explosive it contains. 

The principal advantages of nitro-glycerine as an explosive con- 
sist (I) in its instantAneoua development of force, due to the fact 
that, pound for pound, it produces at least three and a half times. 
* Tnntwlne's Englueer's PocKet-book. 



ovGoQi^lc 



QUARRTING BY BXPL06ITES. 12X. 

as much goe, and twice as much beat, as gaupowder ; and (2) in ita 
high specific graTity, irhich permits the use of small drill-holeB. 

Nitro-gljcerine is rarely used in the liquid state in ordinary 
qnarrjing or blastings owing to the liability of explosion through 
accidental percussion, and owing to its liability to leakage. It ex- 
plodes 80 suddenly that very little tamping is required, the mer& 
weight of moist sand, earth, or water being sufficient. This fact^ 
and the additional one that Ditro-gljcerine is uuaffected by immer- 
sion in water and is heavier than water, render it particularly suit- 
able (or sub-aqueous work, or for holes containing water. If th» 
rock is seamy, the nitro-glycenne must be confined in water-tight 
casings. Such casings, however, necessarily leave some spaces be- 
tween the rock and the explosive, which diminishes the effect of the 
latter. The liquid condition of nitro-glycerine is oseful in causing 
it to fill the drill-hole completely, so that there are no empty spaces 
in it to waste the force of the explosion. On the other hand, the 
liquid form is a disadvantage, because when thus used in seamy 
rock without a containing vessel portions of the nitro-glycerine leak 
away and remain unexploded and unsuspected, and may cause aoci~ 
dental explosion at a future time. 

The price of nitro-glycerine is from 50 to 60 cents per quart. 

176. Dynamite is the name given to any explosive vhich con- 
tains nitro-glycerine mixed with a granular absorbent. If the 
absorbcot is inert, the mixture is called true dynamiie; if the 
absorbent itself contains explosive substances, the mixture is called 
false dynamite. The absorbent, by its granular and compressible 
condition, acts as a cushion to the oitro-glycerine, and protects it 
from percussion and from the consequent danger of explosion, bat 
does not diminish its power when exploded.' \itro-glycerine 
undergoes no change in composition by being absorbed ; and it 
then freezes, burns, explodes, etc., under the same conditions as 
to pressure, temperature, etc., as when in the liquid form. The 
cuahiouing effect of the absorbent merely renders it more difiQcult 
to bring about sufficient percussive pressure to cause explosion. 
The absorption of the nitro-glycerine in dynamite renders the lat- 
ter available in horizontal holes or in holes drilled npward. True- 
dynamite loses only a very small percentage of ita explosive power 
when saturated wiUi water, but is then much more diMcult to ex- 
plode. 



ovGoQi^lc 



322 QLARKTIXO. [CHAP. V. 

True dynamites must contaia at least 50 per cent, of uitro- 
f^ljcerine, otherwise the latter will be too completely cushioned 
by the absorbent, and the powder will be too difficult to explode. 
False dynamitee, on the contrary, may contain as small a percentage 
of nitro-glycerine aa may be desired, gome containing as little as 15 
per cent. The added explosive substances in the false dynamites 
generally contain large quantities of oxygen, which are liberated 
upon explosion, and aid in effecting the complete combuation of 
any noxious gases arising from the nitro-glycerine. The false are 
generally inferior to the trne dynamites, since the bulk of the 
iormer is increased in a higher ratio than the power; and as the 
cost of the work is laigely dependent upon the size of the drill- 
faolee, there is no economic gain. 

Dynamites which contain large percentages of nitro-glycerine 
explode with great suddenness, tending to break the rock into 
small fragments. They are most useful in blasting very hard rock. 
In such rock dynamite containing 75 per cent, of nitro-glycerine 
is roughly estimated to hare about 6 times the force of an equal 
weight of gunpowder ; but in soft rock or clay its power, at equal 
cost, is inferior to that of common gunpowder, because its action 
is akin to a sudden blow, rather than to a continued push. F'^r 
«oft or decompoaed rocks, sand, and earth, the lower grades -:* 
ilynamite, or those containing a smaljer percentage of nitro-glyceh 
ine, are more suitable. They explode -with less suddenness, and 
their tendency is rather to upheave large masses of rock than to 
splinter email masses. 

"Judgment must be exercised as to the grade and quantity of 
explosive to be used in any given case. Where it is not objection- 
able to break the rock into small pieces, or where it is desired to do 
BO for convenience of removal, the higher shattering grades are use- 
ful. Where it is desired to get the rock out in large masses, as in 
quarrying, the lower grades are preferable. For very difficult work 
in hard rock, and for submarine blasting, the highest grades, con- 
taining 70 to 75 per Cent, of nitro-glycerine, are used. A small 
<charge docs the same execntion as a larger charge of lower grade, 
and of course does not require the drilling of so large a hole. In 
submarine work their sharp explosion is not deadened by the 
water. For general railroad work, ordinary tunneling, mining of 
ores, etc., the average grade, containing 40 per cent, of nitro-glycer* 



ovGoQi^lc 



NITBO-GLTOEBIKB EXPL08ITE8. 133 

ine, is used ; (or' quarrying, 35 per cent.; for blnetiog stumps, trees, 
piles, etc., 30 per cent; for sand and earth, 15 per cent." 

177. A great variety of dynainiteB is made. £aoh manufactarer 
usually makes a number of grades, containing different percentages 
of nitro-glycerine, and gives to his powder some fanciful name. 
Dynamite is sold in cylindrical, paper-co^^red cartridges, from ^ of 
an incli to 3 inches in diameter, and 6 to 8 inches long, or longer, 
which are packed in boxes containing 25 or 50 pounds each. They 
are furnished, to order, of any required size. The price per pound 
ranges from 15 cents for 15 per cent, nitro-glycerine to 50 cents for 
75 per cent, nitro-glycerine. 

Table 14 (page 124) gives the names of all the explosives con- 
taining nitro-glycerine, with the per cent in each case.* 

178. HI' By Cbanseling and Wodging, By channeling is meant 
the process of cutting long narrow channels in rock to free the sides 
of large blocks of atone. Quite a large number of machines have 
been invented for doing this work, all of which make the channels 
by one form or the other of the machine drills already described 
{see the second paragraph of § Y!%). The machines are mounted 
upon a track on the bed of the quarry, and can be moved forward 
«8 the work progresses. If the rock is in layers, it is only necessary 
to cut the channels part way through the layer, when the block can 
be detached with wedges, the groove guiding the fracture. If the 
rock is not in layers, after the necessary channels have been cnt 
«round the block, it is necessary to under-cut the block in order to 
release it. This is accomplished by drilling a series of holes along 
the bottom, which process Is called "gadding" by quarry-men. The 
block is then split from its bed by means of wedges. The method 
■of channeling and wedging is much employed in quarrying marble, 
the massive limestones, and the thick-bedded sandstones. The 
method is very economical and expeditious, except in granite and 
the hardest sandstones. For illustrations of the two principal chan- 
neling machines and also quarries being worked by this method, see 
Report on the Quarry Industry, pp. 44-52, in Vol. X. ex the Tenth 
"Census of the United States. 

• W. C, Foster, In Engineering Aew, vol. ili. p. 2H. For a Ust o( all the eiplo- 
tUvea employed as h\aal\ng attenta, together with a description at their composition 
and references to the llteraturo of each, see Englsitering A'aet, vol. xlx. pp. S33-31, 
«Dd vol. xz. pp, 8-10. 



ovGoQi^lc 



[chap. t. 



TABLE 14. 
ZdR or X^PLoama conrAiHiiia NmuMn-TCBBnn. 



Ammonia powder.. . 
AjbeMoa powder. . . . 
Atlu powder, A 

:: :: 2+- 



" B+ 

'■ E 

" F+ 

Brady's dynamite. 

Brain's powder 

Co Ionia powder 

Dualln (Dinmar's) 

Dynamite (Nobel's, Eieael- 
guhr dynamite), 

Old No. 1 
Old No. 3 
Old No. " 

£leclrlc powder 

Explosive gelatine 

PordtB. a grades 

Fulgurite (Holid) 

(liquid) 

Gelatine dynamite, A 

No. 1. 



QeUtine c 



iploslve d< 



Qdlgnlte 

Oiant powder. No. 1 . 
" ■' New" 1. 



■ (Nobel's)! 



Glyxoline 

HecU powder. No, IXX. 

Oun Sawdust 

'■ No. IX..... 



Herculee powder. No. IZX 



Horsley's powder (■ 

TarieUes) 

Jud«oD Giant Pt>wder,No.2 
Judsou powder, FFF. 

FF 

F 

UUP 

Litbofracteur 

Metalllue Nltroleum 

Uica powder. Mo. 1., 

" a 

Miners' Powder Co. "a Dy- 

uamite 

Nepluue powder 

Nitro Tolnol 

Norrbin & OblsaoD's pow> 

Poniopollte 

Porlleia Nitroleum. . . 

Bendrock 

Sebastin, No. I 

■' S 

Selenitic powder 

Vifioriteiif.'s)".!;'.".!! 
I Vitrlle, No. 1 



S 10 C 

sa 



48.& 



ovGoQi^lc 



CHAPTER VL 
stone outtinq. 

Art. 1. Tools. 

179. In order to describe intelligibly the varions methods of 
preparing stonea for use in masonry, it will be necessary to begin 
Tith a description of the tools used in stoue-catting, ae the names 
«f many kinds of dressed stones are directly derived from those of 
the tools used in dressing them. 

With a view to securing uniformity in the nomenclature of 
building stones and of stone masonry, a committee of the American 
Society of Civil Engineers prepared a classification and recommended 
that all specifications should be made in accordance therewith. The 
old nomenclature was very unsystematic and objectionable on many 
grounds. The new system is good in itself, is recommended by the^ 
most eminent authority, has been quite generally adopted by en- 
gineers, and should therefore be strictly adhered to. The following 
description of the Itand tooh used in stone cutting is from the 
report of the American Society's committee.* 

180. Hass Toolb. "The Double Face Hammer, Pig. 9, is a 
heavy tool weighing from 20 
to 30 ponnds, used for roagh- 
1y shaping stones as they 
■come from the quarry and 
for knocking off projections. 
This is used only for the fio- b.— doiibl« Fiat HAMim. 
roughest work. 

"The Face Hammer, Fig. 10, has one blunt and one cutting 
end, and Is used for the same 
purpose as the double face 
hammer where less weight is ( 
required. The cutting end 
is used for roughly squaring 

stones, preparatory to the use fio. id.— Face Ruaan. 

•of finer tools. 

• Tnns. Am. Soo. of C. E., voU tL pp. 297-801 



ovGoQi^lc 



126 STONE CUTTING. [CHAP. VI. 

"The Cavil, Fig. 11, lias one blunt and one pyramidal, or 
,. — T-r~^^^^ pointed, end, and weighs from 15 to 20 pounds. 
'J|__LL--^^ It ie used in quarries for rooghly shaping stone 

^r ^i-*,^^^ for transportation. 

'I — "^ The rick. Fig. 12, somewhat resembles the 

Tia. 11.— Civn. pi^k Used in digging, and is used for rough dress- 
ing, mostly on limestone and sandstone. Its length varies from 
l.*^ to Zi inches, the thickness 
at the eye being about 
inches. 

"The Ax, or Pean ffam-\ 
mer, Fig. 13, has two opposite jOr t 
cutting edges. It is used for \ 
making drafts around the arris, 
or edge, of stones, and in re- 1 
ducing faces, and sometimes '"■ "-— P>ot 

joints, to a level. Its length is about 10 iuches, and the cutting 
edge about 4 inches. It is used after 
the point and before the patent ham* 



"The Tooth Ax, Fig. U, is like 
Fio. IS.— Ax. the as, except that its cutting edges 

are divided into teeth, the number of which varies with the kind 
of work required. This tool 
is not used in granite and 
gneiss cutting. 

"The Bush Hammer, 
Fig. 15, is a square prism of 
steel whose ends are cut into 
a number of pyramidal points. Pro. m.-TootbAi. 

The length of the hammer ia from 4 to 8 inches, and the cutting 

face from 2 to 4 inches square. 

The points vary in number and 
in size with the work to be done. 
One end is sometimes made 
■ ^r-l with a cutting edge like that of 
Fro. IB.— BtiBB HinaB. t)^e ^X. 

" The Crandall, Fig. 16, is a malleable-iroB bar abont two feet 



ovGoQi^lc 



f 



FiQ. IS.— Cruoili. 



ART. 1.] T00L8, 137 

long, slightly flattened at one cod. In this end ia a slot 3 inches 

longand finchwide. Throughthia r~'~^ — 

slot are passed ten double-headed , 

points of i-inch square Bteel, 9 I 

inches long, which are held in "I 

place by a key. 

"The Patent Hammer, Fig. 
17, 18 a double-headed tool bo 
formed as to hold at each end a set of wide thin chisels. The tool 

f- ^m is in two parts, which are held to- 

uWJ gether by the bolts which hold tho 
pi^3 ' chiBels. Lateral motion is prevented 
^^, by four guards on one of the pieces. 

fm.it.— PiramHAiuEK. The tool without the teeth is 

fiiXSJXl^ inches. The teeth are 3^ inches wide. Their thickness 
varies from ^ to ^ of an inch. This tool is _. ^ 
used for giving a finish to the surface of stones. JbW 1 . .. ..-^ 

"The Hand Hammer, Fig. 18, weighing '*U-\J 
from 3 to 5 pounds, is used in drilling holes, Fia. is.— Hahd hihhba. 
ai)d in pointing and chiseling the harder rocks. 

"The Mallei, Fig. 19, is used whei-e the softer limestones and 
— sandstones are to be cut. 

I r' 1 fl n " '^^^ Pitching Chisel, Fig. 30, 

V J fl 11 is OBually of li-inch octagonal steel, 
^ — "^ ^ spread on the cutting edge to a 
pS^i^o rectangle of i X 3^ inches. It is 
Fi8. 11.— HiLUT. cmmi. used to make a well-defined edge to 

the face of a stone, a line being marked on the joint surface to 
which the chisel is applied and the portion of the stone outside of 
the line broken oS by a blow with the hand-hammer on the head 
of the chisel. 

"The Point, Fig. 21, is made of round or octagonal rods of 
steel, from i inch to 1 inch in diameter. It is made about 19 
inches long, with one end bronght to a point. 
It is used until its length is reduced to about C( J > 



inches. It is employed for dressing off the ( J 
irregular surface of stones, either for a perma- 
nent finish or preparatory to the use of the ax. ^'- *•■— Po""- 
According to the hardness of the stone, either the haud-hamn 
or the mallet is used with it. 



ovGoQi^lc 



STONE ODTTINO. 



[chap. VI. 

" The Chisel, Fig. 32, of round etee] of ^ to j inch in diameter 

^_^___^__^^ and about 10 inches long, with one end brought 

C( "1 to a cutting edge from i inch to 2 inches 

-. . ~.^ wide, is used for cutting drafts or margins on 

the &ce of Btonee. 

Fra tt-CmMu "The Tooth Chisel, Fig. 33, is the same 

as the chisel, except that the cutting edge is divided into teeth. 

It is used only on mar- ^ ■ 

bles and sandstones. ^ ^ i ' - t 

"The Splitting CH. 
Chisel, Fig. 34, is used 
chiefly on the softer, Toorac 

stratified stones, and sometimes on fine architectural carvings in 
granite. 

"The Plug, a truncated wedge of steel, and the Feathers of 
'half-round malleable iron. Fig. 25, are used for splitting unstrati- 
■ fled stone. A row of holes is made with the DriU, Fig. 26, on the 



Pig a. '*'■ *■— Dwu*. 

line OD which the fracture is to be made ; In ench of these holes 
two feathers are inserted, and the plugs lightly driven in between 
them. The plugs are then gradually driven home by light blows 
•of the hand hitmmer on each in snccession until the stone splits." 

181. BUCHIBE Tools. In all large stone-yards machines are 
used to prepare the stone. There is great variety in their form, 
but since the surface never takes its name from the tool which 
forms it, it will be neither necessary nor profitable to attempt a de- 
scription of individual machines. They include stone-saws, ston&- 
ontters, stone-planers, stone-grinders, and stone-polishers. 

The saws may be either drag, circular, or band saws ; the cut- 
ting may be done by sand and water fed into the kerf, or by carbons 
or black diamonds. Several saws are often mounted side by side and 
operated by the same power. 

The term ".stone-cutter" is oaoally applied to the machine which 



D.qitizeabyG00l^lc 



ABT. 3.] HBTHOD OF FORMING SCBFACES. 199 

sttackB the rough stone and reduces the inequalitieB somewhat. 
After this treatment the stone goes id Bucceseion to the atooe- 
planer, stone-grinder, and Btone-polieher. 

Those stones which are homogeneous, strong and tough, and 
•comparativeiy free from grit or hard spots, cau be worked by ma- 
chines which resemble those used for iron ; but the harder, more 
brittle stones require a mode of attack more nearly resembling that 
employed in dressing stone bj hand. Stone-ontters and stone- 
planers employing both forms of attack are made. 

Stone-grinders and stone-polisbers differ only in the degree of 
fineness of the surface produced. They are sometimes called " rub- 
bing-machines." Essentially they consist of a large iron plate re- 
Tolving in a horizontal plane, the stone being laid npon it and braced 
to prevent its sliding. The abradent is sand, which is abundantly 
supplied to the surface of the revolving disk. A small stream of 
water works the sand under the stone and also carries away the 
44bris. 

ABT. 3. UffTHOD OF FOHUIXa THE ScBFACES. 

182. It is important that the engineer should understand the 
methods employed by the stone-cutter in bringing atones to any re- 
quired form. The surfaces moat frequently required in stone cutting 
are plane, cylindrical, warped, helicoidal, conical, spherical, and 
sometimes irregular snrfacea. 

183. PtUtS SntTACEB. In squaring up a rough stone, the first 
thing the stone-cutter does is to draw a line, with iron ore or black 
lead , on the edges of the stone, to indicate as nearly as possible the 
required plane surface. Then with the hammer and the pitching- 
tool he pitches off all debris or waste material above the lines, 
thereby reducing the surface approximately to a plane. With a 
chisel he then cnta a draft around the 
edges of this surface, /. e., he forms nar- 
row plane surfaces along the edges of the 
stone. To tell when the drafts are in | 
the same plane, he uses two straight- 
edges having parallel sides and equal 
■widths. See Fig. 27. The projections '^■*'- 

on the aartace are then removed by the pitching chisel or the point, 
antil the straight-edge will just touch the drafta and the inter- 
mediate surface when applied across the stone in any direction. 




ovGoQi^lc 



180 eroNE ccrnNO. [chap. vt. 

The surface is oBoally left a little " slack," t. e., concave, tf' stlov 
room for the mortarj however, the snrtiioe Bbould be but a verf 
little concave. 

The surface is then flmsbed with the ax, patent hammer, bnu 
hammer, etc, according to the degree of smoothness reqaired. 

181. To form a second plane sujface at right angles to the first 
one, the workman draws a line oa the out face to form the inters 
^section of the two planes ; he also draws a line on the ends of th» 
stone approximately in the reqaired plane. With the ax or tha 
chisel he then cuts a draft at each end of the atone until a steel 
square fits the angle. He then joins these drafts by two others afe 
right angles to tfaem, and brings the whole surface to the samfr 
plane. The other faces may be formed in the same way. 

It the surfaces are not at right angles to each other, a berel i> 
tued instead of a square, the same general method being pursued. 

186, CnnnisiOAi BVBFAOBS. These may be either concave or 
oonvez.. The former are frequently required, as in arches; and th» 
latter sometimes, as in the outer end of the face-stones of an arch. 
The stone is first reduced to a paralellopipedon, after which the 
onrved surface is produced in either of two ways : (1) by cutting; 
a oircular draft on the two ends and applying a straight-edfi^e along 
the rectilinear elements (Fig. !28); or (3) by cutting a draft along 
the line of intersection of the plane and cylindrical surface, and 
^)plying a carved templet to the reqaired surface (Fig. : 




186. ConcAL SlTBTAOBB may be formed by a process very sindlar 
to the first one given above for cylindrical surfaces. Such surfaces 
are seldom used. 

187. S?KSSlOAi fluKFAOXB are sometimes employed, as in domes. 
They are formed by essentially the same general method as cylin* 
drioal surfaces. 

188. Wab?BO Bvbfaobs. Under this head are included what 



ovGoQi^lc 




ART. 3.J 1IBTH0D3 OF PINI3HIN0 SUBFACEB. 181 

the Btone-catters call " twiBted earfaoeB, " beliooidal eurfaces, sud 
the general warped surface. None of 
these are common in ordinary stone-work. 

The method of forming a anrtace 
equally twinted right and left will be de- 
scribed ; by obvious modiflcatione the same L 
method can be applied to secure other j: 
forms. Two twist rules are required, the 
angle between the upper and lower edges ^'- "'■ 

being half of the required twist. Draf ta are then cut in the eoda of 
the stone antil the tops of the twist rules, when applied as in Fig. 
30, are in a plane. The remainder of the projecting face ia remored 
until a straight-edge, when applied parallel to the edge of the stoD«^ 
will just touch the end drafts and the intermediate surface. 

If the surface is to be twisted at only one end, a parallel rulo 
and a twist rule are used. 

166. Kunre THX SR&VlXOa. The method of making work- 
ing drawings for constructions in stone will appear in subsequent 
chapters, in connection with the study of the structures them- 
Gelves; but for detailed instructioss, see the text-books on Store- 
otomy or Stone Gutting. 

Aet. 3. Methods of PoiiSFtiio thb Suefacm.* 

190. "All stones used in building are divided into three clasBei^ 
according to the finish of the surface; rjz. : 

I. Bough stones that are used as they come iroai the quarry. 

II. Stones roughly squared and dressed. ■ 

III. Stones accurately squared and finely dressed. 

"In practice, the line of separation between them isnotTery 
distinctly msrlced, but one class gradually merges into the next. 

191. I. "UHBauAKED STons. This class covers all stones 
which are used as they come from the quarry, without other 
preparation than the removal of very acute angles and excessive pro- 
jections from the general figure. The t«rm 'backing,' which ia 
frequently applied to this class of stone, is inappropriate, as it prop- 
erly designates material used in a certain relative position in a wall, 
whereas stones of this kind may be used in any position. 

192. n. "Sqdased Stoheb. This class covers all stones that 

■This Biticle la taken from the report of tile committee o( tbe Amerfcaii Bodttf 
«f CiTil Eogtneera previously referred to. 

D.qitizeabvG00l^lc 



139 



BTONB cirmso. [chap, vl 

are roughly squared and roughly dressed on beds and joints. The 
dressing is usually done with the face hammer or ai, or in soft 
stones with the tooth hammer. In gneiss it may sometimes be 
necessary to use the point. The distinction between this class and 
the third lies iu the degree of closeness of the joints. Where the 
dressing on the joints is such that the distance between the general 
planes of the surfaces of adjoining stones is one half inch or more^ 
the stones properly belong to this class, 

"Three subdivisions of this class may be made, depending OD 
the character of the face of the stones: 

" (n) Quarry-faced stones are those whose faces are left «n- 
tonched as they come from the quarry. 

" {b) Fitch-faced stones are those on which the arris is clearly 
defined by a line beyond which the rock is cut away by the pitching 
'Chisel, so as to give edges that are approKimatelj true, 

" (c) Drafted Btonei are those on which the face is surrounded by 
a chisel draft, the space inside the draft being left rough. Ordi- 
narily, however, this is done only on stones in which the cutting of 
the joints is such as to exclude tliem from this class. 

" In ordering stones of this daes the specifications should always 
etate the width of the bed and end joints which are expected, and 
also how far the surface of the face may project beyond the plane 
of the edge. In practice, the projection varies between 1 inch and 
6 inches. It should also be specified whether or not the faces are to 
be drafted. 

193. IIL " Cut STOITXS. This class covers all squared stones 
with smoothly-dressed beds and joints. As a rule, all the edges of 
cut stones are drafted, and between the drafts the stone is smoothly 
dressed. The face, however, is often left rough where construction 
is massive. 

" In architecture there are a gr^t many ways in which the faces 
of cut stone may be dressed. 




but the following are those 
that will usually be met in 
engineering work; 

" Rough-pointed. When it 
is necessary to remove an inch 
Fw. M.-BOTBH-wni«D. or more from the face of a 

le, it ifl done by the pick or heavy point until the projectiona 



ovGoQi^lc 



ART. 3.] METHODS OF TINISHINO SUBFACES, 



vary from ^ inch to 1 inch. The atone is then said to be rough- 
pointed (Fig. 31). In areesing 



Fio. Si.^FiKi-roiiiTm. 



limestone and granite, this 
operation precedes all otliers. 

"Fine-pointed. (Fig. 32). 
If a smoother finish is desired, 
rongh pointing ie followed by 
fine pointing, which is done 
with a fine point. Fine point- 
ing is used only where the finish made by it is to be final, and never 
as a preparation for a final finish by another tool. 

" Crandalled. This is only a speedy method of pointing, the 
effect being the same as fine pointing, except that the dots on the 
stone are more regular. Tlie variations of level are about i inch, 
and the rows are made parallel. When other rows at right angles 
to the first are introduced, the stone is said to be cross-crandaUed. 
Fig. 33. 




T\a. 33.— CB1HD1U.CD. Fia. 

" Axed, or Pean-hammered, and Patent-hammered. These two 
vary only in the degree of smoothness of the surface which is pro- 
duced. The number of blades in a patent hammer varies from 6 to 
12 to the inch; and in precise specifications the number of cuts to 
the inch must be stated, such as 6-ctit, 8-cut, 10-cut, 12-cut. The 
effect of axing is to cover the surface with chisel marks, which are 
made parallel as far as practicable. Fig. 34. Axing is a final finish, 

"Tooth-azed. The tooth-ax is practically a nnmber of points, 
and it leaves the surface of a stone in the same condition as fine 
pointing. It is usually, however, only a preparation for buiih-liam- 
muring, and the work is then done without regard to effect so long 
as the surface of the stone is sufficiently leveled. 

" BuBb-hammered. The roughnesscsof a stone are pounded off by 



ovGoQi^lc 



STONE CUmNG. [CHAP. 71. 



the bnsh hammer, and the Btone is then eaid to be 'bushed.' 
This kind of finiah ia dtiQgeroas 




Fra. n.— BcBs-n. 



OD sandstooe, as experience has 
ahowD that sandstoDe thus treated 
IB Tery apt to scale. In dreeaing 
limestone which ia to have a buah- 
hammered finish, the usual se- 
qaence of operation ie (1) rough- 
pointing, (3) tooth-azing, and (3) 
boab-hammering. Fig. 35. 

" Rubbed. In dressing sandstone and marble, it ia reij common 
to give the stone a plane aurface at once 
by the use of the atone-aaw [§ 181]. Any 
roughnesaea left by the saw are removed 
by rubbing with grit or sandstone [§ 181]. 
Such atones, therefore, have no margins. 
They are frequently used in architecture 
forstriog-coursGS, lint^la, door-jambs, etc.; fi«- w— bdsbr>. 

and they ai-e also well adapted for use in facing the walls of lock- 
chambers and in other localitiea where a stone surface ia liable to be 
rubbed by veBsela or other moving bodiee. Fig. 36. 

" Diamond Panels. Sometimea the apace between the mai^ns 
ia sunk immediately adjoining them and 
then riaea gradually until the four planea 
form an apex at the middle of the panel. 
In general, such panels are called diamond 
panels, and the one just deacribed, Fig. 
37, ia called a sunk diamoiid panel. 
When the surface of the atone riaea grad- 
ually from the inner linee of the margins 
to the middle of the panel, it is called a 
Both kinds of finish are common on bridge 
The detoila of this method should be 




.— DUMOND PaNCL. 



raised diamond panel, 
quoins and similar work, 
given in the specifications." 



ovGoQi^lc 



CHAPTER VII. 

STONE MA80NBT. 

Iv preparing Bpecificatione, it is Dot safe to depend alone upoo 
ft he tenne in common use to designate the varioas Glasses of masonry; 
but every specification should contain an accurate description of the 
character and quality of the work dsBired. Whenever practicable, 
samples of each kind of cutting and masonry should be prepared 
.beforehand, and be exhibited to the pereous who propose to under- 
take the work. 

194. SBTDimon of Fabts of thx Wau,.* Face, the front 
.snrface of a wall; back, the inside surface. 

Facing, the stone which forms the face or outside of the wall. 
Backing, the stone which forms the back of the wall. Filling, the 
•interior of the wall. 

Batter, The slope of the surface of the wall. 

Course. A horizontal layer of stone in the wall. 

Joints. The mortar-layer hetween the atonu. The horizontal 
joints are called hed-Joinls or simply beds; the vertical joints are 
.sometimes called the builds. Usually the horizontal joints are 
called beds, and the vertical ones joints. 

Coping. A course of stone on the top of the wall to protect it. 

Pointing. A better quality of mortar put in the face of the 
joints to help them to resist weathering. 

Bond. The arrangement of stones in adjacent courses (§ 303). 

Stretcher. A stone whose greatest dimension lies parallel to the 
Jaoe of the wall. 

Header. A stone whose greatest dimenaon liee perpendicnlw 
to the face of the wall. 

Quoin. A comer-stone. A quoin is a header for one face and a 
stretcher for the other. 

Dowels. Straight bars of iron which enter a hole in the upper 
side of one stone and also a hole in the lower side of the stone next 
aboye. 

Cramps. Bars of iron having the ends turned at right angles to 

*The deflnlttonsln this ch&pUr ve In Bccordance irith the recommendatloDH of 

'the CommlttM of tbe AmerlcAo Society of CItU Eoglneen previooBl; referred to, 

. and contocm to- Uw best pracUoe. Uiif ortDiuitel; they are not anlvenwU; adopted. 



jvGooi^le 



136 STOyE MASONRY. [CHAP. VII. 

to the body of the bar, which enter holes in the upper aide of ad- 
jacent stones. 

19fi. Defihiiiorb or Kims of Haboitbt. Stone masonry is 
olaaaified (1) according to the degree of fiuiah of the face of the 
stones, as quarry-faced, pitch-faced, pointed, hnsh-hammered, etc. ; 
(2) according to whether the horizontal joints are more or less con- 
tinuous, as range, broken range, and random; and (3) according 
to the care employed in dressing the beds and joints, as ashlar, 
squared-stone, and rubble. 

196. Quarry-faced Masonry. That in 
which the face of the stone is left as it 
comes from the quarry. Fig. 38. 

Pitch-faced Masonry. That in which 

the face edges of the beds are pitched to- 

a right line. Fig. 39. 

no. as. FM.M. Cuts/one Masonry. That in which 

the face of the stone is finished by one of the methods described in 

g§ 190-193. 

197. Range. Masonry in which a course is of the same thick- 
ness throughout. Fig. 10. 

Broken Range. Masonry in which a course is not continnous 
throughout. Fig. 41, 

Random, Masonry which is not laid in courses at all. Fig. 43. 




' l' l' l' 


^FT^ 


1 1 1 1 


V 1 1 ^ 1 h 


1 1 1 


1 '1-"— ^^i^ 


1 1 1 1 


1 1 1 


H r"-^ 



Fia. 40.— RiNOB, 



Fia. 41.— Bbokih Rimqi. Fib. 4^.— Bani 




Any one of these three terms may be employed to designate the 
coursing of either ashlar {§ 196) or sqnare-stone masonry (§ 197), 
but can not be applied to rubble (g 198). 

198. Ashlar. Cut-stone masonry, or masonry composed of any 
of the various kinds of cat-stone mentioned in g 193. According 
to the Keport of the Committee of the American Society of Civil 
Engineers, " when the dressing of the joints is snch that the dis- 
tance between the general planes of the surfaces of adjoining stones 
is one half inch or less, the masonry belongs to this class." From 



ovGoQi^lc 



DEFINITIONS OF KINDS OF UABONRT. 137 

its derivatioD ashlar apparently meaos large, square blocks; bnt 
practice seems to have made it synooymous with " cnt-stone," and 
this secondary meaning has been retained for convenience. The 
coursing of ashlar is dcBcribed by prefixing range, broken range, 
or random ; and the finiah of the face la described by prefixing th« 
name of the cat-atone (see g§ 190-93) of which the masonry la 
composed. 

Small Aaltlar. Cnt-atone masonry in which the stones are loaa 
than one foot thick. The term is not often nsed. 

Rough Ashlar, A term sometimes given to aqn&red-atone 
masonry (g 197), either qnarry-faced or pitch-faced, when laid aa 
range-work; bnt it is more logical and more expressive to call snch 
work range sqnared-stone masonry. 

Dimension Stones. Cut-stones, all of whose dimensions have 
been fixed in advance. "If the specifications for ashlar masonry 
are so written as to prescribe the dimensions to be need, it will not 
be necessary to make a new class for masonry composed of aach 
atones." 

Squared-ttone Masonry. Work in which the stones are roughly 
squared and roughly dressed on beds and joints (§ 193). The 
distinction between aqaared-stone masonry and ashlar (g 196) 
lies in the degree of closeness of the joints. According to the 
Eeport of the Committee of the American Society of Civil Engineers, 
" when the dressing on the joints is such that the distance between 
the general planes of the snrfoce of adjoining stones is one half inch 
or more, the atones properly belong to this class;" nevertheless, 
euch masonry is often classed as ashlar or cut-atone masonry. 

Rubble Masonry. Masonry composed of nnsqnared stone 
(§ 191). 

Uncoursed Rubble. Masonry composed of unsquared stones 
laid withoDt any at- 
tempt at regular 
conrsea. Pig. 43. 

Coursed Rubble. 
"Cnsqnared-stone ma- 
sonry which is leveled 
off at specified heif hta 
to an approximately 

horizontal surface. It may be speoified that the stone shall be rough- 
ly shaped with the hammer, so as to fit approximately. Fig. j14. 



ovGoQi^lc 



138 STONK UASOKHT. [CMAP. TU. 

199. QXXBBAL Rinis. Baokine gives the following roles to be 
oboerred in the bnilding of all claesee of stone musonry: 

" L Bnild the masonry, as far as possible, in a series of coarses, 
perpendionlar, or as nearly so as possible, to the direction of the 
presBOre which they have to bear; and by breaking joints avoid all 
long continnoDS joints parallel to that pressure. 

" II. Use the largest stones for the foundation course. 

" III. Lay all stones which consist of layers in such a manner that 
the principal preasure which they have to bear shall act in a direction 
perpendicular, or as nearly so as possible, to the direction of the 
layers. This is called laying the stone on its natural bed, and is of 
primary importance for strength and durability. 

" IV. Moisten the surface of dry and porous stones before bed- 
ding them, in order that the mortar may not be dried too fast and 
reduced to powder by the stone absorbing its moisture. 

"V. Fill every part of erery joint, and all spaces between the 
-gtones, with mortar, taking care at the same time that such spaces 
shall be as small as possible." 

Another and very important rule is: the rougher the stones, the 
better the mortar should be. The principal object of the mortar is 
-to equalize the pressure; and the more nearly the stones are reduced 
to closely fitting surfaces, the less important is the mortar. Not 
infrequently this rule is exactly reversed; i. e., the finer the, dressing, 
-the better the quality of the mortar nsed, 

200. A"<"»* Kabovst. For definitions of this class of masonry 
-and its subdiTision, see § 196. 

The strength of a mass of ashlar masonry depends upon the 
eize of the blocks in each course, upon the accuracy of the dressing, 
«nd upon the bond. 

In order that the stones may not be liable to be broken across, 
1L0 soft stone, such as the weaker kinds of sandstone and granular 
iimestone, should have a length greater than 3 times its depth; bnt 
~in harder materials, the length may be 4 or 5 times the depth. The 
breadth in soft materials may range from 1^ to 2 times the depth ; 
■in hard materials, it may be 3 times the depth. 

201. DrMung. The closeness with which stones fit is depend- 
ent solely upon the accuracy with which the surfaces in contact are 
wrought, or dreseed, and is of special importance in the case of 
^bed-jointa. Ifaqyiiui of the sorface projeots beyond the plans 



ovGoQi^lc 



■ AaHLAE MA80NBT. 139 

lit the chiBel-draf t, that projecting part will have to bear an nadne 
share of the pressure, the joint will gape at the edges, — conBtitnting 
what ifi called an open joint, — and the whole will be wanting in 
«tability. On the other hand, if the surface of the bed is concave, 
having been dressed down below the plane of the chisel-draft, the 
pressure is concentrated on the edges of the stone, to the risk of 
splitting them off. Such joints are said to hejluthed. They are 
more difficult of detection, after the masonry has been built, than 
open joints ; and are often executed by design, in order to give a 
nest appearance to the face of the building. Their occurrence 
must therefore be guarded against by careful inspection daring 
the progress of the stone cutting. 

Great smoothness is not desirable in the joints of ashlar masonry 
intended for strength and stability ; for a moderate degree of rough- 
ness adds at once to the resistance to displacement by sliding, and 
to the adhesion of the mortar. When the stone has been dressed 
ao that all the small ridgea and projecting points on its surface are 
reduced nearly to a plane, the pressure is distributed nearly nni- 
iormly, for the mortar serves to transmit the pressure to the small 
depressions. Each stone should first be fitted into its place dry, 
in order that any inaccnraey of figure may be discovered and cor- 
rected by the stone-cutter before it is finally laid in mortar and 
settled in its bed. 

The thickness of mortar in the joints of the very test ashlar 
masonry — for example, the United States post-office and custom- 
house buildings in the principal cities — is about i of an inch ; in 
flrat^^lass railroad masonry — for example, important bridge piera 
and abutments, and large arches — the joints are from i to f 
of an inch. No catting should be allowed after the stone 
has been set in mortar, for fear of breaking the adhesion of the 
mortar. 

A ohiBel-diatt 1 J or 2 inches wide is usually cut at each exterior 
comer. 

202. Bond. No side- joint of any course should be directly above 
a side-joint in the course below ; but the stones should overlap, or 
break joint, to an extent of from 1 to IJ times the depth of the 
coarse. This is called the bond of the masonry. The effect is that 
©ach stone is sapported by at least two stones of the course below, and 
asBists in supporting at least two stones of the course above. The 



ovGoQi^lc 



140 STONE MASONRY. [CHAP. VII. 

object Ts twofold ; first, to distribute the pressure, so that inequali- 
tieBof loud on the upper part of the structure (or of resistance at 
the fonndation) may be traasmitted to and spread over an increas- 
ing ar«a of bed in proceeding downwards (or upwards) ; and second, 
to tie the building together, t. e., to give it a sort of tenacity, both 
lengthwise and from face to back, by means of the friction of the 
stones where they overlap. 

The strongest bond is that in which each course at the face of 
the structure contains a header and a stretclier alternately, the 
enter end of each header resting on the middle of a stretcher of 
the coarse below, so that rather more than one third of the area of 
the face consists of ends of headers. This, proportion may be 
deviated from when circnmstancea require it, but in every case it 
is advisable that the ends of headers should not form less than one 
fourth of the whole area of the face of the structure. A header 
should extend entirely through the wall, and shonld be over the 
middle of the stretcher in the course below. 

A trick of masons is to use " blind-headers," or short stones that 
look like headers on the outside but do not go deeper into the wall 
than the adjacent stretchers. When a conree has been put on top 
of these, they are completely covered up ; and, if not suspected, 
the fraud will never be discovered unless the weakness of the wall 
reveals it. 

Where very great resistance to displacement of the nuksonry is 
required (as in the upper courses of bridge piers, or over openings, 
or where new masonry is joined to old, or where there is danger of 
unequal settlement), the bond is strengthened by dowels or by 
' ciamp-irons (§ 195) of, say, l^inch ronnd iron set with cement 
mortar. 

203. Baoking. Ashlar is usually backed with rubble masonry 
(g 213), which in such caaes is specified as coursed mbble. Special 
care should be taken to secure a good bond between the rubble 
backing and the ashlar facing. Two stretchers of the ashlar fac- 
ing having the same width should not be placed one immediately 
above the other. The proportion and length of the headers in 
the rubble backing should be the same as in the ashlar facing. The 
" tails " of the headers, or the parts which extend into the rubble 
backing, may be left rough at the back and sides; bat their upper 
and lower beda should be dressed to the geoeral plane of Uie bed of 



ovGoQi^lc 



ABHLA.B HASOKRT. 141 

the conrse. These " tails" may taper slightly in breadth, but should, 
not taper in depth. 

The backing should be carried up at the same time with the 
£tce-work, and in courses of the same depth; and the bed of each 
coarse sbonld be carefully built to the same plane vith that of the 
ashlar facing. The rear face of the backing should be lined to a 
jair surface. 

204. Pointing. In laying masonry of any character, whether 
with common or hydraulic mortar, the exposed edges of the joints 
will naturally be deficient in density and hardnese. The mortar in 
the joints near the surface is especially subject to dislodgment, 
since the contraction and expansion of the masonry is liable either 
to separate the stone from the masonry or to crack the mortar in 
the joint, thus permitting the entrance of rain-water, which, freezing, 
forces the mortar from the joints. Therefore it is usual, after the 
niasonry is laid, to refill the joints aa compactly as possible, to the 
depth of nt least half an inch, with mortar prepared especially for 
this purpose. This operation is called pointing. 

The very best cement mortar should be used tor pointing, as the 
best becomes dislodged all too soon. Clear Portland cement mor- 
tar is the bent, although 1 volume of cement to 1 of sand is fre- 
quently used in flrst-ckss work. The mortar, when ready for use, 
should be rather incohercntnndquitedoficiont in plasticity. ■ Before 
applying the pointing, the joint should be well cleansed by scrap- 
ing and brushing out the loose matter, and then be well moistened. 
01 course, the cleansing out of the joints can be most easily done 
while the mortar is now and soft. The depth to which the mortar 
shall bo dug out is not often specified ; it is usually cleaned out 
about half an inch deep, but should be at least an inch. In the 
Brooklyn bridge piers the joints were cleared IJ inches deep. 

The mortar is applied with a mason's trowel, and the joint well 
calked with a calking iron and hammer. In the very best work, 
the joint is also rubbed smooth with a steel polishing tool. Walla 
shontd not be allowed to dry too rapidly after pointing ; therefore, 
pointing in hot weather should be avoided. 

205. Amount of Hortar. The amount of mortar required for 
ashlar masonry varies with the size of the blocks, and also with 
the closeness of the dressing. With f- to i-ioch joints and 12- to 
^O-inch conraes, there will be about 2 cubic feet of mortar per 



ovGoQi^lc 



143 STONE UABONRT. [CHAP. Tll- 

oobic yard; with larger blocks and cloaer joints, i. e., in the best 
masonry, there will he abont 1 cubic foot of mortar per yard ot 
masonry. Laid in 1 to 2 mortar, ordinary ashlar will reqnire i to- 
i of a barrel of cement per cnbio yard of masonry. 

For the qnantities of cement and sand required for a cobic yard 
of mortar of different compositions, see page 88. 

306. When Employed. Ashlar masoniy is nsed for piers, abut- 
ments, arches, and parapets of bridges; for hydranlio worksj for 
facing-qnoins, and string courses; for the coping of inferior kinda 
of masonry and of brick work; and, in general, for works in which 
great strength and stability are required. 

207. Specifloatioas foi Aahlai. The specifications for ashlar,. 
or " fiist-daaa masonry " as employed on the railroads, are abont 
as follows ; * 

A«hlftr Bball coaatet of range pitch-faced muoniy. The stone (hftll be of 
durable quality: aad shall be fr«e from seami, poirdar crack*, diTS, flawi, or 
other Imperfections. 

All foundation counei iball be laid with ulecied, luge, flat stones not less. 

than 1 Inches In thickness, nor of leas superficial surface than fifteen (IS) 

square feet. 

The coutses shall be not less than Inches thick nor more than — 

Inches.^ The courses shall be continuous around and through the wall ; andi 
no course shall be thicker than the one below It, except that the footing- 
course may be thinner than the one next above. Stretchers shall be st least 
twice as wide as thick, and at least four times as long as thick. Headers thtiX 
be, for at least three fourths of their length, not less than twice as wide at . 
thick; tad shall extend entirely through the wall, or have a length not less 
than five times the thickness of the course. The masourr shall consist of 
headers and stretchers alternating; at least one third % of the face of the wall 
shall consist of headers. Stretchers of the some width shall not be placed 
immediately one above Ihe other ; but this shall not apply to the ends of 
stretchetB where headers come centrally between stretchers. Every header 
shall be Immediately over a stretcher of the course next below. Joints on tbo 
face of the wall shall be broken at least three quarters of the thickness of tb» 

The beds and the vertical joints for IS inches back from the face of the 
wall shall be dressed, before being brought io the wall, so as to form mortar 

■ For oompleta speciflcotions tor railroad and also other kinds of masonry, see 
Appendix I. page 529. 

t Frequently 12; sometimes 18. 

j The conrses of the olasees of moaonry referred to above nsnally range from 
U to 30 inohea ; but, of course, may vary according to the cironmstoncaa, and for 
some purposes may be as low as 10 iDohes. 

I Cntten specfflad as one fourth. 



ovGoQi^lc 



BQtJABE-STOHSD UASOHBY. IJS'- 

Joints Dot less than one quattei inch nor mora than one halt Incli In thlckneat. 
All aloaea shall be laid on tha natural bed. No part of a stone shall extend 
beyond the back edge of the under bed. All corueia and batter lines shall 
have a neat chisel-draft one and one half Inches wide on each face. The pn>- 
Jecllons of the rock- face must not exceed four inches beyond the draft-lines ; 
and In tunnel dde-walls, the projectloo must not exceed two Inches. Th* 
face-edge of the Joint shall be pitched to a straight Ibe. 

The backing shall consist of stone of the same thickness as the correspond- 
ing face stone. . When walls exceed four feet In thickness, there shall be aa 
many headers of the same size In the back of the wall as in the face, so ar- 
ranged that a header In the rear of the wall shall be between two headers in 
the front. Tbe backing shall be so laid as to leaTe no spaces between the 
■tones over six Inches wide, wblch spaces shall be filled with spalls set In 
cement mortar. Ho spalls shall be allowed In the bed joints. 

The coping shall be formed of large flat stones, which shall extend entirely 
aoroas the wall when tbe same Is not more than six feet wide. The steps of 
irlng walls shall be capped with stone covering the entire step and extending 
under the step next above at least twelve Inches. Coping and step stones shall 
be at least twelve Inches thick, and have such projections as tbe engineer may 
direct [neually 8 to 6 Inches]. The tops and faces of copings and step atones 
shall be bush- hammered, and their Joints and beds cat to one quarter Inch 
throughout. 

208. BaiTAXXB^Ton Xasohbt. For deflnitiotu of this olaai of 
maaoury and its anbdiriBioiiB, eee § 197. Tbe distinction b«tween 
Bqnarad-atoue masonry and ashlar Iie« in the d«^pree of closeness of 
the joints. According to the Report of the Committee of the 
American Society of CItU Engineers, " when the dressing on the 
joints is such that the distance between the general planes of tbe 
surfaces of adjoining stones is one half inch or more, the stones 
properly belong to this doss ; " howerer, snch masonry is frequently 
classed as ashlar or cat-stone masonry. 

Sqnared-stone masonry is asually quarry •faced, random- work, 
although pitch-faced range-work is not ancommon. Tbe qnoins 
and the sides of openings are nsiially rednoed to a rongb-smooth 
sarface with the face-hammer, the ordinary ax, or the tooth-ax. 
This work is a necessity where door or window frames are inserted; 
and it greatly improves the general effect of the wall, if nsed 
wherever a corner is tamed, 

209. Sqnared-stone masonry is distingnished, on the one hand, 
from ashlar in having less accarately dressed beds and joints, and, on 
the other band, from mbble in being more oarefnily constrncted. 
Id ordinary practice, tbe field covered by this class is not very 
definite. The speoiflcationB for " second-class masoniy" as nsed 



ovGoQi^lc 



144 STONE UASONRY. [CHAP, VII. 

on Tailroads nsnally conform to the above deBcripUon of qnarrj-faced, 
range eqnared-stone maaonrj ; but sometimes this grade of masonry 
is designated " superior rabble. '* 

210. Amoimt of Mortal Beqaired. The amount of mortar 
required for squared-stoue masonry varies with the size of the 
atones and with the quality of the masonry; as a rough average, 
one sixth to one quarter of the mass is mortar. When laid in 1 to 
3 mortar, squared-stone masonry will require ^ to J of a barrel of 
cement per cubic yard of masonry. 

'For quantities of cemeut and sand required for mortars of 
Tarions compositions, see the table on page 88. 

211. Baoking and Foiating. The statements Goncomiog the 
backing and pointing of ashlar (§§ 203 and 204) apply flubstautJally 
to Bqnared-Btone masonry. As the joints of squared-stone masonry 
are thicker than those of ashlar, the pointing should be done pro- 
portionally more carefully; while as a rule it is done much more 
careleesly. The mortar is often thrown into the joint with a 
trowel, and then trimmed top and bottom to give the appearance 
of a thinner joint. Such work ia called ribbon pointing. Trimming 
the pointing adds to the appearance bat not to the durability. 
When not trimmed it is called dashed poiating. 

212. Specifications for Squared-stone Hasonry. Squared-stone 
masonry is employed for the piers and abutmects of lighter bridges, 
for small arches, for box-cnlverts, for basement walls, etc. The 
specifications are about as follows: * 

The stones shall be of durable quality; and shall be free from seams, 
powder crackB, drjs, or other imperfections. 

The couraea shall be not lesa than 10 Incheii thlelc. 

Btretchers sfaall be at least twice as wide as thick, aod at least four times as 
long as thick. Headers shall be at least five times as long as thick, and at least 
as wide at thick. There shall be at least one header lo three atretchen. Joints 
on the face shall be broken at least 8 Inches. 

Tbe beds and verltcal joints for 8 Inches back from the face of the wall 
■hall be dressed to make jointB one half to one inch thick. Tbe front edge of 
the joint shall be pitched 10 a straight line. AU comers and balter-ltnes shall 
be bammer-dressed. 

Tbe backing shall constat of atones not less In thickness than the facing. 
At least one half of the backing shall be stone* containing 3 cubic teeU 
The backing shall be laid in full mortar beds; and the vertical Joints shall 

« Appendix]^ 



ovGoQi^lc 



BUBBLE MASONET. US 

also be filled with mortar. Tbe epacea between the Isrge itones HhaU be filled 
with spalla Bet Id mortar. 

The coping absll be fonaed of large flat Btonea of such tblckneas aa ths 
«ii|;ineer may direct, but in no case to be less than eight inches (8"). The 
upper surface of tbe coping Hliall be bush- hammered, and the Joints and I>eda 
shall be drened to one lialf an inch (1") throughout. Eacb atone must extend 
•ntireiy across llie wall when the wall is not mora than four feet (4) thick. 

213. BUBBIB HaboveT. For defiaitioDs connected with tliis 
'class of masonry, see § 198. 

The BtoneB used for rubble masonry sbonld be prepared by 
eimply knocking off all tbe weak angles of the block. It Bbonldbe 
cleansed from duat, etc., and moistened, before being placed on its 
bed. This bed is prepared by spreading over the top of the lower 
course an ample quantity of good, ordinary-tempered mortar in 
which tbe stone is firmly embedded. The vertical joints should be 
carefully filled with mortar. The interstices between the larger 
masses of stone are filled by thmsting small fragments or chippings 
■of stooe into the mortar. In heavy walla of rubble masonry, the 
precaution should be obeerred to give tbe stones the same position 
in the masonry that they had in the qoarry, t. e., to lay them on 
their "natural bed," since stone offers more resistance topressnre 
in a direction perpendicular to the quarry-bed than in any other. 
Tbe directions of the laminee in stratified stones show tbe position 
of the quarry-bed. 

To connoct the parts well together and to strengthen the weak 
points, Ihroughs or binders should be used in all tbe courses, and 
the angles should be constructed of cut or hammered stono. 

When carefully exocntod with good mortar, rubtjle possesses all the 
strength and durability required in strnctures of an ordinary char- 
acter, and is much less expensive than ashlar. The difbcnity le in 
getting it well executed. The most common defects are (1) not bring- 
ing tbe stones to an even bearing; (3) leaving large vertical openings 
between the several stones; (3) laying up a considerable height of 
tbe wall dry, with only a little mortar on the face and back, and 
then pouring mortar on the top of the wall; (4) using insuEBcient 
cement, or that of a poor qnality. The last defect is usually obviated 
by furnishing the cement to the contractor ; and the second and 
third defects may be detected by probing the vertical joints with a 
«mall steel rod. In order to seonre good mbble, great skill and 



ovGoQi^lc 



146 STONE UASO^TBT. [CRAP. TO. 

oare are required on the part of the mason, and constant vatchial- 
ness on the psrt of the inspector. 

A very stable wall can be built of rubble masonry without any 
dressing, except a draft on the quoins by which to plumb the cor- 
Dere and carry them up neatly, and a few strokes of the hammer to 
spall off any projections or surplus stone. This style of work ia 
not generally advisable, as very few mechanics can be relied upon to 
take the proper amount of care in leveling up the beda and filling 
the joints; and as a coneequence, one small stone may jar loose and 
&11 oat, resulting probably in the downfall of a considerable part of 
the wall. Some of the naturally bedded stones are so smooth and 
oniform as to need no dressing or spalling up; a wall of such stones 
ia very economical, since there is no expense of cutting and no time 
U lost in hunting for the right stone, and yet strong, massive work 
is assured. However, many of the naturally bedded stones have 
inequalities on their surfaces, and in order to keep them level in the 
course it becomes necessary to raise one comer by placing spalls or 
chips of stone under the bed, and to fill the vacant spaces well and 
full with mortar. It is just here that the disadvantage of this styl& 
of work btKJomes apparent. Unless the mason places these spalls so 
that the stone rests firmly, i. e., does not rock, it will work loose, 
particularly if the strnctnre is subject to shock, as the walls of 
cattle- guards, etc Unless these spalls are also distributed so as to 
support all parts of the stone, it is liable to be broken by the weight 
above it A few such instances in the same work may occasion con- 
siderable disaster. 

One of the tricks of masons is to put " nigger-heads" (stones 
from which the 'natural rounded surface has not l»en taken off) 
into the interior of the wall. 

214. Rubble masonry is Bomctimes laid without any mortar, as 
in slope walls (% 316), paving (§ 219), etc., in which case it iscalled 
dry rubble; but as such work is much more frequently designated 
BB slope-wall masonry and stone-paving, it is better to reserve the 
term rubble for undressed stone laid in mortar. Occasionally box 
culverts are built of the so-called dry rubble; but as such construc- 
tion is not to be commended, there is no need of a term to desig- 
nate that kind of masonry. 

21fi. Amount of Hortai Bequired. If rubble masonry is com- 
posed of small and irregular stones, about one third of the mass 



ovGoQi^lc 



BCBBLB HASOKBT. 14T 

vill consist of mortar; if the stonea are larger and more regular, 
one fifth to one quarter will be mortar. I^id in 1 to 2 mortar, 
ordinary rubble requiree from one half to one barrel of cement per 
cubic yard of masonry. 

For the amount of cement and sand required for mortar of Ta- 
rionB compositions, see the table on page 88. 

216. When Smployed. Bubble masonryof the quality described 
•bore is frequently employed for the emalleat sizes of bridge abut- 
ments, small arch culverts, box and open cnlverts, foundations of 
buildings, etc., and for backing for ashlar maaonry (§ 300). 

217. Bpecifloationi for Subble Kwonry.* The following re- 
qnirements, if properly complied with, will secure what is generally 
known among railroad engineers as superior rubble. 

Rubble masoDTy sball consist of couned rubble of good qualtty laid In 
cement mortar. No slone shall be leas than alx lnches(6") in thlckuesB, unleaa 
otherwise directed by the eagineer. No slone sbbll measure teas tbau twelve 
Inches (12") in ila least horizontal dlmeQalon, or leas than its thickness. At 
least one fourth oC the stone In the faCe aball be headers, evenly dlatrlbnted 
throughout the wall. The atones shall be roughly squared on joints, beds, and 
faces, laid so as to break Joints and in full mortar beds. All vertical spaces 
•hall be flusbed with g9od cement mortar and then be packed full with spalls. 
No spalls will be allowed In the beds. Selected atones shall be used at all 
angles, and alkali be neatly pitched to true Hues and laid on hammer-dressed 
beds; draft llnea may be required at the more prominent angles. 

The top of parapet walls, piers, and abutments shall be capped with stonea 
extending entirely acrf'BS the wall, and having a front and end projectton of 
not less than four Inches (4"). Coping stones shall be neatly squared, and laid 
with joints of less than one half Inch (f ). The steps of wlug-walls shall be 
capped with stone covering the entire step, and extending at least six Inches 
(ff) into the wall. Coping and step stones shall be roughly hammer- dressed 
on top, their outer faces pitched to true lines, and be of such thickness (not 
less than six Inches) and have sucii projections as the eti^neer may direct. 

' ' The specifications for rubble maaonry will apply to rubble maaoury laid 
■iry, except as to the use of tlie mortar (see ^ 314)." 

218. Slo?e-wail MauiteT, A slope-Wall is a thin layer of 
masonry nsed to preserve the slopes of embankments, excavations, 
canals, river banks, etc., from rain, waves, weather, etc Theusnal 
Bpecificationa are as follows: — 

The stones must reach entirely through the wall, and be not leas thao four 
inches (4") thick and twelve inches (12") long. They must be laid with brakes 
joiniK; and the joints must be as close and free from spalls as possible. 

•* For complete speciflcatlona for masom^ for varioaa purposes, see Appendix I. 



ovGoQi^lc 



148 STONE HABONRT. [CHAP, VII. 

216. 8T0>z Pati>0. Stone paving is naed for the inverts of arch 
caiverta, for protecting the lower end of archea from nndcrmining, 
and for fonndationa of box culverta and emaU archeB. It is nBoallf 
classed iw dry rubble masonry, although it is occasionally laid with 
cement mortar. The asual specifications are about as follows : 

Stone paving shall be made of flat stones from eight laches (8") to fifteen 
inches (tS"} In depth, «et on edge, closely laid sud well bedded in the soil, and 
shall present an even lop surface. 

320. BtnULP. Kiprsp is stone laid, without mortar, about the 
base of piers, abntmentB, etc, to prevent scour, and on banks to 
prevent wash. When used for the protection of piers, the stones 
are dumped in promiscuously, their size depending upon the 
material at hand and the velocity of the current; stones of 15 to 
25 cubic feet each are frequently employed. When used for the 
protection of banks, the riprap is laid by hand to a uniform thick- 
ness. 

231. STBBSaTH OF Stohe Xasohbt. The results obtained by 
testing small specimens of stone (see § 14) are useful in determine 
ing the relative strength of different kinds of stone, but are of no 
Tftlue in determining the ultimate strength of the same stone when 
-built into a masonry structure. The strength of a mass of masonry 
depends upon the strength of the stone, the size of the blocks, the 
accuracy of the dressing, the proportion of headers to stretchers, 
and the strength of the mortar. A variation in any one of these 
items may greatly change the strength of the masonry. 

The importance of the mortar as affecting the strength of 
masonry to resist direct compression is generally overlooked. The 
mortar acts as a cushion (g 13) between the blocks of stone, and if 
it has ineuflBcient strength it will be squeezed out laterally, pro- 
ducing a tensile strain in the stone; weak mortar thus causes the 
stone to fail by tension instead of by compression. No experiments 
have ever been made upon the strength of stone masonry under the 
conditions actually occurring in masonry structures, owing to the 
lack of a testing-machine of sufBcient strength. Experiments 
made upon brick piers (g 246) 12 inches square and from 2 to 10 
feet high, laid in mortar composed of 1 volume Portland cement 
and 2 sand, show that the strength per sqaare inch of the masonry 
is only about one sixth of the strength of the brick. An increase 
i.ot 50 per cent, in the strength of the brick produced no appreciable 



ovGoQi^lc 



BTRENOTH OF STONB MA80NBT. 149 

«ffect on the atrength of the masonry; bat the substitution of 
cement mortar (1 Portland and 2 sand) for lime mortar (1 lime and 
3 sand) increased the strength of the masonry 70 per cent. The 
method of failure of these piers indicates that the mortar squeezed 
out of the joints and caused the brick to fail by tension. Since the 
mortar is the weakest element, the less mortar used the stronger the 
vail; therefore the thinner the joints and the larger the blocks, the 
stronger the masonry, provided the surfaces of the stones do not 
come in contact. 

It is generally stated that the working strain or stone masonry 
should not exceed one twentieth to one tenth of the atrength of the 
stone; but it is clear, from the experiments on the brick piers re- 
ferred to above, that the strength of the masonry depends upon the 
strength of the stone only in a remote degree. Id a general way it 
may be said that the results obtained by testing small cubes may 
vary 50 per cent, from each other (or say 35 per cent, from the 
mean) owing to undetected difierences in the material, cutting, and 
manner of applying the pressure. Experiments also show that 
stones crack at about half of their ultimate crnshing strength. 
Bence, when the greatest care possible is exercised in selecting and 
bedding the stone, the safe working strength of the stone alonfr 
should not be regarded as more than three eighths of the ultimate 
strength. A further allowance, depending apon the kind of atruo- 
tnre, the quality of mortar, the closeness of the joints, etc., should 
be made to insure safety. Experiments upon even comparatively 
large monoliths give but little indication of the strength of masonry. 
The only practicable way of determining the actual strength of 
masonry is to note the loads carried by existing structures. How- 
ever, this method of investigation will ^ve only the load which doea 
not crush the masonry, since probably no structure ever failed owing 
to the crnshing of the masonry. After an extensive correspondence 
and a thorough search through engineering literature, the following 
list ia given as showing the maximum pressure to which the several 
classes of masonry have been subjected. 

223. Pressure Allowed. Early builders used much more ma^ 
sive masonry, proportional to the load to be carried, than is cus- 
tomary at present. Experience and experiments have shown that 
such great strength is unnecessary. The load on the monolithio 
piers supporting the large churches in Europe does not exceed 3tf 



ovGoQi^lc 



150 STONE MASONRY. [cHAP. V]}. 

toQs per sq. ft. (420 lbs. per eq. iD.)>* <>■' &boat one thirtieth of the 
nitiniate strength of the stone alone. The etone-arch bridge of 140 
ft, span at Pont-j-Prydd, over the Taff, in Wales, erected in 1750, 
is sapposed to have a pressure of 72 tons per sq. ft. (1,000 lbs. per 
sq. in.) on hard limestone rubble masonry laid in lime mortar.f 
Rennie subjected good hard limestone rubble in columns 4 feet 
square to 22 tons per sq. ft, (300 lbs. per sq. in-)4 "^h^ granite piers 
of the Saltaeh Bridge eustaln a pressure of 9 tons per sq. ft. (125 
lbs. per sq. in,). 

The Doaximum pressure on the granite masonry of the towers of 
the Brooklyn Bridge isabout 28( tons per sq. ft. (about 400 lbs. per 
sq. in.). The maximum pressure oq the limestone masoury of this 
bridge is about 10 tone per sq. ft (125 lbs. per sq. in,). The face 
atones ranged in cubical contents from 1( to 5 cubic yards; the 
atones of the granite backing averaged about 1^ en, yds., and of the 
'imestone about IJ- cu, yds, per piece. The mortar was 1 volume 
of Bosendale cement and 3 of sand. The stones were rough-axed, 
or pointed to ^inch bed-joints and |-inch vertical face-joints.g 
These towers are very fine examples of the mason's art. 

In the Rookery Building, Chicago, granite columns nbout S feet 
square sustain 30 tons per sq. ft. without any signs of weakness. 

In the Washington Monument, Washington, D. C, the normal 
pressure on the lower joint of the walls of the shaft is 20.2 tons 
per sq. ft, (280 lbs. per sq. in.), and the maximum pressure brought 
upon any joint under the antion of the wind is S5.4 tons per sq. ft 
(350 lbs. per sq. in.).| 

The pressure on the limestone piers of the St Louis Bridge was, 
before completion, 38 tons per sq. ft. (527 lbs. per sq. in.); and after 
completion the pressure was 19 tons per sq. ft. (273 lbs. persq. in.) 
on the piers and 15 tons per sq. ft. ( I !I8 lbs. per sq. in. ) on the abut- 
ments.^ 

The limestone masonry in the towers of the Niagara Suspension 

* Id this connection It Is convenleDt to remember that 1 ton per oqnare foot la 
•qalnJent neerl; to 14 (exaoUj 13. SB) ponnda per sqo&re incb. 
f The TeabDOKrapb, Unlvereity of IlUnols, No, 7, p. 27. 
t Proo. loat of 0. E., vol. i. p. Sll. 

I F. CoUlngwood, aaat. engineer, In Trans. Am. Boo. of C. E. 
I Bepoit of OoL T. L. Caa«j, V. S. A,, engineer Id charge. 
\ Hlatonr of St. Louis Bridge, pp. 870-71. 



ovGoQi^lc 



VBASVBEMEirC OF UABONBT. 151 

Bridge failed under 36 toos per Bq. ft., and were taken down, — ^how- 
ever, the masonry was not well executed.* 

At the South Street Bridge, Philadelphia, the pressure on the 
limestone rubble masonr; in the pnenmatic piles is 15.7 tons per 
8q. ft. (220 lbs. per sq. in. ) at the bottom and 13 tons per sq. It at 
the top. "This is uDusnall; heary, but there are do signs of weak- 
ness, "f The maximum pressure on the rubble masonry (laid in 
cement mortar) of some of the large masonry dams is from 11 to 14 
toua per sq. ft. (154 to 195 lbs. per sq. in.). The Quaker Bridge 
Dam is designed for a maximum pressure of 16f tons per sq. ft. 
(230 lbs. per sq. m.) on massire rubble masonry in best hydranlio 
cement mortar. I 

223. Bafb FramTe, In the light of the preceding examples 
it may be assumed that the safe load for the diflereat classes of 
jnasonry is about as follows, provided each is the best of its chias : 

Concrete S to 10 toni per square foot. 

Rubble lOtolO •' " 

Squared itoae, 18 to 80 " " " " 

Limeetone ashlar, .... SO to 29 " " " " 

Qnolte ashlar, 80 " " " " 

824. IfunrsncnrT or Huovbt. The method of determining 
the quantity of masonry in a strncture is frequently governed by 
trade rules or local custom, and these vary greatly with locality, 
Kasons have voluminous and arbitrary rules for the measurement 
of masonry; for example, the masons and stone-cutters of Boston 
at one time adopted a code of thirty-six complicated rules for the 
measurement of hammer-dressed granite. As an example of the 
indefinitenees and arbitrariness of all such rules, we quote the follow- 
ing, which are said to be customary in Pennsylvania : " All open- 
ings less than 3 feet vide are counted solid. All openings more 
than 3 feet wide are taken out, but 18 inches is added to the 
running measurement for every jamb built. Arches are counted 
solid from the spring of the arch, and nothing allowed for arching. 
The comers of buildings are measured twice. Pillars lees than 3 feet 
square are counted on three sides as lineal measurement, multiplied 
by the fourth side aud depth; if more than 3 feet, the two opposite 

• Trans. Am. Soc. of C. E., vol. z*U. pp. aM-LL i Ibid., tAiO-pfi. SOB-ft. 

t Slnfftatrrbv ^f*"! voL xlz. p. TB. 



ovGoQi^lc 



15S erONB MABONBT. [cHAP. VII. 

Bides are taken; to each side 18 inches for each jamb is added to- 
lineal measurement thereof; the whole multiplied by the smaller side- 
and multiplied by the depth." 

A well-eetablished custom has all the force of law, unless dua 
notice is given to the contrary. The more definite, and therefore- 
better, method ia to measure the exact solid contents of the masonry, 
and pay accordingly. In " net meaenrement" all openings are de- 
ducted; in "gross measurement" no openings are deducted. 

The quantity of masonry is usually expressed in cubic yards. 
The perch is oocasiooally employed for this purpose; but since the^ 
supposed contents of a peich vary from 16 to 25 cubic feet, the term 
is very properly falling into disuse. The contents of a maaonry 
structure are obtained by measuring to the neat lines of the design. 
If a wall ia built thicker than specified, uo allowance is made for the- 
masoQry outside of the limiting lines of the design; but if the- 
masonry does not extend to the neat lines, a deduction is made for- 
the amount it falls short. Of course a reasonable working allow- 
ance must be made when determining whether the dimensions of 
the masonry meet the specifications or not. 

Id engineering construction it is a nearly uniform custom Uy 
measure ^I masonry in cubic yards; but in architectural construc- 
tion it is customary to measure water tables, string-courses, etc., 
by the lineal foot, and window-silla, lintels, etc., by the square foot. 
In engineering, all dressed or cut-stone work, such as copings, bridge- 
seats, cornices, water-tables, etc, is paid for in cubic yards, with 
an additional price per square foot for the surfaces that are dressed, 
cut, or bush- hammered. 

226. Clasaifloation of Sailroad Sasonry. The stone masonry 
required in the construction of a railroad is usually classified about, 
as follows: first-class masonry, second-class masonry, rubble masonry 
(sometimes called third-class masonry, § 309), rubble maaonry laid 
dry (§ 214), stone paving, slope-waits, and riprap. Firet-class ma- 
sonry is equivalent to ashlar (§§ 200-7) ; thia head generally includes- 
bridge abutments and piers of the lai^r class, and arch culverts of 
greater span than 10 feet. Sometimes second-class masonry is speci- 
fied aaaquared-atone maaonry (§§308-12), and aometimes as superior 
rubble (§§ 313-17); it is used in less important structures than first- 
olass masonry. 

Frequently specifications recognize also the following classifica- 



ovGoQi^lc 



BSnUATES OF COST. Ids' 

tioD : first-class arch masonr;, second^cIasB arch masonry, first-claas 
bridge-pier masonry, seoond-class bridge-pier masonry, and pedestal 
masonrj. The quality of work thus specified is the same as for firat- 
clase and second-class masonry respectively, the only difference 
being peculiar to the form of the masonry etructare, as will be dis- 
cussed in succeeding chapters. The specifications foreacbstmctnre^ 
should give the quantities of each kind of masonry. 

For complete specifications for railroad masonry, see Appendix L 

S26. ESTDUTU OF Cora 07 HuOHBT. The following estimateB- 
of the cost of masonry, from Trautwine's Engineer's Pocket-book,* 
ftre pronounced by experts to be as accurate as such averages can 
be stated, since every item is liable to great variation. The estimates, 
are based on the assumption that a mason receives 13.50 and a 
laborer 12.00 per day of 8 hours, 

237. " <kiiarrTing.f After the preliminary expenses of purchas- 
ing the site of a good quarry, cleaning off the surface earth and 
disintegrated top rock, and providing the necessary tools, trucks, 
cranes, etc., the total net expenses tor getting out the rough stonev 
for masonry ready for delivery may be roughly estimated thus: 
Stones of such size as two men can readily lift, measured in piles, 
will cost per cubic yard from i to J the daily wages of a quarry 
laborer. Large stones, ranging from ^ to 1 cubic yard each, got out 
by blasting, from 1 to 3 daily wages per cubic yard. Larger stones, 
ranging from 1 to IJ cubic yards each, in which most of the work 
must be doue by wedges in order that the individual stones shall 
come out in tolerably regular shape and conform to stipulated dimen- 
sions, from 2 to 4 daily wages per cubic yard. The lower prices ar& 
low for sandstone, while the higher ones are high for granite. Under 
ordinary circumstances, about 1^ cubic yards of good sandstone can 
be quarried at the same cost as 1 of granite — or, in other words, 
calling the cost of granite 1, that of sandstone will be i; hence the 
means of the foregoing limits may be regarded as rather full prices 
for sandstone, rather scant for granite, and about fair for limestonfr - 
or marble. 

228. " Sreanng.t ^^ t;he first place, a liberal allowance should 
be made for waste. Even when the stone wedges out handsomely 
on all sides in large blocks of nearly the required sfaape sud siz^ 
* Publlahed by pennUslon. 
f See Note 1, Appendix II. 
t Bee Nolea 2 and 3, Appendix IL 



ovGoQi^lc 



164 BTONE HABONBT. [CHAF. VIL 

from ^ to :J of the rongb block will generally not more than cover 
waste of dressing. In moderate-sized blocks (say ayersging aboat 
^ a cubic yard each) got ont by blasting, from i to ^ will not be 
too much for atone of medium character ae 'to straight splitting. 
The last allowance is about right for well-ecabbled dressing. The 
smaller the stones the greater must be the allowance for waste. In 
large operations it becomes expedient to have the Btones dressed, as 
far as possible, at the qoany, in order to diminish the cost of tiana- 
portation, which, when the distance is great, constitutes an impor- 
tant item — especially when hj land and on common roads. 

289. " Ashlar. Average size of the stones, say 5 feet long, Z 
feet wide, and 1.4 feet thick — or two such stones to a cubic yard. 
Then, supposing the stone to be of granite or gneiss, the cost per 
«abic yard of ashlar facing will be : 

"Oetting out the stone from Uie quan; by blBstlng, allow- 
ing i for WMte In drsMliig, H cubic yards at $3.00 

per yard (4 00 

Dressing 14 aq. ft of face at 86 ceals, 4 W 

Dresaliig S2 aq. ft. of beds and Joints at 18 cents, ... fi 8S 

Net cost of the dressed stone at the quany, . . . (18 36 

' HaullDg (say 1 mfle), loading, and unloading, .... 1 20 

Mortar, ssy 40 

Laying, including scaffold, hoisting machinery, etc, . 3 00 

Net cost. tai 8ft 

Fn>flt to coDtiactor, aay 10 per cent 8 28 

Total coHt per cubic yard, |SS 14 

"Dreediig will cost more if the faces are to he rounded or 
moulded. If the stones are smaller than we have assumed, there 
will be more square feet per cubic yard to be dressed. If, in the 
foregoing case, the stones be perfectly well dressed on all sides, in- 
eluding the back, the cost per cubic yard would be increased about 
tlO; and if some of the sides be curved, as in arch stones, say 112 
or tl4; and if the blocks be carefully wedged out to given dimen- 
sions, (16 or $1S. Under these conditions the net cost of the 
dressed atone at the quarry will be (36, (31, and (35 per cubic yard, 
respectively. 

"If the stone be sandstone with good natural beds, the getting 
out may be pat at (3.00 per cubic yard. Face dressing at 36 cents 



ovGoQi^lc 



XABKBT PBI02 OF STONE. 155 

pot sq, iL, say t3.S4 per cu. yd. Beds and iointa at 13 centa per 
sq. fC, say t6.76 per cu. yd. The total cost, then, is 119.55 instead 
of t25.14 for granite, and the net cost 117.00 instead of the t21.86 
per en. yd. for granite. The total coet of large, well-scabbled, ranged 
sandBtone masonry in mortar may be taken at about tlO per ca. yd. 

230. " Rubble. With stones ayeraging about ^ cubic yard each, 
and common labor at tl per day. the coat of granite rubble, such 
as is generally used as backing for the foregoing ashlar, will be about 
as follows : 

Oetitng out the Blone from the quany by blasting, allow- 
ing i for vBst« in Bcsbbling, If cu. yds. ® $8.00, . (8 48 

Hauling 1 mile, loading and unloading 1 20 

Uortar (2 cu. ft, or 1.6 struck bushels of quicklime, and 

10 cu. ft. or 8 atruck buahela of sand or gravel, and ' 

mixing) 1 50 

Scabbliug, laying, scafloldlng, hoisting machinery, ola., a SO 

Net coat tS 88 

Profit to contractor, my 16 per cent., 1 80 

Total coat per cubic yard t9 98 

" Common rubble of small stones, the average size being snch as 
two men can handle, costs to get it out of the quarry about 80 cts, 
per yard of pile, or, to allow for waste, say $1.00. Hauling 1 mile, 
^1.00. It can be roughly scabbled and laid for tl.20 more. Mortar, 
as above, tl.50. Total net cost, t4.70; or with 15 per cent, profit, 
(5.40, at the above wages for labor." 

231. HUKXT Fbicx of Stohs. The average market qnotationa 
to builders and contractors for the year 1888 were about as follows, 
f.o.b. (free on board) at the quarry : 

Qranlte— rough (0 40 to $0 60 per cubic foot. 

Limestone — common rubble. ... 1 00 " 1 50 percublcyard. 

" good lauge rubble, . . 1 !» " 3 00 " " " 

bridge stone 08 " 10 per cubic foot 

•■ dlmennIoQ Blone. ... 25 '* 85 " " " 

copings 20 " 85 " " " 

Sandstone, OT " 1 00 percublcyard. 

232. Con OF HuontT.* IT. 8. Public Buildings. The following 
table gives the average contract price during the past few years for 
cutting the stone for the United States government buildings ;t 

• For additional data, Be« Notes 1^ Appendix n,pBgeaBU-(6. ' 

t Amerloan Anbltoot, toL zxit pp. t, 7. 



ovGoQi^lc 



BTONB KASONBT. 



[OHAP. TIL 





TABLE IB. 
Stokb for V. e 


. Public Buiu>n>as. 




KnmovBWAOE. 


Omum*. 








HID. 


Max. 


MiD. 


Mu. 


UIu. 


Hu. 


B«i» and Joints, pw aq. R. .. . 


•«s 


•"15 

OS 
76 

et 

I 10 


•«- 


-iJ 


"ll 


-^ 












































40 
50 


BO 
20 

























The following table shows the coatract price for the maBonry (^ 
the United States public boildiogB : 

TABLE 19. 
Cost or Habosbt in U. S. Public BuiLsraw. 

















Ctt. P». 




Harriabun, Va.... 

Cinclnoatn 

Denver, Col 

Pittsburgh. Pa.... 


168S 


to 90- 




1B84 
1888 
1880 




















leSG 


70 




Columbus, 






•■ " granite 


Pittsburgh. Pa. . . . 


1H8« 


SO 








" andeut-stonegmnfte, BVg. 


IRftO 


1 60 




l«tt« 


a 00 








minga. Stony Point. Mich., aandslone. .[Fort Wayne, Ind.. 


18HB 


1 B» 








Rock-face aehlar, granite, relainlag wall. . . 


MempbU, Tenn... 


1880 


1 00 


Dreaaed coping, " ■• ■■ ... 
White s«i<rBtone.-fumlihed only 








Dallas. Tei 


188.1 


85 












DouDcII Bluffs, la. 


188.1 


1 91 


•■ " " ■■ «7erag8 bid. . . . 




1885 


a 19 


" •• " limestone, lowest bid 




1H8B 


1 87 


" " '• •■ average bid 

Bock-face asblar, cut and moulded trim- 




188S 


asfr 












1884 


a 41 


Cut and moulded, Bedford limestone 


[x)ul8vi!le. Kv. . . . . 


188.-! 


2 0» 




IRW 


a M 


" " " limestone Hannibal, Mo 


18SS 


1 ea 






S 27 


■* " " groDite, superstructure.. 


Pittsburgh. Pa 


'"" 


8 0» 



jvGooi^le 



ACTUAL COST. 1S7 

233. Bailrosd Huonry. The following are the average prices 
-actually paid in the couetraction of the Cincinnati Sonthera Itail- 
road, in 1873-77:* 

Elm-claas bridge maaonty, per CO. 7d. 910 88 

Beccnd-class bridge masonry, Hieenwnt, per cu. yd., . 7 40 

Second-class bridge masonry, dry, per cu. yd., .... 7 08 

FlrsUclasa arcb masoDiy, per cu. yd 11 24 

Second-clBSs arch masonry, in eonanl, per cu. yd., ... 8 01 

Second-class arch masonry, dry, per cu. yd., 7 7S 

Brick-work in tunnels, per cu. yd., 8 SO 

Brick-work in buildings, percu. yd., 7 00 

Baz-culvert masonry, in cem&nl, per cu. yd., 4 89 

Box-culvert masonry, dry, per cu. yd. 483 

Concrete, per cu. yd., SS8 

Slopewalla, percu. yd., 4 41 

Btone paving, per cu. yd. 941 

234. Tunnel Hasonry. The following are the average priceef 
"paid in 1883-S7 on the now Croton Aqnednct tnnnel which supplies 
New York Oity with water. The mortar waB 3 Band to 1 Bosendala 
•Dement. 

1 Dimension-stone masonry (granite), 943 00 

Brick-work lining, per cu. yd., 10 14 

Brick-work backing, per cu. yd., 8 49 

Rubble masonry, lining, per cu. yd. OS 

Concrete lining, SstonetolRosendale cement, percu.yd., S 67 

Concrete lining, 6 stone to 1 Rosendale, per cu. yd., , . S IS 

Concrete backing, 8 stone to I Bosendale, per cu. yd., ,, 4 78 

Concrete backing, 6 stone to I Roaendate, per cu. yd., . 4 23 

Une- hammered face (3-cut) for cut stone, per sq.. ft., . . 84 

Rough-pointed face for cut atone, per sq. ft,, ... . 50 
Additional for all kinds of masonry laid In Portland 

cement mortar, 3 to 1, per cu. yd 1 78 

Additional for all kinds of masonry laid In Rosendale 

cement mortar, 1 to 1, per cu. yd. 1 30 

336. Bridge-pier IDuonry. The following are the details of the 
-Qost, to the contractor, of heavy flrat-class limestone masonry for 
bridge-piers erected in 1887 by a prominent contracting firm : 



ovGoQi^lc 



158 STONE HASONBT. [CHAP, VII. 

Cost of stone (puTCbased), (4 SO 

Sand ukd cemmt, S3 

FreigbC. 1 78 

Laying 140 

: Handling materi&k, 6S 

I^erricks, tools, etc 40 

Buperintendence, oflice expense, eto C8 

Total coat per cubic yard $9 M 

The followmg data coDcerning the cost in 1887 of graoitQ pien 
—two fifths cnt-fitone facing and three fifths rubble backing — are 
farniehed by the same firm. The rock was rerj hard and tough. 
Facing ,'— 

Quarrying, Including opening quarry, %Z 16 

Cutting to dimensions, 6 75 

Laying, 1 7t 

Transportation S mllM, superinteadence, and general ex- 
penses 20S 

Total cost per cubic yard tl4 81 

Backing:— 

Quarrying, $8 10 

I>ressiiig 8 eO 

Laying, , 1 7S 

Bundries, 3 05 

Total coat per cubic yard, |10 50 

The first-class limestoae masonry in the piers of the bridges 
across the Miseouri at Flattsmonth (1879-80) cost the company 
118.60 per cubic yard, exclusire of freight, engineering, expenses, 
and tools.* The cost of flret-clafls masonry in smaller piers usually 
ranges from $12 to tl4 per cabic yard. 

At Chicago in 188? the contract price for the masonry in bridge 
piers and abutments was about as follows : OoDcrate, 1 Portland 
cement, 3 sand, 6 broken stone, 19.00 per en. yd.; oonorete, 1 
natural cement, 3 sand, 5 broken stone, 16.00 per on. yd.; atone 
facing and coping, $30.00 per cu. yd. 

236. Aroh-cnlvert Xasoury. The following are the details of 
the cost of the sandstone arch culvert (613 cu. yds.) at Ifichols 
Hollow, on the Indianapolis, Decatur and Springfield Bailroad, 

* Report ot tlie Chief Engineer, Geo. S. UotIboil 



D.qitizeabyG00l^lc 



ACTUAL COST. 



16t 



bnilt in 1887. Scale of wages per day of 10 hours — foreman, 
♦3.50 ; cutters, $3.00 ; mortar mixer, $1.50 ; laborer, $1.25 ; water- 
boy, 50 cents ; carpenters, $2.50. f 

TABLE 17. 





d® 






Oo«r. 




TowL 


cS^d. 


XatarieUi.— 

Stone— 618 CO. yda. of sandstone 
Oemect— 180 bbls. QeniiMi Portlao 

40 •■ English " 
80 " Louisville " 






$B1BS0 
88 so 




»8 17 = 
96 = 


$419 SO 
ISO 00 
iS8 78 


M 














$1.S39 25 

$1,870 48 
11 00 
SS89 
11 75 


$3 BO 


aaUTis.— 






etc. 






























$1,445 62 


$2 8$ 




$884 87 
4SS 66 
121 72 
11 76 
87 60 
14 68 
7 70 


$0 OS 












































$1082 08 














Prntty 


$80 00 


$0 05 


QXAMD Total : 


$2,507 60 
1,S2S 26 




















$4,086 85 









338. Summary of Cost The following table, compiled from a 
large amoant of data, will be coDvenient for hasty reference. Of 
course any Bach table must be used with caution, since such items 
are subject to great Tariation. 

t D*tB tnrnlBlied by Edwin A. HfU, chief 



ovGoQi^lc 



BTONB HASOKBT. 
TABLE 18. 

BuiOtABT OF COBT OF UaBONBT. 



Arch maaonrr, flnt-ckas 

Arcb masonry, second-class (In cement). 

BoK-culTert masonrv, In cement 

Brick masonry (see g 258) 

Bridge masonry, flrsl-clasa 

Bridge masonry, aecoud-claaa 0" cement^ 

Concrete 

Coping 

Dimenslon-BtoDe masoniy, granite 

Paving 

Slope- wall masoniy 

Stjuared-atone maaoniy 

Riprap , 

Rubble, flrat-claaB 

Hubble, eecond-clam On oerooDt) 




ovGoQi^lc 



- / 

CHAPTEE VIII. 
/ 

BRICK MA80NBT. 

239. HdSTAX. Lime mortar is generally employed for brick 
tnaeonry, porticalarly JQ architectural coiiBtructioDs. Many of the 
leading railroads lay all brick masonry in cement mortar, and the 
practice ehoald be followed more generally. The weakest part of 
a brick structure is the mortar. The primary purpose of the 
mortar is to form an adheBive substance between the bricks ; the 
second is to form a cushion to distribute the pressure uniformly 
over the snrfoce. If the mortar is weaker than the brick, the 
ability of the masonry to resist direct compression is thereby ooa- 
siderably reduced. For the reason, see g 13; for the amoont, see 
the Table 19, page 164. 

If the strains upon a wall were only those arising from a direct 
pressure, the strength of the mortar would in most cases be of 
comparatively little importance, for the crushing strength of aver- 
age quality mortar is far higher than the dead load which under 
ordinary circumstances is put upon a wall ; but, as a matter of fact, 
in buildings the load is rarely that of a direct crushing weight, 
other and more important strains being developed by the system of 
oonstruction. Thus the roof tends to throw the walls out, the raft«ra 
being generally so arranged as to produce a couBiderable outward 
thrust against the wall. The action of the wind also produces a side 
«train which is practically of more importance than either of the 
others. In many eases the contents of a building exert an outward 
thrust upon the walls ; for example, barrels piled against the sides 
of a warehouse produce an outward pressure against the walls. 

In many brick coostructioue the use of cement mortar is abso- 
lutely necessary — as, for example, in tall chimneys, where the bear- 
ing is BO small that great strength of the cementing material is 
required. 

240. The thickness of the mortar-joints should be about i to f 
of an inch. Thicker joints are very common, but should be avoided. 
Jf the bricks are even fairly good, the mortar is the weaker part of 



ovGoQi^lc 



16S BRICK UASON&T. [CHAF. Till. 

the vail ; hence the less mortar the better. Beeides, a thin layer 
of mortar is stronger nnder compreesion than a thick one (see § 16). 
The joints should be ae thin as is consistent with their insuring a uni- 
form bearingandallowingrapidwork in spreading the mortar. Tha 
joints of outside valk should be thin in order to decrease the dis- 
integration by weathering. The joints of inside walls are usoally 
made from f to ^ inch thick. 

Brick should not be merely laid, but every one should be rubbed 
and pressed down in such a manner as to f orca the mortar into the 
pores of the bricks and produce the maximum adhesion ; with quick- 
setting cement this is still more important than with lime mortar. 
For the beet work it ia specified that the brick shall be laid with a 
" shove joint ;" that is, Uiat the brick shall first be laid so as tO' 
project over the one belowj and be pressed into the mortar, and 
then be shoved into its final position. 

Lime mortar is liable to work out of the joints, owing to the 
action of the elements and to changes of tomperatnre. Henoe it 
ia customary either (1) to lay the face in mortar containing more 
lime than that used for the interior, or (2) to lay the 
face in a mortar containing more or less cement, or 
I (3), in rare cases, to point the joints with neat cement 
'' mortar. Whatever thekindof mortarneed,thefiniBh 
of the face of the joint is important The most 
Pm 47 durable joint is finished as shown in Fig. 47, although, 
onfortnnately for durability, it is customary to make the slope in 
the opposite direction. 

841. Since brick have great avidity for water. It is best to 
dampen them before laying. If the mortar is stlfi and the brick 
dry, the latter absorb the water so rapidly that the mortar does 
not set properly, and will cmmble in the fingers when dry. Neglect 
in this particular ia the cause of most of the failures of brick-work. 
Since an excess of water in the brick can do no harm, it is best te 
thoroughly drench them with water before laying. lime mortar ia 
sometimes made very thin, so that the brick will not absorb all tho 
water. This process interferes with the setting of the mortar, and 
partionlarly with the adhesion of the mortar to the brick. Watery 
mortar also contracts excessively in drying (if it ever does dry), 
which causes undue settlement and, possibly, cracks or distortion. 
Wetting the brick before Uying will also remove the duat from tha 
surface, which otherwise would prevent perfect adhesion. 




ovGoQi^lc 




BOKD. 163 

243. Son>. The bricke used in a given wall being of uniform 
size are laid according to a uniform syBtem, which ie called the bond 
of the brick-work. As in ashlar masonry, bo in brick-work, a header 
is a brick whose length lies perpendicuW to the face of the wall; 
and a atretcher is one whose length lies parallel with the face. 
Brick should be made of sach a size that two headers and a mortar- 
joint will ooonpy the same length as a stretcher. 

S4S. Bn g liri i Bond. This consists in laying entire oonrses of 

headers and stretchers, which some- 

; times alternate, as in Fig. 48; but 
■_ generally only one course of headers 
IS laid for every two, three, four, etc., 
1 courses of stretchers. Id ordiziary 
' practice the custom is to lay four to six 
courses of stretchers to one of head- 
ers. The stretchers bind the walls 
together lengthwise ; the headers, crosswise. The proportionato 
numbers of the courses of headers and stretchers should depend on 
the relative importance of transverse and longitudinal strength. 
The proportion of one course of headers to two of stretchers is that 
which gives equal tenacity to the waU lengthwise and crosswise. 

In building brick-work in English bond, it is to be borne in 
mind that there are twice as many vertical or side joints in a course 
of headers as there are in a course of stretchers ; and that anless 
in laying the headers great care be taken to make these Joints very 
thin, two headers will occupy a little more space than one stretcher, 
and the correct breaking of the joints — exactly a quarter of a brick-^ 
will be lost. This is often the case in carelessly built brick.work, in 
which at intervals vertical joints are seen nearly or exactly above 
each other in successive courses. 

344. ^smiih Bond. This consists of a header and a stretcher 
alternately in each course, so placed 
that the outer end of each header '- ~ 
lies on the middle of a stretcher in ' 
the course below (Fig. 49). The : 
number of vertical joints in each : 
conrse is the same, eo that there is no 
risk of the correct breaking of the 
joints by a quarter of a brick being ^°- "-'^'•- »»"■ 

lost; and the wall presents a neater appearance than one built in 




ovGoQi^lc 



164 BKICK HA50NBY. [CHAP. TIIL 

English bond. The Utter, however, vhen correctly built, ia 
stronger and more stable than Flemish bond. 

24S, Hoop-iron Bond. Pieces of hoop-iron are frequently laid 
Sat in the bed-joints of brick-vork t« increase its longitudinal 
tenacity, abont % inches of the ends of each piece being bent down 
and inserted into the vertical joints. Although thin strips of iron 
are generally employed, it would be better to use thicker pieces ; the 
value of the iron for this purpose depends wholly upon the rigidity 
of the ends which are turned down, and this will vary about as 
the square of the thickness. The strip of iron should bo nearly 
«s thick as the mortar-joint. This means of strengthening masonry 
is frequently employed over openings and to connect interior brick 
vails with stone fronts. 

346. C01I7BE88ITB 8TBEV0TK (U BEICK KASOITBT. Experi- 
4nents at Watertown, Uass., with the United States testing-machine, 
upon piers 1% inches square and from 1 ft 4 in. to 10 ft high, gave 
cesalts as follows :* 



TABLE 10. 
Stsenoth or Bbice Mabodht GoicrA.BED vtith that c 

TBS UORTAB. 



' THE BniCK AKD 



i 




1 


r 


1 


or TB« 8T«i(IOIH 


il 

8 




MlB, 


BUz 


HtAO. 






16 

1 

I 
1 
8 


1,908 
l.MB 

1,411 

1.972 
a. 544 


134 

18S 

103 
163 
545 
631 
8,4BS 


.06 


.18 


.10 

.11 

.00 
.13 

.17 




X 


2 mortar (1 lime. 8 sand), 1 Rosea- 




8 


2 mortar (1 lime, 8 saod}, 1 Port- 










1 Hosendale cement. 2 sand 

1 PorllnDd cement, 3 sand 








fi 


.10 


.37 


4.7 






1 


3,875 























"Tests of Uetala, el 



" for the fear ending Jane SO, 1S84, pp. 6e~tai, 



ovGoQi^lc 



COMPRESSIVE STBEIfGTH. 165 

The brick had an average strength of nearly 15,000 lbs. per sq. 
in., tested flatwise between steel. The mortar was 14} months old 
when it was tested. The piers were built by a common mason, with 
only ordinary care; and they were from a year and a half to two 
years old when tested. Their strength varied with their height; 
and in a general way the experiments show that the strength of a 
prism 10 ft. high, laid in either lime or cement mortar, is about two 
thirds that of a 1-foot cube. A deduction derived from so few 
experiments (22 in all) is not, however, conclosive. The dlflerent 
lengths of the piers tested occurred in about equal numberB. The 
piers began to show cracks at one half to two thirds of their ultimate 
strength. 

In attempting to draw conclnsions from any experiments, it 
must be borne in mind continually that the result of a single trial 
may possibly be greatly in error. In this case this pi'ecaution is 
very important, since the difference between experiments apparently 
exactly alike was in some cases as much as 50 per cent. A great 
variation in the results is charaoteristic of all experiments on stone, 
brick, mortar, etc. Except on the ground of a variation in ex- 
periments, it is difBcult to explain why mortar No. 4 ia weaker than 
No. 2, while the masonry is stronger ; or why the masonry of No. 5 
is stronger than that of No. 7. 

Of course the apparent efficiency of the masonry, as given in th© 
table, depends upon the manner in which the strengths of the 
brick and mortar were determined, as well as upon the method of 
testing the masonry. For example, if the brick had been tested on 
end the apparent efficiency of* the masonry would have been con- 
siderably more ; or if the mortar had been tested in thin sheets tho 
strength of the masonry relative to that of the mortar would not 
have been so great. • 

247. Some German experimentsf gave results as in the table 
*It Hhnuld be mentlotiedihat the mortar with which these piera were built appears 
. to be mncb wealier than similar monar under lllie Ronditlotis. (Compare page 79, 
and pages 136. 166, 188, 1B7 of the Keportof Tostsof MotaU, etcmade at Watertown 
tn 1864.) Ordinarily, mortar is eight to- ten times as strong In compreeslon as in 
tension, whereas the flret six mortars In the preceding tabic were but little atrimger 
In compression Uian snch mortar should have been in tension. The officer in charge 
IS " ucubte to offer an; eiplBoatioQ. The cemeut was bought on the market ; the 
maker's name is not known. The cement was not tested." However, the raperi- 
ments are consistent with themBclres, and therefore show relative atretigths correctly. 

t Van Nostrand'B Engin'g Hag., vol. zzxlv. p. iHO, trom the Abstraota of tba 
Inst, of C. E. (London), vol. 79, p, 378. 



ovGoQi^lc 



166 BEICK KA80NBY. [CHAP, VIIL 

below. It is not stated how the strength of the. brick or of the 
masonry was determined.* The term cemeot refers to Portland 
cement. According to the building regalations of Berlin, the safe 
]oad for brick masonry is one tenth of the results in the table. 
TABLE SO. 
RxLATiTB Strenoth OF Bkick ASS Bbice Habohbt. 





IAtuuoi Cbohh- 
DT Buck, im vaa. 




KmovBBioK. 


iffiS: 


TUme, 


SBwd. 


1 Cement, 
IBand. 




6,8«0 
8.6«0 
a,9S0 
3,769 
2,617 
1,190 


8.870 
1.620 
l.SSO 
1,210 
1,150 
680 


2.S0O 
1,760 
1.890 
1,820 
1,260 
570 


2.B60 
2,020 
1.610 
1,620 

1.440 
650 


8.410 






, 1.860 

1,710 

1,060 

760 






Porous perforatod 



Table 1!) shows conclnsiTely that the strength of brick masonry 
is mainly dependent upon the strength of the mortar. An in- 
crease of 50 per cent, in the strength of the brick shows no 
appreciable effect on the strength of the masonry. Notice, 
however, that the masonry in the fifth line of Table 19 is 70 per 
cent, stronger than that in the first, dne to the difference between 
a good Portland cement mortar and the ordinary lime mortar. 
In Table 20 notice that brick laid in a 1 to 3 Portland cement 
mortar is nearly 50 per cent, stronger than in a 1 to 2 lime 
mortar. Similar experiments f show that masonry laid in mortar 
composed of 1 part Bosendale cement and 2 parts aaitd is 56 
per cent, stronger than when laid in mortar composed of 1 part 
lime and i parts sand. A member of the Institute of CiTil Engi- 
neers (London) says^ that brick-work laid in lime is only one fourth 
as strong as when laid in clear Portland cement. Probably the dif- 
ference in durability between cement mortar and lime mortar is 
considerably greater than their difference in strength. 

* U the Btrength ot the brick (In edj line of ths table) be repreeented by 100, that 
ol lbs tUBaoDrf iB 44, 48, 66, uid 63, reapectlrelr, which BhowB that the Talnee In Um 
table were not derived direOtj/ from experiments. 

f Report of BiperlTneDts on Building MatedaU for the Clt; of Philadelphia with 
the U. S. tABttng-machlue at Watertown. Mbbb., pp. 82, S3. 

t Pioc Inat. of C. E., vol. ztU. p. 441. 



ovGoQi^lc 



TBAKBTEBBE 8TB&N0TH. 167 

246. Preainie allowed in Practice. The pressure at the base of 
a brick ehot-tower in Baltimore, 246 feet high, is estimated at 6^ 
tons per sq. ft. (abont 90 lbs. per sq. in.). The preseure at the base 
of a brick chimney at Qlasgow, Scotland, 468 ft. high, is estimated 
at 9 tons per sq. ft. (about 125 Ibe. per sq. in.); and in heavy galra 
this is'increased to 15 tone per sq. ft. (310 lbs. per sq. in.) on the 
leeward aide. The leading Chicago architects allow 10 tons per sq. 
ft. (140 lbs. per sq. in.) on the best brick-work laid in 1 to 2 Port- 
land cement mortar ; 8 tons for good brick-work in 1 to 2 Rosendale 
vement mortar ; and 5 tons for ordinary brick-work in lime mortar. 
Ordinary brick piers have been known to bear 40 tons per sq. ft 
(660 lbs. per sq. in.) for sereral days without any sign of failure. 

Tables 19 and 20 appear to show that present practice is very 
wanBervatiVB with regard to the pressure allowed on brick masonry. 
According to Table 19 (page 164), the ultimate strength of the best 
fonck laid in ordinary lime mortar is 110 tons per sq. ft. ; if laid 
in 1 to 3 Portland cement mortar, 180 tons ; and by Table 30 (page 
166) the strength of ordinary brick in 1 to 2 lime mortar is 100 tons 
per sq. ft., and in 1 to 3 Portland cement mortar 140 to-n?. Prom 
the above, it wonld seem that reasonably good brick laid in good 
lime mortar should be safe under a preseure of 20 tons per sq. ft., 
and that the best brick in good Portland cement mortar should be safe 
under 30 tons per sq. ft. The nominal pressure allowed upon brick 
mosoniy depends upon the kind of materials employed ; the degree of 
care with which it is executed ; whether it is for a temporary or per- 
manent, an important or unimportant structure ; and, it may be 
added, the care with which the nominal maximum load is estimated. 

249. TUHSTZBai StbevoTE 07 Bbios KasOHKT. Masonry is 
seldom employed where any strain except direct compression will 
come upon it, but sometimes it is subject to transverse strain. The 
transverse strength of brick-work depends theoretically upon the 
tensile strength of the brick and upon the adhesion and cohesion 
of the mortar, but practiwilly the strength of the mortar deter- 
mines the strength of the masonry. For example, in the case of 
a high wall whose upper portion is overthrown by a lateral force or 
pressure of any kind, the failure is due either (1) to the breaking of 
the adhesion in the bed -joints and of the cohesion of the side-joints, 
or (2) to the rupture of the mortar in the bed-jointa alone. The 
latter method of failare, however, is improbable, since the cohesion 



ovGoQi^lc 



168 BBICK UASOKBT. [OHAP. TIIL 

of cement mortars is alwajs much greater than their adbeeion (com- 
pare §§ 134 and 137); and hence, in estimating the resistance of the 
wall to OTertnrning, it becomes necessary to fix valnea for hotli the 
coheaive and adhesive strength of the mortar at the time when the 
stracture is first esposed to the action of the lateral force or pres- 
sare, and also to ascertain the relative areas of beds and side-jointa- 
in the assumed section of ruptore. In good brick-work the aggro* 
gate area of the side-joints, in any section parallel to the beds, will 
amount to abont one seventh of the total area of such section. 
Hence, when the masonry is liable to be subjected to transverse- 
strains the adhesive strength of the mortar is more important than 
its cohesive strength. 

The adhesion of mortar to brick or stone has already been dis- 
cussed (§ 137). While the experiments uniformly show a relatively 
- low adhesive power, it is well known that when old walls are de- 
molished the adhesion of even common lime mortar is found to be 
very considerable. Although the adhesive power of mortu may be 
small as compared with its tensile strength, good brick masonry has. 
a considerable transverse strength. 

Experiments made under the author's direction * indicate that 
bnck beams bonded aa regular masonry have a modulns of rapture 
equal to about twice the tensile strength of the mortar when built 
with ordinary care, and about three times when built with great care. 
When the b^ms are ooDstmcted as piers, i. 0., with no interlocking 
action, the modolns of mptnre is about equal to the tensile strength. 
of the mortar. 

260. Application. To illustrate the practical application of thft 
fact that brick-work has a transverse strength, let it be reqnired to 
compute the strain w|iich may come upon a lintel, or girder used 
to support a brick wall over an opening.f 

Let H = the height, in feet, of the wall above the opening ; 
J7h = the height, in feet, of the wall that produces a maxi- 
mum strain on the lintel ; 
H, = the height, in feet, of the masonry when it will just 
support itself over the opening ; 
- ' S= the span, in feet ; 

t = the thickness, in feet, of the wall ; 

* Tek Twcssoa&^PB, CnlTerslt; ot lUlnois, No. T (IBSa-eS), pp. 3&-ST. 

f The principloottbe following compnt&tionB Is from hnoHtoiiBliaJBngiaemitif 
(IiODdou), vol. xlT. pp. Uand T2. 



ovGoQi^lc 



TRAN6TEBSE STRENQTH. 169 

S = the modulus of rupture, in pounds per square inoh, 
of the brick-work ; 

W= the weight, in ponnds, of a cubic foot of the wall. 
H' varies from 100 to 140 pounds, and for conven- 
ience is here assumed to be 144 ; the error is always 
on the safe side ; 

3f, = the bending moment on the lintel, in pounds per 
square inch. 

Oonsider the masonry as a beam fixed at both ends and 
loaded uniformly. Then, by the prinoiples of the resistance of 
materials, when the masonry is just self-supporting, one twelfth 
of the weight of the wall above the opening multiplied by tb» 
span is equal to one sixth of the tensile strength multiplied btf 
tbe thickneBS and also the sqnare of the depth of the wall. 
The weight of tbe wall above the opening is WSH.t. Hence 



■^{WSS,t)8=\{UiR)tH:, (1) 



(3> 



Notice that ' the weight of the wall over any given opening 
increases as tbe height, while tbe resistance increases as the 
square of. the height. The height for which the masonry is 
self supporting is given by equation (2), for a height greater 
than H, the masonry would be more than self-supporting ; and 
for a height less than H, the masonry would need extraneous 
support. 

To find tbe height of the wall producing a maximum stress 
in tbe lintel, notice that the bending moment on ttie lintel is 
equal to the moment of the load mimia tbe moment of the 
resistance of tbe brick-work over the opening ; or, in algebruo 
language, 

Mt = ij(WSffl) S - i{U4R)t3\ 



ovGoQi^lc 



170 BBICS ItASONBT. [CHAP. TIIL 

DiflenntUtiiig the above ^oatioQ, regarding Mi and H as the 
Tariables, and finding the raazimam ralae of ^ in the lunal ira;, 
veget 



The foot that the Talne of H^ in equation (3) ia one half of 
that of H, in eqnation (3), shows that the maximam stress on 
the lintel occurs when the height of the wall is half of its self- 
BUpporting height, at which time one half of the wall will he 
self-snp porting and one half will reqnirs eztraneons support. 
Or, in other words, the greatest stress on a lintel dne to a wall 
of any height will not be greater than that dne to a distributed 
load of 

iH^fi.S/ = iir^S'( = nearly 18 — pounds. . (4) 

SSI. Example^. To apply the above formula, aesume that it ia 
proposed tc out a 10-foot opening through an old brick wall, and 
that it is desirable to know whether the brick-work will be eelf-eup- 
porting, the wall rising 40 feet above the top of the opening. Sub- 
stituting the above data in equation (2) gives 

40 = -^-; or « = 1.25 lbs. per eq. In. 

Hence, to be self-supporting across the opening, the wall must be 
-capable of supporting a tensile strain of 1.25 pounds per sqnare 
inch. It would be poor lime mortar that would not bear eight or 
ten times this. Notice that if the wall were only 4 feet high over 
the opening, instead of 40 feet, as above, the strength required 
would be 13.5 pounds per square inch. 

For another illustration, assume that a brick wall 1 foot thick 
is to be built over a 10-foot opening, and that we wish to know 
whether a timber 10 inchee deep and 12 inches wide will sustain the 
-load. Assuming the beam as being fixed at the ends, the timber 
will sustain a uniformly distributed load of 10 tone with a deflection 
of one twelfth of an inch. This is equivalent to the entire weight 
-of the wall when 14 feet high. If the wall is to be carried higher 



ovGoQi^lc 



TBANBTERSE STBEyflTS. 171 

-th&Q thia, the girder most be supported temporarily, or time must 
be givea for the mortar to set ' 

However, before the wall ia 14 leet abOTe the openiDg, the brick- 
vork at the bottom will hare attained some strength, and therefore 
the load on the girder will not be as great as above. The average 
strength of the brick-work will always be at least the average between 
the strength at the top and the bottom ; that ie, the average strength 
will always be more than half of that at the bottom. Since 10 tone 
is the maximum load allowed on the girder, and since the maximum 
load which comes upon it is half of the entire weight of the masonry 
above the opening,* the timber will receive its maximum load when 
the wall is twice 14 feet, or 28 feet, above the opening. The masonry 
m^y be ran up 28 feet without necessitating any extraneous support 
for the lintel, provided time enough is allowed for the mortar to 
develop the average tensile strength found by substituting in (4) 
the maximum load allowed on the girder, and solving for R. Mak- 
ing thia Bubetitution gives 



, from which R = 0.90 lb. per sq. in. 

With an averse strength of 0.90 lb. per eq. id., the wall will 
'become self-supporting when 55 feet above the opening. 

252. Custom differs as to the manner of estimating the pressure 
•on a girder due to a superincumbent mass of masonry. One extreme 
consists in assuming the masonry to be a fluid, and taking the load 
on the lintel as the weight of all the masonry above the opening. 
The opposite extreme consists in assuming the pressure to be the 
weight of the masonry included in a triangle of which the open- 
ing is the base and whose sides make 45° with this line. Both of 
these methods diSer materially from the one diacussed above ; and 
neither is defensible. As the wall is several days in building, the 
masonry first laid attains considerable strength before the wall is 
-completed; and hence, owing to the cohesion of the tnortar, the final 
weight on the girder can not be equal to or compared with any fluid 
Tolume. 

The principle involved in the second method would be applicable 

* Bm diocuHloD o( equation (8), above. 



ovGoQi^fc 



178 BRICK MA80NET. [OHAP. TIIL 

to a wall composed wholly of perfectly smooth bricks. In a dry 
vail, the angle which the aide linee make with the base would 
depend upon the bond and upon the relative length and breadth of 
the bricks. Assaming the boundary lines to make an angle of 45° 
2^ 
S 

which takee account of the transverse strength of the masonry, i. e., 
the frictional and tensile resistance of the walL If S is relatively 
large and S is email, this fraction will be more than nnity, under 
which conditions the second method is safe. But if R is small and 
ti is large, then this fraction is lesa than one, which shows that 
under these conditions the second method is unsafe. 

The method of § 250 is quite simple and perfectly general. The 
substantial correctness of this method, illustrated in § 351, is 
proven by the fact that large openings are frequently cut through 
walls without providing any extraneous support; and also by the 
fact that walls are frequently supported over openings on timbers 
entirely inadequate to carry the load if the masonry did not have 
considerable strength as a beam. The discussion in § 351 also makes 
clear why frequently a temporary support is sufficient. After the 
masonry has been laid a short time, the strength of the mortar 
causes it to act as a beam. The discussion also shows the advantage 
of using cement mortar (or better, quick-settiog cement mortar) 
when it is desired that the masonry shall early become self'Sup> 
porting. 

253. Keasubekeitt of Bkick-wobk. The method of determin- 
ing the quantity of brick masonry is governed by voluminous trade 
rules or by local customs, which are even more arbitrary than those 
for stone masonry (§ 224, which see). 

The quantity is often computed in perches, but there is no uni- 
formity of understanding as to the contents of a perch. It ranges 
Irom 16} to 25 cubic feet. 

Brick-work is also often measured by the square rod of exterior 
surface. Ko wall is reckoned as being less than a brick and a half 
in thickness (13 or 13} inches), and if thicker the measurement is 
still expressed in square rods of this standard thickness. Unfor- 
tunately the dimensions adopted for a square rod are variable, the 
following values being more or less customary : 16} feet square or 



ovGoQi^lc 



DATA FOB ESTIMATES. 173 

272^ square feet, 18 feet Bqnare or 334 square feet, and 16^ square 
feet. 

The volume of a brick is scmetimes used as a unit in stating the 
contents of a wail. The conteata of the wall ore found by multi- 
plying the number of cubic feet in the wall by the number of brick 
vhioh it is assumed make a cubic foot ; but as the dimenBions of 
brick vary greatly (see § G2), this method is objectionable. A cubic 
ioot is often aesnraed to contain 20 brick, and a cubic yard 600. 
The last two quantities are frequently used interchangeably, although 
the assumed volume of the cubic yard is thirty times that of the 
cubic foot 

Brick-work is also sometimes measured by allowing a certain 
number of brick to each superficial foot, the number varying with 
ihe thickness of the wall. A 4-inch wall (thickness = width of one 
brick) is frequently assumed to contain 7 bricks per sq. ft. ; a 9-inch 
wall (thicknesB = width of two bricks), 14 bricks per aq, ft; a 13- 
inch wall (thickness = width of three bricks), 21> bricks per sq. ft, 
etc; the number of brick per eqoare foot of the face of the wall 
being seven times the thickness of the wall in terms of the width of 
a brick. 

264. The only relief from such arbitrary, uncertain, and indefi- 
nite customs is to specify that the masonry will be paid for by the 
cubic yard, — gross or net measurement, according to the structure 
or the preference of the engineer or architect 

In engineering the uniform custom is to measure the exact solid 
contents of the wall. 

265. Data roB Estduteb. ITumber of Brick Required. Since 
the size of brick varies greatly (g 62), it is impossible to state a rule 
which shall be equally accurate in all localities. If the brick be of 
standard size (8^x4x2^ inches), and laid with i- to |-inch joints, 
a cubic yard of masonry will require about 410 brick ; or a thousand 
brick will lay about 2i cubic yards. It the joints are i- to |-inch, a 
cubic yard of masonry will require about 495 brick; or a thousand 
brick will layabout 2 cubic yards. With face brick (8} x H X H 
inches) and ^-inch joints, a cnbic yard of masonry will require about 
496 brick; or a thousand face brick will lay about 2 cubic yards. 

In making estimatea for the number of bricks required, an al- 
lowance must be made for breakage, and for waste in cutting bnck 
to fit angles, eto. With good brick, in mataiva work this allowance 



ovGoQi^lc 



174 BBIOE HASOKST. [OHAP. TUt. 

need not exceed lotZ per ceot.; but in buildings 3 to 5 per cent, 
is uone too much. 

266. Amonnt of Kortsr Eeqaiz«d. The proportion of mortar 
to brick irill Tsry with tbe size of the brick and with the thickness 
oi the joints. With the standard size of brick (8^x4x2^ inches), 
« cubic yard of masonry, laid with i- to J-inch joints, will require 
from 0.35 to 0.40 of a cnbic yard of mortar; or a thousand brick 
will require 0.80 to 0.90 of a cubic yard. If the joints are ^ to | 
inch, a cubic yard of masonry will require from 0.25 to 0,30 of a 
cubic yard of mortar; or a thousand brick will require from 0.45 to 
0.55 of a cubic yard. If the joints are i of an inch, a cubic yard of 
masonry will require from 0.10 to 0.15 of a cubic yard of mortar; 
or a thousand brick will require from 0.15 to 0.20 of a cubic yard. 

With the above data, and the table on page 88, the amount of 
oemmt and sand required for a specified number of brick, or for a 
given number of yards of masonry, can readily be determined. 

257. Labor Bequlred. " A bricklayer, with a laborer to keep him 
supplied with materials, will lay on an average, in common house> 
walls, about 1,500 bricks per day of 10 working hours; in the neater 
outer faces of brick buildings, from 1,000 to 1,200; in good ordinary 
street fronts, from 800 to 1,000 ; and in the very finest lower-stoiy 
&ces used in street fronts, from 160 to 300 according to the number 
of angles, etc. In phiin massive engineering work, he should aver- 
age about 2,000 bricks per day, or 4 cu. yds. of masonry ; and in 
large arches, about 1,500, or 3 cu. yds." • 

In the United States Government buildings the labor per thou- 
sand, inclading took, etc, is estimated at seven eighths of the wages 
for ten hours of mason and helper. 

Table 21, opposite, f gives the actual labor, per cubio yard, re- 
quired on some large and important jobs. 

268, Cmt. In the conatfnction of the Cincinnati Southern R B., 
during 1873-77, the brick lining of tunnels oost tS. 50 per co. yd. ; 
brick-work in buildings, t7.00.t The average price paid for the 
brick-work in the new Croton Aqueduct tunnel, which supplies New 
York Oity with water, was, including everything, (10.14 per cu. yd. 



*Tnntwlne'B Englneer'a Pooket-Book, p. DTI. 

t Trans. Am. 8oc. ot C. £. 

tBeportof tbe CbleC Engineer, Dec. 1, I3TT, Exhibit & 



ovGoQi^lc 



BPEC1PICATI0N8. 

TABLE 31. 
liAxat BBQCntzD voR Brick MABonrt 



LouTKur AKo DMcnumoH o* nm HMomr. 


DA-nm^me tiM 


Hlgfa Bridge Bulargemeat, N. T. City— 

lining wall and flat arches laid with Teiy cIom Jirinta. 

Wadilngton (D. C.) Aqueduct— 

Cti^lar coDdnit. S feet In diameter with walU IS 


0.714 


8t. LouI« Water Worka— 




New York Oty Storage Reservoir- 
Lining of gate-house walls and archee— rough work. . 


0.8M 



for Hninic, woA |8.4d for baoking. TI10 mortar th oompoeed at 
1 part BoModalo utaral cement and S parts of Band.* 

In Ghic^o in 1887, the price of brick laid in lime in interior 
■walls was about til per tbonsand, equivalent to about $7 per en. yd. 
The vages of masons were from 45 to 60 cent? per hoar, and of 
common labor from 20 to 9.5 cents per hoar. 

268. Spioifioatiohb toe Brick HAamnT. For BoildingL 
There is not even a remote approach to nniformity in the specifica- 
tions for the brick -work of buildings. Ordinarily the specifications 
for the brick masonry are very brief and incomplete. The following 
conform closely to ordinary conetraction. Of course, a higher grade, 
of workmanship can be obtained by more stringent specificationaf 

The brick in the exterior walls most be of good qualitj', hard-burned; fine, 
compact, and uniform In texture ; regular In shape, and uniform In size.}: 
One fourth of the brick In the Interior walU may be what ts known m soft 
or Mlmon hrick (see 3, g (MQ. The brick must be thoroughly wet before 
being laid. Tlie Joints of the exterior walls shall be from ^ to | Inch tldck.g 
The Joints of Interior divl^on-walls may be from { to 1 Inch tldck. The 
mortar shall be composed of I part of freah, well-slaked lime and 3} to 8 parta 

• Report of the Aqnednct CommlBaion, 1888-87, Table 4. 

t For specifications for masonry for VEuloas porpoMS, see AppokUx L 

I8eaf ST, pageST. 

f For the beat work, omit this Item and Insert the following : TJu outMe aolb 
•AoS bifaad ultA Hit bat pnued Mck of uniform color, laid in eolord mortar, vHA 
jeinU nof exaadbig one aghifi i)f an inch in thicktutt. Face brick are made a llttla- 
tuger ^ flS) than oidlaaiy brtck to compensate tor tbe thinner Joints. 



ovGoQi^lc 



176 BRICK MASOITBY. [CHAP. Till. 

of clean, sharp Band.* The lime-paste and the Band shall be thoroughly 
mixed before being uMd. The joints shall be veil filled with the aboTa 
tuortar ; no grout shall be used in the vork. The bond must consiBt of flva 
courses of stretchere to one of headers, and shall be bo arranged aa to thor- 
oughly bind the exterior and Interior portions of the wall to each other. 

The contractor must fumlah, set up, and take away his own scaffolding ; 
he must build In such strips, plugs, blocks, scantling, etc., as are required for 
securing the wood-work ; and must also assist in placing all iron-work, as 
beams, stairways, anchors, bed-plates, etc., connected with the brick-work. 

280. Por Bewen. The following are the qtecifications employed, 
in 1885, in the constrnctioD of brick sewers in Washington, D. G. : 

"The brat quality of whole new brick, burned tiard entirely through, free 
from injurious cracks, with true even faces, and with a cruBhing strength of 
not less tlian 5,000 pounds per square inch, shall be used, and must be thor- 
oughly wet by immersion Immediately before laying. Every brick is required 
Xa be laid in full mortar joints, on bottom, sides, and ends, which for each 
brick is to be performed by one operation. In no case is the joint to be made 
by working in moriar after the brick has been laid. Every second course shall 
be laid with a line, and Joints shall not exceed three eighths of on Incti. The 
brick-work of ttke arches shall be properly bonded, and keyed as directed by 
the en^eer. No portion of the brick-work shall be laid dry and afterwards 
p«uted- 

" The moriar shall be composed of cement and dry sand, fn the proportion 
K)t 800 pounds of cement and 2 barrels of loose sand, thoroughly mixed dry, 
and asufflcient quantity of water afterwards added to form a rather stiff paste 
It shall be used within an hour after mixing, and not at all if once set. 

" The cement shall be of the best quality, freshly burned, and equal in 
«Tery respect lo the Round Top or Shepardslown cement, manufactured upon 
the formula of the engineer-commissioner of the District of Columbia, capable 
of being worked for twenty minutes In mortar without toss of strength, and 
shall be tested in such maimer as the engineer may direct. After being mixed 
with water, allowed to set In air for twenty-four hours, and then immersed in 
vater for rix days, the tensile strength must be as follows : 

Neat cement 95 lbs. per sq. in. 

One part cement and one part sand SS " " " " 

" " " " twoparts " as ' " 

" " " " three" " 12 

"The sand used shall be clean, sharp, free from loam, vegetable matter, Or 
Mher dirt, and capable of giving the above results with the cement, 

'■ Hie water shall be fresh and clean, free from earth, dirt, or sewenge. 

* For masoar; that Is to be snbjected to a heavy pnesnre, omit this Item and 
Insert the following -. Ttn moriar mtut be mntpoitd qf 1 part UiwfiiuCa, 1 part etnwit, 
and 2 parU^dmn, iharp tand. Or, If a heavier prssgnre is to be resisted, spedly 
Uwt some parttonlar grade of oement mortar Is to he osed. iSea liii MS and XtA 



ovGoQi^lc 



8PB0IPICATION8. 177 

" Tigbt mortar-boxes shall be provided by the contractor, and no tnoriar 
■hall be made except Id such boxes. 

" 1'he proportions given are intended to form a mortar In which every 
putlcle of sand shall be enveloped by the cement ; and this result must be 
attained to the satisfaction of the engineer aud under his direction. The 
thorough mixing and Incorporation of all materials (preferably by mscbiue 
labor) will be Insisted upon. If by hand labor, tiie dry cement and sand shall 
be turned over with sliovels by skilled workmen not less than six times befora 
the water is added. Afler adding the waier. the paste shall again be turned 
over and mixed with shovels by billed workmen not leas than three limes be- 
fore it is used." 

S61. for Archet. The spccificatioDs for the brick arch masoDry 
on the AtohiBOD, Topeka and Santa Fi Bailroad are as follows : 

"Tbe bricks must be of the best quality of smooth, hard-burut, paving 
bricks, well tempered and moulded, of the usual size, compact, well Bbaped, 
free from lime, cracks, and olher imperfect ions, and must stand a pressure 
of 4,000 pounds per square inch without crusbing. No tats will be allowed 
In the work except for making necessary closures. All bricks will be culled 
on the ground after delivery, and selected in ftrlct accordance with these 
•peclflcatlons. 

"The mortar must be made of 1 measure of good natural hydraulic cement 
and 3 measures of clean, sharp sand — or such other proportion as may be 
prescribed by the engineer — well mixed together with clean water, In cle«n 
mortar-beds constructed of boards, and must be used immediately after befng 
mixed. 

'- The brick must be laid flush in cement mortar, and must be Iboroughly 
wet when laid. AH Joints and beds must be thoroughly tilled with mortar so 
as to leave DO empty spaces whatever In tbe masonry of the walls and arches, 
which must be solid throughout. The thickness of mortar- joints must be as 
follows 1 Id the walls and In Ibe arch between biicks of tbe same ring, not lew 
than three eighths of an Incb (|") nor more than one half Inch (}"). lu tbe arch 
between rings, not less than one half inch (}") nor more than five eij^bths of 
an inch (("). Each brick Is to be driven into place by blows of a mallet. Tbe 
bricks must be laid in tbe walls with the ordinary English bond, five stretcher 
courses to one header course. They must be laid In the arch In concentric 
rings, each lon^tudlnal line of bricks breaking joints with the adjoining 
linea in tbe same ring and In the ring under it. No headers to be used In 
the arch." 

262. Bbioe n. flitnn XavOIUT. Brick masonry la not mnch 
Dsed, except in the walla of buildings, in lining tnnncU, and in con- 
atrncting severs, the general opinion being that brick-work is in 
every way inferior to stone masonry. This belief may have been 
veil fonnded when brick was made wholly by hand, by inexpert 
operatiTe^ and imperfectly bnmed in the old-time kilns, the prod- 



ovGoQi^lc 



178 BSICK HASONBT. [CHAP. VIIL 

act being then generally poor ; bnt thinga have changed, and sinoe 
the manafactnre of brick haa become a bnBineaa condccted on a 
large ecale by enterprimng men, with the aid of a variety of tnachmee 
and improved kilns, the product is more r^ular in size and qoality 
and stronger than formerly. Brick is rapidly displacing stone for 
the largest and best bnildings in the cities, particnlarly in Chicago 
and St. Peterabnig, where the vicissitndes of the climate try maaonij 
very aeverely. There are many engineering stmoturee in which 
brick oonld be profitably employed instead of atone ; as, for example, 
the walla of boz-cnlTerte, cattle-goards, etc., wad the less important 
bridge piers and abntments, particnlarly of highway bridges. 

Brick-work is superior to stone masonry in several respects, aa 
follows : 1. In many localities brick is cheaper than stone, since 
the former can be made near by while the latter most be shipped. 
2. As brick can he laid by less skillfnl masona than stone, it costs 
less to lay it. 3. Brick is more easily handled than stone, and can 
be laid vrithoat any hoisting apparatas. 4. Brick requires less fit- 
ting at corners and openings. S. Brick masonry is less liable to 
great weakness throogh inaccnrate dressing or bedding. 6. Brick- 
work resists fire better than limestone, granite, or marble, sand- 
stone being the only variety of stone that can compare with brick 
in this respect. 7. Good brick stands the effect of weathering and 
of the acids in the atmosphere better than sandstones, and in dura- 
bility even approaches some of the harder stones (see §§ 31-33). 
8. All masonry tails when the mortar in its joints disintegrates or 
becomes dislodged; therefore brick masonry will endnre the vicisei- 
tndes of the weather as well as stone masonry, or even better, since 
the former nsuaUy haa thinner joints. 

Brick-work is not as strong as ashlar masonry, but costs less ; 
while it is stronger and costs more than ordinary rnbble. 

863. BeiOX XAHmT IxrasTloira to WatBB. It sometimes be- 
comes necessary to prevent the percolation of water through brick 
walls. A cheap and effective process has not yet been discovered, 
and many expensive trials have proved failures. The following 
account* givee the details of two experiments that were entirely snc- 
cessfnl. 

" The face walls of the back bays of the gate-houses of the new 

• Atattaet of A pi^er by Wn. L. DMibom, in Tnu. Am. Soo. of C. E., Ttd. I 
IV.MS-8- 



ovGoQi^lc 



lUSONBT IHPBE7IO08 TO VA.TBB. 179 

Croton reaerroir, located north of Eighty-sixth Street, in Gentnl 
Park, New York City, were bnilt of the best qnality of hard-hnmt 
brick, laid in mortar compoeed of hydranlio cement of New York. 
[Ulster Co. Bosendt^e] and sand mixed in the proportion of on» 
measnre' of cement to two of sand. The space between the walls was. 
4 feet, and was filled with concrete. The face walls were laid np' 
with great care, and every precsation was taken to hare the joints- 
well filled and to insure gooid work. The walls are 12 inches thick, 
and 40 feet high; and the bays, when full, generally have 36 feet of 
water in them. 

"When the reserroir was first filled and the water let into the 
gate-honses, it was foand to filter throngh these walls to a consider- 
able amount As soon as this was discoTered the water was drawn 
ont of the bays, with the intention of attempting to remedy or pre- 
Tent this infiltration. After carefully consideriug several modes ol 
accomplishing the object desired, I [Dearborn] came to the conclu- 
sion to try ' Sylvestei^s Process for Repelling Moisture from Exter 
nal Walls.' 

" The process consists in using two washes or solations for cor- 
ering the surface of the walls — one composed of Castile soap ani^' 
water, and one of alnm and water. The proportions are thre&. 
quarters of a ponnd of soap to one gallon of water, and iialf a pound 
of alnm to four gallons of water, both substances to be perfectly 
dissolved in water before being used. The walls should be perfectly 
clean and dry, and the temperature of the air not below 50° Fahr. 
when the compositions are applied. 

"The first, or soap-wash, should be laid on, when boiling hot, 
with a flat brush, taking care not to form a froth on the brick-work. 
This wash should remain 24 ^ours, so as to become dry and hard 
before the second, or slum, wash is applied, which should be dona 
in the same manner as the first. The temperature of this wash,, 
when applied, maybe 60° or 70° Fahr.; and this also shonld remaini 
24 hours before a second coat of the soap-wash is pnt on.- 
These coats are to be applied alternately until the walls are made- 
imperrions to water. The alum and soap thus combined form an: 
insoluble compound, filling the pores of the masonry and entirely 
preventing the water from entering the walls. 

"Before applying these compositions to the walls of the bays 
tome experiments were made to test the absorption of water by. 



ovGoQi^lc 



180 BEICK MABONBT. [OHAP. Tilt 

bricks nnder pressure after being covered vitii these washes, in 
order to determine boT many coats the walls would require to render 
them imperrions to water. To do this, a strong wooden box large 
enough to bold two bricks was made, pat together with screws, and 
in the top was inserted a 1-inch pipe 40 feet long. In this box 
wore placed two bricks, after being made perfectly dry, which were 
then covered with a coat of each of the waaheB, as before directed, 
and weighed. Tbey were then subjected to a column of water 40 
feet high ; and after remaining a sufflctent length of time they were 
taken ont and weighed again, to ascertain the amount of water they 
had absorbed. The bricks were then dried, and again coated with 
the washes and weighed, and subjected to pressure as before, this 
operation being repeated nntil the bricks were found not to absorb 
any water. Four coatings rendered the bricks impenetrable under 
the pressure of a 40-foot head. The mean weight of the bricks (dry) 
before being coated was 3^ lbs. ; the mean absorption was one half- 
pound of water. A hydrometer was used in testing the solutions. 

"As this experiment was made in the fall and winter (1863), 
after the temporary roofs were put on to the gate<house, artificial 
heat hod to be resorted to to dry the walls and keep the air at a 
proper temperature. The cost was 10 cents per sq. ft. As soon as 
the last coat bad become hard, the water was let into the bays, and 
the walls were found to be perfectly impervious to water, and they 
remain so in 1870, after about 6j years. 

264. "The brick arch of the footway of High Bridge is the arc 
of a circle, 29^ feet radi,ns, and is 13 incbee thick; the width on top 
is 1'' feet, and the length covered is 1,381 feet. The first two 
courses of the brick of the arch are composed of the best hard-burnt 
brick, laid edgewise in mortar composed of 1 part, by measure, of 
hydraulic cement of Kew York [Bosendale natural] and 2 parts of 
sand. The top of these bricks, and the inside of the granite 
coping against which the two top courses of brick rest, was covered, 
when perfectly dry, with a coat of asphalt one half an inch thick, 
laid on when the asphalt was heated to a temperature of from 360° 
to 518° Fahr. On top of this was laid a course of brick fiatwise, 
dipped in asphalt, and laid when the asphalt was hot; and the joints 
were rnn full of hot asphalt. On top of this, a course of pressed 
brick was laid flatwise in hydianlio cement mortar, forming the 
paving and floor of the bridge. 



ovGoQi^lc 



HKFLOBBBOBNOK. 181 

"The area of the bridge covered vithaephalted brick was 23,06fl 
■q. ft There wereiued94,200 Iba. of asphalt^ 33 bsrrelsofcoaltar, 
10 CO. yds. of Band, and 93>800 bricks. The asphalt was the Trini- 
dad rariety ; and was mixed with 10 per cent., b; measure, of ooal 
tar, and 25 per cent, of sand. The time occupied was lOd days of 
masons, and 148 days of laborers. Two masons and two laborers 
will melt and spread, of the first coat, 1,650 sq. ft per day. The 
total cost of this coat was 5^ oents per sq. ft, ezclosiTe of daty on 
asphalt 

" There were three grooves, 2 inches wide by 1 inohes deep, 
made entirely across the brick aroh immediately under the finit 
coat of asphalt thtis dividing the arch into four equal parts. The 
grooves were filled with elastic paint cement. This arrangement 
was intended to guard against the evil effects of the contraction of 
the aroh in winter ; for, since it was expected to yield slightly at 
these points and at no other, the elastic cement would prevent any 
leakage there. The entire experiment has proved a very successful 
one, and the bridge has remained perfectly tight 

" In proposing the above plan for workiog the asphalt with the 
. brick-work, the object was to avoid depending on a large continuons 
snr&ce of asphalt as is usual in covering arches, which very fre> 
quently cracks from the greater contraction of the asphalt than that 
of the masonry with which it is in contact, the extent of the asphalt 
on this work being only about one quarter of an fnch to each brick. 
This is deemed to be an essential element in the success of the im- 
pervious covering." 

366. Xriobiscxhoe. Masonry, particularly in moist climate 
or in damp phices — as cellar walls, — is frequently disfigured by the 
formation of a white efflorescence on the surface. This deposit 
generally originates with the mortar, but frequently spreads over 
the entire face of the wall. The water which is absorbed by the 
mortar dissolves the salts of soda, potash, magnesia, etc., contained 
in the lime or cement, and on evaporating deposits these salts as a 
white efflorescence on the surface. With lime mortar the deposit 
is frequently very heavy, particularly on plastering ; and, usually, 
it is heavier with natnral than with Portland cement. The efflo- 
resoenoe sometimes originates in the brick, particularly if the brick 
wae burned with Bulphnrous coal, or was made from clay contain- 
ing lioa pyritea; aod when the briok gets wet, the water dissolvea 



ovGoQi^lc 



J8S BBICK VABOXRY. [CHAP. VtlL 

-the snlphates of Umo and magnesia, and on evaporatiDg leaves tbe 
oiystals of these salte on the sarface. Frequently the effloreecenee 
<m the brick is due to the absorption by the brick of the impreg- 
nated water from the mortar. 

This efflorescence ia objectionable because of the unsightly a|H 
pearaDce which it often prodnces, and also because the cryatalliza- 
tion of these salts within the pores of the mortar and of the brick 
or stone cansee disint^ration which is iu many respects like frost 

As a preveDtiTe, Gillmore recommends* the addition of 100 lbs. 
of quicklime and 8 to 12 lbs. of any cheap animal fsA to each barrel 
of cement. The lime is simply a vehicle for the fat, which should 
be thoroughly incorporated with the limo before slaking. The ob- 
ject of the fat is to saponify the alkaline salts. The method is not 
entirely satisfactory, since the deposit is only made less prominent 
and less effective, and not entirely remoTcd or prevented. 

The efflorescence may be entirely prevented, whatever its origin, 
by applying Sylvester's washes (see g 263) to the entire external anr- 
foces of the wall ; and, since asually the efflorescence is dne to the 
water absorbed by the mortar, it can generally be prevented, and 
can always be much diminished, by using mortar which is itself im- . 
pervious to water (see g 141). The latter is the cheaper method, 
particularly if the impervious mortar be used only for the face of 
the joints. If the wall stands in damp ground, one or more of the 
horizontal joints just above the surface sbonld be laid in imperrious 
mortar, or better, the brick for several courses should be rendered 
impervious and be laid in impervious mortar to prevent the wall's 
absorbing moisture from below. 

* "IlmM, BjitMlUo CttamSa, and Koitan," p. 9B6> 



ovGoQi^lc 



PARTm. 

FOUNDATIONS, 



CHAPTER IX. 
INTRODHCTORT. 

268, Dimmion. The term foaDdation is ordinarily nsed in- 
differan Jy for either the lower contaee of a strnotiire of maeoiuy or 
th« artiJcial arrangetneot, whatever its character, on which these 
ooursea rest For greater cleamess, the term foundation will here 
be restricted to the artificial arnuigement, whether timber or mason- 
ry, which eapporte the main structure ; and the prepared enrface 
upon wliich this artificial structure rests will be called the bed of 
the foundation. There are many cases in which this distinction 
can not be adhered to strictly. 

267. IMPOETASOE OF THE SUBJECT. The foundation, whether 
for the more important baildings or for bridges and oalverts, is the 
meet critical part of a masonry structare. The failures of works of 
masonry due to fanlty workmanship or to an insufficient thickness 
of the walls are rare in comparison with those due to defective 
foundations. When it is necessary, as so frequently it is at the 
present day, to erect gigantic edifices — as high buildings or long- 
span bridges — on weak and treacherous soils, the highest constrac- 
tire skill is required to supplement the weakness of the Datnial 
foundation by snch artificial preparations as will enable it to snetain 
Buoh massive and costly burdens with safety. 

Probably no branch of the engineer's art requires more ability 
and skill than the conHtrnotion of foundations. The conditions 
gOTeming safety are generally capable of being calculated with 
M much practical accuracy in this aa in any other part of a con- 



ovGoQi^lc 



184 TovsnATioss: niTBODCoroBT. [chap. IX. 

Btrnction ; bnt, nnfortanately, pnctice is freqnentl; based npon 
empirical mles rather than apon a scientific application of fnnda- 
mcDtal principles. It is nnpardonable that any liability to danger 
or loss should exist from the imperfect comprehension of a snbject 
of saohrital importance. Abihty is required in determining the 
conditions of stability ; and greater skill is required in fuLfiiling 
these conditions, that the coat of the foundation may not be pro- 
portionally too great. The safety of a stractore may be imperiled, 
or ite cost unduly increased, according as its foundations are laid 
vith insufficient stability, or with provision for security greatly in 
excess of the requirements. The decision as to vbat general method 
of procedure will probably be best in any particnlar case is a ques- 
tion that can be decided with reasonable certainty only after long 
experience in this branch of engineering ; and after having decided 
npon the general method to be followed, there is room for the 
exercise of great skill in the means employed to secnre the desired 
end. The experienced engineer, even with all the information 
which he can derive from the works of others, finds occasion for the 
use of all hia knowledge and best common sense. 

The determination of the conditions necessary for stability can 
be reduced to the application of a tew fundamental principles which 
may be studied from a text-book ; but the knowledge required to 
determine beforehand the method of constmction best suited to the 
case in hand, together with its probable cost, comes only by personal 
experience and a careful atndy of the experiences of others. The 
object of Part III, is to classify the principles employed in con- 
structing fonndations, and to give such brief accoanta of actual 
practice as will illastrato the applications of these principles. 

288. Pus OF PboPO>ED HVKUmaa. In a general way, soils 
may be divided into three classes : (1) ordinary soils, or those which 
are capable, either in their normal condition or after that condition 
has been modified by artificial means, of sustaining the load that is 
to be brought upon them ; (3) compressible soils, or those that are 
incapable of directly supporting the given pressure with any reason- 
able area of foundatioa ; and (3) aemi-liquid soils, or those in which 
the Suidity is so great that they are incapable of supporting any 
considerable load. Each of the above classes gives rise to a special 
method of coDstructing a foundation. 

1. With a Mil <tf the first class, the bearing power may be in- 



ovGoQi^lc 



PLAJT OF FBOPOSED DISCUSSION. 186 

creased by compacting the Btirface or by drainage ; or the area of 
the foandfttion may be Increased by the uee of masonry footing 
coaraee. Inverted masonry arches, or one or more layers of timbers, 
railroad rails, iron beuns, etc. Some one of these methods is or- 
dinarily employed in constructing foundations on land ; as, for 
example, for buildings, bridge abntmenta, sewers, etc. Usually all 
of these methods are inapplicable to bridge piers, i. e., for fonuda- 
tions under vater, owing to the scouring actioa of the current and 
also to the obstruction of the channel by the greatly extended base 
of the foundation. 

2. With compressible soils, the area of contact may be increased 
by supporting the structure upon piles of wood or iron, which are 
sustained by the friction of the soil on their sides and by the direct 
pressure on the soil beneath their bases. This method is frequently 
employed for both bufldings and bridges. 

3. A semi-fluid soil must generally be removed entirely and the 
structure founded upon s lower and more stable stratum. This 
method is specially applicable to foundations for bridge piers. 

There are many coses to which the above classification is Dot 
strictly applicable. 

For oonvenieace in study, the coostruction of foundations will 
be discussed, in the three sucoeediog chapters, under the heads 
Ordinary Foundations, Pile FoundtUiona, and Foundaiions under 
Water. However, the methods employed in each class are not 
entirely distinct from those used in the others 



ovGoQi^lc 



CHAPTBBX. 
ORDINABY POUNDATIONS. 

269. Is this chapter will be diBcnsaed the method of constnict- 
ing the fouDdstioiiB for bnildiogs, bridge abntment«, caWerta, or, 
in general, for any Btrnctnre founded npon dry, or nearly dry, 
ground. This claw of foundations coald appropriately be called 
Fonndations for BaildingB, since these are the most nomerons of the 
class. 

Jhis chapter is divided into three articles. The firat treats of 
the soil, and includes {a) the methods of examining the site to de- 
termine the nature of the soil, {b) a discnseion of the bearing power 
of different soiU, and (c) the methods of increasing the bearing 
power of the soil. The second article treats of the method of de- 
signing the footing courses, and inclodes (a) the method of deter- 
mining the load to be supported, and {b) the method of iucreasing 
the area of the fonndation. The third contains a few remarks con- 
cerning the practical work of laying the foundation. 

Art. 1. The Soil. 

270. EXAimrATlOB or THI Sir. The nature of the soil to be 
built upon is evidently the first subject for consideration, and if it 
has not already been revealed to a considerable depth, by excava- 
tions for buiidings, wells, etc., it vUl be necessary to make an ex- 
amination of the subsoil preparatory to deciding upon the details 
of the foundation. It will usually be sufficient, after having dug 
the foundation pita or trenches, to examine the soil with an iron 
rod or a post-anger from 3 to 5 feet further, the depth depending 
upon the nature of the soil, and the weight and importance of the 
intended structure. 

In soft soil, soundings 40 or 50 feet deep can be made by driving 
a small (say f-inch) gas-pipe with a hammer or maul from a tem- 
porary scaffold, the height of which will of course depend upon the 
Imgth of the eeotionB of the pipe. If aamplea of the aoil are desired. 



ovGoQi^lc 



AXU. 1.] THE SOIL. 187 

OBO aS-inoh pipe open at the lover end. If much of this kind of 
work is to be done, it is advisable to fit np a hand pile-driving 
machine (see § S35), natng a block of vrood for the dropping weight. 
Borings SO to 100 feet deep can be made very ezpeditionsly in 
common soil or clay with a common wood-aoger turned by men, 
with levers 2 or 3 feet long. The auger will bring up samples snf- 
ficient to determine the nature of the soil, but not its compactness, 
since it will probably be compressed somewhat in being cut oS. 

When the testing must be made through sand or loose soil, it 
may be necessary to drive down an iron tube to prevent the soil 
from &lling into the hole. The sand may be removed from the 
inside of the tube with an auger, or with the "sand-pump" ased in 
digging artesian wells. When the subsoil is composed of varions 
strata and the stmcture demands extraordinary precaution, borings 
must be made with the tools employed for boring artesian wells.* 

271. If the builder desires to avoid, on the one hand, the unnec- 
essarily costly foundations which are frequently constmcted, or, on 
the other hand, those insufficient foundations evidences of which 
are often seen, it may be necessary, after opening the trenches, to 
determine the supporting power of the soil by applying a test load. 

In the case of the capitol at Albany, X. Y., the soil was tested 
by applying a measured load to a square foot and also to a square 
yard. The machine used was a mast of timber 12 inches square, 
held vertical by guys, with a cross-frame to hold the weights. For 
the smaller area, a hole 3 feet deep was dug in the bine clay at the 
bottom of the fonndation, the hole being IS inches square at the 
top and 14 inches at the bottom. Small stakes were driven into 
the ground in lines radiating from the center of the hole, the tops 
being brought exactly to the same level ; then any change in the 
Bur^oe of the ground adjacent to the hole could readily be detected 
and measured by means of a straight-edge. The foot of the mast 
was placed in ^e hole, and weights applied. No change in the 
surface of the adjacent ground was observed until the load reaahed 
5.9 tons per sq. ft., when an nplift of the surrounding earth was 
noted in the form of a ring with an irregularly rounded surface, 
the contents of which, above the prerious surface, measured 0.09 
cubic feet Similar experiments were made by applying the load to 

• For UlaatntlODa of tooU tor thU pnrpoM, Me EngbwertDg News, voL ftl, p.BSL 



ovGoQi^lc 



188 OBDINA.BT rOUNDATIOKS. [CHAP. X. 

a sqoare yard Tith eseentiaJly the same reaulte. The several loads 
were allowed toreniaiD for some time, and the settlementa observed.* 
Similar experimente were made in connection with the cooBtrnc- 
tion of the CoDgreseiooal Libisry Building, Washington, D. C, with 
a fnme which rested upon 4 foot-plates each a foot square. The 
frame could be moved from place to place on wheels, aod the test 
was applied at a number of places. 

272. BUBnra FOWXB OF 80tU. It is scarcely oecessary to sa; 
that soils vary greatly in their bearing power, ranging as they 
do from the condition of hardest rock, through all intermediate 
Btagee, to a soft or semi-liquid condition, as mud, silt, or marsh. 
The best method of determining the load which a specific soil will 
bear is by direct experiment (§ 271); hut good judgment and ex- 
perience, aided by a careful study of the nature of the soil — its com- 
pactness and the amount of water contained in it — will enable one to 
determine, with reasonable accuracy, its probable supporting power. 
The following data are given to assist in forming au estimate of the 
load which may safely be imposed upon different soils. 

273. Book. The ultimate crushing strength of stone, as deter- 
mined by crushing small cubes, ranges from 180 tons per square 
foot for the softest stone — such as are easily worn by running water 
or exposure to the weather — to 1,800 tons per square foot for the 
hardest siones (see page II), The crushing strength of slabs, t. e., 
of prisms of a less height than width, increases as the height de- 
creases. A prism one half as high as wide ia about twice as strong 
as a cube of the same material. If a slab be conceived as being made 
up of a number of cubes placed side by side, it is easy to see why 
the slab is stronger than a cube. Tbe exterior cubes prevent the 
detachment of the disk-like pieces (Fig. 1, page 9) from the sides of 
the interior cubes ; and hence the latter are greatly strengthened, 
which materially increases the strength of the slab. In testing 
cubes and slabs the pressure is applied uniformly over the entire 
upper surface of the test specimen ; and, reasoning from analogy, 
it seems probable that when the pressure is applied to only a small 
part of the surface, as in the case of foundations on rock, the strength 
will be much greater than that of cubes of the same material. 

The table on page 190 contains the results of experimente made 

* W. J. Voiiptsui, the engbMer fn obarge, in Trans. Am. Soe. C. B., vol. U. p- 9B3. 



ovGoQi^lc 



AKT. 1.] THE aOIL. 189 

b; the author, and shove conclneively that a unit of Diaterial htu a 
much greater power of resigtance wjien it forms a portion of a larger 
mass than wfaea isolated in the manner castomary in making ex- 
periments on crnshing strength . 

The ordinary "crushing Btrength" given in next to the last 
column of Table 23 was obtained bj crushing cubes of the identical 
materials employed in the other experiments. The concentrated 
pressure vas applied by means of a hardened eteel die thirty-eight 
Bixty-fourtha of an inch in diameter (area = 0.277 eq. in.). All the 
teste were made between self-adjusting parallel plates of a hydro- 
static testing-machine. Ifo packing was used in either series of 
experiments ; that is to say, the pressed surfaces were the same in 
both series. EoweTer, the block of limestone 7 inches thick (Ex- 
periments Nos. 8 and 13) is an exception in this respect. This 
block had been sawed out and was sligljtly hollow, and it was 
tboaght not to be worth while to dress it down to a plane. As pre- 
dicted before making the test, the block split each time in the di- 
rection of the hollow. If the bed had been flat, the block would 
doubtless have shown a greater strength. The concentrated pres- 
sure was generally applied near the corner of a large block, and 
hence the distance from the center of the die to the edge of the 
block is to the nearest edge. Frequently the block had a ragged 
edge, and therefore these distances are onl; approximate. The 
quantity in the last column — "Batio" — Is the crushing load per 
square inch for concentrated pressures divided by the crushing load 
per square inch for uniform pressure. 

The experiments are tabulated in an order intended to show that 
the strength under concentrated pressure varies (1) with the thick- 
ness of the block and (^) with the distance between the die and 
the edge of the material being tested. It is clear that the strength 
increases veiy rapidly with both the thickness and the distance from 
tho edge to the point where the pressure is applied. Therefore we 
conclude that the corapressiTe strength of cubes of a stone gives 
little or no idea of the ultimate resistance of the same material when 
in thick and extensive layers in its native bed. 

274. The safe bearing power of rock is, certainly not less than 
one tenth of the ultimate crushing strength of cvbe^; that is to say, 
the safe bearing power of solid rock is not less than 18 tons per sq. 
ft. for the softest rock and 180 for the strongest. It is safe to say 



ovGoQi^lc 



OBDINjLET FODHDATIOlrS. 



LCHi.P. X. 



IP* 



k 

fill 



Brick 

Llme«tone. . 
SttndstoDe. . 
LlmeMooe. ■ 



80.100 
n.HOI 
81,046 



01,800 
59,904 
75,810 

7S,8»1 
102,900 
111,188 



1.840 
10.500 
10.100 
2, AH 
8.408 
8.696 
4,671 
8,468 

8,696 



! = 



14.0 
1«.0 
18. S 



16.0 
23.0 
M.O 



Clay, wbtch for yean haa Mltily carried, without appreciable settlemeni. 



that almost any rock, from the hardness of granite to that of a soft 
crnmbling stone easily worn by exposore to the weather or to run- 
ning water, when well bedded will bear the heaYiest load that can 
be brought upon it by any masoDiy conetrnction. 

It scarcely ever occurs in practice that rock is loaded with the 
full amount of weight which it is capable of sustaining, as the extent 
of base necessary for the stability of the structure is generally suffi- 
cient to prevent any undue pressure coming on the rock beneath. 

275. Clay. The clay soils vary from slate or shale, which wilt 
support any load that can come upon it, to a soft, damp clay which 
will squeeze out in every direction when a moderately heavy prea- 



ovGoQi^lc 



AUT, 1.] THE 80II» 191 

Bure ia bronght upon it. Fouadatioas on ol&j slioiild be laid at 
encb depths as to be unaffected by the weather ; since clay, at even 
ooneiderable depths, will gain and lose considerable water as the 
seasons chaDge. The bearing power of clayey soils can be rery 
much improred by drainage (§ ^85), or by preveating the penetra- 
tion of water. If the foundation is laid upon undrained clay, care 
must be taken that ezcavations made in the immediate vicinity do 
not allow the clay under pressure to escape by oozing away from 
under the building. When the clay occurs in strata not horizontal, 
great care is necessary to prevent this flow of the soiL When coarse 
sand or gravel is mixed with the clay, its supporting power is greatly 
increased, being greater in proportion as the quantity of these 
materials is greater. When they are present to such an extent that 
the clay is just sufScient to bind them together, the combination 
will bear as heavy loads as the softer rocks. 

276. The following data on the bearing power of clay will be of 
assistance in deciding upon the load that may safely be imposed 
upon any particular clayey soil. From the experiments made in 
connection with the construction of the capitol at Albany, K. Y., 
as described in § 271, the conclusion was drawn that the extreme 
supporting power of that soil was less than 6 tons per sq. ft, and 
that the load which might be safely imposed upon it was 2 tons per 
sq. ft. " The soil was blue clay containing from 60 to 90 per cent 
of alumina, the remainder being fine siliceous sand. The soil con- 
tuins from 2? to 43, usually about 40, per cent of water ; and vari- 
ous samples of it weighed from 81 to 101 lbs. per cu. ft" In the 
case of the Gongreseional Library (§ 271), the ultimate supporting 
power of " yellow clay mixed with sand " was 13^ tons per sq. ft. ; 
and the safe load was assumed to be 2^ tons per sq. ft Experi- 
ments made on the clay under the piers of the bridge across the 
Missouri at Bismarck, with surfaces 1^ inches square, gave an aver- 
age ultimate bearing power of 15 tons per sq. ft* 

The stifler varieties of what is ordinarily called clay, when kept 
dry, will safely bear from 4 to 6 tons per sq. ft.; bnt the same clay, 
if allowed to become saturated with water, can not be trusted to 
bear more than 2 tons per sq. ft. At Chicago, the load ordinarily 
put on a thin layer of clay (bard above and soft below, resting on a 

■ Beport o( tbe engineer, Qeo. S- Uorlcon. 



ovGoQi^lc 



193 OBDIXABT FOUNDATIONS. [CHAP. Z. 

thick Btnitiim of quicksand) is 1^ to 3 tons per sq. ft; and the set- 
tlement, vhich nsnally reaches a maximam in a year, is about 1 
inch p«r ton of load. Experience in centra] Illinois shows that, if 
the foundation is carried down below the action of frost, the clay 
subsoil will bear 1^ to 3 tons per aq. ft. without appreciable settling. 
Itankine gives the safe load for compreseible soils aa 1^ to 1} tons 
persq. ft. 

277. Sand. The sandy soils vary from coarse gravel to fine sand. 
The former when of sufficient thickness forms one of the firmest 
and best foundations ; and the latter when saturated with water 
is practically a liquid. Sand when dry, or wet sand when prevented 
from spreading laterally, forms one of the best beds for a founda- 
tion. Porous, sandy soils are, as a rule, unaffected by stagnant 
water, but are easily removed by running water ; in the former case 
they present no difficnlty, but in the latter they require extreme 
care at the bauds of the constructor, as will be considered later. 

278. Compact gravel or clean sand, in beds of considerable 
thickness, protected from being carried away by water, may be 
loaded with 8 to 10 tons per sq. ft. with safety. In an experiment 
in France, clean river-sand compacted in a trench supported 100 
tons per sq. ft. Sand well cemented with clay and compacted, it 
protected from wat«r, will safely carry 4 to 6 tons per sq. ft. 

The piers of the Cincinnati Snapension Bridge are founded on a 
bed of coarse gravel 13 feet below low-water, although solid lime- 
stone was only 12 feet deeper; if the friction on the sides of the 
pier* be disregarded, the maximum pressure on the gravel is 4 tons 
pej sq, ft. The piers of the Brooklyn Suspension Bridge are founded 
44 feet below the bed of the river, upon a layer of sand 2 feet thick 
resting upon bed-rock ; the maximum pressure is about 5^ tons 
per sq. ft. 

At Chicago sand and gravel about 15 feet below the surface tat 
Buccessfully loaded with 2 to 2J tons per sq. ft. At Berlin the safe 
load for sandysoil is generally taken at 2 to 2^ tons per sq. ft The 
Washington Monament, Washington, D. C, rests upon a bed of 
very fine sand two feet thick underlying a bed of gravel and bowl- 
ders; the ordinary pressure on certain parts of the fonndatiou is 
not far from 11 tons per sq. ft., which the wind may increase to 
nearly 14 tons per sq. ft. 

• For tlw amount of sacb MoUon, see ij 418-t» and ( ISt. 



ovGoQi^lc 



A.&I. 1.] THS BOIL. 193 

t79. Bami-Liqtiid Soili. With a soil of this claee, u mod, silt, 
or qaickeand, it is cnatomaiy (1) to remove it entirely, or (2) to 
ask piles, tabea, or caieaone through it to a eolid substratum, or 
13) to consolidate the soil by adding saad, earth, stone, etc. The 
method of performing these operations will be described later. Soils 
of a soft or semi-liqnid character should never be relied npon for a 
fonndstion vhen anything better can be obtained ; but a heavy 
Baperstructure may be supported by the upward pressure of a semi- 
iiqnid soil, in the same way that water bears up a floating body.. 

According to Bankine,* a building will be supported when the 

preesare at its base is w k( _ . 1 per unit of area, in which ex- 
pression V) is the weight of a unit volume of the soil, k is the depth 
of immersion, and a is the angle of repose of the soil. It a = 5°, 
then according to the preceding relation the supporting power of 
the soil is 1.4 ro h per unit of area ; if a = 10°, it is 2.0 to A; and 
if cf = 16°, it is 2,9 w k. The weight of soils of this class, i. e., 
mad, silt, and quicksand, varies from 100 to 130 lbs. per cu. ft. 
Bankine gives this formula as being applicable to any soil ; but since 
it takes no account of cohesion, for most soUs it is only ronghly ap- 
proximate, and gives results too small. The following experiment 
seems to show that the error is considerable. " A 10-foot sqnare 
base of concrete resting on mud, whose angle of repose vras 5 to 1 
[a =lli°], bore 700 lbs. per sq. ft."t This is ^ times the result 
by the above formula, using the maximum value of «f. 

lAXg& buildings have been senurely founded on quicksand by 
making the base of the immersed part as large and at the same time 
«s light as possible. Timber in successive layers (§ 309) or grillage 
on piles (g 330} is generally used in such cases. This class of fonn- 
dations is frequently required in constructing sewers in water-bear- 
ing sands, and though apparently presenting no difficulties, such 
foundations often demand great skill and ability. 

280. It is difficult to give results of the safe bearing power of 
soils of this class. A considerable part of the supporting power is 
derived from the friction on the vertical ddes of the foundation ; 
hence the bearing power depends in part upon the area of the aide 
surface in contact vrith the soil. Furthermore, it is difficult to de- 



ovGoQi^lc 



191 OBDIKABT FOUNDAIIOHB. [OHAP. X. 

termine the exact snpporting power of a plastic soil, sinca a coneid- 
ereble eettlemect is oertain to take place with the laiMe at time. 
The experience at New Orleans with alluvial soil tetA a. *«» nxnM- 
menta* that have been made on quickaaad eeeia to indicate tuat 
with a load of ^ to 1 ton per square foot the settlement wil* not oe 



281. Bearing Power: Sommary. Gatheringtogethertfaeresuits 
of the preceding discosBion, we have the following table * 

TABLE 28. 
Batb Bbabinq Fowkb of Soiu. 



Bock— the haidart— In thfck layers, in oatlTe bed <g 374) 

" equal to b«at ashlar Duaonr)' (% i^^i 

" " " " brick " " 

" " ■' poor " " " 

Clay, In thick bed*, always dry (g 376) 

.. .. .. .. moderately dry (§ 376) 

■' aoft(8 376) 

GraTel and coarse Band, well cemented (§378) 

8aod, compact and well cementetl, " 

" clean, dry " 

Quicksand, alluvtal soils, etc. (§ 380) 



282. Condnilon. It is well to notice that there are Bome prac- 
tical considerations that modify the pressure which may safely be 
pat npoD a eoil. For example, the pressare on the fonndation of 
a tall chimney shonld be considerably less than that of the low ma»- 
ure foundation of a fire-proof vault. Id the former c»8e a slight 
meqnality of bearing power, and consequent nneqnal settling, might 
endanger the stability of the strncture ; while in the latter no eeri- 
oua harm wonld reenlt. The pressure per nnit of area shonld be 
leas for a light stractare subject to the passage of heavy loads — as. 




ovGoQi^lc 



UtT. L] THB SOIL. I9b 

for example, a nulroad viaduct — than for a heavy atrnctuni aahjeot 
only to a qniescent load, since the shock and jar of the moving load 
are &r more serions than the heavier quiescent load. 

The determination of the safe bearing power of soils, particalar- 
Ij when dealing with those of a eemi-liquid character, is not the 
only qaestion that mnst receive oaretal attention. In the founda- 
tions tor buildings, it may be neceesary to provide a saf^^nard 
against the soil's escaping by being pressed oat laterally into excava- 
tions in the vicinity. In the foaodations for bridge abntments, it 
may be necessary to consider what the effect will be if the soil around 
the abatment becomes thoroughly satnrated with water, as it may 
during a flood; or what the effect will be if the soil is deprived of 
its lateral support by the washing away of the soil adjacent to the 
abutment. The provision to prevent the wash and nndermining 
action of the stream is often a very considerable part of the cost of 
the structure. The prevention of either of these liabilities is a prob- 
lem by itself, to the solution of which any general discussion will 
contribute but little. 

283. iKFBOViao THi BxABura Fowek of the Boil. When 
the soil directly under a proposed structure is incapable, in its nor 
mal state, of austaining the load that will be brought upon it, the 
bearing power may be increased (1) by increasing the depth of the 
foundation, (2) by draining the site, (3) by compacting the soil, or 
(4) by adding a layer of sand. 

281. iDOTeatlng the Depth. The simplest method of increas- 
ing the bearing power is to dig deeper. Ordinary soils will bear 
more weight the greater the depth reached, owing to their becom- 
ing more condensed from the superincumbent weight. Depth is 
especially important with' cUy, since it is then less liable to be dis- 
placed laterally owing to other excavations in the immediate vicin- 
ity, and also because at greater depths the amount of moisture in it 
will not vary so much. 

In any soil, the bed of the foundation should be below the reach 
of frost. Even a foundation on bed-rock should be below the frost 
line, else irater may get under the foundation through fissures, and, 
freezing, do damage. 

28S. Drainage. Another simple method of increasing the bear- 
ing power of a soil is to drain it. The water may find its way to 
the bed of the foundation down the side of the wall, or by percol»- 



ovGoQi^lc 



196 OEDINABY FOUNDATIONS. [CHAP. X. 

Hon through the soil, or through a seam of sand. In most casea 
the bed can be safflciently drained by covering it with a layer of 
gravel — the thicknesa depending apon the plasticity of the soil, — 
and then surrounding the building with a tile-draiD laid a little 
below the foundation. In extreme cases, it is ucccssury to enclose 
the entire site with a puddle-wall to cut oft drainage water from a 
higher area. 

286. Spriuga. In laying foundations, springs are often met 
with, and sometimes prove very troublesome. The water may be 
excluded from the foundation pit by driving sheet piles, or by plug- 
ging the spring with concrete. If the flow is so strong as to wash 
the cement out before it has set, a heavy canvas covered with pitch, 
etc., upon which the concrete is deposited, is sometimes used ; or 
the water may be carried away in temporary channels, until the 
concrete in the artificial bed shall have set, when the water-ways 
may be filled with semi-fliiid cement mortar. Below is an account 
of the method of stopping a very troublesome spring encountered 
in laying the foundation of the dry-dock at the Brooklyn Navy 
Yard. 

"The dock is a basin composed of stone masonry resting on 
piles. The foundation is 42 foet below the surface of the gronnd 
and 3? feet below mean tide. In digging the pit for the founda- 
tion, springs of fresh water were discovered near the bottom, which 
proved to be very troublesome. The upward pressure of the water 
was BO great as to raise the foundation, however heavily it was loaded. 
The first indication of undermining by these springs was thescittUng 
of the piles of the dock near by. In a day it made a cavity in which 
a pole was run down 20 feet below the foundation timbers. luto 
this hole were thrown 150 cubic feet of stone, which settled 10 feet 
during the night ; and 50 cubic feet more, thrown in the following 
day, drove the spring to another place, where it burst through a 
bed of concrete 2 feet thick. This new cavity was filled with 
concrete, but the precaution was taken of putting in a tube so as to 
permit the water to escape ; stili it burst through, and the opera- 
tion was repeated several times, until it finally broke out through a 
heavy body of cement 14 feet distant. In this place it uudermined 
the foundation piles. These were then driven deeper by means of 
followers ; and a space of 1,000 square feet around the spring was 
then planked, forming a fioor on which was laid a layer of brick in 



ovGoQi^lc 



ART. 1.] -I'HE SOIL. 197 

dry cement, and on that s layer of brick set in mortar, and the 
fouDdation vaa completed orer alL Several vent-holes were left 
through the floor and the foandatioa for the escape of the water. 
The work was completed in 1851, and has stood well ever eince." * 

287. Conaolidating the Soil. A soft, clayey boU can be greatly 
improved by spreading a thin layer of sand, dry earth, or broken 
stone over the bed of the foundation and pounding it into the soil. 
If the soil is very soft, compacting the surface will be insufficient ; 
in this case the soil may be consolidated to a considerable depth by 
driving short piles into it. For this purpose small piles — say 6 
feet long and 6 inches in diameter — serve better than large ones ; 
and they can be driven with a band-maul or by dropping a heavy 
block of wood with a tackle attached to any nmple frame, or by a 
hand pile-driver (§ 335). They may be driven as close together a» 
neceasary, although 2 to 4 feet in the clear is nsnally Bufficient. 
The latter method of compacting the soil is &r more efficient .than 
poanding the surface. In the case of impact upon earth, the im- 
mediate layers are compressed at once, and by their inertia and 
adhesion to the surrounding soil they intercept the effect of the 
blow, and thus prevent the consolidation of the lower strata. Even 
though the effect of a blow is not communicated to any conEiderable 
depth, the heavy masses of masonry make themselves felt at great 
depth, and hence for heavy buildings it is necessary to consolidate 
the lower strata. This can be done most easily and most efficiently 
by driving piles (see Art. 3). 

In this connection it is necessary tp remember that clay is com- 
pressible, while sand is not. Hence this method of consolidating 
soils is not applicable to sand, and is not very efficient in soils 
largely composed of it. 

288. flaad Files. Experiments show that in compacting th& 
noil by driving piles, it is better to withdraw them and immediately 
fill the holes with sand, than to allow the wooden piles to remain. 
This advantage is independent of the question of the dnrability of 
the wood. When the wooden pile is driven, it compresses the soil 
an amount nearly or quite equal to the volume of the pile, and 
when the latter is withdrawn this consolidation remains, at least: 
temporarily. If the hole is immediately filled with sand this com- 

* DelkfleM'B Fonndatloni In CompreBdble Soils, p. It — a pamphlet pnbllsbed tij 
Uw Englaem'B UepaitmeBt of ttic D. 8. Annr- 



ovGoQi^lc 



198 OBDINABT FOCNDATIOXa [CKAP. t 

preeeion ib retained permanently, and the consolidation may be still 
farther increased by ramming the sand in in thin layers, owing to 
the ability of the latter to traosmit preaanre laterally. And further, 
the aand pile will support a greater load than the wooden pile; for, 
einco the sand acts like innumerable small arches I'eacbing from 
one side of the hole to the other, more of the load is transmitted to 
the soil on the sides of the hole. To secure the best resalts, the 
sand should be fine, sharp, clean, and of uniform size. 

889. When the piles are driven primarily to compact the soil, 
it is customary to load them and also the soil between them, either' 
by cutting the piles oS near the surface and laying a tight platform 
of timber on top of them (see § 320), or by depositing a bed of con- 
crete between and over the heads of the piles (see g 319). 

If the soil is very soft or con |iused largely of sand, this method 
is ineffective; in which case long piles are driven as close together 
as is necessary, the supporting power being derived either from the 
resting of the piles upon a harder substratum or from the buoyancy 
due to immersion in the semi-liquid soil. This method of securiug 
a foundation by driving long piles is very expensive, and is seldom 
resorted to for buildings, since it is generally more economical to 
increase the area of the foundation. 

290. Lsyen of Sand. If the soil is very soft, it may be ex- 
cavated and replaced by sand. The method of using sand for piles 
has been described in § 388, which see. The opportunities for the 
use of sand in foundations are numerous, and the employment of 
it would, in many constructions, promote economy and stability. 
The simplest method of using sand for this purpose is to excavate 
a trench or pit to the proper depth, and fill it by depositing succes- 
sive layers of sand, each of which should be thoroughly settled by 
a heavy beetle before laying the next. To cause the sand to pack 
firmly, it should be slightly moistened before being placed in tho 
trench. 

Sand, when used in this way, possesses the valuable property of 
assuming a new position of equilibrium and stability should the 
soil on which it is laid yield at any of its points ; not only does this 
take place along the base of the sand bed, but also along its edges 
or sides. The bed of sand must be thick enough to distribute the 
pressure on its upper surface over the entire base. There is no way 
of telling what this thickness should be, except by trial. 



ovGoQi^lc 



A.BT. 3.] DXBIOHINO THE FOOTIKG. 199 

291. The foUowicg exsmples, cited by Trantwine,* are interest- 
ing as shoving the enrpriBiog effect of even a thin la;er of sand 
or gravel : 

"Some portions of the circular brick aqaednct for supplying 
Boston with water gave a gieat deal of trouble when its Inches 
passed through running quicksaiids and other treacherous soils. 
Concrete was tried, but the wet quicksand mixed itself with it and 
killed it. Wooden cradles, etc., also failed ; and the diflBcnlty was 
overcome by aimply depositing in the trenches about two feet in 
depth of strong gravel.. 

" Smoatou mentions a stone bridge built upon a natural bed of 
gravel only about 3 feet thick, overlying deep mud so soft that an 
iron bar 40 feet long sank to the head by its own weight. Althongh 
a wretched precedent for bridge building, this example illnstrates 
the bearing power of a thick layer of well-compacted gravel." 



Art. 2, DBSiONiifQ the Footinq. 

S92. Load to bi Sottoktss. The first atep ie to aecertain the 
load to be supported by the foundation. This load consists of three 
parts : (1) the building itself, (2) the movable loads on the floors 
and the snow on the roof, and (3) the part of the load that may be 
transferred from one part of the foondation to the other by the 
force of the wind. 

293. The weight of the bnilding is easily ascertained by calcu- 
lating the cubical contents of all the various materials in the struct- 
ure. If the weight is not equally distributed, care must be taken 
to ascertain the proportion to be carried by each part of the foun- 
dation. For esample, if one vertical section of the wall is to con- 
tain a number of large windows while another will consist entirely 
of solid masoury, it is evident that the pressure on the foundation 
under-the first section will be less than that under the second. 

In this connection it must be borne in mind that concentrated 
pi-essu res are not transmitted, undiminished, through a solid mass 
in the line of application, but spread out in successively radiating 
lines ; hence, it any considerable distance intervenes between the 
foundation and the point of application of this concentrated load, 

I Pooket-book (od. 1861^ p. SS*. 



ovGoQi^lc 



200 



OEDIXA.BT FOUNQATIONB. 



[chap. z. 



the preBSure will be nearly or quite aniformly distribated over the 
entire area of the base. The exact diatribntion of the presBure can 
not be compnted. 

The following data will be nsefal in determining the weight ot 
the BtrnctDte ■ 

TABLE 24. 
WxiGBT or Uabokkt. 



Brick-work, pressed brick, llita Jolnta 

" ordloar; quality. 

Krft brick, thick Jofnta 

COQCrete, 1 cement, Bund, and 6 broken statie 

Granite— 6 per cent, more ihan the correapondlng Umestoae. . . 

Llmeetoae, ashlar, largest blocks and thinnest Joints 

" 13"lo20"coursesand|-tai'inch}olnts..„ 

" squared-stone (see § M6) 

" rubble, best 

Hortar, 1 Rosendale cement and S sand 

" common lime, dried 

Ssudstone— 14 per cent, less than the corre^tondiog limestone. 



Ordinary lathing and plastering weighs abont 10 lbs. per sq. ft. 
The weight of floors is approximately 10 lbs. per sq. ft for dwell- 
ings ; 25 lbs. per sq. ft. for public bnildinga ; and 40 or 50 lbs. per 
sq. ft. for warehouses. The weight of the roof varies with the kind 
of covering, the span, etc. A shingle roof may be taken at 10 lbs. 
per sq. ft., and a roof coTered with slate or corrugated iron at 25 
lbs. per sq. ft. 

2S4. The movable load on the floor depends upon the nature ot 
the building. For dwellings, it does not exceed 10 lbs. per sq. ft.; 
for large office buildings, it is usually taken at 30 lbs. per sq. ft. ; 
for chnrches, theatres, etc., the maximum load — a crowd of people 
— may reach 100 lbs. per sq. ft ; for stores, warehonses, factories, 



ovGoQi^lc 



ART, 2.] DSBIGNIXa THE FOOTINO. SOI 

«tc., the load will be from 100 to 400 Iba. p«r sq. ft., according to 
the porpoees for which they are used. 

The preceding loada are the ones to be need in determining the 
atrength of the floor, and not in designing the footings; for there 
is no probability that each and every square foot of floor will hare 
its maximnm load at the same time. The amount of moving load 
to be aesigned in any particalar case is a matter of jndgmeat. At 
Chicago in designing tall steel-skeleton office bnildings, it is the 
practice to assume that nearly all of the maximnm lire load reaches 
the girders, that a smaller per cent, reaches the colnmns, and that 
no live load reaches the footings. In many cities the bnilding law 
■peoifiee the live load to be assamed as reaching the footing. 

Attention must be gi?en to the manner in which the weight of 
the root and floors is traiuderred to the walls. For example, if the 
floor joists of a warehoase ran from back to front, it is erident that 
the bock and front walls alone will carry the weight of the floors 
and of the goods placed upon them, and this will make the pressnre 
npOQ the foundation under them considerably greater than under 
the other walls. Again, if a Btone<front is to be carried on an arch 
or on a girder having its bearings on piers at each side of the bnild- 
ing, it is manifest that the weight of the whole BuperiQCumbent 
structure, instead of being distributed equally on the foundation 
under the front, will be concentrated on that part of the founda- 
tion immediately under the piers. 

296. The pressure of the wind against towers, tall chimneys, 
etc, will cause a concentration of the weight of the structure upon 
one side of the foundation. The maximum horizontal pressure of 
the wind is usually taken as 50 lbs. per sq. ft. oa a flat surface per- 
pendicular to the wind, and on a cylinder at about 30 lbs. per sq. 
ft. of the projection of the surface. The pressure upon an inclined 
Bortace, as a roof, is about I lb. per sq. ft. per degree of inclination 
to the horizontal. For example, if the roof has an inclination of 
30° with the horizontal, the pressure of the wind will be about 30 
lbs. per sq. ft 

The effect of the wind will be considered in §g 301-1. 

299. Abx& BwtlTIBBD. Having determined the pressure which 
may safely be brought upon the soil, and having ascertained the 
weight of each part of the structure, the area required for the foun- 
dation is easily determined by dividing the latter by the former. 



ovGoQi^lc 



303 ORDINABY FODNDATIONa [CHAP. JL. 

Then, having found the area of fonndation, the base of the etruct- 
nre must be extended by footings of masonry, conorete, timber, 
etc, BO as to (1) cover that srea and (2) distribute the pressure uni- 
formly over it. The two items will be considered in inverse order. 
297. Cevteb 07 Pbebsoxb m Cevteb op Basb. In construct- 
ing a foundation the object is not so much to secure an absolutely 
unyielding base as to secure one that will settle as little as possible, 
and uniformly. All soils will yield somewhat under the pressure of 
any building, and even masonry itself is compressed by the weight 
of the load above it. The pressure per square foot should, there- 
fore, be the same for all parts of the building, and particularly of 
the foundation, so that the settlement may be uniform. This can 
be secured only when the axis of the load (a vertical line through 
the center of gravity of the weight) passes through the center of 
the area of the foundation. If the axis of pressure does not coincide 
exactly with the axis of the base, the ground will yield most on the 
side which is pressed most ; and as the ground yields, the base as- 
sumes an inclined position, and carries the lower part of the struct- 
ure with it, thus producing unsightly cracks, if nothing more. 

The coincidence of the axis of pressure with the axis of resist- 
ance is of first importance. This principle is self-evident, and yet 
the neglect to observe it is the most frequent cause of failure in the 
foupdations of buildings. 

Fig. 50 is au example of the way in which this principle is 
violated. The shaded portion 
represents a heavily loaded exte- 
rior wall, and the light portion a 
lightly loaded interior wall. The 
foundations of the two walls are 
rigidly connected together at 
their intersection. The center 
of the load is under the shaded 
Fia.so. section, and the center of the 

area is at some point farther to the left ; consequently the exterior 
wall is caused to incline outward, producing cracks at or near the 
comers of the building. Doubtless the two foundations are con> 
nected in the belief that an increase of the bearing surface is of firat 
importance ; but the true principle is that the coincidence of the 
axis of pressure vrith the axis of resiBtance is the most important. 



ovGoQi^lc 




12T. 3.] B^smmsa the pootiko. 303 

Fig. 51 is another illostrstion of the same priociple. The foao- 

dation is continDOua nnder the opeDiog, 

and hence the center of the fdnDdation is to 

the left of the center of preseore ; conse- 
quently the wall inclines to the right, pro- 

dncing cracks, usually over the opening.* 
298. The center of the load can he made 

to tall inside of the center of foundation by 

extending the footings outwards, or by c 

tailing the foundations on the inside. The - 

latter finds exemplification in the properly Fia.m. 

constructed foundation of a wall containing a number of openings. 

For example, in Fig. 52, if the foundation is uniform under the 
entire front, the center of pressure must be 
outside of the center of the base ; and conse- 
quently the two side walls yill incline outward, 
and show cracks over the openings. If the 
width of the foundation under the openings 
be decreased, or if this part of the foundation 
be omitted entirely, the center of pressure 

n. will fall inside of the center of base and the 
walls will tend to Incline inwards, and hence 
'•-\ be stable. 

289. CoudniiouB. One conclusion to be 

na.Ba. drawn from the above examples is ttiat the 

foundation of a wall should never be connected with that of anotlier 

wall either much heavier, or much lighter than itself. Both are 

equally objectionable. 

A second conclusion is that the axis of the load should strike a 
little inside of the center of the area of the base, to make sure that 
it will not be outside. Any inward inclination of the wall is ren- 
dered impossible by the interior walls of the building, the floor- 
beams, etc. ; while an outward inclination can be counteracted only 
by anchors and the bond of the masonry. A slight deviation of the 
axis of the load outward from the center of the base has a marked 
effect, and is not easily counteracted by anchors. 

■ For an acconnt showing the Tlolotlon of thla principle Id the constmctJon ol 
the Cooper InatltDte Bnfldlng, New York Citjr, and the method used to retnedf lt,aee 
Hmtforv Bngtiieir, ToL XlL pp. 460-68. 




ovGoQi^lc 



204 



OBDtNABT FOUNDAnOSB. 



[OHAP. X. 

The above coucIiuioiiB may be Banunarized in the folloiring 
principle : All foundations should be so constructed as to compress 
the ground slightly ooitcatb upwards, raiher than oomvex up- 
wards. On even eligbtly compressible Boils, a Bmall difference in 
the presanre on the foundation will be eafficient to cause the bed to 
become conves upwards. At Chicago, an omission of 1 to 2 per 
cent of the weight (by leaving openings) nsoaUy caasee sufficient 
convexity to prodace unsightly cracks. With very dight differences 
of pressure on the foundation, it is safflcient to tie the building 
together by careful bonding, by hoop-iron built in over openings, 
and by heavy bars bnilt in where one \to1I joins another. 

300. TXBStVxman Fibbs. The art of constructing foanda- 
tions on compressible soil has been brought to a high degree of 
development by the architects of Chicago. The special feature of 
the practice in that city is what is called " the method of independ- 
ent piers ;" that is, each tier of columns, each pier, each w^l, etc., 
has its own independent foundation, the area of which is propor- 
tioned to the load on that part.* The interior walls are fastened to 
the exterior ones by anchors which slide iu slots. For a detailed 
account of the methods employed in one of the best and largest 
buildings erected there, see Sanitary Engineer, Dec. 10, 1885. 

301. Sttect or the Wtn. Overturning. The preceding dib- 
onssion refers to the total weight that is to come upon the foun- 
dation. The pressure of the wind against towers, 
tall chimneys, etc., transfers the pointof applica- 
tion of the load to one side of the foundation. The 
method of compnting the position of the center 
of the pressure on the foundation under the action 
of the wind is illustrated in Fig. 53, in which 

ABED represents a vertical section of the 
tower; 

d is a point horizontally opposite the center of 
the surface exposed to the pressure of the 
wind and vertically above the center of grav- 
no.u. ity of the tower; 

* Thia metfaod wh lint made koown to the pnblia bf Frederick Bonnun, at Chi- 
cago, tn a pamphlet entitled "The Method ot ConitnicUiig FonndftUoDa oolscdated 
Pfen," published b; him In 1872. The above examples and principles amfrom that 




ovGoQi^lc 



ABT. 8.] DESiaKING THE FOOTINQ. 200 

C is the position of the center of pressure Then there ia no wind ; 

jV IS the center when the wind is acting. 
For convonieiice, let 

P = the maxim am preesare on the fonndation, per unit of area; 

p = the pressure of the wind per unit of area (see % 295); 

H = the total preasare of the wind against the exposed surface ; 

W = the weight of that part of the stracture above the sectioa 
considered, — in this case, A B ; 

3 = the area of the horizontal cross section ; 

/ = the moment of inertia of this section j 
I = the distance A B ; 

h = the distance a C\ 

d = the distance N C; 
if = the moment of the wind. 

When there is no horizontal force acting, the load on ^ fi is 
uniform ; but when there is a horizontal force acting — as, for ex- 
ample, the wind blowing from the right, — the pressure is greatest 
near A and decreases towards B. To find the law of the variation 
of this pressure, consider the tower as a cantilever beam. The 
masimum pressure at A will be that due to the weight of the tower 
plus the compression due to flexure ; and the pressure at B will be 
the compression due to the weight minus the tension doe to flexnre. 

The antform pressure dne to the weight is -^. The strain at A dns 

to flexnre is, by the principles of the resistance of material^ -aj. 
Then the mazimnm preeenre per unit of area at ^ ii 
._ W Ml 

'^-s+n- « 

aod the minimTini pteenure at B is 

-=^-fJ- m- 

Equations (1) and (S) are pertectl; general ; tb^ are applioabls 
to any cross section, and also to any system of horizontal and Te^ 
tical forces. In succeeding chapters they will be employed in 
finding the unit preesare in masonry dams, bridge piers, arohe^ 
etc 



ovGoQi^lc 



806 OBDUTABT BOCSDi^OSB. \CBXr. X. 

The valne of 7 in the above lormnlas is girsD in Fig. 54 for the 
sections occurring most frequently in practice, Notice that / is the 
dimension parallel to the direction of the wind, and b the dimen- 
mon perpendicular to the directioQ of the wind. 




302. If the ares of the Bectiou A B, Fig. 53, is a rectangle 
B=lb, »XiAI=-^bV. Substituting these values in equation (1) 
gives 

'■=n+'^ (') 

The moment of the wind, M, is eqnal to the product of its total 
pressure, S, and the distance, h, of the center of pressure above 
the horizontal section considered ; ot M= H.h. H ia equal to 
the presenre per nnit of area, p, multiplied by the area of the sur- 
face exposed to the pressure of the wind. Substituting the above 
value of Jfin equation (3) gives 

To still further simplify the above formula, notice iiiat Fig. 63 
gives the proportion 

S: W::NCiaC, 
from which 

E.aC^ W.NC; 
or, ohanging the nomenclature, 

H\= Wd. 
Kotioe that the last relation can also be obtained directly hj the 
principle of moments. Substituting the value ot H. h, as sJimve, in 
equation (4) gives 

wllioh in a ofrnmuent form for practic&l application. 



jvGooi^le 



AXF. S.] DB8IGVLN0 THS TOOTIHO. 207 

Ad exsmiuation of eqnation (6) shows that when d = Jf C = ^t, 
the maximiim presanre at A is twice the average. Notice also that 
Doder theae conditionB the pressure at £ is zero. This is equiva- 
lent to what is knowB, in the theory of arohes, as the principle of 
the middle third. It shows that ae long as the center of pressnre 
lies in the middle third, the msximnm pressnre is not more than 
twice the average pressure, and that there is no tension at B. 

The above discussion of the distribution of the pressure on the 
foundation is amply sufQcient for the case in hand ; however, the 
subject is discosaed more fully io the chapter on Stability of Masonry 
Dame (see Chapter XIII). 

303. The average pressure per unit an A B has already been 
adjusted to the safe bearing power of the soil, and if the maximum 
pressnre at A does not exceed the ultimate bearing power, the occa- 
sionat maximnm pressure due to the wind will do no barm ; but if 
this maximnm exceeds or is dangerouBly near the ultimate strength 
of the soil, the base must be widened. 

304. Sliding, llie pressure of the wind is a force tending to 
elide the foundation horizontally. This is resisted by the friction 
caneed by the weight of the entire structure, and also by the earth 
around the base of the foundation, and hence there is no need, in 
this connection, of considering this manner of failure. 

306. DznOKiae TEZ PoOTJve. The term footing is nsnally no- 
deretood as meaning the bottom course or coarses of masonry which 
extend beyond the &ces of the wall. It will be need here as apply- 
ing to the material — whether maeonry, timber, or iron — employed 
to.increaee the area of the base of the fonndation. Whatever the 
character of the soil, footings should extend beyond the face of the 
wall (1) to add to the stability of the structare and lessen the dan- 
ger of the work's being thrown out of plumb, and (3) to distribute 
the weight of the stmcture over a larger area .and thus decrease 
the settlement due to the compression of the ground. To serve 
the first purpose, footings must be securely bonded to the body of 
the wall ; and to produce the second effect, they most have sufficient 
strength to resist the transverse strain to which they are exposed. 
In ordinary bnildings the distribution of the weight ie more impor- 
tant than adding to the resistance to overturning, and hence only 
the former will be considered here. 

The area of the foundation may be increased until the inherent 



ovGoQi^lc 



208 OBDINABT FOUNDATIONS. ^CUAP. X. 

bearing power of the area covered is sufficient to support the load 
(1) by extending the bottom courses of niaaonry, or (2) by tlie sae 
of one or more layere of timbers, railroad rails, or steel I-beams, or 
(3) by resting the strnctare upon inverted masonry arches. 

306. Off'Seti of Moionry Footings. The area of the fonndstion 
having been determined and its ceuter having been located with 
reference to the axis of the load (g 297), the next step is to deter- 
mine how much narrower each footing course may be tlutn the one 
next below it. The projecting part of the footing resists as a beam 
fixed at one end and loaded unUormly. The load is the pressure 
on the earth or on the courne next below. The oS-set of such a 
couree depends upon the amount of the pressure, the transverse 
strength of the material, and the thickness of the course. 

To deduce a formula for the relation between theae qnantitieB, 
let 

P = the pressure, in tons per square foot, t^t the bottom of the 
footing courae under consideration ; 

S = the modulus of rupture of the material, in pounds per 
square inch ; 

p = the greatest possible projection of the footing course, in 

inchea ; 
/ = the thicknees of the footing course, in inches. 

The part of the footing course that projects beyond the one above 
it, is a cantilever beam uniformly loaded. From the principles of 
the resiatauce of materials, we know that the upward pressure of the 
earth against the part that projects multiplied by one half of the 
length of the projection is equal to the continued product of one 
sixth of the modulus of rupture of the material, the breadth of the 
footing course, and the square of the thickness. Expressing this 
relation In the above nomenclature and reducing, we get the for- 
mula 






or, with Bufflcient acouracj, 

p=ii/^ m 

Hence the projection available with any given thicknesa, or the 
thickness required for any given projection, ma; easily be computed 



ovGoQi^lc 



ABT. 2.] 



DESmSIKQ THE FO0TI2fQ. 



209 



by equation (7). Notice that with the off-set given by the above 
formala the stone would be on the point of breaking. 

SOT, The margin to be allowed for safety will depend npon the 
care used in computing the loads, in solocting the materials for the 
footing courses, and in bedding and plucing them. If all the loads 
have been allowed for at their probable maximum value, and if the 
material is to be reasonably uniform in quality and laid with care, 
then a comparatively small margin for safety is sufficient ; but if 
all the loads have not been carefully computed, and if the job is to 
be done by an unknown contractor^ aud neither the material nor 
the work is to be carefully inspected, then a large mai^in is neces- 
sary. As a general rnle, it is better to assume, for each particular 
case, a factor of safety in accordance with the attendant conditions 
of the problem than blindly to use the results deduced by the 
application of some arbitrarily assumed factor. The following table 
is given for the conrenience of those who may wish to use 10 as a 
factor of safety. 

TABLE 25. 
Save Off^aet tob Hasomrt Foorras Coitrsbs, in Tsbub of thz Thick- 

MESS OF THE CotlBSE, USINO 10 AS A FaCTOB OF BAyKTT. 

For limitations, see § 808. 



B1u»«toDe flagglag iaee page 18). . 

Onniie(EeepBge 18) 

Linieslone (ai-e |uig« IS) 

SiiDtlBtoDe (see piige 18) 

Slate (He pnge 18) 



Bent Hard Brick (we page* 40. 41) 

Huril Brick (see pages 4U, 41) 

1 1 Forlland ) 

Coucrete (see page llSt) i'i sand )■ 
(8 pebbles ) 
1 1 Fonltuid > 

Concrete (see page IISp) -(B sand [■ 

< 5 pebbles ) 



To illoBtrate the method of using the preceding table, assume 
that it is desired to determine the off-set for a limestone footing 
course when the pressure on the bed of the foundation is 1 ton per 
square foot, using 10 as a factor of safety. In the table, opposite 



ovGoQi^lc 



310 OBOmABT FOUITDATIONS. [CHA.P. X. 

limestone, ia next to the lut ccdnmn, ve find ih» qnantit; 1.9. 
This BhowB that, ander the conditions stated, the ofl-Bet may be 1.9 
times the thickneaa of the conne. 

The Talnea in the table a^ree Tery veil with the practioe of th« 
principal architects and engineen for hammer-dreeaed stones laid 
in good cement mortar. 

If it ie desired to use any other factor of safety, it is only neces- 
sary to BubstitaU for R, in the preceding formula, the desired frac- 
tional part of that quantity as given in the second colnnm of the 
above table. For example, aasnme that it is necessary to use lime- 
stone in the foundation, and that it is required to draw in the foot- 
ing courses as rapidly as possible. Assome also that the pressun, 
P, on the base of the foundation is 3 tone per square foot. If the 
limestone is of the best, and if it is laid with great care, it will be 
sufficient to use 1 as a factor of safety. Under these condition^ 
equation (7) as above gives 

That is, the projection maybe 2.3 times the thickness of the course. 

308. Strictly, the above compntations are correct only for the 
lower ofF-set, and then only when the footing is composed of stones 
whose thickness is equal to the thickness of the course and which 
project less than half their length, and which are also well bedded. 
The resistance of two or more conrses to bending varies as the square 
of their depth, and the bending due to the aniform pressure on the 
base will also increase as the square of the sum of the projections, 
and therefore the suocessive ofE-sets should be proportional to the 
thickness of the course; or, in other words, the Talnes as above are 
applicable to any course, provided no stone projects more than half 
ito length beyond the top couise. 

The preceding results will be applicable to built footing courses 
only when the pressure above the coarse is lees than the safe strength ' 
<d the mortar (see § 133 and g 157a). The proper projection for 
rubble masonry lies somewhere between the values given for stone 
and those given for concrete. If the rubble consists of large stones 
well bedded in good strong mortar, then the values for this class of 
masonry will be but little less than those given in the table. If the 
rabble oonsistB of amall irregular stones laid with PortUnd or nat- 



ovGoQi^lc 



ART. 2.J DESIQITINQ THE FOOTOra. ZU 

nral cement mortar, the projection should not much exceed that 
giren for concrete. It the mbble is laid in lime mortar, the pro- 
jection of the footing coarse sboold not be more than half that 
alloved when cement mortar is used.' 

309. Timber Footing. In very soft earth it would be inexpe- 
dient to use masonry footiugB, since the foundation would be Tury 
deep or occupy the space usually devoted to the cellar. One method 
of OTercoming this difficulty consista in conBtrncting a timber grat- 
ing, sometimes called a grillage, by setting a aeries of heavy timbers 
firmly into the soU, and laying another series transversely on top of 
these. The timbers may be fostened at their intorsectionB by spikes 
or drift-bolts (g 381) if there is any possibility of sliding, which is 
unlikely in the class of foundations here considered. The earth 
should be packed in between and around the serera] beams, A 
flooring of thick planks, often termed a platform, is laid on top of 
the grillage to receive the lowest course of masonry. In extremtt 
cases, the timbers in one or more of the courses are laid close to- 
gether. Timber should never be used except where it will be alvrays 
wet. 

The amount that a course of timber may project beyond the one 
next above it can be determined by equation (7), page 208. Making 
R in that equation equal to 1,000 — the value ordinarily used, — and 
solving, we obtain the following results for the safe projection: If 
the pressure on the foundation is 0.5 ton per square foot, the safe 
projection is 7.5 times the Uiickness of the coarse ; if the pressure 
is 1 ton per square foot, the safe projection is 5.3 times the thick- 
ness of the coarse ; and if the pressure is 2 tons per square foot, the 
safe projection is 3.7 times the thickness of the course. The above 
values give a factor of safety of about 10. To use any other factor, 
insert in equation (7), above, the corresponding fractional part of the 
ultimate transverse strength of the particular timber to be used, 
and solve. 

The above method of computation is not applicable to two or 
more conrses of timber, if one is transverse to the other. 

310. This method of increasing the area of the footing is much 
osed at New Orleans. The Custom-house at that place is founded 
npon a 3-inch plank flooring laid 7 feet below the street pavement, 
A grillage, consisting of timbers 13 inches square laid side by side, 
is laid npon the floor, over which similar timbers are placed tranr 
Tersely, 2 feet apart in the clear. 



ovGoQi^lc 



SIS OBDINAKT P0UHDATI0N8, [CHAP. X. 

Most of the bnildinga of the World's GolninbUn Exposition, 
Chicago, 1893, were tonnded npon timber footings. 

311. Steel Footing. Veiy recently, steel, usnallj in the form 
of railrottd rails or I-beams, baa been need instead of timber ia 
fonndatioQB. The rails or I-boams are placed side by side, and 
concrete is rammed in between them. 

Steel '\a superior to timber for this pnrpoee, in that the latter 
oan be used only where it is always wet, while the former is not 
affected by Tariations of wetness and dryness. Twenty years' ex- 
perience in this nse of steel at Chicago shows that after a short time 
ihe snrfane of the metal becomes encased in a coating which pre- 
vents further oxidation. The most important sdvant^e, howeTor, 
in this use of steel is that the oS7set can be much greater with steel 
than with wood or Btone; and hence the foandations may be shal- 
low, and still not occupy the cellar space. 

The proper projectioos for the steel beams can be oompnted by 
ft formula somewhat similar to that of g 306 ; bnt the steel footing 
Ib appropriately a part of the steel-skeleton constmction, and hence 
will not be considered here. For a presentation of the method 
of compntations formerly employed in Chicago, see Engineering 
Newt, Tol. xxri. page 133; and for adverse criticisms thereon, see 
md.f pages SfSS, 312, 415, and toI. xxxii. page 387. Concerning 
the effect of the strength of the base of the column, see Johnson's 
"Modern Framed Structures," pages 444—46. For a diHcnssioti 
which considers the deflection of Ihe severul beams, see Engineering 
Record, Tol. xxxix. pages 333-34, 354^56, 383, 407-8. The last 
is the most exact method of analysis, and also seonres the greatest 
economy of material. 

31S. XuTerted Aroh. Inrerted arehes are frequently bnilt nnder 
and between the bases of piers, as shown in Fig. 55. Employed in 
this way, the arch simply distributes 
the preeenre over a greater area; but 
it is not well adapted to this use, for 
Ut'TKaxi::'Ti^'' [jvt r''-- ji'-''^^« (^) '' ^^ nearly impossible to prevent 
^^r-^^^^^^'^'f^^^^4^h ^^ ^^^ piers of a series from being 
pushed outward by the thrust of the 
"^"^ arch, and (2) it is generally impos- 

sible, with inverted arches, to make the areas of the different parts 
of the foondatioi! proportional to the load to be supported (see g 



ovGoQi^lc 



AST. 3.] FBBPASISS TSB BED, 21S 

297). The only advantage the inverted arch has over msBonij 
footingB is in the Bhallover foundation obtained. 

313. In a few caeee masonry pierB have been ennk to a solid snb- 
stratam by excavating the material from the inside, and then resting 
arches on these piore. This is an ezpeosive method, and has essen- 
tially the same objectiona as the inverted arch. 

A&T. Z. Feepabino the Bed. ' 

314. Ok Bock. To prepare a rock bed to receive a foundation 
it is generally only necessary to cnt away the loose and decayed por- 
tions of the rock, and to drees it to a plane snr&ce as nearly perpen- 
dicular to the direction of the pressure as is practicable. If there 
are any fiaaareB, they should be filled with concrete. A rock that 
is very mnch brokeo can be made amply secnre tor a foundation b^ 
the liberal use of good cement concrete. The piers of the Niagara- 
Caotilever Bridge are founded upou the top of a bank of bowlders^ 
which were first cemented together with concrete. 

Sometimes it is neQeesary that certain parts of a structure 
start from a lower level than the others. In this case care should 
be taken (1) to keep the mortar-joints as thin as possible, (2) to lay 
the lower portions in cement, and (3) to proceed slowly with the 
work ; otherwise the greater quantity of mortar in the wait on the 
lower portions of the slope will canse greater settling there and a 
consequent breaking of the joints at the stepping- places. The 
bonding over the off-sets should receive particular attention. 

316. Ok Pixk Eabte. For foundations in such earths as hard 
clay, clean dry gravel, or clean sharp sand, it is only necessary to 
dig a trench from 3 to 6 feet deep, so that the foundation may be 
below the disintegrating effect of frost. Provision should also be 
made for the drainage of the bed of the foundation. 

With this class of foundations it often happens that one part of 
the structure starts from a lower level than another. When this is 
the case great care is required. All the precantions mentioned in 
the second part^raph of § 314 should be observed, and great care 
should also be taken so to proportion the load per unit of area that 
the settlement of the foundation maybe uniform. This is difficnlt 
to do, since a variation of a few feet in depth often makes a great 
difference in the supporting power of the soiL 



ovGoQi^lc 



'iU OBDINABT VOUVDATIOKB. [CHAP. Z. 

316. Ik Wn OBono. The difflcalty in boiIb of tbie class is in 
dispoaing of the water, or in prevuntiag the semi-liquid soil from 
rnnDing into the excavation. The difficulties are similar to those 
met with in couBtnicting ronndations under water — see Chapter XIL 
Three general methods of laying a foundation in this kind of soil 
will be hriefly deacribed. 

317. Coffer-Sam. If the soil is only moderately wet — ^not satu- 
rated, — it is sufficient to inclose the area to be ezcayated with sheet 
pilee (boards driren vertically into the ground in contact with each 
other). This curbing is a simple form of a coffer-dam (Art. 1, 
Chap. XII). The boerds should- be sharpened wholly from one 
aide; this point being placed next to the last pile driven causes 
them te crowd together and make tighter joints. The sheeting may 
be driven by hand, by a heavy weight raised by a tackle and then 
dropped, or by an ordinary pile-driver (§§ 336-36). Unless the 
amount of water is quite small, it will be necessary to drive a double 
row of sheeting, breaking joints. It will not be possible to entirely 
prevent leaking. The water that leaks in may be bailed out, or 
pumped — either by hand or by steam (see g 395). 

To prevent the sheeting from being forced inward, it may be 
braced by shores placed horizontally from side te side and abutting 
against walee (horizontal timbers which rest against the sheet piles). 
The bracing is put in snccessiyely from the top aa the excavation 
proceeds ; and as the masonry is built up, short braces between the 
sheeting and the masonry are substituted for the long braces which 
previously extended from side te side. Iron screws, somewhat 
similar to jack-screws, are used, instead of timber shores, in exca- 
vating trenches for the foundations of buildings, for sewers, ete. 

If one length of sheeting will not reach deep enough, an addi- 
tional section can be placed inside of the one already In position, 
when the excavation has reached a sufficient depth to require it. 
Ordinary planks 8 to 13 inches wide and H or 3 inches thick are 
used. 

For a more extended account of the use of coffer-dams, see 
Chapter XII — Foundations Under Water, Art. 1 — Coffer-Dams. 

316. In some oases the soil is more easily excavated if it is first 
drained. To do this, dig a hole — a sump — inte which the water will 
drain and from which it may be pumped. If necessary, several 
anmpa may be sunk, and deepened as the excavation proceeds. 



ovGoQi^lc 



JlBI. 3.] PRBPABISQ THE BED, 216 

Qaicksand or soft alluTinm may aometimee be pamped out along 
with the water bj a centrifugal or a mud pump (g 395 and g 448). 
Od large jobs, each material is sometimea taken out with a clam- 
Bhell or orange-peel dredge (§ 412). 

319. Concrete. Concrete can frequently be used advantage- 
ously in fouudations in wet soils. If the water can be removed, the 
concrete should be deposited in continuous layers, about 6 inches 
thick, and gently rammed until the water begins to ooze out on the 
upper surface (see g 158/). If the water can not be removed, the 
concrete may be deposited under the water (see § IMq), although it 
is more difficult to insure good results by this method than when 
the concrete is deposited in the open air.* 

380. Orillage. If the semi-liquid soil extends to a considerable 
depth, or if the soft soil which overlies a solid substratum can not be 
removed readily, it is customary to drive piles at uniform distances 
over the area, and construct a grilli^e (see § 380) on top of them. 
This construction is very common for bridge abutments (Chapter 
XV). The piles should be sawed off (§ 378) below low-water, which 
usually necessitates a coSer-dani (g 317, and Art. 1 of Chapter XII), 
and the ezcavatiou of the soil a little below the low-water line. 

For a more extended account of this method of laying a founda- 
tion, see §§ 360-82. 

321. In excavating shallow pita in sand conteining a small 
amount of water, dynamite cartridges have been soccesef nlly used to 
drive the water out A hole is bored with an ordinary auger and 
the cartridge inserted and exploded. The explosion drives the wator 
hack into the soil so far that, by working rapidly, the hole can be 
excavated and a layer of concrete placed before the water returns. 

322. CovcLiruoir. It is hardly worth while here to discuss this 
subject further. It is one on which general instruction can not be 
given. Each case must be dealt with according to the attendant 
circumstances, and a knowledge of the method best adapted to any 
given conditions comes only by experience. 



ovGoQi^lc 



CHAPTER XL 
PILE FOUNDATIOHS. 

3S3. DllunriOHL File. Although a pile is generally tmdeh 
stood to be a roand timber driveD into the soil to support a load, 
the term has a variety of applications which it will be well to explain. 

Staring Pile. One used to sustain a vertical load. This is the 
ordinai'y pile, and usually is the one referred to when the word pile 
is employed without qualificatios. 

Sheet Piles. Thick boards or timbers driven in close contact 
to inclose a space, to prevent leakage, etc. Generally they are con- 
siderably wider than thick; but are sometimes square, in which case 
they are often called close piles. 

False Pile. A timber added to a pile after driving, to supple- 
ment its length. 

Foundation Pile. One driven to increase the supporting power 
of the soil under a foundation. 

Screw Pile. An iron shaft to the bottom of which is attached 
a broad-bladed screw having only one or two turns. 

Disk Pile. A bearing pile near the foot of which a disk is keyed 
or bolted to give additional bearing power. 

Pneumatic Pile. A metal cylinder which is sunk by atmoB-- 
pherio pressure. This form of pile will be discussed in the next 
chapter (see g 431). 

Abt. 1. DESCKiFnoNs, AND Ukthodb of Driving. 

SH. Isov Fueb. Both cast and wrought iron are employed for 
ordinary bearing piles, sheet piles, and for cylinders. Iron cylin- 
ders are generally sunk either by dredging the soil from the inude 
(g 415), or by the pnenmstic process (eee the next chapter, particu- 
larly g§ 431-35). For another method of employing iron cylinders, 
see gg 364r^. 



ovGoQi^lc 



iBT. 1.] DBsoBiPnoira, and xsthods ov dbitinq. 'in 

C^t-iron piles are b^nniag to be nsed as eabstitatea for com- 
mon wooden ones. Lnge or flanges are Qsuallj cast on the udes of 
the piles, to which bracing may be attached for secaring them in 
position. A wood block is laid upon the top of the pile to receive 
the hammer used in driving it; and, after being driven, a cap with 
a socket in its lower side is placed upon the pile to receive the load. 
The supporting power la sometimes increased by keying on an iron 
disk. The advantages claimed for cast-iron piles are: (I) they are 
not subject to decay; (2) they are more readily driven than wooden 
ones, especially in stony ground or stifl clay; and (3) they possess 
greater cmshing etrength, which, however, is an advantage only 
when the pile acts as a column (see § 355). The principal disadvan- 
tage is that they are deficient in transverse resistance to a suddenly 
applied force. This objection applies only to the handling of the 
piles before being driven, and to such as are liable, after being driven, 
to sudden lateral blows, as from floating ice, logs, etc 

Recently, rolled sections of wroaght-iron have been employed to a 
limited degree for bearing-piles, but present prices prohibit an ex- 
tended use of wrought-iron piles. It is possible that iron may take 
the place of wood for piles where they are alternately wet and dry, 
or where they are ditBcalt to drive; but where the piles are always 
wet — aa is usually the case in foundation work, — wood is as dnrable 
as iron; and hence, on account of cheapness, is likely to have the 
preference. 

326. BOBEW Puis. Tbesearegenerally wholly of iron, although 
the stem is sometimes wood. The screw pile usually consists of a 
rolled-iron shaft, 3 to 8 inches in diameter, having at its foot one or 
two turns of a cast-iron screw, the blades of which may vary from 1^ 
to S feet in diameter. The piles ordinarily employed for light- 
honsea exposed to moderate seas or to heavy fields of ice have a 
shaft 3 to 5 inches in diameter and blades 3 to 4 feet in diameter, 
the screw weighing from 600 to 700 pounds. For bridge piers, 
the shaf fas are from 6 to 8 inches and the blades from 4 to 6 feet in 
diameter, the screw weighing from 1,500 to 4,000 pounds. 

Screw piles were invented by Mitchell of Belfast, and are largely 
need in Europe, but not to any great extent in this country. They 
have been nsed in founding small light-houses on the sea-shore, for 
signal stations in marine surveying, for aochonge for buoys, and 
for various purposes inland. 



ovGoQi^lc 



218 PILE POONDATIOSB. [CHAP. XI. 

For founding beacons, etc. , the screw pile has the. Bpecial advaQ- 
tage of not being drawn out by the upward force of the waves against 
the euperstructare, Eren when all cohesion of the ground is de> 
Btroyed in screwing down a pile, a conical mass, with its apex at the 
bottom of the pile and its base at the surface, wonld have to be 
lifted to draw the pile out. The supporting power also is consider- 
al 1 ; owing to the increased bearing surface of the screw blade. 
Screw piles have, therefore, an advantage in soft soil. They could 
also be used advant^eously in situations where the jar of driving 
ordinary piles might disturb the equilibrium of adjacent atructurea. 

326. These piles are usually screwed into the soil by men work- 
ing with capstan bara Sometimes a rope is wound around the 
shaft and the two ends pulled in opposite directions by two capstans, 
and sometimes the screw is turned by attaching a large cog-wheel to 
the shaft by a friction-clutch, which is rotated by a worm-screw 
operated by a hand crank. Of course steam or horse>power conld 
be used for this purpose. 

The screw will penetrate most soil . It will pass through loose 
pebbles and stones without much difficulty, and push aside bowlders 
of moderate size. Ordinaryclay does not present much obetmction; 
cbeein, dry sand gives the moat difficulty. The danger of twisting 
off the shaft limita the depth to which they may be sunk. Screw 
piles with blades i feet in diameter have been screwed 40 feet into 
a mixture of clay and sand. The resistance to sinking increases 
very rapidly with the diameter of the screw; but under favorable 
(drcumstances an ordinary screw pile can be sunk very quiCkly. 
Screws 4 feet in diameter have, in less than two hoars, been sunk 
by hand-labor 90 feet in sand and clay, the surface of which was 
20 feet below the water. For depths of 15 to 20 feet, an average of 
4 to 8 feet per day is good work for wholly hand-labor. 

For an illnstrated and detailed account of the founding of a rail- 
road bridge pier on screw piles, see Engineering News, Vol. XIIL 
pp. 310-13. 

327. Due Filbs. These differ but little from screw piles, a 
flat disk, instead of a screw, being keyed on at the foot of the iron 
stem. Disk piles are sank by the water-jet (g 343). One of the few 
cases in which they have been used in this country was in founding 
an ocean pier on Coney Island, near Kew York City. The shafts 
were wrought-iron, lap-welded tubes, 8f inches outside diameter, in 



ovGoQi^lc 



ABT. 1.] DBSCBIPTIONS, AND UETHODS OF DBITIXO. 219 

sectioQB 12 to 20 feet loag ; the diska were 2 feet Id diameter and 
9 inches thick, and were fastened to the shaft by eet-screvB. Many 
of the pUes were 57 feet long, of which 17 feet was in the sand.* 

328. Sah Pzlxs. For an account of the method of nsing sand 
as piles, see g 3S8. 

929. Shut Filh. These are flat piles, which, being driven 
saccessively edge to edge, form a vertical or nearly vertical sheet 
for the purpose of preventing the materials of a foundation from 
spreading, or of guarding them against the undermining action of 
water. They may be made either of timber or iron. Ordinarily 
sheet piles are simply thick planks, sharpened and driven edge to 
«dge. Sometimes they have a tongue on one edge and a correspond- 
ing groove on the other, to aid in guiding them into place while 
driviug. When heavy timbers are employed as sheet piling, wooden 
blocks or iron lugs are fastened on the edges to assist in guiding 
them into position. Sheet piles should be sharpened wholly, or at 
least mainly, from one side, and the long edge placed next to the 
pile already driven. This causes them to crowd together and 
make comparatively close joints. 

When a space is to be inclosed with sheet piling, two rows of 
gnide piles are first driven at regular intervals of from 6 to 10 feet, 
and to opposite sides of these, near the top, are notched or bolted a 
pair of parallel string-pieces, or wales, from 6 to 10 inches square, 
so fastened to the guide piles as to leave a space between the wales 
equal to the thickness of the sheet piles. If the sheeting is to stand 
more than 8 or 10 feet above the ground, a second pair of wales is 
required near the level of the ground. The sheet piles are driven 
(§§ 334-45) between the wales, working from both ends towards 
the middle of the space between a pair of guide piles, so. that the 
last or central pile acts as a wedge to tighten the whole. 

330. WODDEV BEAUVe Pubb. Spruce and hemlock answer 
very well, in soft or medium soils, for foundation piles or for piles 
always under water ; the hard pines, elm, and beech, for firmer 
soils ; and the hard oaks, for still more compact soils. Where the 
pile is alternately wet and dry, white, or post oak and yellow or 
southern pine are generally used. 

* For a dotted uid lUostrated deacrlptfon of tbia work, aee an Mtlole hj Cbarlet 
Maodonald, C.E., In Tnns. An. Soo. of C. E., Vcd. Vm. pp. 837-87. 



ovGoQi^lc 



220 PILE FOUNDATIONS. [CHAP. H 

Piles shonld never be less than 8 inches in diameter at the small 
end and never more than 18 inches at the large end. Specifications 
nenally require that these dimeneiona shall not be less than 10 nor 
more than 14 inches respectively. Piles should be straight-grained, 
should be trimmed close, and should have the bark removed. 

331. Specifloations Ua Filet. The ordinary specifications are 
abont as follows :* 

Pile*, whether used In foundatloos, tr«8tle-vork, or pile bridges, eb&ll be 
of good quality, Musd, while oak^or such other timber as tbe engineer may 
direct, not less than i«n inches (10") in diameter at the soialler end and 
14 inches (14") at the larger, and of such lengths m the engineer may require. 
Tbey must be stroigbt-giatned, must be trimmed close, and must have all the 
bark taken oB before being driven. They must be cut oS square at the butt, 
and be properly sharpened. If required by the engineer, the point shall be 
■liod with iron shoes [tee g 882], and the bead hooped with iron bands of ap- 
, proved dze and form [see § 883], which will be p^d for by tbe pound. 

332. Pile Caps and Shoes. To prevent braising and splitting 
in driving, 2 or 3 inches of the bead is usually chamfered oft. As 
an additional means of preventing splitting, the head is often 
hooped with & strong iron band, 2 to 3 inches wide and ^ to 1 inch 
thick. The expense of removing these bands and of replacing the 
broken ones, and the consequent delays, led to the introduction, 
recently, of a cap for tbe protection of the head of the pile. The 
cap consiste of a cast-iron block with a tapered recess above and 
below, the chamfered head of the pile fitting into theJower recess 
and a coshion piece of hard wood, upon which the hammer falls, 
fitting into the upper one. The cap preserves the head of the pile, 
adds to the effectiveness of the blows (§ 361), and keeps the pile head 
in place to receive the blows of the hammer. 

A further advantage of the pile cap is that it saves piles. In 
hard driving, without the cap the head is crushed or broomed to 
such an extent that the pile is adzed or sawed off several times 
before tt is completely driven, and often after it is driven a portion 
of the head must be sawed off to secure sound wood upon which to 
rest the grilhige or platform (§ 380). In ordering piles for any ' 
special work where the driving is bard, allowance most be made for 
this loss. 

Files are generally sharpened before being driven, and some- 

■Bee also "Piling" in the general apeclflcations (or railway masonir, as given Id 
Appendix L 



ovGoQi^lc 



AKT. 1.] DSSCRIPnOKB, AND MBTH0D8 OP DRIVING. 231 

times, parttcnlarty iu etODj ground, the point is protected by an 
iron shoe. The shoe may be only two V-shaped loops of bar iron 
placed over the point, in planes at right angles to each other, and 
spiked to the piles ; or it may be a wrought or cast iron socket, of 
which there are a nnmber of forms on the market. 

839. Bplioing PUei. It frequently happens, in driving piles in 
iwampy places, for false-works, etc., that a single pile is not long 
enough, in which case two are spUoed together. A common method 
of doing this is as follows :* after the first pile is driven its head is 
cat off sqnare, a hole 2 inches in diameter and 13 inches deep is 
bored in its head, and an oak treenail, or dowel-pin, 33 inches 
long, is driven into the hole ; another pile, similarly squared and 
bored, is placed upon the lower pile, and the driving continued. 
Spliced in this way the pile is deficient in lateral stiSnese, and the 
npper section is liable to bounce off while driving. It is better to < 
reinforce the splice by flatting the sides of the piles and nailing on, 
with say S-inch spikes, four or more pieces 3 or 3 inches thick, 4 or 
b inches wide, and 4 to 6 feet long. In the erection of the bridge 
over the Hudson at Foughkeepsie, H. Y., two piles were thus 
spliced tc^^her to form a single one 130 feet long. 

Piles may be made of i;ny required length or oross-aection by 
bolting and fishing together, sidewise and lengthwise, a number of 
squared timbers. Such piles are frequently used as guide piles in 
sinking pneumatic caissons (g 436). Hollow-built piles, 40 inches 
in diameter and SO feet long, wore used for this purpose iu con- 
structing the St. liouis Bridge (§ 457). They were sunk by pump- 
ing the sand and water from the inside of them with ^ sand pump 
(§ «8). 

384. Pnx-DBnmrs Haoeihes. Pile-driving machines may be 
classified according to the chamcter of the driving power, which 
may be (1) a falling weight, (3) the force of an explosive, or (3) the 
erosive action of a jet of water. Plies are sometimes set in holes 
bored with a well>anger, and the earth rammed around them. This 
is quite common in the construction of small highway bridges in 
the prairie States, a 10- or a 12-inch auger being generally used. 
The varions pile-driving machines will now be briefly described and 
compared. 

is tot RaUioad Haaamr, aa glvcD In 



ovGoQi^lc 



223 PILE FOmTDATIONS. [CHAF. ZI. 

336. Srop-hiamer File-diiTer. The ubusI method of driving 
pilea is by a encceeaion of blows given with s heavy block of wood 
or iron — called a ram, monkey, or hammer — which is carried by a 
rope or chain passing over a pulley fixed at the top of an upright 
frame, and allowed to fall freely on the head of the pile. The 
machine for doing this is called a drop-hammer pile-driver, or a 
monkey pile-driver — aenally the former. The machine is generally 
placed upon a car or scow. 

The frame consists of two uprights, called leaders, from 10 to 60 
feet long, placed about 2 feet apart,, which guide the falling weight 
in its descent. The leaders are either wooden beams or iron chau- 
uel-beams, usnally the former. The hammer is generally a mass of 
iron weighing from 500 to 4,000 pounds {usoaUy about 2,000) with 
grooves in its sides to fit the guides and a staple in the top by which 
it is raised. The rope employed in raising the hunmer is usually 
wonnd up by a steam-engine placed on the end of the scow or car, 
opposite the leaders. 

A car pile-driver is made especially for railroad work, the 
leaders resting upon an aniiliary frame, by which piles may bo 
driven 14 to 16 feet in advance of the end of the track ; and the 
frame is pivoted bo that piles may be driven on either side of the 
track. This method of pivoting the frame carrying the leaders is 
also sometimes applied to a machine need in driving pUee for foun- 
dationa 

In railroad construction, it ie not possible to use the pile-driving 
car with ita steam-engine in advance of the track ; hence, in this 
kind of work, the leaders are often set on blocking and the ham- 
mer is raised by horsee hitched directly te the end of the rope. 
Portable engines also are sometimeB used for this purpose. Occa- 
sionally the weight is raised by men with a windlass, or by pulling 
directly on the rope. 

A machine used for driving sheet piles differs from that de- 
scribed above in one particular, viz. : it has but one leader, in front 
of which the hammer moves up and down. With this construction, 
the machine can be brought close up to the wall of a coffer-dam 
(g 317 and § 390], and the pile already driven does not interfere 
with the driving of the next one, 

336. There are two methods of detaching the weight, i. e., of 
letting the hammer fall : (1) by a nipper, and (2) by a friction-clutch. 



ovGoQi^lc 



ABT. 1.] DBBCKIFTI0K5, AXD UETHODS OP DRXTINO. 233 

1. The nipper congista of a block which slides freely between 
the leaders and which carries a pair of hooke, or tongs, projecting 
from ita lower side. The tongs are so arranged that when lowered 
on to the top of the hammer they aatomatically catch in the staple 
in the top of the hammer, and hold it while it ie being lifted, until 
they are disengaged by the upper ends of the arnu striking a pair of 
inclined surfaces in another block, the trip, which may be placed 
between the leaders at any elevation, according to the height of fall 
desired. 

With this form of machine, the method of operation is as fol- 
lows : The pile being in place, with the hammer resting on the bead 
of it and the tongs being hooked into the staple in the top of the 
hammer, the rope is wonnd np until the upper ends of the tongs 
strike the trip, which disengages the tongs and lets the hammer 
fall. As the hoisting rope is unwound the nipper block follows the 
hammer, and, on reaching it, the tongs automatically catch in the 
staple, and the preceding operations may be repeated. This method 
is objectionable owitig to the length of time required (o) for the 
nipper to descend after the hammer has been dropped, and (A) to 
more the trip when the height of fall is changed. Some manufac- 
turers of pile-driving machinery remove the last objection by making 
an adjustable trip which is raised and lowered by a light line pass- 
ing over the top of the leaders. This is a valuable improvement. 

When the rope is wound up by steam, the maximum speed is 
from 6 to 14 blows per minute, depending upon the distance the 
hammer falls. The speed can not be increased by the skill of the 
operator, although it could be by making the nipper block heavier. 

2. The method bynsing s, frtction-clutch, or friction-drum, as it 
is often called, consists in attaching the rope permanently to the 
staple in the top of the hammer, and droppmg the hammer by set- 
ting free the winding drum by the use of a friction-clutch. The 
advantages of this method are (a) that the hammer can be dropped 
from any height, thns secaring a light or heavy blow at pleasure; 
and {b) that no time is lost in waiting for the nipper to descend and 
in adjusting the trip. 

When the rope is wound up by steam, the speed is from 20 to 
30 blows per minute, but is largely dependent upon the skill of the 
man who controls the friction-clutch. The hammer is caught on 
the rebound, is elevated with the speed of a falling body, and hence 



ovGoQi^lc 



284 PILE FOUNDATIONS. [CHAP. II, 

the absolute maximum speed is attained. The rope, by which the 
hammer is elevated, retards the falling weight ; and hence, for an 
equal effect, this form requires a beafier hammer than when the 
nipper is used. Although the frictioo-dram pile-driver is much 
more efScient, it is not as generally nsed as the nipper driver. The 
former is a little more expensiTe in first cost. 

337. Btcam-hammer File-driveT. As regards frequency of use, 
the next machine is probably the steam-hammer pile-driver, invented 
by Naamyth* in 1839. It consists essentially of a steam cylinder 
(stroke about 3 feet), the piston-rod of which carries n weight of 
about 3,500 pounds. The steam-cylinder is fastened to and between 
the tops of two I-beams about 8 to 10 feet 
long, the beams being united at the bottom by 
a piece of iron in the shape of a frustum of a 
cone, which has a hole through it. The under 
side *>{ this connecting piece is cut out so as to 
fit the top of the pile. The striking weight, which 
works up and down between the two I-beams 
as guides, has a cylindrical projection on the 
bottom which passes through the hole in the 
piece connecting the feet of the guides and 
strikes the pile. The steam to operate the ham- 
mer is conveyed from the boiler through a flex- 
ible tube. Fig. 56 shows the striking weight of 
the latest form of steam-hammer. It differs 
from that described above in having four rods 
for guides, instead of the two I-beams. 

The whole mechanism can be raised and 
lowered by a rope passing over a pulley in the 
top of the leaders. After a pile has been placed 
in position for driving, the machine is lowered 
upon the top of it and entirely let go, the pile 
being its only support. When steam is admitted 
below the piston, it rises, carrying the striking 
weight with it, until it strikes a trip, which 
'^- •*■ cuts off the steam, and the hammer falls by its 

own weight. At the end of the down stroke the valves are again 

• It ia ordinarily called Naamytb'R hammer, but Bourdon ahonld at least ahaio 
Ifae credit (see Bng^titrbig Xem. vol. xlii. pp. 59, eO). 



ovGoQi^lc 



AST. 1.] DBBCEIPTIOHS, AND METHODS OP DEIVIHO. 386 

aatom&tically reversed, aod the stroke repeated. Sy altering the 
adjustment of tUa trip-piece, the length of stroke (and thus the 
force of the blows) can be increased or diminiBhed. The admiBsion 
and escape of steam to and from the cylinder can also be controlled 
directly by the attendant, and the nnmber of blowa per minute 
is increased or diminished by regulating the supply of steam. The 
machine can give 60 to 80 blows per minato. 

3S8. Drop-hammer ti. Steam-hammer, The drop-hammer is 
capable of driving the pile against the greater resistance. The 
maximum fall of the drop-hammer is 40 or 50 feet, while that of 
the steam-hammer is about 3 feet. The drop-hammer ordinarily 
weighs about 1 ton, while the striking weight of the Bteam-hammer 
usually weighs about H tons. The energy of the maximum blow 
of the drop-hammer is 45 foot-tons (= 45 ft. X 1 ton), and the 
energy of the maximum blow of the steam-hammer is 4.5 foot-tons 
(= 3 ft. X 1} tons). The energy of the maximum blow of the 
drop-hammer is, therefore, about 10 times that of tho steam- 



However, the eflectiveness of a blow does not depend alone upon 
its energy. A considerable part of the enei^ is invariably lost by 
the compression of the materials of the striking surfaces, and the 
greater the velocity the greater this loss. An extreme illustratioD 
of this would be trying to drive piles by shooting rifle-bullets at 
them. A 1-toa hammer falling 45 ft. has 10 times the energy of a 
l|-ton hammer falling 3 ft., but in striking, a far larger part of th« 
former than of the latter is lost by the compression of the pile head. 
In constracttDg the foundation of the Brooklyn dry dock, it wac 
practically demonstrated that "there was little, if any, gain in 
having the fall more than 45 feet." The loss due to tbe comprea- 
sioD depends upon the material of the pile, and whether the head of 
it is bruised or not. The drop-ltammer, using the pile-cap and the 
frictioD-dmm, can drive a pile against a considerably harder reslaU 
ance than the steam-hammer. 

It is frequently claimed that the steam-hammer can drive a pile 
i^^inst a greater resistance than the drop-hammer. As compared 
with the old style drop-hammer, t. «., without the friotion-drum 
and the pile^Mp, this is probably true. Tbe striking of tbe weight 
upon the head of the pile splits and brooms it Tory much, which 
materially diminishes tbe eflectiTenees of the blow. In hard driving 



ovGoQi^lc 



21S6 PILE TOVTSDATIOSB. [CHAP. ZI, 

with the drop-hammer, without the pile-cap, the heads of the piles, 
even when hooped, will crush, bulge ont, and freqneatly split for 
many feet below the hoop. For this reason, it is sometimes speci- 
fied that piles shall not be driven with a drop-hammer. 

The rapidity of the blows is an important it«m as affecting the 
efficieaoy of a pilenlriver. If the blows are delivered rapidly, 
the soil does not have sufficient time to recompact itself about 
the pile. AVith the steam-driver the blows are delivered in such 
quick snocession that it is probable that a second blow is de- 
livered before the pile has recovered from the distortion produced 
by the first, which materially increases the efleottTeness of the 
second blow. In this respect the steam-hammer is superior to the 
drop-hammer, and the friction-clntch driver is snperior to the 
nipper driver. 

Id soft soils, the steam-hammer drives piles faster than either 
form of the drop-hammer, since after being placed in position on 
the head of the pile it pounds away without the loss of any time. 

9S9. In a rough way the first cost of the two drivers — excluEdve 
of scow or car, hoisting engine, and boiler, which are the same in 
each — is about tSO for the drop-hammer driver, and about tSOO for 
the steam-driver. Of course these prices will vary greatly. The per 
cent, for wear and tear is greater for the drop-hammer than for the 
steam-hammer. For work at a distance from a machine-shop the 
steam-driver is more liable to cause delays, owing to breakage of 
some part which can not be readily repaired. 

3M.. Onnpowder Pile-driver. This machine was invented by 
Shaw, of Philadelphia, in 1870. The expansive force of gunpowder 
is utilized both in driving the pile and ia raising the ram. The 
essential parts of the machine are the ram and gtin. The former 
consists of a mass of iron weighing generally about 1,500 pounds, 
which terminates below in a sort of piston ; this piston fits tightly 
into a chamber in another mass of iron, the gun. The ram travels 
between vertical guides much as ia the other machines ; and the 
gun and ram are hoisted as is the steam-hammer. The ram having 
been raised to the top of the guides, and the gun placed upon the 
top of the pile, a cartridge of from 1 to 3 ounces of gunpowder 
is placed in the cylinder, or gun, and the ram is allowed to descend. 
The piston enters the cylinder, compresses the air, and generates 
heat enough to ignite the cartridge, when the ezpansive force of 



ovGoQi^lc 



XBT. 1.] DESCBIPTIONB, AKD KBTBODS OF DBIYIKO. 3S7 

the powder forces the pile down and the ram up. A cartridge is 
thrown into the gon each time as the nun ascends. The rapidi^ 
of the blows is limited h; the akill of the operator sod by the heat- 
ing of the gon. Thirty to forty blows, of from S to 10 feet each, 
can be made per minute. 

341. The only adTantage of this machine is that the hammer 
does not come in contact with the head of the pile, and hence does 
not injure it The disadvantages are (1) that it is of no assistaiice 
in handling the pile ; (2) that it is not economical ; (3) that the 
gases soon destroy the gnn ; (i) that a leakage of gas occnrg aa the 
gan gets hot, which renders it less efficient as the rapidity of firing 
is increased ; and (5) that the gnn gets so hot as to explode the 
cartridge before the descent of the ram, which, of coarse, is an 
entire loss of the explosive. Its first cost is great. It is not now 
nsed. 

342. Driving PilM with Dynamite. It has been proposed to 
drive piles by exploding dynamite placed directly npon the top of 
the pile. It is not known that this method has been need except 
in a few instances. It would be a slow method, but might prove 
valuable where only a few piles were to be driven by saving the 
transportation of a machine ; or it might be employed in locations 
where a machine could not be operated, llie higher grades of 
dynamite are most suitable for this purpose.* 

343. Driving Files with Water Jet Although the water jet 
is not strictly a pile-driving machine, the method of sinking piles 
by its use deserves careful attention, because it is often the cheapest 
and sometimes the only means by which piles can be sunk in mud, 
silt, or Band. 

The .method is very simple. A jet of water is forced into the 
soil just below the point of the pile, thus loosening the soil and 
allowing the pile to sink, either by its own weight or with very light 
blows. The water may be conveyed to the point of the pile through 
a flexible hose held in place by staples driven into the pile ; and 
after the pile is sunk, the hose may be withdrawn for use again. 
An iron pipe may be substituted for the hose. It seems to make 
very little difference, either in the rapidity of the sinking or in the 
accuracy with which the pile preserves its position, whether the 
nozzle is exactly nnder the middle of the pile or not 

" F<x s brief dcMclpticm of exploalTM, Me pp. U9-M. 



ovGoQi^lc 



tSS nut rorsDATiovs. [chap, xl 

The mter jet aeenu to hare beep first naed in aigmeering in 
1852, at the taggeation of Genenl Geo. R McCkUui. It hw been 
extennTely employed on the nndy shores of the Golf and Sooth 
Atfauitic States, where the compactnesa of the nnd makes it diiE' 
cult to obtain suitable foundations for ligbt-bonaes, wbarres, etc 
Another reason for its use in that section is that the palmetto piles 
— the onlj ones that will reaist the isTSgea of the teredo— are too 
soft to withstand the blows of the drop-hammer pile-driTer. Bj 
emplojiDg the water jet the neceaaitj for the nse of the pile-haminer 
is remored, and conaeqaentlj palmetto pilea become aTailabl& 
The jet has also been employed in a great rarie^ of ways to &cili> 
tate the passage of common piles, screw and disk pfles, cylinders, 
caissons, etc., etc, through earthy materiaL' 

844. The efficiency of the jet depends upon the increased fluidity 
given to the material into which the piles are snnk, the actnal dis- 
placement of material being small. Hence the efficiency of the jet is 
greatest in clear sand, rand, or soft clay ; in gravel, or in saud con- 
taining ft large percentage of giarel, or in hard clay, the jet is almost 
useless. For these reasons the engine, pamp, hose, and nozzle 
should be arranged to delirer large quantities of water with a mod- 
erate force, rather than smaller quantities with high initial Telocity. 
In graTel, or in sand containing considerable gravel, some benefit 
might result trora a velocity sufficient to displace the pebbles and 
drive them from the vicinity of the pile ; but it is evident that 
any practicable velocity wonid be powerless in gravel, except for a 
very limited depth, or where circumstances favored the prompt 
removal of the pebbles. 

The error most frequently made in the application of the water 
jet is in using pumps with insufficient capacity. Both direct-acting 
and centrifngal pumps are frequently employed. The former 
afkrds the greater power ; but the latter has the advantage of a less 
first cost, and of not being damaged aa greatly by sand in the water 
used. 

The pumping plant used in sinking the diak^pilee for the Coney 
Island pier (see g 327), " consisted of a WorthiDgton pump with a 
13-inch steam cylinder, Sj-inch stroke, and a water cylinder 7^ 
inches in diameter. The suction hose was 4 inches in diameter, 

* B«e » pampbkl— " TIm WbWt Jet ma mi Aid to Englneeriiig Coiwtnictlitt"— 
pnUtalwd (UBl) bj the Engtaieer DepwUneot wt the U. S. Arm;. 



ovGoQi^lc 



AKT. 1,] DEBCEIPTI0ir8, AND METHODS OF DEITING. 22> 

and the diechsrge hose, which was of fonr-plj gum, waa 3 inches. 
The boiler was upright, iZ inches in dinmeter, 8 feet high, and 
contained 63 tubes 2 inches in diameter. An abundtince of eteam 
was supplied by the boiler, after the exhaust had been turned into 
the smoke-stack and soft coal used as fuel. An average of about 
160 pounds of coal was consumed in sinking each pile. With the 
power above described, it was found that piles could be driven in 
clear sand at the rate of 3 feet per minute to a depth of I'Z feet ; 
after which the rate of prt^ese gradually diminished, until at 18 
feet a limit waa reached beyond which it waa not practicable to 
go without considerable loss of time. It frequently happened that 
the pile would ' bring up ' on some tenacious material which waa 
assumed to be clay, and through which the water jet, unaided, 
could not be made to force a passage. In such cases it was found 
that by raising the pile about 6 inches and allowing it to drop sud- 
denly, with the jet still in operation, and repeating as rapidly as 
pOBsiblc, the obstruction was finally overcome ; although in some in- 
stances five or six hours were consumed in sinking as many feet." * 

In the shore-protection work on the Great I^kes, under the 
direction of the United States Army engineers, the pumping plant 
"consisted of a vertical tubular boiler, with an attached engine 
having an 8 K 12-inch cylinder, and giving about 130 revolutions per 
minute to a 43-inch driving-wheel. A No. 4 Holly rotary pump, 
with 18-inch pulley, was attached by a belt to the driving-wheel of 
the engine, giving about 300 revolutions per minute to the pump. 
The pump was supplied with a 4-inch suction pipe, and discharged 
through a 3-inch hose about 50 feet in length. The hose waa pro- 
vided with a nozzle 3 feet in length and 2 inches in diameter. The 
boiler, engine, pump, and pUe-driver were mounted on a platform 
12 feet iu width and 24 feet in length," f 

345. Jtt VI. Hammer. It is hardly possible to make a compari- 
son between a water- jet and a hammer pile-driver, as the conditions 
most favorable for each are directly opposite. For example, sand 
yields easily to the jet, but o£Fers great resistance to driving with 
tile hammer ; on the other hand, in stiff day the hammer is much 

*Chw. McDonald, in Tniu. Am. Soc otC. E.,toI. vlll. pp. 227-87. 
t " The Wst<a^«t u ui Aid to En^neeriDg Congtmction," p. II ;— • pampblM 
pabllihed (1B81) by the BngiiMei Department ol the U. 8. Annr. 



ovGoQi^lc 



tSO FILE TOUHDATIOHS. [CHAP. XL 

more ezpeditioiu. For inland work the hunmer is better, owing to 
the difficnttf of obtaining the large qnantitiea of water required for 
ihe jet ; bnt for river and harbor work the jet is the most adran- 
tageouB, Under eqnall; favorable conditions there is little or no 
difference In cost or speed of the two methods.* 

The jet and the hammer are often adTantageonaly need together, 
especially in stiff els;. The efficiency of the water-jet can be greatly 
increased by bringing the weight of the pontoon npon which the 
machinery is placed, to bear npon the pile by means of a block and 
tackle. 

346. Con or Pnxa. At Chicago and at points on the Missis- 
sippi above SL Xjoais, piiiepiU- cost from 10 to 15 cents per lineal 
foot, according to length and location. Soft-wood piles, including 
rock elm, can be had in almost any locality for 8 to 10 cents per 
foot. Oak pile* 20 to 30 feet long cost from 10 to 1'^ cents per 
foot ; 30 to 40 feet long, from 13 to 14 cents per foot ; 40 to 60 
feet long, from 30 to 30 cents per foot. 

347. Cor or Fni Sstma. There are many items that affect 
the cost of work, which can not be inclnded in a brief sommaty, bat 
which mnst not be forgotten in using such data in making *i*™'^faia. 
Below are the det^ls for the seven^ classes of work. < 

848. Eallroad Conitoiietion. The following table is a snnunaiy 
of the cost, to the contractor, of labor in driving piles (ezclnsiTe of 
hauling) in the construction of the Chicago branch of the Atchiaon, 
Topeka and Santa F^ R. R. The piles were driven, ahead of the 
track, with a horse-power drop-hammer weighing 2,200 pounds. 
The average depth driven was 13 feet. The table includes the 
cost of driving piles for abutments for Howe trass bridges and 
for the false work for the erection of the same. These two items 
add considerably to the average cost. The contractor received 
the same price for all classes of work. The work was as varied as 
such jobs usually are, piles being driven in all kinds of soil. Owing 
to the large amount of railroad work in progress in 1887, the cost 
of material and labor was about 10 per cent, higher than the aver- 
age of the year before and after. Cost of labor on pile-driver : 1 
foreman at $4 per day, 6 laborers at %%, 2 teams at 13.50; total cost 
of labor = ^3 per day. 

• Beport of Chlet of Bngliwen U. 8. A., ISSS, i^ ia64-» 



ovGoQi^lc 



ABl. I.] DXSOBIPnONS, AND 1IRTH0D8 OF DBITUro. 



231 



Covr OF PiiiB Ditimifl in Railkoad CoimntnmoiT. 

Number of pllci included In UiU report 4,409 

" Uueal feet lucluctnt In thU report 109,S68 

ATersffe lenph of tbe piles, in feet.... 24.8 

Number of da^ emptoyed iu driviDg 494 

" lineal feet driven per day 881.8 

Cost of driving, per pile 93. 68 

• ' foot 10.4 cent*. 

349. Bailroad Kspain. The following are the data of pile 
driving for repatra to bridges on the ludianapolis, Decatar and 
Springfield R. R. The work was done from December 21, 1885, to 
oannary 5, 1886. The piles varied from 12 to 32 feet in length, 
the aTerage being a little oTer 21 feet. The average distance driveo 
waa about 10 feet. The hammer weighed 1,650 pounds ; the last 
&11 was 37 feet, and the corresponding penetration did not exceed 
2 inches. The hammer was laised by a rope attached to the draw- 
bar of a locomotire — comparatiTely a very expensiTe way. 
TABLE a». 
Cost or Pusa wok Bkiime Rxpaibb. ' 



iTun or Exrnn*. 


Total. 


PaaPiu. 


PUFOM. 


'^■■K"5;.t^"rt'5:5!"ii^'".'^..v.v::,:;::: 

Z^„.^™w. .™o.p^t«^ ^ .H^. « dap. 

**^'"-f^teSJ^^ipi.ii.:::::::::::.:;;::::;:::: 


TI.U 


so.oe 














iw.w 


$1.™ 














•SM.M 











On the same road, 9 piles, each 20 feet long, were driven 9 feet, 
for bnmping-poata, with a 1,650-ponnd hammer dropping 17 feet. 
The hammer waa raised with an ordinary crab-winch and single 
line, with double crank worked by four men. The cost for labor waa 
8.3 cents per foot of pile, and the total expense was 31.6 sen is per foot. 

350. Bridge Constmotion. The following table gives the cost 
of labor in driving the piles for the Northern Pacific R. R. bridge 
Over the Bed River, at Grand Forks, Dakota, constructed in 1887. 
The soil was sand and clay. The penetration nnder a 3,350>poimd 
hammer falling 30 feet was from 2 to 4 inches. The foreman r»> 
eeired $5 per d»y, the stationary engineer t3.50, and laborerB t2. 



ovGoQi^lc 



FILB rocaDATiovs. 



[chap. zi. 



TABLE ST. 
Govt at Labob oi DBmMS Pilbb di Bmdcb CoHnBOcnoB. 




r 1 1 



i»«>l lajH 



IB I W 



■.ScM. IMMH 



« Id i»dK to get M llM Mwlns. 

861. Fonadation Tilet. The contract price for the fonndatioii 
p0ee — ^white oak — for the railroad bridge over the Mimoiui BiTer, at 
Sibley, Ho., waa 22 cente per foot for the piles and 28 cents per foot 
for driring and sawing oS below water. They were 50 feet Ic j& 
and were driven in aand and grsrel, in a coffer-dam 16 feet deep, 
by a drop-hammer weighing 3, 303 ponnds, blling 36 feet. The b&m' 
mer was raised by steam power. 

3S2. In the coustrnction of a railroad in southern Wisconsin 
daring 1886-87, the contract price — the lowest competitiTe bid — ^for 
the piles, in place, under the piers of several large bridges averaged 
as in the following table. The piles were driven in a strong current 
and sawed off under water, hence the comparatively great ezpenae, 
TABLE 28. 
OoKTUCT Pbicb at FocHDAnoir Pnxs. 



HinauLorPn* 


KnmorDatnM. 


Comucr FHoa rck Lmu, Foot. 




PorPll>HMdi8Mn>d 


Oak 
Oak 


OrdlnsiT 
Bird 


40cenu 
40 " 
48 " 
00 " 


IScenU 
M " 
3S ■■ 
80 " 



jvGooi^le 



AXt. 3.] BXARUrO POWEB OF FILES. 83S 

353. In 1887 the contract price for piles in the foaDdations ol 
bridge piers in the river at Chicago vas 36 cents per foot of pile 
left in the foundation. This pricn covered cost of timber (10 to 15 
cents), driving, and cutting off IS to 14 feet below the surface of 
the water, — about 17 feet being left in the foondatioQ. 

The cost of driving and sawing off may be estimated about 
OS follows : (17 + 13) feet of pile at 13 cents per foot = 13.90 ; 17 
feet of pile, left in the structure, at 35 cents per foot = (5.95. 
t6.95 — 93.90 = 92.05 = the cost per pile of driving and sawing oil, 
which is equivalent to near); 7 cents per foot of total length of pile. 
In this case the waste or lose in the pile heads cut off adds consider- 
ably to the cost of the piles remaining in the structore. In mak- 
ing estimatfis this allowance should never be overlooked. 

361. Harbor and &inr Work. In the shore-protection work at 
Chicago, done in 1883 by the Illinois Central R B., a crew of 9 
men, at a daily expense, for labor, of (17.35, averaged 65 piles per 10 
hoars in water 7 feet deep, the piles being 31 feet long and being 
driven 14 feet into the sand. The cost for labor of handling, sharp- 
ening, and driving, was a little over 36 cents per pile, or 1.9 cents 
per foot of distance driven, or 1.1 cents per foot of pile.* Both 
steam-hammere and water-jets were used, but not together. Notice 
that this is very cheap, owing (1) to the use of the jet, (3) to little 
loss of time in moving the driver and getting the pDe exactly in the 
predetermined place, (3) to the piles not being sawed off, and (4) 
to the skill gained by the workmen in a long job. 

On the Mississippi River, under the direction of the U. 8. 
Army engineers, the cost in 1883 for labor for handling, sharpen- 
ing, and driving, was 13.11 per pile, or 30 cents per foot driven 
The piles were 35 feet long, the depth of water 15.5 feet, and the 
depth driven 13,6 feet. The water-jet and drop-hammer were used 
together. The large cost was due, in part at least, to the current* 
which was from 3 to 6 miles per hour, f 

Aet. 3. Bearisg Poweb of Piles. 
36S. Two cases must be distinguished ; that of columnar piles or 
those whose lower end rests upon a hard stratum, and that of ordi- 
nary bearing piles or those whose supporting power is due to the 



ovGoQi^lc 



284 PILE FOCNDATIONa [CHAP. XL 

triction of the earth on the sides of the pile. In the first cose, the 
hearing power is limited by the strength of the pile considered ae a 
column ; and, since the earth prevents lateral deflection, at least to 
a conaidenible degree, the strength of anch a pile will approximate 
closely to the crushing strength of the material. This class of piles 
needs no further consideration here. 

3fie. Xetkoim of DBTiBMoraia Svpfobtdtq Fovxb. There 
are two general methods of determining the supporting power of 
ordinary bearing piles: first, by considering the relation between the 
supporting power and the length and size of the pile, the weight of 
the hammer, height of tall, and the distance the pile was moved by 
the last blow ; or, second, by applying a load or direct pressure to 
each of a number of piles, observing the amount each will support, 
and expressing the result in terms of the depth driven, size of pile, 
and kind of soil. The first method ia applicable only to piles driven 
by the impact of a hammer ; the second is applicable to any pile, 
no matter how driven. 

1. If the relation between the supporting power and the length 
and size of pile, the weight of the hammer, the height of fall, 
and the distance the pile was moved by the last blow can be stated 
in a formula, the supporting power of a pile can be found by insert- 
ing these quantities in the formula and solving it. The relation 
between these quantities must be determined from a consideration 
of the theoretical conditions involved, and hence such a formula is 
a rational formnh. 

3. By applying the second method to piles under all the con- 
ditions likely to occur in practice, and noting the load supported, 
the kind of soil^ amount of surface of pile in contact with the soil, 
etc., etc., data could be collected by which to determine the sup- 
porting power of any pile. A formula ezpresaing the supporting 
power in terms of these quantities is an empiricat formula. 

357. KATroiTAL TOEMITLAB. The deduction of a rational for- 
mula for the supporting power of a pile is not, strictly, an appro- 
priate subject for mathematical investigation, as the conditions can 
not be expressed with mathematical precision. However, ae there 
is already a great diversity of formulas in common use, which give 
widely divergent results, a careful investigation of the subject is 



The present practice in determining the bearing power of piles is 



ovGoQi^lc 



ABT. 2.] BBi.BINQ POWEB OF PILES. 235 

neither scientific nor creditable. Many engineers, instead of Jn- 
quiring into the relative meritfl of the different formnlaB, take an 
average of all the formnlaa they can find, and feel that they have a 
result based on the combined wisdom of the profession. This prac- 
tice is exactly like that of the ship's surgeon who ponred all his 
medicines into a black jag, and whenever a sailor was ailing gave 
him a spooufnl of the mixture. Other engineers, knowing the great 
diversity and general unreliability of the formulas, reject them 
all and trust to their own experience and judgment. The self- 
reliant engineer UBually chooees the latter course, while the timid 
one trusts to the former. 

To correctly discriminate between the several formulas, it ia 
necessary to have a clear anderstanding of all the conditions in- 
volved. The object of the following discussion is to discover the 
general principles which govern the problem. 

~ - 358. When the ram strikes the bead of the pile, the first effect 
^ IS to compress both the head of the pile and the ram. The more 
the ram and pile are compressed the greater the force required, until 
finally the force of compression ia sufficient to drive the pile through 
the soil. The amount of the pressure on the head of the pile when 
it begins to move, ia what we wish to determine. 

To produce a formula for the pressure exerted upon the pile by 
the impact of a descending weight, let 
^' = the weight of the ram, in tone ; 
w = " " " pile " 
S = the aectiou of the ram, in sq. ft; 
8 = " " " pile " " 
L = the length of the ram, in feet ; 
1= " " " pile " 

E_= the co-efficient of elasticity of the ram, in tona per sq. ft ; 
e= " " " " " pile " " " '* 

i = the height of fall, in feet ; 

d = the penetration of the pile, t. e., the distance the pile is 
moved by the last blow, in feet. The distance d ia the 
amount the pile as a whole moves, and not the amount 
the top of the head moves. This can be fonnd accu- 
rately enough by measuring the movement of a point, 
say, 2 or 3 feet below the head. 
P =1 the pressure, in tone, which will juat move the pile theivery 



ovGoQi^lc 



SS6 PILB TODHDATIONB. [CHAP. XL 

Btnall distance d, — ^that is to eay, the pressara prodnoed 
bj the last blow; or, briefly, P maj be called the snp 
porting pover. 
Then Wh is the accnmnlated energy of the ram at the instant it 
Btrikea the head of the pile. This energy is spent (1) in compresB- 
ing the ram, (2) in compressing the head of the pile, (3) in moving 
the pile as a whole against the resistance of the soil, (4) in overcom- 
ing the inertia of the pile, (5) in overcoming the inertia of the soil 
at the lower end of the pile, and (6) by the friction of the ram 
against guides and air. These will be considered in order. 

1. The energy consumed in compressing the hammer is repre- 
eented by the product of the mean preaanre and the compreBsion, or 
shortening, of the ram. The pressure at any point in a striking 
weight varies as the amount of material above that point ; that is to 
say, the pressure at any point of the hammer varies inversely as its 
distance from the lower anr&ce. The pressure at the lower surface 
is P, and that at the upper one is zero ; hence the mean pressure 
is i i*. From the principles of the resistance of materials, the com- 
pression, or the shortening, is ^^t in which p is the uniform pres- 
sure. From the above, p = ^ P. Consequently the shortening is 
I PL 

If the fibers of the face of the ram are not seriously crushed, the 
mean pressure will be one half of the maximum pressure due to im- 
pact ; or the mean pressure during the time the ram and pile are 

being compressed is ^P. Then the euergy consumed \s-j~^y,. 

The yielding of the material of the ram is probably small, and might 
be omitted, but as it adds no complication, as will presently appear, 
it is included. 

2. The mean pressure on the head of the pile is ^ P, as abovei 
For simplicity assume that the pile is of uniform section through- 
ouL To determine the shortening, notice that for the part of the 
pile above the ground the maximum pressure is uniform through- 
out, but that for the part under the surface the maximum pressure 
varies as some function of the length. If the soil were homogeneous, 
the pressure would vary about as the length in the ground ; and 



ovGoQi^lc 



AXr. 3.] BEABIITQ FOWEB Of FILES. S3? 

1 PI 
hence the shorteniiig wonld be -^ — . Bnt, remembering that the 

resistance ia gBnerally greater at the lower end than at the upper, 

and that any swaying or Tibration of the upper end will atill further 

diminish the resistance near the top, it is probable tbat the mean 

pressure is below the center. It will here be assumed that the mean 

pressure on the fibers of the pile is two thirds of that on the liead, 

2 PI 
which is equivalent to assuming that the shortening is — — , when 

the pile is wholly immersed. If only a part of the pile is in contact 

PI' ^ PI P / &\ 

with the soil, the shortening will be [- - — * = — [l' -f - / ), 

86 i 86 86\ o J 

in which I' is the exposed portion and /, the part Immersed. For 
simplicity in the following discussion the shortening of the pile 



the top projects above the ground, it will only bo necessary to sub* 

stitute (I' -f i'l) fot* I ii> equations (1) and (3) below. 

1P*1 
Then the energy lost in the compression of the pile is - — . 

3. The energy represented by the penetration of the pile is Pd. 

4. In the early stage of the contact between the ram and the 
pile, part of the energy of the ram is being used up in oyercoming the 
inertia of the pile ; but in the last stage of the compression, this 
energy is given out by the stoppage of the pile. At most, the effect 
of the inertia of the pile is small ; and hence it will be neglected. 

5. The energy lost in overcoming the inertia of the soil at the 
lower end of the pile will vary with the stiffness of the soil and with 
the velocity of penetration. It is impossible to determine the amount 
of this resistance, and hence it can not be included in a formula. 
Omitting this element causes the formula to give too great a support- 
ing power. The error involved can not be very great, and is to be 
covered by the factor of safety adopted. 

6. The friction of the ram against the guides and against the air 
diminishes the effect of the blow, but the amount of this can not be 
computed. Omitting this element will cause the formula for the 
supporting power to give too great a result. The friction against 
the air increases very rapidly with the height of fall, and hence the 



ovGoQi^lc 



S88 riLH P0DNDATI0N8. [CHAP. XL, 

OQAller the fall the more Dear!; will the formula give the true sap- 
porting power. 

8S9. Eqoating the energy of the falling weight with that con* 
mmed in compreaeing the pile and tarn, and in the penetiatioQ of 
tlw pile, aa discnwed in paiagrsphs 1, 3, and 3 above, we have 



Joiving equation (1) gives 



36 d" 5" £'«•«• 



iLs9 + iiS£^{ZL»«-\-ilSBy 

3Lf6 + 4lSS' 



V) 



(2) 



An examination of eqnation (3) shows that the prassnre upon the 
pile ■rnnvB with the height of fall, the weight, aection, length, and 
coefficient of elaaticitj of both ram and pile, and with the penetra- 
tion. It is easy to see that the weight of the ram and the height 
of the fall sbonld be included. The penetration is the only element 
which varies with the nature of the soil, and bo of conrse it also 
should be included. It is not so easy to see that the length, section, 
and co-efficient of elasticity of the material of the pile and ram 
should be included. If any one will try to drive a large nail into 
hard wood with a piece of leather or rubber intervening between 
the hammer and the head of the nail, he will be impressed with the 
fact that the yielding of the leather or rubber appreciably diminishes 
the effectiveness of the blow. Esseutially the same thing occnrs in 
trying to drive a large nail with a small hammer, except that in this 
case it is the yielding of the material of the hammer which dimin- 
ishes the effect of the blow. In driving piles, the materials of the 
pile and ram act as the rubber in the first illustration; and, reason- 
ing by analogy, those elements which determine the yielding of the 
materials of the pile and ram should be included in the formula. 
Obvionsly, then, the pressure due to impact will be greater the 
harder the material of the pile. Notice also that if the head of the 
pile is bruised, or "broomed," the yielding will be increased; and, 
conseqnently, the pressure due to the blow will be decreased. 



ovGoQi^lc 



ISC. 3.] BXASUTQ FOWEB OF FILES. ^39 

860. Tbfl Anthor'i FonniilA for Pnotloe. To simpli^ eqwttin 
(2), pat 

e 3S»« 

and tlieii equation (2) beoomes 

P = V2q Wh + q'd'-qi. .... (3) 

Equation (3) can be Bimplified still further by compatiiig q tot 
tbe conditions as the; ordinarily occur in practica Of conrse, in 
this case it will only be possible to aasnuie some arerage ralne for 
the various quantities. Aesnme tbe section of the pile to be 0.6 sq. 
ft; tbe section of the ram, 3 sq. ft.; the length of the ram, 3.5 ft; 
the length of the pile,* 25 ft. ; the co-efiBcient of elasticity of the 
ram, 1,080,000 tons per sq. ft; and tbe co-efficient of elasticity of 
the pile, 108,000 tons per sq. ft (an average valne for oak, elm, 
pine, etc, but not for palmetto and other soft woods). Oompnting 
tbe corresponding value of q, we find it to be 5,160; but to secure 
round numbers, we may take it at 5,000, which also gives a little 
additional security. 

Equation (3) then becomes 

P = 10O{VWh+{5Ody-50d), ... (4) 

which ia the form to be tised in practice. 

Equation (4) is approximate because of the assumptions made in 
deducing equation (1), and also because of the averse valne taken 
for q; bnt probably the error oooaeioned by these approximations ia 
not material. 

361. Notice that, since the co-efficient of elasticity of sound 
material was used in deducing the value of q, equation (4) is to be 
applied only on condition that the last blow is Btrnck upon sound 
wood; and therefore the head of the test pile should besawed off so 
as to present a solid surface for the last, or test, blow of the hammer. 
{This limilaiion is exceedingly important.) Since the penetration 
per blow can be obtained more accurately by taking tbe mean dis- 
tance for two or three blows than by measuring the distance for a 
single one, it is permissible to take the mean penetration of two or 



ovGoQi^lc 



240 FILE FOL'NDATIOXS. [CHAP. XI. 

three blows; but their number and force should be such as not to 
crnsh the bead of the pile. 

In this oonnectioD the foUoiring table, giren b; Don. J. Vhitte- 
more, in the Transactions of the American Society of Civil Engi- 
neers, Tol. xiL p. 4i'i, to show the gain in efficiency of the driving 
power by cutting off the brniaed or broomed head of the pile, is very 
instmctivB. The pile vraa of green Norway pine; the ram was of 
the Naamyth type, and weighed 2,800 ponnds. 

Qaih IK EmctKHCT aw thh Dsiyine Powkr bi 
Cdttdiu oft tbb Broomed Head or tex Fu^. 

.... S blows. 



Sd ft. of penetraliou required 



Olh 



10th 

llth " " " 109 

12th " '• " 168 

18th " " " 237 

14th ■' " " «M 

Head of ilie pile adzed off. 

ISth ft. of penetration required 37S 

16th ■' " ■• 6J3 

ITth ■' " " 882 



Total number of blows 6,238 

TTotice that the average penetration per blow was 8 J times greater 
during the 15th foot than during the 14th; and nearly i times 
greater in the 19th than in the 18th. It does not seem unreason- 
able to believe that the first blovrs after adzing the head off were 
correspond ill gly more effective than the later ones; consequently, 
it is probable that the first blows for the 15th foot of penetratioa 
were more than & times as efficient as the last ones for the 14th foot, 
and also that the first blows for the 19th foot were 8 or 10 times 
more efficient than the last ones for the 18th foot. Notice also that 
aince the head was only "adzed off," it is highly probable that the 
spongy wood was not entirely removed. 



ovGoQi^lc 



ilXI, 2.] BEAfilNO POWER OF PILES. S41 

II the penetration for the last blow before the head was adzed ofF 
■were used in the formula, the apparent supporting power would be 
Tery much greater than if the penetration for the first blow after 
adzing off is employed. This Bhows how anscientific it is to pre- 
«cribe a limit for the penetration without specifying the accompany- 
ing condition of the head of the pile, as is ordinarily done. 

362, Weiabaoh'a Formnla, Equation (2), page 238, ip essentially 
equivalent to Weisbach's formula for the supporting power of a pile. 
Weisbach assumes that the pressure is uniform throughout, and 
■obtains the formula* 
ffff. 



P=(^^)(/^^^7^r^_4 . („ 



, . , „ SE , _ se 
m which H = —^, and R^ = -y. 

363. Bankine's Formula. Equation (2), page 238, is also essen- 
tially eqnivalent to Rankiue's formula ; and differs from it, only 
because he assumes the pressure to Tary directly as the length of 
the pile, and neglects the compression of the ram. Bankine'g 
f ormnls is \ 

p = /t£^+i^-^f-', ... ,6) 

EqaatioD (2) differs from WeiBbach's and Bankine's on the safe 
«ide. 

364, Ekpisical Fosmtua. General Prinoiples. (1) An empiri- 
•cal formula should be of correct formj (2) the constants in it should 
be correctly deduced ; and (3) the limits within which it ia applica- 
ble should be stated. 

For example, suppose that it were desired to determine the 
equation of the straight line A B, Fig. 57. 
Since the given line is straight, wc will as- 
sume that the empirical formuiii is of the 
form y = m X. Wo might find wi by m 
ing the ordinates 1, 2, 3, and place n 
to th^ir mean. If 1, 2, 3, be the numerical 
ralues of the respective ordinates, the for- 
mula becomes y = %x, which gives the line "1^ 
C. The mean ordinate to C is equal to r^- "■ 

-the mean ordinate ia A B, but the two are not by any means tha 
's TniMlKloi))! p. 701' 



ovGoQi^lc 



i4& PILE P0CNDATI0S8. [CHAP. XI. 

aame line. It is evident that this empirical formula is of the ^rrong: 
form. 

For another illnstration, assume that some law is correctly repre- 
sented by the curve A B, Fig. 58. The form 
of the empirical formula may be such as to 
give the curve CD. These curves coincide 

a J exactly at two points, and the mean ordinate 

\^^ "~^>/ to the two is the same. To use a com- 
S \ mon eipreseion, we may say that, "on the 

average, the empirical formula agrees exactly 

Fio. u. with the facts ;" but it is, nevertheless, not. 

even approximately true. The constants were not correctly de- 
duced. 

Even if of the correct form and correctly deduced, an empirical 
formula can be safely applied only within the 
limits of those values from which it was deter- , 
mined. For esample, a law may be repre- 
eent«d by the curve A B, Fig. 59. From Tf|- 
obaervatious made in the region C E, the em- Fio. lo. 

pirical formula lias been determined, which gives the curve C E D^ 
which between the limits C and E is all that can be desired, but 
which is grossly in error between the limits fand D. To use an 
empirical formula intelligently, it is absolutely necessary that the 
limits within which it is applicable should be knowu. 

Of course, the observations from which the empirical formula 
was deduced can not be used to test the correctness of the formula; 
such a procedure can check only the mathematical work of deriving^ 
the constants. 

Elementary as the preceding principles are, many empirical 
formulas are worthless owing to a disregard of these conditions in 
deducing them. 

365. Comparison of Empirical Tormolas. We will now briefly 
consider the empirical formulas that are most frequently employed 
to determine the supporting power of piles.* 

HamcrlVs formula for the dynamic effect of a falling body is f 
■ P = 4.426 W V, "as deduced from experiments." 

The experiments consisted in letting a weight of a few onnces 



ovGoQi^lc 



ART. 2.] BEAKtNa POWBE OF PILES. 243 

&U a few inches npon a coiled spring ; and henoe the formula is 
entirely inapplicable to pile driving. 

Beavfoy's fommla is P = 0.5003 ff" F", "as determined by 
experiment." This fonnala was deduced under the same conditions 
as Hasweli's, and hence is useless for pile driving. The difference 
between the formulas is due to the fact that Hoswell used only on© 
weight and one spring, and varied the height of the fall, while Beau- 
foy employed one weight and Bprings of such relative etifiness as 
would etop the weight in nearly the eame distance for different 
heights of fall.* Notice that Haswell'e, and also Beanfoy's formula, 
would give the same bearing power for all soils, other things being 
the same. 

Nystrom'8form'ula\\sP = -7-^^-Y — ttj- In a later book,J Nys- 

trom gives the formala P = j ~j~> assuming that "about 25 per 

cent, of the energy of the ram is lost by the crushing of the head of 
the pile." Both of these formulas are roughly approximate, theo- 
retical formulas, although frequently cited tis " practical formulas." 

Mason's formula % ie P = j^^=-~ — r-v As' in the precedinff 

cases, this is frequently referred to as a " practical formula ;" but an. 
examination of the original memoir shows that it is wholly a theo- 
retical formula with no pretensions of being anything else. It is 
also sometimes referred to as having been " tested by a series of 
experiments ;" but apparently the only basis for this is that th© 
piles upon which Fort Mon^omery (Rouse's Point, N- Y.) stood 
from 1846 to 1850 without any sign of failure, when tested by this 
formula, showed a co-efiBcient of safety of 3^^- '^^'^ evidence is not 
conclusive: (1) the factor is large enough to cover a considerable 
error in the formula; (2) since the foi-mula assumes that all of the 
energy in the descending ram is expended in overcoming the resist- 
ance to penetration, the compnted bearing power is too small, and 
consequently the co-efficient of safety is oven greater than aa stated; 

• Van Nostrancl's Bngln'g Mag., vol xvil. p. 836. 
t NjMroni's Pooket-Book, p. 158. 
t New HecbsniCi, p. 184. 

j Reflistauce at Piles, J. L. Mason, p. 8; Ho. 5 of Papers on FrsotlOttl BnglneeTtng, 
published by tbe KngiiieeriDg Deportinent of the U. B. Armf. 



ovGoQi^lc 



244 PILE FOUNDATIONS. [CHAP. XI. 

and (3) it is probably safe to say that after a pile has stood a short 
time its bearing pover is greater than at the moment the drivlDg 
ceased, owing to the settlement of the earth about it. 

Sander's formula* is P' = -o~r' *° 'l^ich P' is the safe bear- 
ing power. This formula was deduced on the assumptions that the 
energy of the falling weight was wholly employed in forcing the 
pile into the ground, — i. e., on the aeeumptiou that Pd= Wb, or 

P = —J-, — and that the safe load waa one eighth of the ultimate 

supporting power. It is therefore a roughly approximate, theoreti- 
cal formula. 

Notice that, since some of the enei^y is always lost, Pd, the 
«nergy represented by the movement of the pile, must always be 
less than Wh, the energy of the hammer; hence, P is always lees 

than— T-; or, in mathematical language, P<—t~- This relation is 
very useful for determining the greatest possible value of the sup- 
porting power. P will always be considerably leas than-T— ; and 

this difference is greater the lighter the weight, the greater the 
foil, the softer the material of the pile, or the more the head is 
bruised. When d is very small, say i inch or less, the difference is 
so great as to make this relation useless. 

Trautwine's formula,^ in the nomenclature of page 235, is 

P= ■ It was deduced from the observed supporting 

power of piles driven in soft soil. Strictly speaking, it is applicable 
only under conditions similar to those from which it was deduced ; 
and hence it is inapplicable for hard driving and to piles whose 
heads are not bruised about the same amount as were the experi- 
mental ones. No formula can be accnrate which does not, in some 
way, take cognizance of the condition of the liead of the pile. For 
example, experiments Nos. 3 and 4 of the table ou page 346 are 
the same except in the condition of the heads of the piles, and yet 



ovGoQi^lc 



ART. 2.1 BEAHINQ POWEft OF PILES. HiS 

the load supported by the former was 3| times that sapported by the 

latter. This formula ia not applicable to piles drireii with a eteam 

hammer, since according to it the energy represented by the sinking 

of the pile is greater than the total energy in the deecendinf^ weight. 

For example, if W = IJ tons, A = 2 feet, and d=l inch = ^^ of a 

Wit 
foot, the formnia P < —r- becomes P < 36 tons. Trautwine'a 
a 

formula gives /* = 49 tons; that is to say, Trantwine's formala 

makes the sapporting power one third more than it would be if >(9 

energy were lost. 

Engineering News formula,* the most recent and the most 

popular, is f" = ., , in which P ' is the safe load in tons; aud 

d' is the penetration, iu inchta, under tbe last blow, which is 
assumed to be appreciable and at an approximately naiform rate. 

366. Tlie Author's Empirical Formula. Certain aaBamptJons 
and approximations were maije in deducing equation (3), page 339. 
If it is thought not desirable to trust entirely to theory, thea the 
formnia 



P=V%qWh-^q*d'-qd . . . . (7) 

may be considered as giving only the form which the empJiica) 
formula should have. Under this condition q becomes a numerical 
co-efficient to be determined by experiment, which must be mad» 
by driving a pile and measuring d, after which the BUBtaining power 
must be determined by applying a direct pressure. The last, or 
test, blow should be struck on sound wood. 

367. Table 29 gives all the experiments ou the supporting 
power of piles for which the record is complete. Unfortunately 
these experiments do not fulfill the conditions necessary for a propel 
determination of q in equation (T). It is known that in some of th« 
cases the head of the pile was contiderably broomed, and there is 
internal evideuce that this was so in the others. 

The data of the following table substituted in equation (7) give 
values of q from 1.5 to 337, with an average of 130. The range of 
these results ghows the inconsistency of the experiments, and the 
Bmallneea of the average shows that the last blow was not Btrncl* in 
sound wood. This value of q is of no practical use 

• Enginrrring Xnrt. vol. xx. pp. 511. B12 (Doe. M, 1888,. 



ovGoQi^lc 



KLB FOCKDATIOira. 
TABLE 39. 

OK TBK SuPFOBTQIa 



[chap. XI. 



PowKB OF PhiH, 



s^ 



CircnUr of the Office o( Chief of Eaglneen 

U. 8. A.. Nov. 13, ■81, pp. 2, 8. 
Trautwiue'ft Po(Aet-Book, H. 188S, p. 64& 
Jour. Frank. Inst., vol. SS, p. 101. 
DelafleM's "Foiuukiioiu In Campreadble Sotli,' 



pp. 1 
ginee 



leen' bepaitmeiit o: 
TniitwlDS In Baiiivad Oattlt. July 8, 1887, p. 



368. Aa confirming the reliability of the /orm of eqnatioDB (3), 
^4), and (7), it is intereeting to not)be that A. C. Hertiz* fonnd, 
from the records of the driving and afterwards pulling np of nearly 
■400 piles, the following relation : 



d- 



Wh 



P_ 
500' 



-which may be pnt in the form 

P= 4/500 Wh + {250 rf)' - 250 d. . 



(8) 



Equation {8) has exactly the form of equation (3), page 239, 
although deduced in an entirely different way. The value (250) of 
the constant q in equation (8) is lees than that in equation (4), 
page 239, which shows that the heads of the piles were broomed. 
The value of q in equation (8) ia greater than that deduced from tho 
data of Table 29, which shows that the piles from which equation 
(8) was determined were not bruised as much a& those in the above 
table. 

369. SuFPOBime Fowbe DETEBinirED bt BXFERUixirT. It is 
not certain that the bearing power of a pile when loaded with a con- 
tinued quiescent load will be the same as that during the very short 



ovGoQi^lc 



^BT. 2.] BEABINO POWER OF FILES. M7 

period of the blow. The friction on the sides of the pile will have 
a greater effect in the former case, nhile the reaiBtance to penetra- 
tion of the point will be greater in the latter. This, and the fact 
that the supporting power of piles sank by the water-jet can be 
determined in do other way, shows the necessity of ezperimentB to 
determine the bearing power under a steady load. 

Unfortunately no extended experiments have been made in this 
'direction. We can give only a collection of as many details as pos- 
.sible concerning the piles under actual structures and the loads 
which they sustain. In this way, we may derive some idea of the 
sustaining power of piles under various conditions of actual practice. 

370. intimate Load. In constructing a light-house at Proctors- 
Tille, La., in 1856-57, a test pile, 12 inches square, driven 29.5 feet, 
bore 29.9 tons without settlement, but with 31.2 tons it "settled 
slowly." The soil, as determined by borings, had the following 
character : " For a depth of 9 feet there was mud mixed with 
sand ; then followed a layer of sand abont 5 feet thick, next a layer 
of sand mixed with clay from 4 to C feet thick, and then followed 
^e clay. By draining the site the surface was lowered about 6 
inches. The pile, by its own weight, sank 5 feet 4 inches." The 
above Ibad is equivatent to a frictional resistance of 600 lbs. per 
sq. ft. of surface of pile in contact with the soil. This pile is No. 
1 of the table on p^e 246. 

At Philadelphia in 1873, a pile was driven 15 ft. into "soft river 
mud, and 5 hours after 7.3 tons caused a sinking of a very small 
fraction of an inch ; under 9 tons it sank i of an inch, and under 
15 tons it sank 5 ft." The above load is equivalent to 320 lbs. 
per sq. ft. of surface of contact. This pile is No. 2 of the table on 
page 246. 

In the construction of the dock at the Pensacola navr yard, a pile 
.driven 16 feet into clean white sand austained a direoBvpnll of 43 
tons without movement, while 45 tons cHused it to rise slowly; and 
-46 tons were required to draw the pile. This is equivalent to a fric- 
tional resistance of 1,900 lbs. per sq. ft. This pile is No. 4 of the 
table on p. 246. 

" In the construction of a foundation for an elevator at Buffalo, 
N. Y., a pile 15 inches in diameter at the large end, driven 18 ft., 
bore 25 tons for 27 hours without any ascertainable effect. The 
■weight was then gradually increased until the total load on the 



ovGoQi^lc 



248 PILE FOUNDATIONS. [OHAP. XI- 

pile WM 37i tons. Up to thb weight there had been no depreesioa 
of the pile, bat vith 37^ tone there was a gradual depreaaioQ which 
aggregated ^ of an inch, bejond which there waa no depression 
until the weight was increased to 50 tons. With 50 tons there wa^ 
a further depression of | of an inch, making the total depression 
1} inches. Then the load was increased to 75 tons, under which 
the total depression reached 3^ inches. The experiment was not 
carried beyond this point. The soil, in order from the top, was- 
as follows : 3 ft of blue clay, 3 ft of grarel, 5 ft of stiff red clay, 
2 ft of quicksand, 3 ft of red clay, 2 ft of gravel and sand, and 
8 ft of very stiff bine clay. All the time during this experiment 
there were three pile-drivers at work on the foundation, thus keep- 
ing up a tremor in the ground. The water from Lakfe Erie had 
free access to the pile through the gravel."* This is equivalent 
to a frictional resistance of 1,850 lbs. per sq. ft This is pile No. &■ 
of the table on page 246. 

S71. In m^ng some repairs at the Hull docks, England,, 
several hundred sheet-piles were drawn out They were Vi X 10 
inches, driven an average depth of 18 feet in stiff blue clay, and 
the average force required to pull them was not less than 35. S 
tons each. The frictional resistance was at least 1,875 lbs. per sq. 
ft of surface in contact with the soil, f 

372, Safe Load. The piles under the bridge over the Missouri 
at Bismarck, Dakota, were driven 33 ft. into the sand, and suataia 
20 tons each— equivalent to a frictional resistance of 600 lbs. per sq. 
ft. The piles at the Flattsmouth bridge, driven 28 ft into the 
sand, sustain less than 13} tons, of which about one fifth is live- 
load, — equivalent to a frictional resistance of 300 lbs. per sq. ft 

At the Hull docks, England, piles driven 16 ft into " alluvial 
mud " sustain at least 30 tons, and according to some 25 tons ; for 
the former, the friction is about 800 lbs. per sq. ft The piles 
under the Koyal Border bridge " were driven 30 to 40 ft into sand 
and gravel, and sustain 70 tons each," — the friction being about 
, 1,400 lbs. per sq. ft 

373. "The South Street bridge approach, Philadelphia, fell by 
the sinking of the foundation piles under a load of 24 tons each. 

• By coarttsr of John C. Tr»atwliie, Jc, tzom prlnte correspondsnce of John &. 
Payne and W. A. Haven, «DglD«ere Id charge, 
t Proa. Iiut of C. £., vol. Ixlv. pp. 311-19. 



ovGoQi^lc 



Art. 3.] bkabing powbb op piles. 24» 

They were drivea to an abiolate stoppage by a l-ton hammer fall- 
ing 33 feet. Their length was from 24 to 4L feet. The piles vere 
driTSQ through mud, then tough olay, and into hard gravel."* 
According to Trantwine's formula their ultimate supporting power 
was 164 tona, and according to the Engineering News formula th« 
safe load was 64 tona. It is probable that the last blow was strnck 
on a broomed head, which wonld greatly reduce the penetration, 
and that consequently their enpporting power was greatly over- 
estimated. If the penetration when the last blow was struck on 
eound wood were 2 inches, then according to Trantwine's formula 
Uie ultimate supporting power was 64.7 tona, and according to the 
Engineering Newt formula the safe load was 31.3 tone. 

374. SUTPOBTDtO FOTEB 07 SOBZW LSli DiBE FiLBS. The sup- 
porting power depends upon the nature of the soil and the depth to 
which the pile is sunk. A screw pile " in soft mud aboYe clay and 
«and " supported 1.8 tons per sq. ft of blade.t A disk pile in 
" quicksand " stood 5 tons per sq. ft. under vibrations. X Charles 
McDonald, in constructing the iron ocean-pier at Coney Island, as- 
sumed that the safe load upon the flanges of the iron disks sunk into 
the sand, was 5 tons per sq. ft. ; but " many of them really snpport 
as much as 6.3 tons per sq. ft. continually and are subject to occa- 
sional loads of 8 tons per sq. ft., without causing any settlement 
that can be detected by the eye."§ 

376. Faoiob or Satitt. On account of the many uncertainties 
in connection with piles, a wide margin of safety is recommended by 
all authorities. The factor of safety i-anges from 2 to 12 according' 
to the importance of the structure and according to the faith in the 
formula employed or the experiment taken as a guide. At best, 
the formulas can give only the supporting power at the time when 
the driving ceases. If the resistance is derived mainly from fric- 
tion, it is probable that the supporting power increftses for a time 
after the driving ceases, since the co-efficient of friction is usually 
greater after a period of rest If the supporting power is derived 
mainly from the resistance to penetration of a stifl substratum, the 
bearing power for a steady load will probably be smaller than the 

• TraoB. Am. Soc. of C. E., voL vIL p. aM. 
t Proc. InBt of C. E„ vol xvll. p. «L 

X Ibid., p. 443. ' 

ITrans. Am. Soo. C. E., vol. Till. p. 236. 



ovGoQi^lc 



350 PILE FOCNDATIONS. [CHAP. XI. 

force reqaired to drive it, as most materials require a less force to 
change their form slowly than rapidly. K the soil adjoining the 
piles becomes wet, the supporting power will be decreased; and 
vibrations of the Btrncture will have a like effect. 

The formnlas in use for determining the supporting power of 
piles are so unreliable, that it is quite impOBaible to determine the 
factor of safety for any existing structure with anything like accu- 
racy. 

The factor to be employed should vary with the nature of the 
fltmcture. For example, the abutments of a etone arch should be 
fioustrncted eo that they will not settle at all ; but if a railroad pile 
trestle settles no serious damage is done, since the track can be 
flhimmed up occasionally. In a few cases, a small settlement has 
taken place in a railroad trestle when the factor of safety was 3 or 
4, as computed by equation (4), page 239. 

Art. 3. Arbanqeueitt of the Foundation. 

376. SiBPOSITIOR OF THE PiLEB. The length of the piles to be 
used is determined by the nature of the soil, or the conveniencea 
for driving, or the lengths most easily obtained. The safe bearing 
power maybe determined from the data presented in §g 370-73, or, 
better, by driving a teat pile and applying equation (4), page 239. 
Then, knowing the weight to be supported, and having decided 
upon the length of piles to be used, and having ascertained their 
safe bearing power, it isan easy matter to determine how many piles 
are required. Of course, the number of piles under the different 
parts of a structure should be proportional to the weights of those 
parts. 

If the attempt is made to drive piles too close together, they are 
liable to force each other up. To avoid this, the centers of the 
piles should be, at least, SJ or 3 feet apart. Of course, they may 
be farther apart, if a less number will give sufficient supporting 
power, or if a greater area of foundation is necessary to prevent 
overturning. 

When a grillage (§ 380) is to be placed on the head of the piles, 
great care must be taken to get the latter in line so that the lowest 
course of grillage timber, in this case called capping, may rest 
squarely upon all the piles of a row. In driving. under water, a 



ovGoQi^lc 



ART. 3.] ABRANOEHEXT OF THE FOUNDATION. S51 

convenieDt waj of markiug the poaitiooB of the piles is to conBtmct 
a light frame of narrow boards, called a spider, in which the posi- 
tion of the pilog is indicated hy a small square opening. This frame 
may be held in place by fastening it to the sides of the coffer-dam, 
or to the piles already driven, or to temporary supporte. Under 
ordinary circnmstances, it is reasonably good work if the center 
of the pile is under the cap. Files frequently get considerably out 
of place in driving, in which case they may sometimes be forced 
back with a block aud tackle or a jack-screw. When the heads of 
the piles are to be covered with concrete, the exact position of the 
piles is comparatively an unimportant matter. 

In close driving, it is necessary to commence at the center 
of the area and work towards the sides ; for if the central ones are 
left antil the last, the soil may become so consolidated that they 
can scarcely be driven at all. 

377. Butt TB, Top Down. According to Kankine* all piles 
should be driven large end down, having first been sharpened to a 
point 1^ to 3 times as long as the diameter of the pile. This is at 
least of doubtful utility. If the pile is supported wholly by fric- 
tion, then the supporting power will be greater when the small end 
is down. If the soil Is semi-liquid, the buoyancy would be slightly 
greater when the large end is down ; but the buoyancy constitutes 
but a very small part of the supporting power, and the difference 
in buoyancy between top and bottom down is still less. If the pile 
derives its support mainly from a solid substratum, then its bearing 
power would be greater with the large end down ; but, in this case, 
it should not be sharpened. For close driving, it is frequently 
recommended that, to prevent the piles from forcing each other up, 
they should be driven butt end down. K"otice, however, that if 
the soil is non-compressible, as pure sand, or if the piles are driven 
80 close as to compress the soil considerably, it will rise acd carry 
the piles with it, whether they were driven with the big or the little 
end down. Piles are generally driven small end down, but never- 
theless practical experience shows that there are conditions in which 
it is apparently impossible to drive them in this way, even in 
comparatively isolated positions. These conditions appear to occur 
most frequently in swamps, and in connection with quicksand. 

• " civil Englnearing, " p. 603. 



ovGoQi^lc 



S62 PILE FOUNDAnONS. [CHAP. SI. 

378. SATraa-OFT THX Files. When piles are driven, it ia. 
geoerall; necessary to e&w tbem off either to bring them to tbe^ 
eame height, or to get the tops tower than they can be driyen, or tO' 
Becnre Bound wood upon which to rest the timber platform that 
carricB the masonry. When above water, pilee are nenally sawed off 
by hand ; and when below, by machinery — usually a circular saw on 
a vertical shaft held between the leaders of the pile driver or mounted 
upon a special frame, and driven by the engine used in driving the 
piles. The saw-Bhaft ia eometimea attached to a vertical shaft held 
between the leaders by parallel bars, by which arrangement the saw 
can be swung in the arc of a circle and several piles be cut off with- 
out moving the machine. The piles are sometimes sawed off with 
what is called ft pendulum saw, i.e., a saw-blade fastened between ' 
two arms of a rigid frame which extends into the water and is free- 
to swing about an axis above. The saw is swung by men pushing- 
on the frame. The first method is the better, particularly when 
the piles are to be sawed off under mud or silt. 

Considerable care is reqnired to get the tops cnt off in a hori- 
zontal plane. It is not necessary that this shall be done with mathe-^ 
matical accuraoy, since if one pile does stand up too t&r the excess- 
load upon it will either force it down or crush the cap until the 
other piles take part of the weight. Under ordinary conditions, It 
ia a reasonably good job if piles on land are sawed within half an 
inch of the same height ; and under water, within one inch- When 
a machine is used on land, it is usually mounted upon a track ^nd 
drawn along from pile to pile, by which device, after having leveled 
up the track, a whole row can be sawed off with no farther atten- 
tion. When sawing under water, the depth below the surface is. 
indicated by a mark on the saw-shaft, or a tai^t on the saw- 
shaft is observed npon with a leveling instrument, or a leveling rod 
ia read upon some part of the saw-frame, etc. In sawing piles off 
under water, from a boat, a great deal of time ia consumed (par- 
ticularly if there is a current) in getting the boat into position, 
ready to begin work. 

Piles are frequently sawed off under 10 to 15 feet of water, and 
occasionally under 20 to 25 feet. 

379. FnniHIHO THB FoinniATIOir. There are two cases: (1> 
when. the heads of the piles are not under water ; and (2) when they 
are under wat«r. 



ovGoQi^lc 



ABT. 3.] AERANGEMEinr OF THE FOUNDATION. 263 

1. When the piles are not under water there are t^in two cases : 
i(fl) when a timber grillage is used ; and (i) when coitcreit alone u 
mBed. 

2. When the piles are eawed ott under water, the timber Htmct* 
ure (in this case called a crib) which intervenes between the piles 
and the masonry ia put together first, and then sunk into place. The 
construction is esBentially the same as when tbe piles are not under 
water, but differs from that case in the manner of getting the tim- 
ber into its final resting place. The methods of constructing foun- 
dations under water, including that by the use of timber cribs, will 
be discussed in Art. 2 of the next chapter. 

380. Pilei and Grillage. This ia a stout frame of one or more 
courses of timber drift-bolted or pinned to the tops of the piles 
and to each other, upon which a floor of thick boards is placed to 
receive the bottom courses of masonry. For illustrated examples, 
see Fig. 84, page 3C3, Fig. 86, page 380, and Fig. 90, page 386. 

The timbers which rest upon the beads of the piles, called caps, 
are usually about I foot square, and are fastened by boring a hole 
through each and into the head of the pile and driving into the 
hole a plain rod or bar of iron having a slightly larger cross-section 
"than the hole. 

381. These rods are called drifi-hoUt, and are usually either 
-ttrod 1 inch in diameter (driven into a j-inch anger hole) , or a 
hax 1 inch square (driven into a l-inch hole). Formerly jag-bolts, 
■or rag-bolts, i. e., bolts whose sides were jagged, or barbed, were 
used for this and similar purposes ; but universal experience shows 
that smooth roda hold much tbe better. In some oxperimenta 
made at tbe Poughkeepsie bridge (g 414), it whs found that a I-inch 
rod driven into a ^|-inch hole in hemlock required on the average 
s force of 3( tons per linear foot of rod to withdraw it; and a 1-inch 
rod driven into aj-inch hole in white or Norway pine required 
5 tons per linear foot of rod to withdraw it.* The old-style j^- 
bolt was square becanse it was more easily barbed ; and probably 
this is the reason why square drift-bolts are now more common. 
Another advantage of the round drift-bolt, over the square one, is 
that the former does not cut or tear the wood as much as the latter. 
The ends of the rods should be slightly rounded with a hammer. 

Transverse timbers are put on top of the caps and drift-bolted 
lo them. Old bridge-timbers, timbers from false works, etc., are 
■ ForMdltlonal data, see Note 8, page HT. 



ovGoQi^lc 



254 PILE FOrNDATIOJfS. [CHAP. XI. 

frequently used, and are ordinarily as good for this purpose aa new. 
As many courses may be added as is necessary, each perpendicular 
to the one beVow it. The timbers of the top course si-e laid close 
together, or, as before stated, a floor of thick boards is added on top 
to receive the masonry. 

This form of construction is very common in the fonndAtions ot 
bridge abutments. Of course no timber should be used in a foun- 
dation, except where it will always be wet 

382. FiLEB AND CoVCRETE. A thick layer of concrete, resting; 
partly on the heads of the piles and partly on the soil between 
them, is frequently employed instead of the timber grillage aa above. 
Objection is sometimes made to the platform (§ 380) as a bed for a 
foundation that, owing to the want of adhesion between wood and 
mortar, the masoury might slide off from the platform if any un- 
equal settling should take place. To obviate this, the concrete is 
frequently substituted for the grillage and platform. 

However, there is but alight probability that a foundation will 
ever fail on account of the masonry's sliding on timber, since, ordi- 
narily, this could take place only when the horizontal force is 
nearly half of the downward pressure.* This could occur only 
with dams, retaining walls, or bridge abutments, and rarely, if 
ever, with these. One of the fundamental principles of all masonry 
construction is to build the courses perpendicular to the line of 
pressure, which condition alone would prevent slipping. Any pos* 
sibility of slipping can be prevented also by omitting one or mora 
of the timbers in the top course — the omitted timbers being per- 
pendicular to the direction of the forces tending to produce sliding; 
— or by building the top of the griilage in the form of steps, or by 
driving drift-bolts into the platform and leaving their upper ends 
projecting. 

Although the use of concrete, as above, may not be neoeesary to 
prevent sliding, it adds materially to the supporting power of th« 
foundation; it utilizes the bearing power of the soil between the 
piles as well aa the supporting power of 'the piles themselves, 
which is a very important consideration in soft soils. Another ad- 
vantage of this form of construction is that the concrete can be lai(i 
without exhausting- the water or sawing off the piles. Frequentli 

•See labia 86, page 819. 



ovGoQi^lc 



ABT. 3.] ABBAKGEUEITC OF THS F0UNDA.T10N. 3S& 

concrete can also be need advaatageonsly in conoection with timber 
grillage to paclc in aroand the timbers. 

383. LATEXAI. TnLDDtO. Notico that, although the masonry 
maj not elide off from the timber platform (g 38:^), the foundation 
may yield laterally by the piles themselves being pushed over. If 
the piles reach a firm subsoil, it will help matters a little to remoTO 
the upper and more yielding soil from aroand the tops of the piles 
and fill in with broken stone ; or a wall of piles may be driven 
around the fonndation — at some distance from it, — and timber 
braces be placed between the wall of piles and the foundation. 
When the foundation can not be buttressed in front, the structure 
may be prevented from moving forward by rods which bear on the 
face of the wall and are connected with plates of iron or blocks of 
stone imbedded in the earth at a distance behind the wall (se^ 
§ 551), or the thrust of the earth against the back of the wall may 
be decreased by supporting the earth immediately behind the 
foundation proper upon a grillage and platform resting on piles, or 
the same result may be attained by constructing relieving arches 
against the back of the wall (see g 55^). 

384. CDSHDro'B File Foohbatios. The desire to -utilize the 
cheapness aud efficiency of ordinary piles as a foundation for bridge 
piers and at the same time secure greater durability than is pos- 
sible with piles alone, led to the introduction of what is known aa 
Goshing's pile foundation, first used in 1868, at India Point, Hhode 
Island. It consists of square timber piles in intimate contact with 
each other, forming a solid mass of bearing timber. Surrounding 
the pile claster is an envelope of cast or wrought iron, sunk in the 
mud or silt only enough to protect the piles, all voids between piles 
and cylinders being filled with hydraulic concrete. 

Several such foundations have been used, and have proved 
satisfactory in every respect. The only objection that has ever 
been urged against them is that the piles may rot above the water 
line. If they do rot at all, it will be very slowly ; and time alone 
can tell whether this is an important objection. 

In making a foundation according to the Gushing system, the 
piles may be driven first and the cylinder sunk over them, or the 
piles can be driven inside the cylinder after a few sections are 
in place. In the latter case, however, the cylinders may be sub- 
jected to undue strains and to subse<}uent damage from shock and 



ovGoQi^lc 



•256 PILE rODNDATlOKS. [CHAP. XI. 

vibration; and besides, the sawiug off of the piles would be very 
-difficult and inconveuient, and they wonld have to be left at irreg- 
ular heights and with battered tops. On the other hand, if the 
piles are driven first, there is danger of their spreading and there* 
by interfering with the sinking of the cylinder. 

The special advanti^es of the Cnshing piers are : (I) cheapness, 
(2) ability to resist scour, (3) small coDtiHction of the water way, 
and (4) rapidity of construction. 

369. Example. The railroad bridge over the Tenas River, near 
Mobile, rests on Gushing piers. There are thirteen, one being a 
pivot pier. Each, excepting the pivot pier, is made of two cast- 
iron cylinders, 6 feet in exterior diameter, located 16 feet between 
centers. The cylinders were cast in sections 10 feet long, of metal 
1^ inches thick, and united by interior flanges 3 inches thick and 
-3 inches wide. The sections are held together by 40 bolts, each 
1^ inches in diameter. The lower section in each pier was pro- 
vided with a cutting-edge, and the top section was cast of a length 
sufficient to bring the pier to its proper elevation. 

The pivot pier is composed of one central cylinder 6 feet in 
-diameter, and six cylinders 4 feet in diameter arranged bexagonally. 
The radios of the pivot circle, measuring from the centers of cylin- 
ders, is l'Z\ feet. Jilach cylinder is capped with a cast-iron plate 
8J inches thick, secured to the cylinder with twenty 1-inch bolts. 

The piles are sawed pine, not less than 10 inches square at the 
-small end. They were driven first, and the cylinder sunk over 
them. In each of the large cylinders, 13 piles, and in each of the 
.smaller cylinders, 5 piles, were driven to a depth not less than 30 
feet below the bed of the river. The piles had to be in almost per- 
fect contact for their whole length, which was secured by driving 
their points in contact as near as possible, and then pulling their 
tops together and holding them by 8 bolts 1^ inches in diameter. 
In this particular bridge the iron cylinders were sunk to a depth 
not less than 10-feet below the river bed ; but usually they are not 
sunk more than 3 to 7 feet. The piles were cnt off at low water, 
the water pumped out of the cylinder, and the latter then filled to 
the top with concrete. 



ovGoQi^lc 



CHAPTER XIL 
FOUOTJATIONS UMDER WATER. 

386. The clasa of fotrndationB to be diBcusBed in this chapter 
iiould appropriatol; be called Foundations for Bridge Piers, since 
the latter are aboiit the only ones that are laid under water. In this 
class of work the chief difficulty is in excluding the water prelim- 
inary to the preparation of the bed of the foundation and the con- 
«truction of the artificial structure. This nsnally requires great 
resources and care on the part of the engineer. Sometimes the 
preservation of the foundation from the scouring action of the cur- 
rent is an important matter. 

PreTenting the undermining of the foundation is generally not a 
matter of much difficulty. In quiet water or in a sluggish stream 
bat little protection is required ; in which case it is sufficient to de- 
posit a mass of loose stone, or riprap, aronnJthe base of the pier, 
if there is danger of the riprap's being undermined, the layer must 
be extended farther from the base, or be made so thick that, if 
undermined, the stone will fall into the cavity and prevent further 
damf^e. A willow mattress sank by placing stones upon it is an 
economical and efficient means of protecting a structure against 
ecour. A pier may be protected also by inclosing it with a row of 
piles and depositing loose rock between the pier and the piles. In 
minor structures the foundation may be protected by driving sheet 
piles around it. 

If a bage qnanti^ of stone be deposited around the base of the 
pier, the velocity of the current, and consequently its scouring 
action, will be increased. Such a deposit is also an obstruction to 
navigation, and therefore is seldom permitted. In many cases the 
•only absolute security is in sinking the foundation below the scour- 
ing action of the water. The depth necessary to secure this adds to 
the difficulty of preparing the bed of the foundation. 

387. The principal difficulty in laying a foundation under water 
consists in excluding the water. If necessary, masonry can be laid 
under water by divers ; but this ia very expensive and is rarely re- 
ported to. 



ovGoQi^lc 



S68 FOUNDATIONS CHDEB WATBB. [CHAF. XH. 

There are five methoda in nae for laying foundatioDB under water: 
(1) the method of excluding the water from the bed of the founda- 
tion by the uee of a coffer-dam; (3) the method of founding the 
' pier, without excluding the water, by means of a timber crib but- 
mounted by a water-tight box in which the masonry ia laid; (3) the 
method of sinking iron tubes or masonry wells to a solid substratnm 
by excavating inside of them; (4) the method in which the water ia. 
excluded bythe presence of atmospheric air; and (5) the method of 
freezing a wall of earth around the site, inside of which the excava- 
tion can be made and the masonry laid. These several methods will 
be discussed separately in the order named. 



Abt. 1. The Goffeb-Dau Pbocebb. 

388. A coffer-dam is an inclosure from which the initer is pumped 
and in which the masonry is laid in the open air. This method con- 
eists in constructing a coSer-dam around the site of the proposed 
foundation, pumping out the water, preparing the bed of the fonn-- 
datioQ by driving piles or otherwise, and laying the masonry on the- 
inside of the coffer-dam. After the masonry is above the water the- 
coffer-dam can be removed. 

389. ConTBUCnm of ths Sak.* The construction of coffer- 
dams varies greatly. In still, shallow water, a well-built bank of* 
clay and gravel is sufficient. If there is a slow current, a wall of 
bags partly filled with clay and gravel does fairly well; a row of 
cement barrels filled with gravel and banked up on the outside haa- 
also been used. If the water is too deep for any of the above- 
methods, a single or double row of sheet piles may be driven and 
banked np on the outside with a deposit of impervious soil sufBcient. 
to prevent leaking. If there ia much of a current, the puddle on 
the outside will be washed away; or, if the water ia deep, a large- 
quantity of material will be required to form the puddle-waM; and 
hence the preceding methods are of limited application. 

390. The ordinary method of conatructing a coffer-dam in deep- 
water or in a strong current is shown in Fig. 60, The area to bo 
inclosed ia first aurrounded by two rows of ordinary piles, m, m. On 
the outside of the main piles, a little below the top, are bolted two- 

« See also I SIT, page 214. 



ovGoQi^lc 



ABT. I.] THE COFFEB-DAU PBO0E8S. Z59 

loDgitadinal pieces, w, w, called walea; and on the inside are fastened 
two similar pieces, g, g, which serve as guides for the sheet piles, s, x, 
while being driven. A rod, r, connects the top of the opposite 
maia piles to prevent spreading when the puddle is put in. The 
timber, /, is pat on primnrily to carry the footway,/, and is some- 
times notched over, or otherwise fastened 16, the pieces w, w to pre- 
vent the puddle space from spreading, b and b are braces extend- 
ing from one side of the coffer-dam to the other. These braces are 
put in position Buccessively from the top as the water is pumped 



ont; and as the masonry is built up, they are removed and the sides 
of the dam braced by short struts rcBting against the pier. 

The resistance to overturning is derived principally from the 
main piles, m, m. The distance apart and also the depth to which 
they should be driven depends upon the kind of bottom, the depth 
of water, and the danger from floating ice, \q^, etc. Roles and 
formulas are here of but tittle use, judgment and experience being 
the only guides. The distance between the piles in a row is usually 
from 4 to 6 feet. 

The dimensions of the sheet piles (§ 329) employed will depend 
npon the depth and the number of longitudinal waling pieces used. 
Two thicknesses of ordinary 2-inch plank are generally employed. 
Sometimes for the deeper dame, the sheet pilesare timbers lO or 1:^ 
inches square. 

The thickness of the dam will depend upon (1) the width of gang^ 
iray required for the workmen and machinery, (2) the thickness re- 



ovGoQi^lc 



360 ftJCNDATlONS CSDEE WATER. [CHAP. XII. 

quired to prevent OTtfrturnitig, and (3) th<i tliickaess of puddle 
DecessBrj to prevent leakage through the wall. The thickness of 
shallow dams will UBually \>b determined by the first consideration ; 
but for deep dams the thickness will be governed by the second or 
third requirement. If tht< braces, b, b, are omitted, aa is sometimes 
done for greater Gonvenience in working in the coffer-dam, then the 
main piles, m, m, must \m stronger and the dam wider in order to 
resist the lateral pressure of the water. A rule of tiiumb frequently 
used for this case is: "For depths of less than 10 feet make the 
width 10 feet, and for depths over 10 feet give an additional thick- 
ness of 1 foot for each additional 3 feet of wall." Trautwine's rule 
is to make the thickness of the puddle-wall three fourths of its 
height; but in no case is the wall to be less than 4 feet thick. If 
the coffer-dam is well braced across the inclosed area, the puddle- 
wall may vary from 3 feet for shallow depths to 10 feet for great 
depths; the former width has been successfully employed for depths 
of 18 to 30 feet, although it is considerably less than is customary. 

The puddle-wall should be constructed of impervious soil, of 
which gravelly clay is best. It is a common idea that clay alone, or 
«lay and fine sand, is best. With pure clay, if a thread of water ever 
ao small finds a passage under or through the puddle, it will steadily 
wear a larger opening. On the other hand, with gravelly clay, if 
the water should wash out the clay or fine sand, the larger particles 
will fall into the space and intercept fii'st the coarser sand, and 
next the particles of loam which are drifting in the current of water; 
and thus the whole mass puddles itself better than the engineer 
could do it with his own hands. An embankment of gravel is com- 
paratively safe, and becomes tighter every day. While a clay em- 
bankment may be tighter at first than a gravelly one, it is always 
liable to breakage. Before putting in the puddling, all soft mud 
and loose soil should be removed from between the rows of sheet 
piles. The puddling should be deposited in layers, and compacted 
as much as is possible without causing the sheet piles to bulge so 
much as to open the joints. 

391. Coffer-dams are sometimes constructed by building a strong 
crib, and sinking it The crib may be composed either of uprights 
framed into caps and sills and covered on the outside with tongued 
and grooved planks, or of squared timbers laid one on top of the 
-other, Ic^-house fashion, and well calked. The outer uprights are 



ovGoQi^lc 



ART, l.J THE COFPER-])A» PBOCEBS. ii6\ 

braced against the meide uprighte and sills to prevent cmshing 
inwards. This crib may be bnilt on land, launched, towed to itB 
final place, and snnk by piling stones on top or by throwing tbem 
into cella of the crib-work which are boarded up for that pnrpoBe. 
The bottom of the stream may be leveled off to receive the crib by 
dredging, or the dam may be made tight at the bottom by driving 
sheet piles around it. The crib must be securely bolted together 
(see § 381) vertically, or the buoyancy of the water will lift off the 
upper courses. 

A movable coffer-dam is sometimes constructed in the same 
general way, except that it is made in halves to allow of removal 
from around the finished pier. The two halves are joined together 
by fitting timbers between the projecting eourses of the crib, and 
then passing long bolts vertically through the several courses. Some 
of the compartments are made water-tight to facilitate the move- 
ment of the crib from place to place." 

Coffer-dams are also built by sinking an open crib, similar to the 
above, and then sheeting it on the outside by driving piles around 
it after it is sunk. For shallow depths, this method is very efficient. 

392. Sometimes two coffer-dams are employed, one inside of the 
other, the outer one being used to keep out the water, and the inner 
one to keep the soft material from flowing into the excavation. The 
outer one may be constructed in any of the ways described above. 
The inner one is usually a frame-work sheeted with boards, or a crib 
of squared timbers built Ic^-house fashion with tight joints. The 
inner crib is sunk (by weighting it with stone) as the excavation 
proceeds. The advantages of the use of the inner crib are (1) that 
the coffer-dam is smaller than if the saturated soil wore allowed to 
take its natural slope from the inside of the dam to the bottom of 
the excavation ; (3) the space between the crib .and the dam can be 
kept full of impervious material in case of any trouble with the out- 
side dam ; (3) the feet of the sheet piling are always covered, which 
lessens the danger of undermining or of an inflow of water and mud 
under the dam ; and (4) it also reduces to a minimum the material 
to be excavated. 

393. Iron has been used in a tew instances as a sheeting for cof- 
fer-dams. Plates are riveted together to form the walls, and stayed 

* For ftn Ulnetnted eismple, gee Froc Engineer's Club of Phlladelpbla, vol. tv- 



ovGoQi^lc 



363 FOUNDATIONS UNDER WATEE. [CHAP. XH. 

oD the inside by horizontal rings made of angle iron. Wood is 
cheaper and more easily wrought, and therefore generally preferred. 

394. Leaka&E. a serious objection to the use of cofFer-dams is 
the difficulty of preventing leakage under the dam. One of the 
simplest devices to prevent this is to deposit a bank of gravel around 
the outside of the dam ; then if a vein of water escapes below the 
aheet piling, the weight of the gravel will crush down and fill the 
hole before it can enlarge itself enough to do serious damage. If the 
coffer-dam is made of crib-work, short sheet piles may be driven 
around the bottom of it ; or hay, willows, etc., may be laid around 
the bottom edge, upon which puddle and stones are deposited ; or 
a broad flap of tarpaulin may be nailed to the lower edge of the 
crib and spread out loosely on the bottom, upon which stones and 
puddle are placed. A tarpaulin is frequently used when the ^ 
bottom is very irregular, — in which ease it would cost too much to 
level oft the site of the dam ; and it is particularly useful where the 
ttottom is rocky and the sheet piles cau not be driven. 

When the bed of the river is rock, or rock covered with but a 
few feet of mud or loose soil, a coffer-dam only sufficiently tight to 
keep out the mud is constructed. The mud at the bottom of the 
inclosed area is then dredged out, and a bed of concrete deposited 
under the water (§ 154). Before the concrete hae set, another coffer- 
dam is constructed, inside of the first one, the latter being made water- 
tight at the bottom by settling it into the concrete or by driving 
sheet piles into it. However, the better and moi-e usual method is 
to sink the masonry upon the bed of concrete by the method de- 
scribed in Art 2 (pages 266-71). 

It is nearly impossible to prevent considerable leakage, unless the 
"bottom of the crib rests upon an impervious stratum or the sheet 
piles are driven into it. AVater will find its way through nearly any 
^iepth or distance of gravelly or sandy bottom. Trying to pump a 
river dry through the sand at the bottom of a coffer-dam is expen- 
sive. However, the object is not to prevent all infiltration, but only 
to so reduce it that a moderate amount of bailing or pumping will 
keep the water out of the way. Probably a coffer-dam was never 
built that did not require considerable pumping; and not infre- 
quently the amount is very great, — so great, in fact, as to make it 
clear that some other method of constructing the foundation should 
Jiave been chosen. 



ovGoQi^lc 



ABI. I.] THE COFFEB-DAH PBOCESS. £63 

Seams of aand are veij troublesome. Logs or stones under the 
edge of the dftm are also a cause of considerable annoyance. It is 
fiometimes beat to dredge away the mud and loose soil from the site 
of the proposed coffer-dam ; but, when this is necessary, it is usu- 
ually better to construct the foundation without the use of a coffer* 
dam, — see Art. 2 of this chapter (page 366). Coffer^ma should 
be used only in very shallow water, or when the bottom is clay or . 
some material impervious to water. 

395. Pofflpi. Incontrtructingfoundations, it iefrequentlynecee- 
ssry to do considerable bailing or pumping. The method to be em- 
ployed in any particular case will vary greatly with the amount of 
water present, the depth of the excavation, the appliances at hand, 
etc. The pumps generally used for this kind of work are (1) the ordi- 
nary wooden haud-pump, (3) the steam siphon, (3) the pulsometer, 
and (4) the centrifugal pump. Rotary and direct-acting steam 
pumps are not suitable for use iu foundation work, owing to the 
deleterious effect of sand, etc., in the water to be pumped. 

1. Hand Power. When the lift is small, water can be bailed 
out faster than it can be pumped by hand ; but the labor is propor- 
tionally more fatiguing. The ordinary hand found ation-pnmp con- 
sists of a straight tube at the bottom of which is fixed a common 
flap valve, and in which works a piston carrying another valve. The 
tube is either a square wooden box or a sheet-iron cylinder, — usually 
the latter, since it is lighter and more durable. The pump is oper- 
ated by applying the power directly to the upper end of the piston- 
rod, the pnmp being held ii position by stays or ropes. There are 
more elaborate foundation-pumps on the market, 

2. The gleam siphnn la the simplest of all pumps, since it has 
DO movable parts whatever. It coiisista essentially of a discbarge 
pipe — open at both ends— through the side of which enters a smaller 
pipe having its end bent up. The lower end of the discharge pipe 
dips into the water ; and the small pipe connects with a steam boiler. 
The steam, in rnshing out of the small pipe, carries with it the air 
in the upper end of the discharge pipe, thus tending to form a 
vacuum in the lower end of that pipi.> ; the water then rises in the 
discbarge pipe and is carried out with the steam. Although it is 
possible by the use of large quantities of steam to raise small quan- 
tities of water to a great height, the steam siphon is limited prac- 
tically to lifting water only a few feet. Its cheapnese and simplicity 



ovGoQi^lc 



364 POUXDATIOXS USDEB WATBR. [CHAP. XIE 

are recommendations in its &Tor, and itB efficiency is oot much less 
than that of other forms of pumps. A common form of the steam 
siphon resembtes, in external appearance, the Eada sand-pnmp 
represented in Fig. 66 (page 293). 

3. The pulso?neler is an improved form of the steam siphon. It 
may properly be called a steam pump which dispenses with all mov- 
able parts except the valves. The height to which it may lift water 
is practically unlimited. 

4. The centrifugal pump* consists of a set of blades revolving in 
a short cylindrical case which connects at its center with a saction 
(or inlet) pipe, and at its circumference with a dischai^ pipe. The- 
blades being made to revolve rapidly, the air in the case is carried 
outward by the centrifugal force, tending to produce & vacuum in 
the suction pipe ; the water then enters the case and is discharged 
likewise. The distance from the water to the pump is limited by 
the height to which the ordinary pressure of the air will raise the 
water ; I but the height to which a centrifagsl pomp can lift the 
water is limited only by the velocity of the outer ends of the revolv- 
ing blades. "When a quick application with a discharge of lai^ 
quantities of water is the most important consideration, the cen- 
trifugal pump is of great value. Since there arc no valves in action 
while the pump is at work, the centrifugal pump will allow sand 
and large gravel — in fact almost anything that can enter between 
the arma — to pass. Pumps having a 6-inch to 10-inch discharge 
pipe are the sizes most frequently used in foundation work. 

396. Phepabisb the Fouhdatios. After the water is pumped 
out, the bed of the foundation may be prepared to receive the 
masonry by any of the processes described in §§ 2^3-91, which see. 
Ordinarily the only preparation is to throw out, nsually with hand 
shovels, the soft material. The masonry may be started directly 
upoa the hard substratum, or upon a timber grillage resting on 
the soil (§§ 309-10) or on piles (§ 380). 

397. Cost. It is universally admitted that estimates for th& 
cost of foundations under water are very unreliable, and none are 
more so than those contemplating the use of a coffer-dam. The 
estimates of the most experienced engineers frequently differ greatly 



ovGoQi^lc 



AET. 1.] THE COPPEB-BAM PROCESS. 266 

from the actual cost. The difQculties of the case have already bees 
discuased (§ 394). 

For the cost of piles and driving, see §§ 346-5i. .The timber 
will cost, according to locality, anywhere from $15 to $25 per 
thoneand feet, board measure. The coat of labor in placing the 
timber can not be giTen, since it varies greatly with the design, size, 
depth, etc. The iron in drift-bolts, screw-bolts, and spikes, is 
usually estimated at 3^ to 5 cents per pound in place. Excavation ' 
in coffer-dams frequently costs as high as tl to 91.50. per cubic 
yard, including the neceaeary pumping. 

398. Example. The following example is interesting as show- 
ing the cost under the most favorable conditions. The data are for 
a railroad bridge across the Ohio River at Point Pleasant, W. Va.* 
There were three 250-foot spans, one 400-foot, and one 200-foot. 
There were two piers on land and four in the water ; and all ex- 
tended about 90 feet above low water. The shore piers were 
founded on piles — driven in the bottom of a pit — and a grillage, con- 
crete being rammed in around the timber. The foundations under 
water were laid by the use of a double coffer-dam (§ Z92). The 
water was 10 feet deep ; and the soil was 3 to 6 feet of sand and 
gravel resting on dry, compact clay. The foundations consisted of 
a layer of concrete 1 foot thick on the clay, and two courses of 
timbers. The quantities of materials in the siz foundations, and 
the total cost, are as follows : 

PlDettmberin cribs insldeof coffer-dams, and in fotiDdMione, 378,310 ft. B.U. 

Oak timber in cofler dams, mAia and gheet piliog 344,412 ■■ " 

Poplar timber in cofferdams , 8.587 " " 

Rouod piles in fouaJutlon and coffer-dams I8,S71 Hd. ft. 

ExcBvation In foundaliona 4,843 cu. yds. 

Concrete " " 649" " 

Riprap 687 " " 

The total cost of fouodalions, including labor of all kinds, derricks, barges, 
engines, pumps, iron, tools, ropes, and everyttiiDg necessaiy for tbe rapid com- 
pletion of tbe work, was f<M,e53.63. 

In the construction of the bridge over the Miasonri River, near 
Plattamouth, Neb., a concrete foundation 49 feet long, 21 feet 
wide, and 32 feet deep, laid on shore, the excavation being through 
clay, bowlders, ahale, and soapstone, to bed-rock (33 feet below 

• Sngbuering Mat, vol. xitL p. 388. 



ovGoQi^lc 



'266 FODNDATIONS CNDEB WATER. [CHAP, 311. 

Enrface of the water), coat 139,607.23, or (43.81 per yard for the 
■concrete laid,* 

399. For the relative coat of foundations, see Art. 6, page 309. 

400. CoHCllFBiOH. Uncertainty as to what trouble and expeose a 
coffer-dam will develop usually causes engineers to choose some other 
method of laying the foundations for bridge piers. Cofler-dams 
are applicable in shallow depths only ; hence one objection to found- 
ing bridge piers by this process, particularly in rivers subject to 
scour or liable to ice gorges, is the danger of their being either un- 
dermined or pushed off the foundation. When founded in mud or 
sand, the first mode of failure is most to be feared. This danger is 

■diminished by the use of piles or large quantities of riprap; but 

■such a foundation needs constant attention. When founded on 
rock, there is a ^wssibility of the piers being pushed off the founda- 

' tion ; for, since it ia not probable that the coffer-dam can be pumped 
perfectly dry and the bottom cleaned before laying the masonry or 
depositing the concrete, there is no certainty that there is good 
anion between the base of the pier and the bed-rock. 

Coffer-dams are frequently and advantageously employed in 
laying foundations in soft soils not under water, as described in 

.§§316-21 (pages 214-15). 

AaT. 2. The Geib and Open-Caisson Procbss. 

401. ])EranTI05fl. Unfortunately there is an ambiguity in the 
nse of the word caisson. Formerly it always meant a strong, water- 
tight box having vertical sides and a bottom of heavy timbers, in 
which the pier is built and which sinks, as the masonry is added, 
until its bottom rests upon the bed prepared for it. With the in- 
troduction ot the compressed-air process, the term caisson was ap- 
plied to a strong, water-tight box — open at the bottom and closed 
at the top — upon which tlie pier is built, and which sinks to the 
bottom as the masonry is added. At present, the word oaiBson gen- 
erally has the latter meaning. In the pneumatic process, a water- 
tight box — open at the top — is usually constructed on the roof of 
the working chamber {" pneumatic chamber"), inside of which the 
masonry is built ; this box also is called a caisson. The caisson 



o.GoQi^lc 



AST. it-l THE CfilB AND OPEN-CAISSON PROCESS. 267 

open at the bottom is liometimeB called an inverted caisBOD, and the 
■one open at the top an erecl caisson. The latter when built over 
■an inverted, or pnenm&tic, caisson, is somctimeB called a coSer-dam. 
For greater cleameBS the term caisson will bo used for the inverted, 
or pneumatic, caisson ; and the erect caisson, which is built over a 
pneumatic caisson, will be called a coffer-dam. A caisson employed 
in other than pneumatic work will be called an open caisson. 

402. Feikciple. This metliod of constructing the foundation 
consists in building the pier in the interior of an open caisson, 
which sinks as the masonry is added and finally rests upon the bed 
prepared for it. The masonry usually extends only a foot or two 
below extreme low water, the lower part of the structure being com- 
posed of timber crib-work, called simply a crib. The open caisson is 
built on the top of the crib, which is practically only a thick bottom 
for the box. The timber is employed because of the greater facil- 
ity with which it may be put into place, as wilt appear presently. 
Timber, when always wet, is as dorabte as masonry ; and ordinarily 
there is not much difference in cost between timber and stone. 

If the soil at the bottom is sott and unreliable, or if there is 
-danger of scour in case the crib were to rest directly npon the bot- 
tom, the bed is prepared by dredging away the mud (§ 407) to a 
«ufficient depth or by driving piles which are afterwards sawed off 
{§ 3T8) to a horizontal plane. 

403. COMBTBircTlOH 07 THE CAIB80E. The constructioD of the 
caisson differs materially with its deptli. The simplest form is 
made by erecting studding by toenailing or tenoning them mto 
the top course of the crib and spiking planks on the outside. For 
a caisson 6 or 8 feet deep, which is about as deep as it is wise to 
try with this simple construction, it is sufRcient to use studding 6 
inches wide, 3 inches thick, and 6 to 8 feet long, spaced 3 feet apart, 
mortised and tenoned into the deck course of the crib. The sides 
and floor (the upper course of the crib) should be thoroughly calked 
with oakum. The sides may be braced from the masonry as the 
sinking proceeds. When the crib is grounded and the masonry is 
above the water, the sides of the box or caisson are knocked off. 

When the depth of water is more than 8 to 10 feet, the caisson 
is constructed somewhat after the general method shown in Fig. 61. 
The sides are formed of timbers framed together and a covering of 
thick planks on the outside. The joints are carefully calked to 



ovGoQi^lc 



S68 F0CSDATI0N8 CNDER WATEB. [CHAP. Xn. 

make the caissoQ water-tight. In deep caissoOB, the ddes can he 
bailt up as the masoary progresBeB, and thna not be in the waj ol 
the masons. The sides and bottom are held together onl; bj thd 
heavy vertical rods ; and after the caisson has come to a bearing 
Qpon the soil and after the masonry is above the water, the rods are 
detached and the sides removed, the bottom onJy remaining aa a 
part of the permanent structnre. 

For an illustration of the form of caisson employed in sinking a 
fonndatioD by the compressed-air process, see Plate I. 

404. The caisson shonld be so contrived that it cao be 



groanded, and afterwards raised in case the bed is fonnd not to 

be accurately leveled. To effect this, a small sliding gate is some* 
times placed in the side of the caisson for the parpose of filling it 
with water at pleasure. By means of this gate, the caisson can be 
filled and grounded; and by closing the gate and pumping out the 
vrater, it can be set afloat. The same resnlt can be accomplished by 
putting on and taking ofE stone. 

Since the caisson is a heavy, unwieldy mass, it is not possible to 
control the exact poeition In which it is sunk ; and hence it should 
be larger than the base of the proposed pier, to allow for a little ad- 
justment to bring the pier to the desired location. The margin to 



ovGoQi^lc 



6.BT. 2.] THB CBIB AND OPEN-CAISSON PEOCBSS. 268 

be allowed will depend upon the depth of water, size of cuBson^ 
iacilitioB, etc A foot all round Ib probably none too much under . 
favorable conditions, and generally a greater margin should b6 
allowed. 

406. COMSTKUCnoir or THS Csib. The crib is a timber struct- 
ore below the caissoQ, which tranBmite the pressure to the bed of 
the foundation. A crib is esaentially a grillage (see g 309 and § 380) 
which, instead of being built in place, ia first constructed and then 
sunk to its final resting place in a single mass. A crib is usually 
thicker, i, e., deeper, than the grillage. If the pressure ia great, the 
-crib is built of successive courses of squared timbers in contact; but 
if the pressure ia small, it is built more or less open. In either 
case, if the crib is to rest upon a soft bottom, a few of the lower 
courses are built open so that the higher portions of the bed may 
be squeezed into these cells, and thus allow the crib to come to an 
«ven bearing. If the crib ia to rest upon an uneven rock bottom, 
the site is first leveled up by throwing in broken stone. If the bot- 
tom is rough or sloping, the lower courses of the crib are sometimes 
made to conform to the bottom as nearly as pbssible, as determined 
from soundings. This method requires care and judgment to pre- 
vent the crib from sliding off from the inclined bed, and should be 
used with gi'oat caution, if at aU. 

The crib is usually built afloat. Owing to the buoyancy of the 
water, about one third of a crib made wholly of timber would pro- 
ject above the water, and would require an inconveniently lai^ 
weight to sink it ; therefore, it is best to incorporate considerable 
stone in the crib-work. If the crib ia more or less open, this Ib 
-done by putting a floor into some of the open spaces or pockets, 
which are then filled with stone. If the crib is to be solid, about 
every third timber is omitted and the space filled with broken stone. 

The timbers of each course should be securely drift-bolted (§ 381) 
to those of the course below to prevent the buoyancy of the upper 
portion from palling 'the crib apart, and also to prevent anj possi- 
tiility of the upper part's sliding on the lower. 

406. Tdoeb IK FomrsATloiiB. The free use of timber in' 
foundations is the chief difference between American and European 
nuthods of founding masonry in deep water. The consideration 
that led to its introduction in foundations was its cheapness. Many 
of the more important bridges built some years ago rest upon cril>- 



ovGoQi^lc 



S70 FOUNDATIONS UNDER WATEB, [CHAP. XTU 

work of Toniid logs notched at their intersection and secured by- 
drift-boltB. At preeeni, cribs are always built of squared timber. 
As a rnle, there is now but very little difference between the cost 
of timber and masonry in foundations. The principal advantage- 
in the use of the timber in fonndations under water is the facility- 
witli which it is put into position. Soft wood or timber which 
in the air has comparatively little durability, is equally as good 
for this purpose as the hard woods. It has been oonclnsiyely proved 
that any kiud of timber will last practically forever, if completely 
immersed in water. 

407. EXCATATDTQ THS BITB. When a pier is to be founded in 
a sluggish stream, it is only neceeeary to excavate a hole m the^ 
bed of the stream, in which the crib {or the bottom of the caisson)' 
may rest. The excavation is usually made with a dredge, any form 
of which can be employed. The dipper -dredge is the best, bnt thfr 
clam-shell or the endless chain and bucket dredge are sometimes 
used. If the bottom is sand, mud, or silt, the soil maybe removed 
(1) by pumping it with the water through an ordinary centrifugal 
pump (§ 395), — the aiiction hoae of which Js kept in contact with, 
or even a little below, the bottom, — or (2) by the Eads sand-pnmp' 
(§ 448). With either of these methods of excavating, asimple frama 
or light coffer-dam may be sunk to keep part of the loose soU fronk 
running into the excavation. 

408. If the stream is shallow, the current swift, and the bottom 
soft, the site may be excavated or scoured out by the river itself. 
To make the current scour, construct two temporary wing-dams, 
which diverge up stream from the site of the proposed pier. The- 
wings can be made by driving stout stakes or small piles into tho 
bed of the stream, and placing solid panels—made by nailing ordi- 
nary boardsto light uprights — against the piles with theirloweredgs 
on the bottom. The wings concentrate the current at the location 
01 the pier, increase its velocity, and cause it to scour out the bed of 
the stream. This process requires a little time, usually one to three 
days, but the coat of construction and operation is comparatively 
slight. 

When the water is too deep for the last method, it is sometimes 
possible to suspend the caisson a little above the bed of the stream, 
in which case the current will remove the sand and silt from under 
it. At the bridge over the Mississippi at Quincy, 111., a hole ID feet 



ovGoQi^lc 



SKI. 3.] DEBDQING THBOCQH WBLLfl. 371 

deep was thus sconred out. If the water is already heavily charged 
with sedimeDt, it may drop the aedimeat od striking the crib and 
thus fill up instead of scour out. Notwithstanding the hole ia 
liable to be filled up by the gradoal action of the current or by a 
sudden flood, before the crib has been placed in its final position, 
this method ia frequently more expeditious and less expensive than 
using a coffer- dam. 

409. It the crib should not rest squarely upon the bottom, it. 
can BOmetimee be brought down with a water-jet (§ 34j) in the. 
hands of a diver. However, the engineer should not employ a 
diver unless absolutely necessary, aa it is very expensive. 

410. If the soft soil extends to a considerable depth*, or if the 
necessary spread of foundation can not be obtained without an un- 
desirable obstruction of the channel, or if the bottom ia liable to- 
scour, then piles may be driven, upon which the crib or caisson maj 
finally rest. Before the introduction of the compreaeed-air process, 
this was a very common method of founding bridge piers in onr- 
westem rivers ; and it is still frequently employed for small piers. 
The method of driving and sawing oft the piles has already been 
described — see Chapter XI. 

The mud over and around the heads of the pilea may be sucked 
off with a pump, or it may be scoured out by the current (g 408). 
The attempt ia sometimes made to increase the bearing power of the- 
foundation by filling in between the heads of the piles with broken 
stone or concrete ; but this is not good practice, as the stone does, 
but little good, is difficult to place, and ia liable to get on top of the- 
piles and prevent the crib from coining to a proper bearing. 

Abt. 3. Dredoinq Throuqh Wells. 

411. A timber crib is frequently sunk by excavating the material 
through apartments left for that purpose, thn^ undermining the 
crib and causing it to sink. Hollow iron cylinders, or wells of 
masonry with a strong curb, or ring, of timber or iron beneath them^ 
are sunk in the same way. 

This method is applicable to foundations both on dry land and 
under water. It is also sometimes employed in sinking shafts ia 
tunneling and mining. 

41S. XXCATATOSB. The soil is removed from under the crib. 



ovGoQi^lc 



27S FOtniDATIOlIS UXDEB WATER. [CHAP. XII. 

vith a clom-ahell dredge, or with ao endless chain and bncket 
dredge, or with the Eads sand-pump, or, for small jobs, with the 
sand-pump employed in driving artesian wells. 

The clam-shell dredge coasists of the two halves of a hemi- 
spherical shell, which rotate about a horizontal diameter ; the edges 
of the shell are forced into the soil by the weight of the machine 
itself, and the pull upon the chain to raise the excavator draws the 
two halves together, thus forming a hemispherical bucket which ^ 
incloses the material to be excavated. The Morris and Gumming 
dredge consists of two quadrants of a short cylinder, hinged and 
operated similarly to the above. The orange-peel dredge (shown atf 
A in Fig. 63, page 374) appears to have the preference for this kind 
of work. It consists of a frame from which are suspended a num- 
ber of spherical triangular spades which are forced vertically into 
the ground by their own weight; the pull upon the excavator to 
lift it out of the mud draws these triangles leather and encloses 
the earth to be excavated. There are several forms of dredges 
similar to the above, bat differing from tbem in details. 

For a description of the Eads sand-pump, see §-448. 

413. In one case in France, the soil was excavated by the aid of 
iwmpressed air. An 8-)nch iron tube rested on the bottom, with its 
top projecting horizontally above the water ; and compressed air was 
discharged through a small pipe into the lower end of the g-inch 
tube. The weight of the air and water in the tube was less than 
an equal height of the water outside ; and hence the water in the 
tube was projected from the top, and carried with it a portion of the 
mud, sand, etc. Pebbles and stones of considerable size were thus 
thrown out. See § 447. 

414. Voted Exaicplebl— Pooghkeepiie Bridge. The Pongh- 
keepsie bridge, which crosses the Hudson at a point about 75 miles 
above New York City, is founded upon cribs, and is the boldest ex- 
ample of timber foundation on record. It is remarkable both for 
the size of the cribs and for the depth of the foundation. 

There are four river piers. The crib tor the largest is 100 feet 
long, 60 feet wide at the bottom and 40 feet at the top, and 104 
feet high. It is divided, by one longitudinal and six transverse 
walls, into fourteen compartments through which the dredge worked. 
The side and division walls terminate at the bottom with a 12" X 
I'Z" oak stick, which served as a cutting edge. The exterior walls 



ovGoQi^lc 



ABT. 3.] DBBDGINO THROUGH WELLS, 273 

and the loDgitndinal diTisiou vail were built solid, o! triaugnlar 
cross section, for ^0 feet above the cutting edge, and above that 
they were hollow. The gravel used to sink the crib weis deposited 
in these hollow walls. The longitudinal walls were securely tied to 
each other by the end and cross division walla, and each course of 
timber was fastened to the one below by 450 1-inch drift-bolts 30 
inches long. The timber was liemlock, I'i inches square. The 
fourteen compartments in which the clam-sbell dredges worked 
wei-e 10 X 12 feet in tlie clear. The cribs were kept level while 
sinking by excavating: from fii'st one and then the other of the com- 
partments. Gnwel was added to the pockets as the crib sunk, 
When hard bottom was reached, the dredging pockets were filled 
with concrete deposited under water from boxes holding one cubic 
yard each and opened at the bottom by a latch and trip-line. 

After the crib was in position, the masonry was started in a 
floating caisson which finally rested upon the top of the crib. 
Sinking the crib and caisson separately is a departure from the 
ordinary method. Insteadof usinga floating caisson, it isgenei-ally 
considered better to construct a coffer-dam on top of the crib, in 
which to start the masonry. If the crib is sunk first, the stones 
which are thrown into the pockets to sink it are liable to be loft ' 
projecting above the top of the crib and thus prevent the caisson 
from coming to a full and fair bearing. 

The largest crib was snnk through about 53 feet of water, 20 
feet of mud, 45 feet of clay and sand, and 17 feet of sand and 
gravel. It rests, at 1:I4 feet below liigh water, upon a bed of gravel 
16 feet thick overlying bed-rock. The timber work is 110 feet high, 
including the floor of the caisson, and extends to 14 feet below high 
water (7 feet below low water), at which point the masonry com- 
mences and rises 39 feet. On top of the masonry a eteel tower 100 
feet high is erected. The masonry in pUn is 25 x 87feet. and has 
nearly vertical faces. The lower chord of the channel span is 130 
feet and the rail is 212 feet above high water. 

The other piers are nearly as large as the one here described. 
The cribs each contain an average of 2,500,000 feet, board meusur*?, 
of timber and 350 tons of wronght iron. 

415. Atoha&laya Bridge. This bridge is over the Atcbafaluya 
bayou or river, at Morgan City, La., about 80 miles west of New 
Orleans. The soil is alluvial to an unknown depth, and is subject 



ovGoQi^lc 



£74 TOOITDATIOKS UKDER WATER. [OHAP. Zn. 

to rapid and extensive scour ; and no stone suitable for piers could 
be found within reasonable distance. Hence iron cylinders were 
adopted. They are fonndation and pier combined. The cylinders 
were snnk 120 feet below high water — from 70 to 115 feet below the 
mud line — by dredging the material from the inside with an orange- 
peel ezcavator. Fig. 63 shows the excavator and the appliances 
for handling the cjUndera. 



Fid. (B.— Bixccd Iboh Piut bt Dbidqimi. 

The cylinders are 8 feet in ontside diameter. Below the lerel 
of the river bed, they are made of cast iron 1^ inchee thick, in 
lengths of lOJ feet ; the sections were bolted together throngh in- 
side flanges with 1-inch bolts spaced 5 inches apart. Above the 
river bottom, the cylinders are made of wronght-iron plates ^ inches 
thick, riveted together to form short cylindrical sectidne with angle 
iron flanges. The bolts and spacing to unite the sections are the 
same as in the cast-iron portions. 

The cylinders were fflled with concrete and capped with a heavy 



ovGoQi^lc 



AET. 3.] DBEDOINO THBOnOH WELLS. 27S 

caet-iron plate. Two each cylinderB, braced together, form the pier 
between two 250-feet apaiis of a railroad bridge. 

The only objection to Buch piers relates to their Btabilitj. These 
have Btood satisfactorily since lttti3, 

416. Eavkeibnry Bridge. The bridge over the Hawkesbury 
River in Bouth-eaatern Australia is remarkable for the depth of the 
foundation. It ia founded upon elliptical iron caissons 48 x 30 feet 
at th6 cutting edge, which rest upon a bed of hard gravel 126 feet 
below the river bed, 185 feet below high water, aii.\ i^'J feet below 
the tirack on the bridge. The soil penetrated was mud and sand. 
The caisBonB were sunk by dredging through three tubes, 8 feet in 
diameter, terminating in bell-monthed eztenaiouB, which met the 
catting edge. The spaces between the dredging tabes and the 
onter shell were filled with gravel as the sinking progressed. The 
caissons were filled to low water with concrete, and above, with cut- 
stone masonry. 

417. Brick Cyliadera. In Germany a brick cylinder was sunk 
SSG feet for a coal shaft. A cylinder 35} foet in diameter was snnk 
76 feet through sand and graTel, when the frictional reBistance 
became so great that it could be sunk no further. An interior 
cylinder, 15 feet in diameter, was then started in the bottom of the 
larger one, and Bunk 180 feet further through running quicksand. 
The soil was removed without exhansting the wator. 

A brick cylinder — outer diameter 46 feet, thickness of wait 3 
feet — was sunk 40 feet in dry sand and gravel without any difficulty. 
It was built IS feet high (on a wooden curb 31 inches thick), and 
weighed 300 tons before the sinking was begun. The interior earth 
was excavated slowly, bo that the sinking was about 1 foot per day, 
— the walls being built up as it sank. 

In Europe and India masoury bridge piers are sometimes sunk 
by this procesa, a sufficient number of vertical openings being left 
through which the material is brought up. It is generally a tedioua 
and slow operation. To lessen the friction a ring of masonry is some- 
times built inBide of a thin iron shell. The last was the method em- 
ployed in putting down the foundations for the new Tay bridge, " 

418. I^CTIOVAL Bebutasoe. The friction between cylinders 
Knd the soil depends upon the nature of the soil, the depth sunk, 
and the method osed in sinking. If the cylinder ia aunk by either 

* For in QlitatnUd Mooont, aee SKgingiring Acu*, voL xlT. pp. 66-68. 



ovGoQi^lc 



876 



FODNDATIONS UNDEB WATBE. 



[chap. in. 



of the pnenmatic procesaea (§§ 425 and 4S6), the flow of the water 
or the air along the sidea of the tabe greatly dimtDiehee the fric- 
tion. It ie impoBBible to give any very defiaite data. ' 

The following table * gives the Talues of the co-efBcient of fric- 
tion f for materials and surfaces which occur in sinking foundations 
■ for bridge piers. Each result is the average of at least ten experi- 
mente. "All materials were rounded off at their face to sledge 
shape and drawn lengthwise and horizontally over the gravel or 
sand, the latter being leveled and bedded as solid as it is likely 
to be in its natural position. The riveted sheet iron contained 
twenty-five rivets on a surface of 2.53 x 1-67 = 4.23 square feet; 
the rivet-heads were half-round and \^ inch in diameter." Notice 
that for dry materials and also for wet gravel and sand, the frictional 
resistance at starting is smaller than during motion, which is con- 
trary to the ordinary statement of tlie laws of friction. 
TABLE 80. 

Co-KFTICIKNT OF FfilCTlOS OF MATERIALS AITD SOBFACEfl trSBD IH FOOK- 
SATIONB, 



ElSD ov IUtsbmu. 



m 



si 



Pioe (BBwed) on gravel acd Mod. . 
Sbect iroD 'without rivets on s&nd. . . 
" with " " " ... 

Cast iron ( un planed) on s&nd 

Qraolte (roughly worked) on sand. . 
Pine (Bswed) on sand 



419. Valuei from Actual Praotloe. Cast Iron. During the 
construction of the bridge over the Seine at Orival, a caetriron 

' * Bf A. SchmoU In " Zeltschrltt des Verelnea DentBCher iDgenlenre," aa repnb- 
Habed In Selected Abstiacte of loBt. ot C. E., vol. 111. pp. 298-303, 

t The co-efflotent ot friction Is equal to the total trictlon divided bj/ the touu 
Dornul presBore; that ia to ear, It la the friction per unit ot praaanie perpendlcnlnr 
lo the soilMea In contact. 



ovGoQi^lc 



AKT. S.] DBEDOINQ THBODOH WELLS. 377 

cylinder, BtaDding in an extenBive and rather uniform bed of graToI, 
and having ceased to move for thirty-two honrs, gave a frictional re- 
BiBtaDce of nearly 200 lbs. per aq. ft.* At a bridge over the Dannbe 
near Stadlau, a cylinder Bunk 18.75 feet into the soil (the lower 3.75 
feet being "solid clay") gave a frictional resistance of 100 lbs. per 
sq. ft.* According to some Enropean experiments, the friction of 
cast-iron cylinders in sand and river mnd was from 400 to 600 lbs. 
per sq. ft for small depths, and 800 to 1,000 for depths from 20 to 
30 feet.t At the first Harlem River bridge, New York City, the 
frictional resistance of a cast-iron pile, while the soil around it was 
still loose, was 5S8 lbs. per. sq. ft. of surface ; and later 716 lbs. per sq. 
ft. did not move it. From these two experiments, McAlpine, the en- 
gineer in charge, concluded that " 1,000 lbs. per sq. ft. is a safe value 
for moderately fine material." J At the Omaha bridge, a cast-iron 
pile sunk 87 feet in sand, with 15 feet of sand on the inside, could not 
be withdrawn with a pressare equivalent to 254 lbs. per sq. ft. of 
surface in contact with the soil ; and after removal of the sand from 
the inside, it moved with SOO lbs. per sq. ft.§ 

WrougM Iron. A wrought-iron pile, penetrating 19 feet into 
coarse sand at the bottom of a river, gave 280 lbs. per sq. ft; an- 
other, in gravel, gave 300 to 335 lbs. per sq. ft.| 

Matonry. In the silt on the Clyde, the friction on brick and 
concrete cylinders was about 3j tons per sq. ft.^ The friction on 
the brick piers of the Duflerin (India) Bridge, through clay, was 
WW lbs. per sq. ft** 

Fneumalic Caissons. For data on the frictional resistance of 
pneumatic caissons, see § 455. 

Piles. For data on the frictional resistance of ordinaiy piles, 
see g§ 370-71. 

420. Con. It is difficult to obtain data under this head, 
since but comparatively few foundations have been put down 
by this process. Furthermore, since the cost varies so much with 

* Tu Noatraod's Engln'g Hag., toL xx. pp. 121-22. 
t Proc. Insl, of C. E., vol. L p. 131. 

t McAlploe In Jour. Fraak. InsL, voL Iv. p. 106 ; alao Proc Init. of C. S., nL 
XXVU. p. 286. 

f y*a Nostrand'B Engfn'g Hog., vol. vllL p. 47L 
I Proc Inst, of C. E., vol. iv. p. 200. 
T Ibid., vol. iiziT. p. HG. 
** Snffinetrinff JVmi, voL zti. p. IBO. 



jvGooi^le 



278 FOUKDATIOITS UNDER WATBK. [CHAP. XIL 

the depth of water, stren^h of cnrreiit, kind of bottom, danger of 
floods, requirementa of DavigatioQ, etc, etc., no such data are valu- 
able onlees accompanied by endless details, 

Crlbi. The materials in the cribs will cost, in place, aboat as 
follows : timber from $30 to 140 per thousand feet, board measure ; 
drift and screw bolts from 3( to 5 cents per pound ; concrete from 
$4 to (6 per cnbic yard. Under ordinarily favorable conditions, the 
sinking by dredging will cost about (1 per cubic yard. 

Iron Tabei. Wrought-irou plate work will cost, exclusive of 
freight, from 3 to 4^ cents per pound ; cast-iron tubes, exclusive of 
freight, 1^ to 2 cents per pound. 

4S1. For the relative cost of different methods, see Art. 6 
of this chapter. 

422. CovcLunoH. A serious objection to this method of sink- 
ing foundations is the possibility of meeting wrecks, logs, or other 
obstructions, in the underlying materials ; but unlesa the freezing 
process (see Art. 5 of this chapter) shall prove all that is claimed 
for it, the method by dredging through tubes or wells is the only 
one that can be applied to depths which much exceed 100 feet— the 
limit of the pneumatic process. 

Abt. 4. Pnkumatio PaOOHSB. ■ 

424. The principle involved is the utilization of the difference 
between the pressure of the air inside and outside of an air-tight 
chamber. The air-tight chamber may be either an Iron cylinder, 
which becomes at once foundation and pier, or a box— open below 
and a:i' -tight eisewhere — upon the top of which the masonry pier 
rests. The former is called a pneumatic pile; the latter a pneu- 
matic caisson. The pneumatic pile is seldom used now. There 
are two processes of utilizing this difference of pressure, — the 
vacuum and thspUnum. 

426. VAOmnc Pbockh. The vacuum process consists in ex- 
hausting the air from a cylinder, and using the pressure of the at- 
mosphere upon the top of the cylinder to force it down. Exhausting 
the air allows the water to flow past the lower od^ into the air- 
chamber, thus loosening the soil and causing the cylinder to sink. 
By letting the air in, the water subsides, after which the exhaustion 
may be repeated and the pile sunk still farther. The vacuum 



D.qitizeabyG00l^lc 



AWr. 4.] PNBtTMATIC PR0CK8S. 279 

slioald be obtained Buddenly, so thnt the pressare of the atmosphere 
shall have the effect of a blow ; hence, the pile is connected by a 
large flexible tube with a large air-chamber — naually monnted upon 
a boat, — from which the air ia exhausted. When communication is 
opened between the pile and the receiver, the air rushes from the 
former into the latter to establish equilibrium, and the external 
pressure causes the pile to sink. 

To increase the rapidity of sinking, tbe cylinders may be forced 
down by a lever or by an extra load applied for that purpose. In 
case the resistance to sinking is very great, the material may be re- 
moved from the inside by a aand-pniap (g 448), or an orange-peel 
or clam-shell dredge (g 413) ; bat ordinarily no earth is removed 
from tbe inside. Cylinders have been sunk by this method 5 or 6 
feet by & single exhaustion, and 84 feet in 6 hours. 

The vacuum proeeBs has been superseded by the plenum process. 

426. Flbhux, ok Coxtbebhes-ais, Pbocsbb. The plenum, or 
compressed-air, process consists in pumping air into the air-chamber, 
eo ae to exclude the water, and forcing tbe pile or caisson down by 
a load placed upon it. An air-lock (§ 431) is so arranged that the 
workmen can pass into the caisson to remove the soil, logs, and 
bowlders, and to watch the progress of the sinking, withont re- 
leasing the pressure. The vacuum process is applicable only in mud 
or sand; but the compressed-air process can be applied in all kinds 
of soil. 

427. HiSTOBT 07 TezuMATIC FBOCEMEa. It is said that Papin, 
the eminent physicist — bom at Blois in 1647, — conceived the idea 
of employing a continued supply of compressed air to enable work- 
men to build under a large diving-bell. In 1779, Coulomb pre- 
sented to the Paris Academy of Science a paper detailings a plan for 
executing all sorts of operations under water by the use of com- 
pressed air. His proposed apparatus was somewhat like that now 
in general use. 

In England in 1831, Earl Dnndonald, then Lord Cochrane, took 
out a patent for a device for sinking tubular shafts through earth 
and water, by means of compressed air. His air-lock was much like 
modern ones, and was to be placed at the top of the main shaft. 
Hie invention was made with a view to its use in tunneling under 
the Thames, and in similar enterprises. In 1841, Bush also took 
out a patent in England for a plan of sinking foundations by the 



ovGoQi^lc 



260 FOUNDATIOKS CHDBB VATBB [CHAP. XH. 

aid of compressed air. A German, by name Q. Pfaon Mailer, made 
a somewhat similar design for a bridge at Majence, in 1850 ; but as 
hia plan was not executed, it was, like the patents of Cochrane and 
Bush, little known till legal controrersiee in regard to patent-rights 
dragged them from obscurity. 

42B. The first practical application of the plenum process waA 
made in France in 1841 by M. Triger. In order to reach a vein of 
coal on a sandy island in the Loire, opposite to Chalons, he sunk 
an iron tube about 40 inches in diameter, some 60 feet, by the 
blows of heavy weights. The fine sand was removed from the in- 
terior by means of a scoop bucket. On reaching a layer of coarse 
gravel, he could not force the tube through. He therefore capped 
his tube with an air-lock, and by compressed air forced out the 
water which bad all the while filled the tube, and sent workmen to 
the bottom. The pressure he used was never greater than two at- 
mospheres. The water was discharged through a small tube, into 
which, several feet from the bottom, a jet of air was allowed to 
enter, thus diminishing the specific gravity of the column till it 
was rapidly blown out. In 1845, Triger read a paper on the sinking 
of a tube about 6 feet in diameter to a depth of 82 feet by the same 
method, and suggested the use of it for the conetmction of deep 
foundations for bridges. 

Dr. Potts, of England, generally has the credit of inventing the 
vacuum process, for which he took out a patent in 1848. Many 
times in sinking foundations by the vacuum process, the com- 
pressed-air process was resorted to so that men could enter the pile 
to remove obetractions ; and finally the many advantages of the 
compressed-air process caused it to entirely supersede the vacuum 
process. At present the term "pneumatic process" is practically 
BynoDymone with compressed-air process. 

429. The first foundations sunk entirely by the compressed-ur 
process were the pneumatic piles for the bridge at Bochester, Eng- 
land, put down in 1851. The depth reached was 61 feet. 

The first pneumatic caisson was employed at Kehl, on the east- 
ern border of France, for the foundations of a r^boad bridge across 
the Rhine. 

480. The first three pneumatic pile foundations in America 
were constructed in South Carolina between 1856 and 1860. Im- 
mediately aftor the civil war, a number of pneumatic piles were 



ovGoQi^lc 



&BT. 4.] PKEnUATIO FKOCBSS. 381 

ronk in vestem rivera for bridge piers. Tbe first pneumatic cais- 
sons in America were those for the St. Louis bridge (g 457), put 
down in 1870. At that time th^se were the largest caissons ever 
constructed, and the depth reached — 109 feet 6^ inches — ^has not 
yet been exceeded. 

Of late years, the pnenmatio caisson has almost entirel; saper- 
seded the pneumatic pile ; in fact the plenom-pnenmatic caisson 
has almost entirely saperseded, except in very shallow water or in 
water over about 80 cr 100 ft. deep, ill other methods of founding 
bridge piers. 

431. 'PhzOIUTIO PII2B. Although pneumatic cylinders are now 
rarely employed, they will be briefly described because of their 
historic interest. 

The cylinders are made of either wrought or cast iron. The 
wrought-iroQ cylinders are composed of plates, about half an inch 
thick, riveted together and strengthened by angle irons on the in- 
side, and reinforced at the cuttting edge by plates on the outside 
both to increase the stiffness and to make the hole a little larger so 
as to diminish friction. The cast-iron cylinders are composed of 
sections, from 6 to 10 feet long and 2 to 8 fe^t in diameter, bolted 
together by inside flanges, the lower section being cast with a sharp 
edge to facilitate penetration. Two of these tubes, braced together, 
are employed for ordinary bridge piers ; and sis small ones around 
a large one for a pivot pier. They are filled with concrete, with a 
few courses of masonry or a heavy iron cap at the top. 

Fig. 63 shows the arrangement of the essential parts of a pneu- 
matic pile. The apparatus as shown is arranged for sinking by the 
plenum process ; for the vacuum process the arrangement differs 
only in a few obvious particnlars. The upper section constitutes 
the aif'lock. The doors a and h both open downwards. To enter 
the cylinder, the workmen pass into the air-lock, and close the 
door a. Opening the cock d allows the compressed air to entor tbe 
lock ; and when the pressure is equal on both sides, the door b is 
opened and the workmen pass down the cylinder by means of a ladder. 
To save loss of air, the air-lock should be opened veiy seldom, or 
made very small if required to be opened often. 

The air-supply pipe connects with a reservoir of compressed air 
on a barge. If the air were pumped directly into the pile without 
(he intervention of a storage reservoir, as was done in tiie early ap* 



ovGoQi^lc 



S62 vomsmxTiose usdeb wamb. [ckap. m. 

plications of the plenam process, even a momentaiy stoppage of tiie 
imgine Tonld endanger the lives of the workmen. 

432, The soil may be excavated by ordinary hand tools, elevated 
to the air-lock by a windlass and bucket, and passed out through 
the main air-lock. Sometimes a double air-lock with one large and 
sne small compartment ia nsed, the former being opened only to let 
gangs of workmen pass and the latter to allow the passage of the 




-k^'i 



Fm. et.— FmnuTic Piu. 

skip, or bucket, containing the excavated materiaL Sometimes an 
auxiliary lock, gf, is employed. The doors / and g are so con- 
nected by parallel bars (not shown) that only one can be opened at 
I time. The excavated material is thrown into the chute, the 
door/ is closed, which opens g, and the material discharges itself 
on the outside. 

Mud and sand are blown out with the sand-lift (g 447} or sand- 
pump (§ 448) without the use of any air-lock. 

433. The cylinders are guided in their descent by a frame-work 
resting upon piles or upon two barges. One of the chief difficoltiee in 



D.qitizeabyG00l^lc 



ART. 4.] PNEUKATIC PKOCBSS. 2S3 

sinking pneumatic piles is to keep them vertioal. If the cylinder 
becomes inclined, it can generally be righted (1) b; plaoiog wooden 
wedges ander the lower side of the cutting edge, or .f3) by excavat- 
ing under the upper side bo that the air may escape and loosen the 
material on that side, or (3) by drilling^ holes through the upper- 
most side of the cylinder through which air may escape and loosen 
the soil, or (4) by straining the top over with props or tackle. If 
several pneumatic piles are to form a pier, they should be sunk one 
at a time, for when sunk at the same time they are liable to mn 
together. 

434. Bearing Power. The frictional resistance of iron cylinders 
has been discussed In §§ 418-19, page 375-77, which see. 

McAlpine, in sinking the piers of the Harlem bridge. New York 
City, devised a very valuable but simple 
and cheap method of increasing the bear- 
ing power of a pneumatic cylinder (see 
fig. 64). He attached to the lower end ; 
of the cylindrical colnmn a hollow conical ■ 
iron section, the large end of which is , 
much larger than the main cylinder. V 
The base of the pier was still further in- 
creased by driving short sheet piles '; 
obliquely under the lower edge of the 
conical base and removing the soil from Pio. M. 

under them, after which the whole WEB filled in with concrete.* 

Id cold climates the contraction of the iron cylinder upon the 
masonry filling might rupture the former; hence, it is sometimes 
recommended to fill the pile below the frost line with asphaltic con- 
crete. It has also been proposed to line the cylinders with thick, 
soft wood staves, which will compress under the contraction of the 
Iron cylinder. However, the danger from this cause is not very 
serious; for, after the concrete has set, it is strong enough to 
support the load if the iron case were removed. 

43fi. After the cylinder has reached the required depth, concrete 
enough to seal it is laid in compressed air; and when this has 
set, the remainder can be laid in the open air. A short distance 
at the top is usually filled with good masonry, and a heavy iron cap 
put over all. 

* Jour. Frank. lost., vol. Iv. pp. 06 uid 1T7. 



ovGoQi^lc 



284 FOUKDATIONB USDBB WATER. [CHAP. XIL 



436. PxinuTlo Caubobi. A pDenmatic caiseoD is an ii 
box — opea below, bnt air-tight and watet^ tight elsewhere, — upon the 
top of which the masonry pier is built. The eseeiitial difference 
hetweea the pneniDatic pile and the pneumatic caisson is one of de- 
grce rather than one of quality. Sometimes the caisson envelope 
the entire masonry of the pier ; bnt in the usual form t!ie masonry 
eoTelops the iron cylinder and rests upon an enlargement of the 
lower end of it The pneamatic pile is sunk to the final depth be- 
fore fbeing filled with concrete or masonry; but with the caisson 
the masonry is built upward while the whole pitir is being sunk 
downward, the masonry thus forming the load which forces the 
caisson into the soiL A pneumatic caisson is, practically, a gigantic 
diving bell upon the top of which the masonry of the pier rests. 

Fig. 65 IB a section of a pier of the bridge across the Missouri 
Birer near Blair, Neb.,* and shows the general arrangement of the 
pier and pneumatic caisson. The tube extending through the mid- 
dle of the caisson and pier, known as the air-ahaft, is for the ascent 
and descent of the men. The air-lock — situated at the junction of 
the two cylinders which form tho air-ehaft — consists of a short sec- 
tion of a lai^e cylinder which envelops the ends of the two sections 
of the air-shaft, both of which communicate with the air-lock by 
doors as shown in Fig. 65. The apartment in which the men are 
at work is known as the working chamber or air-chamber. The 
small cylinders shown on each side of the air-shaft are employed in 
supplying concrete for filling the working chamber when the sinking 
is completed. The pipes seen in the air-chamber and projecting 
above the masonry are employed in discharging the mud and sand, 
as will he described presently. The timbers which appear in the 
lower central portion of the working chamber are parte of the trusses 
which support the central portions of the roof of the caisson. 

The masonry is usually begun about 2 feet below low water, the 
space intermediate between the masonry and the roof of the working 
chamber being occupied by timber crib-work, either built solid or 
filled with concrete. In Fig. 65 the masonry rests directly upon 
the roof of the air-chamber, which construction was adopted for the 
channel piers of this bridge to reduce to a minimum the obstrnctioa 
to the flow of the water. 

Frequently a coffer-dam is built upon the top of the crib (see 
* ITrom the report of Oeo. 8. Morlson, chief engliieer of the bridge. 



jvGooi^le 



AKT. 4.] PNECMATIC PROCESS. 285 

Plate I); but in this particular cage the maeopry was kept aboTe the 
enrface of the water, hence no coffer-dam was emplojed. When 



Pio. W.— PmnMiTic Ciii 



the coffer-dam ia not used, it is necessary to regulate the rate of 
sinking by the speed with which the masonry can be built, which is 
liable to cause inconvenience and delay. When the coffer-dam is 



ovGoQi^lc 



S86 FOUNDATIONS UNDBB WATBB. [CHAP. ZU. 

dispensed with, it is Qecessary to go on with the construction of the 
masonry whether or not the additional weight is needed in sinking 
tne caisson. 

437. The details of the conatmction of pnenmatic caisBons can 
be explained beat by the description of a puiicnlar case. 

438. tovtntknm o? the Eatbi db Okacx Bbidgi. Foldine 
Plate I * shows the details of the conBtrnction of the caisson, crib, 
and coffer-dam employed in 1884 in sinking pier No. 8 of the 
Baltimore and Ohio R. R. bridge across the Susquehanna Rirer at 
flavre de Grace, Ud. The timber work of Fig. 66 (page 293) also 
shows some of the detuls of the constmction of the walls of the 
working chamber. 

439. The CaiMon. The details of the construction of the caisson 
areas follows: Six courses of timber, 12 X 12 inch, one lying on top 
of tbe other, formed the skeleton of the walls of the working cham- 
ber. These timbers were first put up witb a batter of $ of an inch 
horizontal to 1 foot vertical; they were not halved at the corners, 
but every alternate piece was carried through with a full section, 
"log-house" feshion. These timbers were fastened at the comers, 
intersections, and several intermediate points, with drift-botts (g 381) 
1 inch square and S2 inches long. Inside of this timber shell, three 
courses of 3-inch plank, placed diagonally, were spiked to the hori- 
zontal timbers and to each other by 6-inch and 7-inch boat-spikes. 
Inside of the diagonal planking was another conree of 3-inch plank 
placed vertically and well spiked, the head of each spike being 
wrapped with oakum to prevent leakage. The vertical seams were 
thoroughly calked. 

A strong and thoroughly braced trues (see also Fig. 66, page 393) 
was next erected longitudinally through the center of the working 
chamber. The first course in the deck of the working chamber was 
then placed in position on the central truss and side watts. The work- 
ing chamber was 9 feet 3 inches high from bottom of shoe to the 
underside of deck, which was higher than required for working, bnt 
was adopted so as to permit greater depth of the central truss. Out^ 
side of the horizontal timbers, after tbey had been adzed to a true 
surface, were then placed the 12- by 14-incli sticks (shown at the ct- 

* Compiled from the oii^aa] working dfawlDga. The accompuiriDg deecrlpUon 
It Irom personal Inspection sided bj an article In Si^n»ering Seat by CoL Wm. V 
PauoD, engineer in charge. 



ovGoQi^lc 



ABT. 4.J PNErMATIC PROCESa 281 

treme left of Fig. 66) 15 feet long, exteDdiDg 2 feet below the bottom 
horizoatal timber and having their lower ends beveled as shown. 
These timbers extended 6 feet above the horizontal members, 
and were shouldered at the upper end so that three of the deck - 
courses rested upon them. Four screw-bolts were passed through 
each outside post and through the entire wall; and, in addition to 
these, two drift-bolts, 1 inch square and 30 inches long, in each ver- 
tical served to more thoroughly bind the wall t<^tber. This com- 
. pound of timber and planking formed the walls of the working 
chamber. After the first deck coarse was in place, a few pieces of 
the second course were laid diagonallj to give it stiffness; the under- 
side of this deck or roof was then lined with planks and thoronghlj 
calked, and a false bottom put into the working chamber prepara- 
tory to launching it. 

After the caisson was launched the deck courses, eight in all, 
were pat on. The fiivt course was made of single-length timbers, 
reaching from inside to inside of the vertical wall poets, and resting 
on top of the horizontal timbers aad Inside plauking and also on the 
top chord of the central truss, and being fastened to these members 
by 22-inch drift-bolts. The second course was laid diagonally and 
was made of varying lengths of timbers. The third course was laid 
from side to side across the caisson, and the fourth course lon^- 
tadinally and resting on the shoulders of the 13 X 14 inch verticals. 
The fifth course was laid across, the siith diagonally — croBBing the 
second course, — and the seventh and eighth conraes extended to the 
extreme outside limits of the caisson and rested on the heads of the 
vertical posts. This general arrangement of the top courses, resting 
as they did on the heads and shoulders of the outside verticals, gave 
ft direct bearing on the poets and relieved the wall bolts of the great 
shearing strain to which they would otherwise have been subjected. 

The outside poets were bolted to the deck conraes by one 3-foot 
screw-bolt and two 30-inch drift-bolta, fastening them to the longi- 
tudinal and diagonal courses respectively. The several deck courses 
were bolted to each other by 22-inch drift-bolts (not shown in the 
illustrations), spaced 5 feet apart along each stick. All the timbers 
in the deck were bedded in cement mortar and the vertical joints 
were grouted, so as to give a full and uniform bearing for aach stick 
and also decrease the leakage and danger from fire. 

The center truss (see also Fig. 66) was constructed to bear s unt 



D.qitizeabyG00l^ic — 



88d FOUNDATIONS UNDER WATER. [CHAP. XIL 

formly (tistributed load, or to act aa a cantileTer. It was composed 
of a t<o^ and bottom chord, each made of two 12 x 12 inch eticks, 
with posts and diagonals of wood, and Terticat and dit^na] tie-roda 
IJ inchiD diametec; the iron vertical rods extended tbroagh the 
tint deck courses, and the top chord was also bolted to the deck 
with drift-boltB. The object of this was to unable the truss to act 
as a stiffening rib to the deck, independently of its action ae a 
girder. The bottom chord was also extended to the ends, and by 
means of straps and bolts acted both as a strut and tie>brace for the 
ends of the caisson, and constituted the only end bracing. 

The sides of the caissons were braced against outside prcssnres by 
16 X 16 inch timbers abutting against the walls and bottom chord of 
the center truss, and against preesnre from the inside by S-inch iron 
tie-rods extending from out to out of the caisson, none of which are 
shown. All the timber used, except the planking and outside pot^ts 
and the bracing in the working chamber, was 12x12 inch. Iron 
straps, ext«ndiDg 6 feet on tliu sides and ends, were placed at the 
comers and bolted to the caisson timbers. These straps were made 
of bar-iron 3x1 inch and prevented spreading of the walls of the 
caisson under excessive pressure within. Planks were spiked to the 
lower part of the posts ; and also a narrow plank, called a shoe, was 
spiked under the bottom of the posts (see Fig. 66). 

440. " The construction was simple and strong ; in no case vms 
there any bending or springing of the walls. The arrangement of 
the cutting edge with square shoulders was a departure from the 
ordinary V-ahape (compare Figs. 65 and 66, pages 286 and 293), 
and was found to possess many advantages. It enabled the men to 
better regulate the sinking of the caisson by giving an increased bear- 
ing surface. With this support, the material could be cleaned out 
from under one side or end ; the caisson could be leveled ; and, if 
the material was softer in one spot than another, the caisson could 
be prevented from tipping. It further afforded a good surface for 
blocking np when it vras found desirable to support the caisson 
during the removal of the material ; and it gave also greater security 
in case of a 'blow-out' hr the failure of air-pressure,"* 

When it is anticipated that gravel or bowlders will be met with 
in sinking, the cutting edee is usually shod with iron. The iron 
cutting edge was omitted in all the caissons for this bridge, and it is 
* Col. Wta. H. Patlon, engUieer Id cbarge tor tbe nllroad companr. 



ovGoQi^lc 



ART. 4.] FKEUKATTC FBOCEBS. 389 

claimed that the experience here shows that " in no case is an iron 
shoe either adrantageouB or necessary." 

441. The Crib. The construction of the crib ia shown very fully 
in Plate I. The timbers were all 12 X IS inches square, bolted to 
each other by 33-inch drift-bolta — spaced 5 or 6 feet apart, — and 
were dovetailed at the corners and connections. The parts of all 
the walla of the crib were firmly bolted to the deck of the caiason. 

Ordinarily the division walls of the crib are bailt vertically from 
top to bottom ; but in this case, they were off-set, as shown, to 
secure a better bond in the mass of concrete. If the walls are built 
solid from top to bottom, the concrete filling ia thereby divided into 
a number of separate monolithic columns ; bat in the construction 
as above, the concrete forma practically a single solid mass. The 
walls are built solid, owing to the difficulty of getting the concrete 
thoroughly packed in around so many timbers. Large stones, such 
as could be handled by one man, were bedded in mortar as the sno- 
cesstve layers of concrete were formed, and over and around these 
another layer of concrete was rammed. In most localities there is 
but little difference in cost between a solid timber crib and one with 
timber pockets filled with concrete. 

442. The Coffer-dam. Uprights were first placed at intervals of 
about 5^ feet, and connected by mortise and tenon to caps and 
silla. This frame-work was held down to the crib by rods 3 inches 
in diameter, having hooks at the lower end which passed into eye- 
bolts in the sides of the crib. On the sides of the dam, the upper 
end of these rods passed through 13 X 13 inch timbers resting on 
the sides of the dam and projecting about 3 feet outside ; and at the 
ends of the dam, they passed thruugh short pieces bolted to one of 
the cross timbers and projecting beyond the end of the dam. 

Owing to the great depth required, the coffer-dam waa built in 
sections, the connecting rods being made in sections with swivel 
connections. The second section was not added until the depth 
sunk required it. When the top section of the dam was put on, 
the projecting ends of the timbers through which the connecting 
rods passed were sawed off. The bottom section was sheeted with 
thre? courses of 3-inch plank, and the top section with two thick- 
neeseg. The joint between the coffer-dam and the crib, and also 
'he sheeting, were well calked. 

The sides of the coffer-dam were braced against the pressure of 



ovGoQi^lc 



S90 FOimDATIONB CNDSB WATBK, [CHAP. XII. 

the water, by 13 X 12 inch timbers restiog on the top of e&ch seo- 
tioQ, and by a syBtem of bracing in the middle of each eection. 
When the maaonry woa completed, the coSer-dam was remored by 
disconnecting the vertical rods. 

443. Maehi&ery Bai^ e. The machinery barge was an ordinary 
flat-boat fitted ap for the purpose. At one end of the barge there 
were three boilers each of fifty horse-power. In the middle were 
two large air-compressom, designed by the contracting engineer. 
Gen. Wm. Sooy Smith. One furnished all the compressed air re- 
quired, the other being ready for nse in case of any accident or 
break-down. At the other end of the boat were two Worth ingtou 
steam pumps to furnish water for the excavating plant naed in the 
caiBSOtt. There were also a small engine and a dynamo which fnr- 
nisbed the current for the electric lamps used in the caisson and, at 
night, on the outside. 

444. Katerial Keqnired. Table 31 gives the dimeneionB and 
quantities of materials in the pneumatic foundations of this bridge, 
and Table 32 (page ZOi) gives the cost 

TABLE 81, 

DniBKBioKS Am* QuANTiTiEB OF Matrrialb ih Fouhdatiohb or 

Havre de Grace Bridoe.* 



DncuFTUur. 


Hdhub or TO» Pira, 


n. 


IIL 


IV. 


vm. 


IX 




M.3 
tS.B 

;!| 

♦0.0 

K03,47S 

in,sss 

SIOSB 
1.84B 

iS 


tn.» 

EB.S 

«:o 

'bt'oT 
jijn? 

1,SW 
111.BI 

siBn 

TOO 
S,fiM 


n.4 
u.s 

IT.S 

IIB.WB 
10M18 

J 

a.aa 


JD.fl 

a£.8 

•li 

si! 


Tin 


'■mut '■ " ■ " ■■ 


tt.B 










Tlmbar In Uie c«<M>n, Imt, board meuun. 
" crib. ■hufUi. etc., •■ - ... 





















* Tba data br oootUa; of SonjsmlUi A Co., omlrmclorm for the pneumatic (oundatloni. 

445. Fonriiur 07 thi Aix-logk. Before the construction of 
the St. Louis bridge the air-look had always been placed at the tap 



ovGoQi^lc 



ART. 4.] FNSUICATIO PB00BB8. 291 

of the air-Bh&ft, and was of ench conBtrnotion th&t to leog^lieD the 
shaft, as the caisson sunk, it was necessary to detac'j the lock^ add 
a seotion to the shaft, and then replace the lock on top. This vas 
not only inconyenient and an interrnptioii to the other work, bnt 
reqoired the men to climb the entire distanoe under compressed 
air, Thich is exceedingly fotigaing {see g 460). To overcome theee 
objections, Eads placed the air-lock at the bottom of the shaft. 
This position is objectionable, since in case of a " blov-ont," i'. e., 
a rapid leakage of air, — not an nnfretjnent occurrence, — the men 
may not be able to get into the lock in time to escape drowning. If 
the lock is at the top, they can get ont of the way of the water by 
climbing up in the shaft. 

At the Havre de Grace bridget the air-shaft was constructed of 
wrought iron, in sections 15 feet long. The air-lock was made hj 
placing diaphragms on the inside flanges of the opposite ends of the 
top section. A new section and a third diaphragm could be added 
without disturbing the air-lock ; and when the third diaphragm 
was in place, the lower one was removed preparatory to using it 
again. Some engineers compromise between these two positions, 
and leave the air-lock permanently at some intermediate point in 
the pier (see Fig.- 65, page 285). 

446. EXCATATOBS. In the early application of the pnenmatio 
method, the material was excavated with shovel and pick, elevated 
in buckets or bags by a windlass, and stored in the air-lock. When 
the fur-Iock was full, the lower door was dosed, and the air in the 
lock was allowed to escape until the npper door could be opened, 
and then the material was thrown ont. This method was expensive 
and slow. 

In the first application of the pneumatic procees in America 
(g 430), Qen. Wm. Sooy Smith invented the auxiliary air-lock, jr/. 
Fig. 63 (page 283), through which to let out the excavated mate- 
rial. The doors, / and g, are ho connected together that only one 
of them can be opened at a time. The excavated material being 
thrown into the chute, the closing of the door / opens g, and the 
material slides out. This simple device is said to have increased 
threefold the amount of work that could be done. 

447. Band-lift. This is a device, first nsed byCten. Wm. Sooy 
Smith, for forcing the sand and mud out of the caisson by means 
of the pressore in the working chamber. It oonsiHts of a pipe. 



ovGoQi^lc 



292 FOUNDATIONS UNDER WATEK. [CHAP. XII. 

rescblDg from the working chamber to the Barface (see Fig. 63 aod 
Plate I), controlled by a valve in the working chamber. The sand 
is heaped np aroand the lower end of the pipe, the valve opened, 
and the pressure foroes a continnons stream of air, sand, and water 
ap and ont. For another application of this principle, see g 413. 

In sand, this method of excavating is very efficient, being eight 
to ten times as expeditions ae the aaxiliary air-look. Of course, 
the efficiency varies with the depth, i. e., with the pressure. When 
the 8oil 18 so impervioas that the water in the working chamber can 
not be forced out under the edge of the caisson, it is made to pass 
through the sand-lift pipe. 

The "goose-neck," or elbow at the top of the discharge pipe, is 
worn away very rapidly by the impact of the ascending sand and 
pebbles. At the Havre de Grace bridge, it was of chilled iron 4 
inches thick on the convex side of the curve, and even then lasted 
only two days. At the Brooklyn bridge, the discharge pipe ter- 
minated with a straight top, and the sand was discharged against a 
block of granite placed in an inchned position over the upper end. 

Although the sand-lift is efficient, there are some objections to 
it : (1) forcing the sand ont by the pressure in the cylinder de- 
creases the pressure, which causes, particularly in pnenmatic piles or 
small caissons, the formation of vapors so thick as to prevent the 
workmen from seeing; (2) the diminished pressure allows the 
water to flow in under the catting edge ; and (3) if there is much 
leakage, the air-compressors are unable to supply the air fast 
enough. 

448. Xnd-pomp. During the construction of the St. Louis 
bridge, Gapt. Eads invented a mud-pump, which is free from the 
above objections to the sand-lift, and which in mud or silt is more 
efficient than it. This device is generally called a sand-pump, but 
is more properly a mnd-pump. 

The principle involved in the Eads pump is the same as that 
employed in the atomizer, the inspirator, and the injector; riz., the 
principle of the indnced current. This principle ia utilized by dis- 
charging a stream of water with a high velocity on the outside of a 
small pipe, which prodnces a partial vacuum in the latter ; when 
the pressure of the air on the outside forces the mud through the 
small pipe and into the current of water by which the mud is 
Ukrried away. The current of water is the motive power. 



ovGoQi^lc 



PNEUUATIC PROCESS. 



Fig. 66 ie an interior view of the caieaon of the Baltimore and 
Ohio R. R. bridge at Havre de Grace, Md., and shows the general 
arrangement of the pipee and mud-pnmp. The pnmp itself is a 



ovGoQi^lc 



394 FODNDATIOSS UlTDEB WATEB. [cHAP. XIL 

hollow pesr-shaped casting, about 15 inches in diameter and 15 
inches long, a section of which is shown in the comer of Fig. 66. 
The water is forced into the pump at a, impingee against the coni- 
cal casing, d, flows around this lining and escapes upwards through 
a narrow anonlar space, /. The interior casing gives the water an 
even distribution around the end of the suction pipe. The flow of 
the water through the pump can be regulated by screwing the suc- 
tion pipe in or out, thus closing or opening the annular space, y^ 
To prevent the too rapid feeding or the entrance of lumps, which 
might choke the pipe, a strainer — simply a short piece of pipe, 
plugged at the end, having a series of ^-inch to J-inch holes bored 
in il — was put on the bottom of the suction pipe. The discharge 
pipe of the mud-pump terminates in a "goose-neck" through 
which the material is discharged horizontally. 

The darkly shaded portions of the section of the pump wear 
away rapidly ; and hence they are made of the hardest steel and 
conatructed ao aa to be readily remored. Different engineers bare 
different methods of providing for the renewal of theae parts, the 
outline form of the pump varying with the method employed. The 
pump used at the St. Louis bridge was cylindrical in outline, but 
otherwise essentially the same as the above. 

449. In order to use the mud-pump, the material to be exca- 
vated is first mixed into a thin paste by playing upon it with a jet 
of water. This pump is used only for removing mud, silt, and soil 
containing small quantities of sand ; pure sand or soil containing 
large quantities of sand ia " blown out " with the sand-lift. 

The water is delivered to the mud-pump under a pressure, ordi- 
narily, of SO or 90 pounds to the square inch. At the St. Louis 
bridge, it was found that a mud-pump cf 3J-incb bore waa capable 
of raising 20 cubic yards of material 120 feet per hour, the water 
pressure being 150 pounds per square inch.* 

450. Hirater-colnmiL A combination of the pneumatic process 
and that of dredging in the open air through tubes has been em- 
ployed extensively in Europe. It seems to have been used first at 
the bridge ocrosa the Rhine at Kehl. The same method woa uaed 
at the Brooklyn bridge. The principle is rudely illustrated in 

* HlBtorf of th« St. LonU Bridge, p. Sia 



ovGoQi^lc 





J<> 








Jtr 










^ 






^~ 


/~H^r*i4 


& 


CAjimitr' \ 






N 


B 


W^ 







ABT. 4.] PSSVUXJIO PR00K88. 29S 

Fig. 67. The central shaft, which is open top and bottom, project! 
a little below the cutting edge, 
and is kept fnli of water, the 
greater height of water in the 
coIantQ balanciDg the pressure 
of the air in the chamber. The 
workmen simply push the mate- 
rial under the edge of a water- 
shaft, from whence it ia exca- 
vated by a dredge (§ 412). 

461. Blasting. Bowlders or 
points of rock may be blasted in 
compressed air without auy ap- 
preciable danger of a "Wow- 
out" or of injuring the ear- 
drums of the workmen. This '^ "'■ 
point was settled in sinking the foundations of the Brooklyn bridge ; 
and since then blasting has been resorted to in many cases. Bowl- 
ders are sometimes " carried down," i, e., allowed to remain on the 
surface of the soil in the working chamber as the excavation pro- 
ceeds, and subsequently imbedded in the concrete with which the 
air-chamber is filled. 

4fi2. Kate OV Smcnra. The work in the oaisBon usually con- 
tinues day and night, winter and summer. The rate of progress 
varies, of course, with the kind of soil, and particularly with the 
number of bowlders encountered. At the Havre de Grace bridge, 
the average rate of progress was 1.37 ft. per day ; at Plattsmouth, 
3.22 ft. ; and at Blair, 1.75 ft. per day. 

453. OuiDnie THE Cajbbos. Formerly it was the custom to 
control the descent of the caisson by suspension screws connected 
with a frame-work resting upon piles or pontoons. In a strong 
current or io deep water, it may be necessary to support the caisson 
partially in order to govern its descent ; but ordinarily, the suspension 
is needed only until the caisson is well imbedded in the soil. The 
caisson may be protected from the current by constructing a break- 
water above and producing dead water at the pier site. 

After the soil has been reached, the caisson can be kept in its 
course by removing the soil from the cutting edge on one side or 
the other of the caisson. In case the cuseon does not settle dovm 



D.qitizeabyG00l^lc 



296 POCNDAHOKS DNDEE WATBE. [CHAP. XII. 

after the soil has been remoyed from under the cutting edge, a re- 
duction of a few pounds in the air pressure in the working chamber 
is usually sufficient to produce the desired reenlt. At the Havre de 
Grace bridge, it was found that by allowing the discharged mate- 
rial to pile np against the outside of the caisson, the latter could 
bemored laterally almost at will The top of the caisson was made 
3 feet larger, all round, than the lower course of masonry, to allow 
for deriatioD in sinking. The deviation of the caisson, which was 
founded 90 feet below the water, was less than 18 inches, even 
though neither suspension screws nor guide piles were employed. 

In sinking the foundations for the bridge over the Missouri 
Biver near Sibley, Ho., it was necessary to move the caisson con- 
siderably horizontally without sinking it much farther. This was 
accomplished by placing a number of posts — 13 inches square — 
in an inclined position between the roof of the working chamber 
and a temporary timber platform resting on the ground below. 
When these posts bad been wedged np to a firm bearing, the 
air pressure was released. The water flowing into the oaisson 
loosened the soil on the outside, and the weight of the caisson com- 
ing on the inclined posts caused them to rotate about their lower~ 
ends, which forced the caisson in the desired direction. In this 
way, a lateral movement of 3 or 4 feet was secured while sinking 
about the same distance. 

A caisson is also sometimes moved laterally, while sinking, 
by attaching a cable which is anchored off to one side and kept 
taut. 

4M. A new method of controlling the descent of the caisson has 
been recently introduced, which is specially valuable in swift cur- 
r«nta or in rivers subject to sudden riEcs. It was used first in the 
construction of the piers for a bridge across the Yazoo Kiver near 
Ticksbui^ Miaa. A group of 73 piles, each 40 feet long, was driven 
into the river bed, and sawed off under the vrater ; the caisson was 
then floated into place, and lowered until the heads of the piles 
rested against the roof of the working chamber. As the work 
proceeded, the piles were sawed off to allow the caisson to sink. 
One of the reasons for employing piles in this case, was that, if the 
caisson did not finally rest upon bed-rock, they would assist in sup- 
porting the pier. 

That such ponderous masses can be so certainly guided in their 



ovGoQi^lc 



AKT. 4.] PNEUMATIC PROOB88. 297 

descent to bed-rock, is not the least valtiable nor least ioterestiiig 
&Gt connected with this method of sinking fonndatione. 

456. TSICTIOVAI. BziUTAiros, At the Havre de Qrace bridge, 
the normal friotional reeistanee on the timber sides of the pneamatic 
caisson vas 280 to 350 lbs. per sq. ft. for depths of 40 to 80 feet, 
the soil being silt, sand, and mud ; when bowlders were encoun- 
tered, the resistance was greater, and when the air escaped in large 
qnantities the resistance was less. At the bridge over the Missouri 
Kiver near Blair, Xeb., the frictional resistance usually ranged be- 
tween 350 and 450 Iba. per sq. ft., the soil being mostly fine sand 
with some coarse sand and gravel and a little clay. At the Brook- 
lyn bridge the frictional resistance at times was 600 lbs. per eq. ft. 
At Cairo, in sand and gravel, the normal friction wiui about 600 lbs. 
per sq. ft. 

For data on the friction of iron cylinders and masonry shaftA, 
see |g 418-19, pages 2V5-77; and for data on the friction of ordi- 
nary piles, see §§ 370-73, pages 347-48. 

166. FlLUFe THE Ais-OHAHBEK. When the caisson has 
reached the required depth, the bottom ts leveled off — by blasting, 
if necessary, — and the working chamber and shafts are filled with 
concrete. Sometimes only enough concrete is placed in the bottom 
to seal the chamber water-tight, and the remaining space is filled 
with sand. This was done at the east abutment of the St Louis 
bridge, the sand being pamped ia from the river with the sand- 
pump previously used for excavating the material from under the 
caisson. 

467. Voted Exahfibl The St. Louis Bridge. The founda- 
tions of the steel-arch bridge over the Mississippi at St. Louis are 
the deepest ever sunk by the pneumatic process, and at the time of 
construction {1870) they were also very much tbe largest. The 
caisson of the east abutment was an irregular hexagon in plan, 
83 X 70 feet at the base, and 64 x 48 feet at the top — 14 feet above 
the catting edge. The working chamber was 9 feet high. The 
cutting edge finally rested on the solid rock 94 feet below low 
water. . The maximum emersion was 109 feet 8^ inches, the greatest 
depth at which pneumatic work has yet been done. The othui' 
caissons were almost as large as the one mentioned above, but were 
not sunk as deep. 

The caissons were constructed mainly of wood ; but the sido 



ovGoQi^lc 



S9S FOITHDATIONa DNDEB WATEB. [CHAF. XII. 

wsIIh and the root were covered with plate iron to prevent leakage, 
and strengthened by iron girders on the insida - T%iB was the first 
jmemnatic caisson coiiBtracted in America ; and the ase of large 
qnsntitiea of timber was an important innovation, and has become 
one of the distinguishing characteristics of American practice. In 
all subsequent experience in this country (except as mentioned in 
§ 458), the iron lining for the working chamber has been dispensed 
with. The masonry rested directly upon the roof of the caisson, 
i. e., QO crib-work was employed. In sinking the first pneumatic 
foundation an iron coffer-dam was built upon the top of the caisson ; 
but the last — the largest and deepest — was sunk without a coffer- 
dam, — a departure from ordinary European practice, which is occa- 
sionally followed in this country (see § 436). 

468. The Brooklrii Bridge. The foundations of the towers of 
the suspension bridge over the East Eiver, between New York City 
and Brooklyn, are the largest ever sunk by the pneumatic process. 
The foundation of the New York tower, which was a little larger 
and deeper than the other, was rectangular, 173 X 102 feet at the 
bottom of the foundation,- and 157 x 77 feet at the bottom of the 
masonry. The caisson proper was 31J feet high, the roof being a 
solid mass of timber 22 feet thick. The working chamber was 9^ 
feet high. The bottom of the foundation is 78 feet below mean 
high tide, and the bottom of the masonry is 46^ feet below the 
same. From the bottom of the foundation to the top of the 
balustrade oit the tower is 354 feet, the top of the tower being 376 
feet above mean high tide. 

To make the working chamber ui^-tight, the timbers were laid 
in pitch and all seams calked ; and in addition, the sides and the 
roof were covered with plate iron. As a still further precaution, 
the inside of the air-chamber was coated with varnish made of rosin, 
menhaden oil, and Spanish brown. 

For additional details see the several annual reports of the en- 
gineers in charge, and also numerous articles in the engineering 
newspapers and magazines from 18C9 to 1873. 

469. Forth Bridge. For an illnBtrated account of the pueumatu 
foundation work of the bridge across the Frith of Forth, Eng- 
land, see Engineeriitfi Newf, vol. xiii. pages 242-43. The caissons 
employed there differed from those described above (1) in being 
made almost wholly of iron, (3) in an elaborate system of cages for 



ovGoQi^lc 



ABT. 4.] FKEUHATIC PB0CES3. 209 

hoisting the material from the ineide, and (3) in the use of inter- 
locked hydraulic apparatns to open and close the air-locks. Each 
of the two deep-water piers consists of four cylindrical caissons 
?0 feet in diameter the de^>eBt of which rests 96 feet below high 
tide. 

460. Phtkolobioal XvnoT W COXPBMBED AlK. In the ap- 
plication of the compressed-air process, the qaestion of the ability 
of the hnman system to bear the increased pressure of the air be- 
comes very important. 

After entering the air-lock, ae the pressure increases, the first 
sensation experienced is one of great heat. As the pressure is still 
further increased a pain is felt in the ear, arisiog from the abnormal 
pressure upon the ear-drum. The tubes extending from the back 
of the month to the bony cavities Over wbich this membrane is 
stretched, are so very minute that compressed air can not pass 
through them with a rapidity sufficient to keep up the equilibrium 
of pressure on both sides of the drum (for which purpose the tubes 
were designed by nature), and the excess of pressure on the ontaide 
causes the pain. These tubes can be distended, thus relieving the 
pain, by the act of swallowing, or by closing the nostrils with the 
thiimb and finger, shutting the lips tightly, and intlatiag the 
cheeks. Either action facilitates the passage of the air through 
these tubes, and establishes the eqailibrinm desired. The relief is 
only momentary, and the act must be repeated from time to time, 
as the pressure in the air-lock increases. This pain is felt only 
while the air in the lock is beiug "equalized," i. e., while the air is 
being admitted, and is most severe the first time compressed air is 
encountered, a little experience geuerully removing all unnleaaant 
sensations. The passage through the lock, both going in and com- 
ing out, should be slow ; that is to say, the compressed air should 
be let in and out gradually, to give the pressure time to equalize 
itself throughout the various ports of the body. 

When the lungs and whole system are filled thoroughly with 
the denser air, the general effect is rather bracing and exhilarating. 
The increased nmonnt of oxygen breathed in compressed air very 
much accelerates the organic functions of the body, and hence labor 
in the caisson is more exhaustire than in the open air : and on get- 
ting outside again, a reaction with a general f aeling of prostration 
sets in. At moderate depths, however, the laborers in the caisson. 



ovGoQi^lc 



300 FOUNDATIOITS UNDEB WATBK. [OBAP. XII. 

after a little experience, feel no bad effects from the compressed air, 
either while at work or afterwards. 

Bemainlng too long in the working chamber cansee a form of 
paralysis, recently named caisson disease, which is sometimes fatal 
The injarions effect of compressed air is much gr«iter on men ad- 
dicted to the nse of intoxicating liquors than on others. Only 
sound, able-l^ied men should be permitted to work in the cusson. 

In passing throngh the air-lock on leaving the air-chamber, the 
workman experiences a great lose of heat owing (1) to the expan- 
sion of the atmosphere in the lock, (2) to the expansion of the free 
gases in the cavitiea of the body, and (3) to the liberation of the 
gases held in solution by the liquids of the body. Hence, on com- 
ing out the men should be protected from currents of air, sbonld 
drink a cup of strong hot coffee, dress warmly, and lie down for a 
short time. 

461. No physiolo^cal difficulty is encountered at small depths ; 
bat this method is limited to depths not mnch exceeding 100 feet, 
owing to the deleterious effect of the compressed air upon the work- 
men. At the east abutment of the St. Louis bridge (§ 457), the 
caisson was sunk 110 feet below the surface of the water. Except 
in this instance, the compressed-air process has never been applied 
at a greater depth than about 90 feet. Theoretically, the depth, in 
feet, of the lower edge of the caisson below the surface divided by 
33 is equal to the number of atmospheres of pressure. The press- 
ure is never more than this, and sometimes, owing to the fric- 
' tional resistance to the flow of the water through the soil, it is a 
little less. Therefore the depth does not exactly indicate the 
preaanre ; but the rule is sufQciently exact for this purpose. At St. 
Louis, at a depth of 110 feet, the men were able to work in the 
compressed air only four hours per day in shifts of two hours each, 
and even then worked only part of the time they were in the air- 
chamber. 

With reasonable care, the pneumatic process can be applied at 
depths less than 80 or dO feet without serious consequences. At 
great depths the danger can be greatly decreased by observing the 
following precautions, in addition to those referred to above : (1) In 
hot weather cool the air before it enters the caisson ; * {'i) in cold 



ovGoQi^lc 



ART. 4.] PHEUICATIC PH0CES8. 301 

weather warm the air Id the lock vben the men come out ; and 
(3) raiee and lower them b; machinery. 

For as exhaustive acconnt of theTarione aspects of this sabject, 
see Dr. Smith's article on the " Phydological Effect of Comprened 
Air," in the Report of the Engineer of the Brooklyn Bridge.* 

462. Cost. The contract for pneumatic foundation is usually 
let at specified prices per nnit for the materish left permanently in 
the structure and for the materia) excavated, inclnding the neces- 
sary labor and tools. The prices for material in place are about as 
follows : Timber in caisson proper, from (40 to 150 per thousand 
feet, board measure, according to the locality in which the work is 
done ; and the timber in the crib-work and coffer-dam about (5 to 
t7 per thousand less. The concrete, which'is usually composed of 
broken stone and sufficient I to 2 or 1 to 3 Portland cement mortar 
to completely fill the voids, costs, exclusive of the cement, from (5 to 
t? per cubic yard for that in the crib, and about twice this sum for 
that in the air-chamber and under the cutting edge, The wrought- 
iron spikes, drift-bolts, screw-bolts, and cast-iron washers cost from 
3| to 6 cents per pound.f The caisson and filling costs from (14 to 
(30 per cubic yard ; and the crib and filling from (8 to (10. 

The price for sinking, incladtng labor, tools, machinery, etc., 
ranges, according to the kind of soil, from 18 to 40, or even 60, 
cents per cubic foot of the volume found by multiplying the area 
of the caisson at the cutting edge by the final depth of the latter 
below low water. In sand or silt the cost is 18 to 20 cents, and in 
stiff clay and bowlders 40 to 50 cents. 

463. Examples. The table on page 302 gives the details of the 
coat of the pneumatic foundation of the Havre de Grace bridge, as 
fally described in gg 438-44. 

The t»ble on page 303 gives the details of the cost of the pnen- 
matic caissons of the bridge across the Missouri River near Blair, 
Neb. Thecaieeons (Fig. 65, page 285) were54feetlong, 24 feet wide, 
and 17 feet high. In the two shore piers, Nos. I and IV of the 
table, the caissons were surmounted by cribs 20 feet high ; but in 
the channel piers, the masonry rested directly upon the roof of the 

■ PrlM EsM7 of tlie Alnmnl Anodstloii of the College of Pbrsiclaiu and anc- 
BMDS ot New Totk atf, 1S7S. 

t There Me nanallr from 140 to ISO ponnda of Iron per thonund feet (board iimm- 
nie) of timber. 



ovGoQi^lc 



FODNDAnONS UNDEB 



[cHAF. xn. 



TABLETS. 
Con, to THS B. B. Co., or Foukdattoiis or Hatsb db Gk&cb Budsk." 





Kmon or Tn Pio. 




tL 


UL 


rv. 


vm. 


IX. 


Diptt^fll outUiic «!«• Mow knr wu<r. 


U.S 

•»,««■« 

n,8a 

MS1.14 
I.B)1.H 

1tOJS.M 


ro.T 

B«.T 

idbIsos 

110.088.44 

1.M7.15 

r.oiT.so 

IB8T 
S.tU.19 
1,4MS6 

■11 
S;SS 


S0.> 

n.s 

„..,- 

•ffiS 
■iSS:S 


in.sn 

IIS.ir8.0T 
!.»».40 

"S 

1.T4S.H 
S1,M1.III> 




Depth of eutUuc sdee below mod lliM. 
(oUl COM, pw DM on. yd 


3&8 

m.ssi 

10T.8B8 
ttl.TeT.86 




gj:a^lJi.^,i«oiyd:: 


•^•s-fs 


C«( a( ^<ikl^?'^Ml^pErc!!! ft .rf 
OooonM below cnulng vils*, a »li.» - - 


sg».tt 

io,au,Qa 




es,oig.4T 
isn 


T1.MS,1S 

ti.isa 


wjM.aB 


10»,848.« 




tow muuirr, iDoludtng coSor^tuni. 


M.M 



ATsnge loMJ com of tbe loundaUoo, to R. K. Co., par d« cubic yard ma. 

caisson. The' work waa done, in J883-83, by the bridge companj^s 
men under the direction of the engineer. 

464. Id 1869-72, thirteen cylinders were sunk by the plennm- 
pnenmatic process for the piers of a bridge orer the Schuylkill 
Eiyer at South Street, Philadelphia. There were three piers, one 
of which waa a piTot pier. There were two cylindere, 8 feet ia 
diameter and 83 feet long, sunk through 32 feet of water and 30 
feet of " sand and tough compact mud intermingled with bowlders;" 
two cylinders, 8 feet in diameter and 57 feet long, sunk through 22 
feet of water and 5 feet of soil as above ; one cylinder, 6 feet in 
diameter and 64 feet long, sunk through 'JS feet of water and 18 
feet of soil as above ; and 8 columns, 4 feet in diameter and aggre- 
gating 507 feet, sunk through 22 feet of water and 18 feet of soil 
OS above. A 10-foot section of tbe 8-foot cylinder weighed 14,600 
pounds, of the 6<foot, 10,800 pounds, and of the 4-foot, 6,800 
ponnds. The cylinders rested upon bed-rock, and were bolted to 
* Data bf emuteey ot So^amlUt A Co., coDtractdDK engineers tor the pnenmmtio 



ovGoQi^lc 



PHBUMATIO PEOCESa. 



303 



AKT. 4.] 

It. The actual cost to the contractor, exclneJTe of tools and ma- 
chineiy, was as in Table 34 (page 304). 
TABLB 88. 
Con or Phbdhatio Fduhtoltiohb aw Blaix Bbidob,* 



tt oktaMin vu hnrerad afWr oomple- 
ib it coMdk «dsa bekiir nirface of 




Crib and mllnc. 

Air-lock, ibftfu, 

SIdUdk oainoo. ooal o(, tnalivUns anicllon aiid re- 

■aofkl ol macblaenr 

SlaUnc caUaod, coat of, par eqbia foot of dlapUce- 

ment below position of cuUin( edite wbeo 

oalMonwHOompistal 

Stalling calnon, coac of. per cobio foot o( diaplaoe- 

menC below aiufacB of water 

._. . of.peroublo toot at < 






$II.T58. 
7.MS.IB 
I, (81 .10 



"1 



l.UdBO 
0.888.18 



Arerass ooet t of tha foundaUona, per CQbto j-ard tVl.TO. 

465. "ExcaTatlon in the Brooklyn caiesou J cost for labor 
only, including the men on top, about (5.25 per cubic yard 
[19 cents per cubic foot]. Runoiog the six air-compressora 
. added to this $3.60 per hour, or about 47 cents per yard ; lights 
added tO.56 more; and these with other contingencies nearly 
equaled the cost of labor. The great cost was due to the ezceBsiTe 
hardness of the material over much of the surface, the caisson finally 
resting, for nearly its whole extent, on a mass of bowlders or bard- 
pan^ The concrete in the caisson cost, for every expense, about 

* Complied from the report of Geo. S. Horlaon, oUef eugtneer of tbe biidgg. 

t Rxclnslve ot engineering eipenaea and coat ot tools, macUuery, and bnlldlngs. 
In a note to the anthor, Hr. Horison, tbe engineer of the bridge aafa : " It la Impoe- 
■Ible to divide tbe bnlldlnga, tool*, and engineering Bipenam betveen the Bubetmct- 
nre and other portions of the work. The balk of Uie items ot tools and machlcerr 
[(lS,a(N.S8}, however, relates to tbe foondatlonB." The engineering ezpenaea and 
bnlldlnga were nearly 3 per cenL of the total cost of the entire bridge. The ooet ot 
tools and machinery was equal to a little over IS per oent. of tbe ooat ot tbe tonnda- 
tiona as above. Inclnding these Items would add neariy one lizUt t( 
the laat thne 11dm. 

1 For a brief deacriptlon, see f 456. 



ovGoQi^lc 



FOUNDATIONS UNDEB WATEE. 



[chap. XIL 



TABLE U. 
Cost or Precxatio Piles at Phtladslphu nr 1889-73.* 



Com oCcul Iron, a tfiO-Wper ton 

" " bolu. &IH oonW per lb 

" " grouied rubble maaonry (ezdiulTe of labor^, i^ $tl.M 

" " (InlclnK.andlByiDg nuBOurr 

ToM oonof (be eyllodert in ptaos 

Com of iron per line*l root or orlldder 

** " mMeri«)i tor Diaaop.T per lineal tooC arcrlioder. .. 
" " ■InklntcaadlajIngmuoDrrperllneBirootoIayliiidei 

Total oo«t,t per lineal foot, of crllnder Id place . . 



4m.84 



li(.ivrr,«o 
s:o.i» 

S.77B.KT 






tl5.50 per cabic yard. The caisson aod filling together aggregated 
16,898 cnbic yards ; and the approximate cost per yard for every 
expense was $20.71." ( The foundation therefore cost about t30 
per cnbio yard. 

The pneumatic fonndations for the channel piers of the bridge 
over the Missouri at Plattsmouth, Neb., cost as follows: One 
foundation, consisting of a caisson 50 ft. long, 20 ft. wide, and 15.5 
ft. high, surmonnted by a crib 14.15 ft. high, snnk through 13 ft. 
of water and 20 ft. of soil, cost-H9.29 per cubic yard of net volume. 
Another, consisting of a caisson 50 ft. long, 20 ft. wide, and 15.5 
ft. high, surmounted by a crib 36.25 ft. high, snnk through 10 
ft. of water and 44 ft. of soil, cost 914.45 per cnbic yard of net 
volume, g 

468. European Examples. The following^ is interesting as 
showing the cost of pneumatic work in Europe : 

"At Moolins, cast-iron cylinders, 8 feet 2^ inches in diameter, 
with a filling of concrete and sunk 33 feet below water into marl, 
cost $62.94 per lineal foot, or (29.71 for the iron work, and 933.23 



* Compiled from an article bf D. HoN. Stanfler, engtneer Id abwge. In Tmiu. Am. 
800. o( C. E., ¥ol. yll. pp. 3S7-S09. 

t Ezcltulve of tools and machinery. 

I F. CoUlngwood, aaalBtaDt en^oeer Brooklni bridge, In Tntns. Am. Boc. of C. S. 

( Complied from the report of Geo. B. Morl«ni, chief engliieer of the bridge. 

T Bjr Jnlea OaDdard, as tmnslnted from the French by L. F. VemoD-Haroonrt for 
the Proceedings of the Iiutituce of Civil Eagtneers (London). 



ovGoQi^lc 



ABT. 4.] PITBUIIATIO PROCESS. 309 

for sinking and concrete. At Argenteail, with cylindere 11 feet 10 
inches in diameter, the sinking atone cost (42.13 per lineal foot 
[nearly (10 per cubic yard], where a cylinder was sunk 53} feet in 
three hundred and ninety hours. [The total cost of this founda- 
tion was (34.09 per cubic yard, see table on page 310.] At Orival, 
where a cylinder was suuk 49 feet in twenty days, the coat of sinking 
was (36.83 per lineal foot. At Bordeiiux, with the same-sized 
cylinders, a gang of eight men conducted the sinking of one cylin- 
der, and UBually 34 cubic yards wei-e excavated every twenty-four 
hours. The greatest depth readied was 55| feet below the ground, 
and 71 feet below high water. In tJie regular course of working, a 
cylinder was sunk in from nine to fifteen days, and the whole opera- 
tion, including preparations and filling with concrete, occupied on 
the average 35 days. One cylinder, or e half pier, cost on the aver- 
age (11,S98.40, of which (1,461 was for sinking. M. Morandi^re 
estimates the total cost of a cylinder sunk like those at Argenteuil, 
to a depth of 50 feet, at (7,012.80. 

467. "Considering next the cost of piersof masonry on wrought- 
iron caissons of excavation, the foundations of the Lorient viaduct 
over the Scorff cost the large sum of (24. 11 per cubic yard, owing to 
difficulties caused by the tides, the labor of removing the bowlders 
from nnderaeath the caisson, and the lai^ coet of plant for only 
two piers. The foundation of the Kehl bridge cost stiil more, about 
(38.23 * per cubic yard ; but this can net be regarded as a fair iu- 
Btance, being the first attempt [see g 429] of the kind. 

"The foundations of the Nantes bridges, sunk 56 feet below 
low-water level, cost about (14.84 per cubic yard. The average 
coet per pier was as follows : 

CalsMD (41 fe«t 4 fochea by 14 feet S inches), 50 loiu of wrougbl 

iron®(n6.88 (5.844 

CoSer-dBm, 8 toos of wrought iron ® (58.44 175 

Excavation. 916 cubic fards @ (4.47 4.0nt 

Concrete 4,188 

Haacmry, plant, etc 1.8T0 

Average coat per pier (16,166 

"One pier of the bridge over the Mouse at Rotterdam, with a 

■ Notice the iltghl inoonslirteiiof between tbia qnuitltr and tbe one Id tba thlid 
line from the laat of tbe table on page 810, both being from tbe same anfcle. 



ovGoQi^lc 



306 FOUNDATIONS UNDSS WATER. [CQAP. XII. 

caisBon of 222 tons and a cofler-dam casing of 94 tons, and sank 75 
feet below high water, cost $70,858, or $13.97 per cabic yard. 

" The Yiohy bridge has five piers bailt on caissons 34 feet by 13 
feet, and two abutments on caissons 26 feet by 34 feet. The foun- 
dations were sunk 23 feet in the gronnd, the upper portion con- 
Biating of ahingle and conglomerated gravel, and the laet 10 feet of 
marl. The cost of the bridge was aa followa : 

lolerest far eight moDOw, and depredation of plant worth (19,480. . t8,8W 

Coat of prepsratioDB, approach bridge, and atagiog 4,IKH 

OOMona (tereD), 150} lou ® IU&.B8 17,106 

BinkiDg 9,8i8 

Concrete and maaonry B,80S 

Contrartor'a bonus and genenl ezpenaeB 6,107 

Total coat of five foundattona 947,141 

The cost per cubic yard of the foundation below low water was 
tl6.69, of which the sinking alone cost $3.50 in gravel, and $4.37 
in marl 

"At St. Maurice, the cost per cubic yard of foundation was 
$15.94, exclusive of staging." 

468. CORCLnnoH. Except in very shallow or very deep water, 
the compressed-air process has almost entirely snperseded all others. 
The following are some of the adrantages of this method. 1. It is 
reliable, since there is no danger of the caisson's being stopped, 
before reaching the desired depth, by sunken logs, bowlders, etc., 
or by excessive friction, as in dredging through tubes or shafts in 
cribs. 2. It can be used regardless of the kind of soil overlying the 
rock or ultimate foundation. 3. It is comparatively rapid, since 
the sinking of the caisson and the building up of the pier go on at 
the same time. 4. It is comparatively economical, since the weight 
added in sinking is a part of the foundation and is permanent, and 
the removal of the material by blowing out or by pumping is aa 
uniform and rapid at one depth as at another, — the cost only being 
increased somewhat by the greater depth. 5. This method allovrs 
ample opportunity to examine the ultimate foundation, to level the 
bottom, and to remove any disintegrated rock. 6. Since the rock 
can be hiid bare and be thoroughly washed, the concrete can be com- 
menced upon a perfectly clean surface ; and hence there need be no 
qaestioD as to the stability of the foundation. 



ovGoQi^lc 



AST. 6.] THE FEEEZING PBOCESS. 



Aet. 5. The Fbebzinq Proobsb. 

469. TanoiPIX The presence of water has alirays 1)6611 the 
great obat&cle in foundation work and in shaft sinking, and it 
is only very recently that any one thought of transforming the 
liquid soil into a solid wall of ice aboat the space to be excavated. 
The method of doing this consists in inclosing the site to be ex- 
caTated, by driving into the ground a number of tubes through 
which a freezing mixture ie made to circulate, Thrae consist of a 
large tube, closed at the lower end, inclosing a smaller one, open at 
the lover end. The freezing mixture is forced down tho inner 
tube, and rises through the outer one. At the top, these tubes 
connect with a reserroir, a refrigerating machine, and a pump. 
The freezing liquid is cooled by an ice-making machine, and theo. 
forced through the tubes ontil a wall of earth is frozen around: 
them of sufficient thickness to stand the external pressure, when 
the excaTation can proceed as in dry ground. 

470. HiSTOBT. This method was iavented by F. K. Poetscb, 
i/l.D,, of Aschersleben, Prussia, in 1883. It has been applied in. 
but three cases. The first was at the Archibald colliery, near 
Sohweidlingen, Prussia, where a Toia of quicksand, 20 feet thick, waff 
encountered at a depth of about 150 feet below the surface. Her& 
twenty-three pipes were used, and 3fi days consumed in the freezing 
process, under local difficulties. The second was at the Centrum 
mine, near Berlin, where about 107 feet of quicksand, etc., was 
penetrated. Engineers had been baffled for years in their attempts, 
to sink a shaft here ; bat in 33 days Mr. Poetsch had, with only 16 
freezingtubes, secured a 6-foot wall of ice uround the shaft area, and 
the shi^t was excavated and curbed without difficulty. The third 
piece of work was at the Eimilia mine, Fensterwalde, Austria, in. 
1885, where an Sj-foot shaft was sunk through 115 feet of quick- 
sand.* 

471. DSTAILB 07 TEE Pbocksb. In the last case mentioned 
above, " 12 circulatory tubes were used, sunk in a circle about 14. 
feet in diameter, from 13 to 15 days being required to sink them t^ 
depth of about 100 feet. The outside tubes were 8^ inches in 

* As this TolnmB IB going throDf^ the pre«8, this msthod ii belog applied In two 
plaoes la tlila ooiuaT^IicD Uonntaln, Hlch., and Wyoming, Penit.— In »tniHng 



ovGoQi^lc 



808 FODNDiiTIONB UKDEB WA.TEB. [CHA.F. XIL 

diameter, and made of plate iron 0.15 inch thick. The tabes were 
sunk by aid of the water-jet. They were giveo a very slight incli- 
nation outward at the bottom to avoid any deviation in Binking 
that might interfere with the line of the shaft. The freezing 
liquid employed was a solution of chloride of calcium, which cun- 
gc»lB at a temperature of —85" 0. ( —31° F.). The circulation of 
the liquid through the tubes was secured by a -small pump with 
a pistoa 6 inches in diameter and a 12-inch stroke. At the begin- 
ning of the operation, this pump made 30 double strokes per min- 
ute, which was equivalent to the passage of 0.6 gallon of the liquid 
through each tube per minute ; at the end of the operation, when 
it was only aeceBsary to maintain the low temperature, the pump 
strokes were reduced to 15 per minute. The refrigerating machiue 
employed was one of a model guaranteed by the maker to produce 
1,100 pounds of ice per hour. The motive-power was supplied by 
a small engine of about 5 horse-power. The ammoniac pump had 
a piston 2.8 inches in diameter and a d.3-inch stroke, and made 30 
strokes per minute. The pressure maintained was about 10 atmos- 
pheres. The quantity of ammoniacal liquid necessary to charge 
the apparatus was 381 gallons ; and under normal conditions the 
daily consumption of this liquid was 0.78 gallon. 

'* The actual shaft excavation was commenced 53 days after the 
freezing apparatus had been set in motion. The freezing machine 
was in operation 340 days. The work was done without difficulty, 
and a progress of 1.64 feet per day was made. The timbering was 
v^ry light, but no internal pressure of any kind was observed. The 
brick masonry used for finally lining the shaft was about 11 inches 
thick. When the shaft was finished, the tubes were withdrawn 
without diflScnIty, by circulating through them a hot, instead of a 
cold, solution of the chloride of calcium, thus thawing them loose 
from the surrounding ice. The tubes were entirely uninjured, and 
could be used (^^in in another similar operation. 

472. "The material in the above plant is estimated to have cost 
tl5,000, and 94,800 more for mounting and installation. The daily 
expense of conducting the freezing process is estimated at (11, llie 
total expense for putting down the shaft is estimated at 9138.66 per 
linear foot."* The last is equivalent to about 92.25 per cubic foot. 
). at, 2S, tnuulatod from Li Binlt Viva nt Juw U, 



ovGoQi^lc 



ABT. ti.J GOHFARIBON 07 HETHODS. 309 

473. Kodlfloation for Foundationi under Tater. For sinking 
foondationB under water, two methods of applying this process hare 
been proposed. One of these consists in combining the pnenmatio 
and freezing processes. A pneumatic caisson is to be snnk a short 
distance into the river-bed, and then the congealing tnbee are 
applied, and the entire mass between the caisson and the rock is 
frozen solid. When the freezing is completed, the caisson will be 
practically sealed against the entrance of water, and the air-lock can 
be removed and the masonry built up as in the open air. 

The other method consiBtB in sinking an open caisson to the 
river-bed, and putting the freezing tubes down through the water. 
When the congelation is completed, the water can be pumped out 
and the work conducted in the open air. 

471. AsvAKTAaBB Clahced. It is claimed for this process that 
it is expeditious and economical, and also that it is particularly 
valuable in that it makes possible an accurate estimate of the total 
cost before the work is commenced, — a condition of affairs unat- 
tainable by any other known method in equally difficult ground. It 
has an advantage over the pneumatic process in that it is not limited 
by depth. It can^ applied horizontally as vifell as vertically, and 
hence is specially nseful in sub-aqueous tunneling, particularly in 
soils which, with compressed air, are treacherous. 

475. SIFFICULTIEB AITTICIFATED. So far it has been used only 
in sinking shafts for mines. Two difficulties are anticipated in ap- 
plying it to sink foundations for bridge piers in river beds ; viz., 
(1) the difficulty in sinking the pipes, owing to striking sunken logs, 
bowlders, etc.; and (3) the possibility of encountering running 
water, which will thaw the ice-wall. These difficulties are not in- 
snrmonntable, but experience only can demonstrate how serious 
they are. 

476. Cost. See g 473, and compare with table on page 310l 

Art. 6. GouPARiaoN of Methods 

477. The fallowing comparison of the different methods is from 
an article by Jules Qaudard on Foundations, as translated by L. F. 
Vernon- Harcourt for the proceedings of the Institute of Civil En- 
gineers (London). Except as showing approximate relative costs, in 
Europe, it is not of much value, owing to improvement made since 
the article was written, to the differences between European and 



ovGoQi^lc 



310 



TOUVDATIOSS UKDKB 



[chap. xn. 



AmericBn practice, *Bd to differences in cort of nwteittk in the two 

«onntrie& 

478. " 1L Cntizette Desnoycn has framed a cbMification of the 
metkods of foondationa most soitable for different depths, and also 
an estimate of the cost of each. Theae estiioBtes, howerer, most be 
considered merely approximate, as nnforeseen circumstances pro- 
dace conndeisble Tariations in works of this nature. 




"When the fonndations consist of disconnected pillars or piles, 
the above prices most be applied to the whole cubic content, includ- 
ing the interrals between the parts ; but of course at an eqoal cost 
■olid piers are tbe best. 

479. " For pile-work foundations the square yard of base is prob- 
ably a better unit than the cubic yard. Thus the foundationa of 
the Vernon bridge, with piles from 24 to 31 feet long, and with 
cross-timbering, concrete, and caisson, cost t70 per square yard of 
base. According to estimates made by HIL Picquenot, if the fonn- 
dations had been pnt in by means of oompreseed sir, the cost would 
have been (159.64 ; with a caisson, not water-tight, sunk down, 
(66.37 ; with concrete ponied into a space inclosed with sheeting, 
(62,33 ; and by pnmping, (83.56 per square yard of baae." 



ovGoQi^lc 



PART IV, 

MASONRY STRTJCTUREa 



CHAPTEB Xia 

HASONRT DAU8. 

480. It is not the intention here to disctisH erery featare of 
masonry dams ; that has been done in the special reports and arti- 
cles referred to in g 520, page 334. The fundamental principles 
Till be considered, particularly vith reference to their applica- 
tion in the subsequent stady of retaining walls, bridge abutments, 
bridge piers, and arches. The discnasioDs of this chapter are 
applicable to masonry dams, reservoir walls, or to any wall wbioh 
counteracts the pressure of water mainly by its weight. 

There are two ways in which a masonry dam may resist the 
thrust of the water ; viz., (1) by the inertia of its masonry, and 
(ti) as an arch. 1. The horizontal thrust of the water may be held 
in equilibrium by the resistance of the masonry to sliding forward 
or to overturning. A dam which acte in this way is called a gravity 
dam, Z. The thnui of the water may be resisted t^ being trans- 
mitted laterally to the side-hilU (abutments) by the arch-like action 
of the masonry. A dam which acts in this way is called an arched 
dam. 

Only two dams of the pure arch type have ever been built. The 
almost exclusive use of the gravity type is due to the uncertainty 
of onr knowledge concerning the laws governing the stability of 
masonry arches. This chapter will be devoted mainly to gravity 
dams, those of the arch type being considered only incidentally. 
Arches will be diaouised fully in Chapter XVUL 

ni 



ovGoQi^lc 



MA80NET DAlfS. [CHAP. XIU. 



AST, 1. Stability of Gaatitt Daub. 

481. FsnoiTLXB. By the prmciplee of hydroetatics we know 
(1) that the preesure of a liquid upon any aurtaco is eqoal to the 
weight of a Tolume of the liquid whose base is the urea of the im- 
mersed surface and whose height is the vertical distance of the center 
of gravity of that surface below the upper surfftc© of the water ; (2) 
that this pressure is always perpendicular to the pressed surface ; 
and (3) that, for rectangular surfaces, this pressure may be con- 
sidered as & aiugle force applied at a distance below the upper 
surface of the liquid equal to f of the depth. 

482. A gravity dam may fail (I) by sliding along a horizontal 
joint, or (3) by overturning about the front of a horizontal joint, 
or (3) by crushing the masonry, particularly at the front of 
any horizontal joint. However, it is admitted that by far the 
greater number of failures of dams is due to defects in the founda- 
tion. The method of securing a firm fonndation has already been 
discussed in Part III ; and, hence, this subject will be considered 
here only incidentally. There is not much probability that a dam 
will fail by sliding forward, but it may fail by overturning or by the 
crnehing of the masonry. These three methods of failure will be 
considered separately and in the above order. 

483. In the discussions of this article it will be necessary to 
consider only a section of the whII included between two vertical 
planes — a unit distance apart — perpendicular to the face of the 
wall, and then so arrange this section that it will resist the loads and 
pressure put upon it ; that is, it is sufficient, and more convenient, 
to consider the dam as only a unit, say 1 foot, long. 

484. HO][Sircu.TVBE. The following nomenclature will be used 
throughout this chapter : 

ff = the horizontal pressure, in pounds, of the water against a 
section of the back of the wall 1 foot long and of a height 
equal to the height of the wall. 
TTa: the weight, in pounds, of a section of the wall 1 foot long. 
w = the- weight, in pounds, of a cubic foot of the masonry. 
A = the height, in feet, of the wall ; t. e., h = EF, Pig. 66. ' 
/ = the length of the base of the cross section ; i. e., I ^ A it- 
Fig. 68. 
t = the vridth (rf the wall on top ; i. e.,t=I)E, Fig. 68. 



ovGoQi^lc 



AKT. 1.] 8TA.BIIITT OF QBATITT DAKB. SI'S 

b = the batter of the wall, i, e., the inclinatioQ of the sar&ce 
per foot of rise — b' being nsed for the batter of the op- 
stream face and b, for that of the down-stream face. 
K = AC= the diatance from the dowD-stream face of any joint to 
the point in which a Teitical through the center of gran^ 
of the wall pierces the plane of the base, 
i = the distance the center of pressure deviates from the center 
of the base. 
62.K = the' weight, in pounds, of a cubic foot of water. 

48fi. Stasilitt aoaivbt Busnre. The horizontal pressure of 
the water tends to slide the dam forward, and is resisted by the 
friction due to the weight of the wall. 

486. Sliding Foree. The horizontal pressure of the water 
against an elementary section of the wall, by principle (1) of § 481, 
is equal to the area of the section multiplied by Imlf the height of 
the wall, and that product by the weight of a cubic nnit of water; or 



ir=kx ixihxe 



= 31.25 h'. 



• (1) 



Notice that H is the same whether the pressed area is inclined or 
vertical ; that is to say, H is the horizontal component of the total 
pressure on the surface. 

187. Betisting Fonwi. The weight of an elementary section of 
the wall is equal to the area of the vertical ^ 

cross section multiplied by the weight of s 
cubic nnit of the masonry. The area of 
tha cross section, ABED, Fig. 68, equals 

SFxJ)£+iBFxFS+iDOxA0 

= ht+ih'b'+ih'b, ... (2) 

Then the weight of the elementary sec- 
tion of the wall is 

W=v,(ht + ib'b' + ik'b,) . (3) ^^-jj- 

The vertical pressure of the water on »io. en. 

the inclined face increases the pressure on the foundation, and, 
cousequeutly, adds to the resistance against slidiDg, The vertical 
pressure on EB is equal to the horizontal projectioQ of that area 
multiplied by the distance of the center of gravity of the surface 
below the top of the water and by the weight of a oubic unit of 




ovGoQi^lc 



SU ItlSOKKT DA.HB. [cHAP. XIII. 

water, or, the Tertical presBnre = FB X 1 X i A X 62.5 = A i' x 
\h X 62.8 = 31.25 A" A'. 

488. If the eartlt reete against the heel of the dam (the bot- 
tom of the inside face), it will iacrease the preaanre on the fonn> 
dation, dace earth weighs more than water ; on the other hand, the 
horizontal pressnre of the earth will be a little greater than that of 
an eqnal height of water. Howerer, since the net resistaDce with 
the earth upon the heel of the wall is greater than with an equal 
depth of water, it will be asenmed that the water extends to the 
bottom of the walL 

If the water finds ita way under and around the foundation tA 
the wall, even in very thin sheets, it will decrease the preaaure of 
the wall on the foundation, and, consequently, decrease the 
stability of the wall. The effective weight of the submerged por- 
tion of the wall will be decreased 62^ lbs. per ca. ft However, the 
assumption that water in hydrostatic oonditiou finds its way under 
or into a daxn is hardly admissible ; hence the effect of baoyanoy 
will not be considered.* 

489. Co-effl«]ent of Triotion. The Talnes of the co-efficient of 
friction most frequently required in masonry computations are given 
in the table on p^e 315. There will be frequent reference to this 
table in subsequent chapters ; and therefore it is made more full 
than is required in this connection. The raluee have been collected 
from the best authorities, and are believed to be &ir averages. See 
also the table on page 276. 

490. Condition for Equilibrium. In order that the wall may 
not elide, it is necessary that the product found by multiplying the 
co-efficient of friction by the sum of the weight of the wall and the 
vertical pressure (rf the water shall be greater tluui the horisontal 
pressure of the water. That is to say, in order that the dam may 
not slide it is necessary that /i (TF-f 31.25 A*d') shall be greater 
than if; or, in mathematical language, 

S 81.25 A' 

f*> w + 31.25 A' 4' -^ w (A f -H t A' A' -f- 1 A' ft,) -|- 31.26 A' V ' 

■ SliiM the above wh written, Jas. B. Francla prewDtsd a paper (Mar IB, 18B9 
before tbe American Socielr ot CItII Enftlneen, wblcb Metas to ahow that wata 
premire la oommiuilaated, almoet nndlTnlnlahed, through a larer of PortJand osotnt 
mortar (I part cement and 3 parta eand) 1 toot tblok. 



ovGoQi^lc 



BTABIIITT Of OBATITT DAHS. 



TABLE W. 
Co-amctEHTB or Fsiction tob Dbt HAaasRi. 



Soft liamtone on soft llmMtone, both well dreased 

Brick-work OD brick-work, with illghtlr damp morter 

Hard brick-workoD hard brick-work, with allgbtlj damp morUur. 

PoiDt-dresaed gnulte on like naaite. 

" " " " well-arened graolle 

Common brick on common tjrick 

" " " hard llmealniie 

Hard llmmtoae on bard llmesloDe, with moist mortar 

BeloQ bkxika (preMed) on Ifke beton blocki 

Fiiiti-cut grnnite on pressed " ■' 

Well (Irewed ffDinlte on well drasned gninite 

Polished limestone on polished limestone 

Well-dresxed granite on like prnntte. with rresh mortat 

Common brick on common bnck, with wet mortar 

Polished marble on common brick 

PolDt-drened granite on Krarel 

'■ ■• " " dry clay. 

" "sand 

'■ " " " moist claj. 

Wnwght IroD on well-dreaaed Umentone 

" •• •' haid, well dressed limeatoDe, wet 

Oak, flatwise, on limestone 

" endwise, on limestone 



0.75 
0.7S 
0.70 . 
0.70 
0.60 

o.as 

0.6S 
0.«S 
O.flS 
0.60 

o.eo 
o.«o 
o.so 
o.so 

0.4S 



0.S5 

o.os 

0.40 



which rednced becomes 



o*.o " * .\ 

'' -* v,{2t + h{b' + A,)) + 62.5 A b'' • ' ■ W 

The weight of a cubic foot of masonry, w, Tariee between 135 lbs. 
for concrete or poor brick-work, and 160 Ibe, for granite ashlar. 
Dams are asnally boilt of mbble, whrch weighs about 150 lbs, per 
cu. ft. To simplify the formula, we will assume that the masonry 
weighs 125 lbs, per oo. ft.; i. e,, that the weight of a cubic foot of 
masonry is twice that of water. This assumption is on the safe side, 
whatever the kind of masonry.* Making this snbstitntioD in (4) 



> Imiiiaiiiiil satetr genenllf reqniree Increaaed cost of cosBtrncitlon, and benoe 
It Is not permlaslblti to nse approximate data BlmpI; beoanae the error la on the side 
toward Mfet}. It will be shown that there is no probabllltr of any dam'i falling by 
dldlnc and that the Mae, and ooDaeqnenU; the Toloine and ooat, are detennlnod tij- 



ovGoQi^lc 



H^BONBT DAXB. [CHAP. XIII, 



Other thiDga being the same, the tbinner the vail at the top, 
the euier it will slide. If the section of the wall is a triangle, t. «., 
if ^ = 0, then b; equation (5) ve see that the dam ia safe against 
sliding Then 

"^"((vV^) '°' 

An examination of the table on page 315 shows that there is no 
probability that the co-efficient of friction will be less than 0.5 ; and 
inserting this TcJne of m in (6) shows that sliding can not take 
place if {J b' + ft,) > or = 1. To prevent oTertnming, (ft' + S,) 
is nsuall; = or > 1 (see Fig. 72, page 338) ; and, besides, a con- 
siderable thickness at the top (see § 509) is needed to resist the 
shook of waves, etc. Hence there ia no probability of the dam's 
failing by sliding forward. Farther, the co-efficient of friction in 
the table on page 315 takes no acoonnt of the cohesion of the mor- 
tar, which may have a possible maximnm value, for best Portland 
mortar, of 36 tons per sq. ft. (500 Iba. per sq. in.); and this gives 
still greater security. Aj^in, the earth on, and also in front of, the 
toe of the wall adds greatly to the resistance against sliding. Fi- 
nally, it is customary to buUd masonry dams of uncoui-sed rubble 
(gg 213-17), to prevent the bed- joints from becoming channels for 
the leakage of water; and hence the stones are thoronghly inter* 
locked, — which adds still further resistance. Therefore it is certain 
that there is no danger of any masonry dam's failing by sliding for- 
ward under the pressure of still water. 

491. It has occasionally happened that dams and retaining walls 
have been moved bodily forward, sliding on their base; but such an 
occurrence is certainly unusual, and is probably the result of the 
wall's having been founded on an unstable material, perhaps on an 
inclined bed of moist and nacertain soil. In moet that was said is 
Part UI oosoeming tonndatioQS, it was amnmed that the foand» 



ovGoQi^lc 



ART. 1.] , BTAfllLITT OP GRAVITY DAMS. 317 

tion iras required to support only & vertical load. When the etraci- 
ore is enbjected also to a lateral pressure, as in dams, additional 
means of security are demanded to prevent lateral yielding. 

When the foundation reeta upon piles a simple expedient is to 
drive pUee in front of and against the edge of the bed of the founda- 
tion; but obvioUBly this is not of much value except when the piles 
reach a firmer soil than that on which the foundation dire^ly reErts. 
If the piles reach a firm subsoil] it will help matters a little if the 
upper and more yielding soil is removed from around the top of the 
pile, and the place filled with broken atone, etc Or a wall of piles 
may be driven around the foundation at some distance from it, and 
timber braces or horizontal bnttreBses of masonry may be placed at 
intervals from the foundation to the piles. A low masonry wall is 
sometimes used, instead of the wall of piles, and connected with the 
foot of the main wall by horizonlal buttresses, whose feet, on the 
counter-wall, are connected by arches in a horizontal plane in order 
to distribute the pressure more evenly. 

In founding a dam upon bed-rock, the resistance to sliding on 
the foundation may be greatly increased by leaving the bed rough ; 
and, in case the rock quarries out with smooth surfaces, one or more 
longitudinal trenchesmaybeeicavated in the bed of the foundation, 
and afterwards be filled with the masonry. 

In the proposed Quaker Bridge dam the maximum horizontal 
thrust of the water is equal to 0.597 of the weight of the masonry. 

492. Stasiutt AOAnrsT OvEBTiruiniQ. The horizontal pres 
sure of the water tends to tip the wall forward about the front o. 
any joint, and is i-eaigted by the moment of the weight of the wall. 
For the present, it will be assumed that the wall resta upon a rigid 
base, and therefore can fail only by overturning as a whole. 

The conditions necessary for stability against overtuming can be 
completely determined either by considering the moments of the 
several forces, or by the principle of resolution of forces. In the 
following discussion the conditions will be first determined by mo- 
ments, and afterward by resolution of forces. 

463. A. Br Vokemte. The Orerturning Xomsnt The pressure 
of the water is perpendicular to the pressed snr&ce. If the water 
presses against an inclined face, then the pressure makes the same 
angle with the horizontal that the surface does with the verticaL 
Since there is a little ditKcnlty in dnding the aim of this force, it is 



ovGoQi^lc 



S18 - IU80NBT DAHS. [cHAP. ZUL 

more convenient to deal with the horizontal and vertical componenti 
of the presanre. 

The horizontal preaanre of the water can be foand b; equation 
(1), page 313. The arm of this force is eqnal to '^ A (principle 3, 
g 481). Hence the moment tending to overturn the wall is equal to 

iHk = i 31.25 V = 10.43 A', . . . . (7) 

vhicb, for convenience, represent by M, . 

494, The Heiisting Xomenti. The forces resisting the over- 
taming are (1) the weight of the wall and (2) the vertical pressure 
of the water on the inclined face. 

The weight of the wall can be computed by equation (3), page 
313, It acts vertically through the center of gravity of the cross 
section. 

The center of gravity can be fonnd algebraically or graphically. 
There are several ways in each case, but 
the following graphical solution is the sim- 
plest. In Fig. 69, draw the diagonals D B 
and A E, and lay off A J = E I; then 
draw D J, and mark the middle of it Q. 
The center of gravity, 0, of the area 
A B E D \^a^& distance from Q towards 
B equal \a\Q B. This method ia appli- 
cable to any four-sided figure. 

The position of the center of gravity can 
Q F also be fonnd algebraically by the principle 

Fio. a*. that the moment of the entire mass about 

Any point, as A, is eqnal to the moment of the part ADO, plus 
^he moment of the portion D E F 0, pins the moment of the part 
E B F,— all about the same point, A. Stating this principle alge- 
braically gives 

t A 6, (4 A* S.) + A / (t ( -I- 4 J,) + 1 A' J* (J A 6' + f + ft 5,) 

= aA*&' + Af + JA'5.)i. . ... (8) 

in which x = the distance A C. Solving (8) givee 

'= 14(S' + »,) + i • <"' 



jvGooi^le 



ABT. 1.] BTABILIIT OP GBATITT DAMB. 319 

The arm of the weight a A C {= x), and therefore the mo- 
ment is 

WxAC=u[ht + ih'{i' + b,)}x.. . . (10) 

which, for conTenieace, represent by JT, . 

496. The vertical pressure of the water on the inclined face, 
JS B, has been computed in g 487, which eee. This force acts ver- 
tically between F and B, at a distance from B equal \a \ F B; the 
armof thi8forcoi8-a£-4^5 = / — iAJ'= hb,-\- 1 -\-\hb'. 
Therefore, the moment of the vertical pressare on the inclined 
face is 

81J35 A* S' (i A, -I- * + I A J'), .... (11) 

which, for convenience, represent by M, . Of conree, if the presaed 
face is vertical, M^ will be eqoal to zero. 

496. The moment to resist overturning is equal to the sum of 
(10) and (11) above, or Jl^, + JT, . 

The moment represented by the sum of Jtf, and M^ can be deter- 
mined directly by considering the pressure of the water as acting 
perpendicular ia B Ba,t\ E B from B; the arm of this force is a 
line from A perpendicular to the line of action of the pressure. If 
the cross section were known, it would be an easy matter to measure 
this arm on a diagram; but, in designing a dam, it la necessary to 
know the conditions requisite for stability before the cross section 
can be determined, and hence the above method of solution is the 
better. 

497. Condition for Equilibrium. In order that the wall may 
not turn about the front edge of a joint, it is neceasary that the 
overtaming moment, JV, , as found by equation (7), shall be less 
than the snm of the resisting moments, M, and M^ , as found by ■ 
equations (10) and (II); or, in other words, the factor against over- 
^ . M.+ M, 

turnmg = jy (13) 

498. Factor of Saf^ againft Orertuming. In computing the 
stability against overturning, the vertical pressure of the water 
against the inside face is frequently neglected; t.e., itls assumed 
that M,, u above, is zero. This assumption is always on the safe 
side. Computed in this way, the factor of safety against overturn- 
ing for the proposed Quaker Bridge dam, which when completed 



ovGoQi^lc 



330 HASONBY DAKS [OHAP. Zm. 

will be considerably the largest dam in the world, Tariee between 
3,07 and 3.68. Krantz,* who included the vertical component in 
his compatatione, considera a factor of 2.6 to 5.65 aa safe, the larger 
value being for the largest dam, owing to the more seriona conse- 
qnencea of failure. The greater the factor of safety provided for, 
the greater is the first cost; and the less the &ctor of safety, the 
greater the expense of maintenance, including a possible reconstruc- 
tion of the structure. 

489. A Bt ItSaOI.VTlOH Of TOBOBS. In Fig. 70, f is the center 
of preseore of the water on the back of 
the wall. K B =^\ E B. o is the center 
of gravity of the wall, — found as already 
described. Through K draw a line, K a, 
perpendicular ta E S; through o draw a 
vertical line o a. To any convenient 
scale lay off ab eqnal to the total pressure 
of the water against E B, and to the 
same scale make af equal to the weight 
of an elementary section of the wall. 
Complete the parallelogram a b ef. The 
'■"■ ™ diagonal ae intersects the base of the 

wall at N. 

500. On the assumption that the masonry and foundation 
are absolutely incompressible (the compressibility will be considered 
presently), it is clear that the wall will not overturn as long as the 
resultant ae intersects the base AB between A and B. The factor 

against overturning is -jj-pj, which b the equivalent of equation (13), 

The wall can not slide horizontally on the base, when the angle 
NaC is lees than the angle of repose, t. e., when tan NaOit less 
than the co-efficient of friction. The factor against sliding is equal 
to the co-efBcieut of friction divided by Ian XaC, which is only 
another way of stating the conclusion drawn from equation (4), 
page 315. 

501. Btabuitt AOAnrsT CBirBHnrei. The preceding diecnssion 
of the stability against overturning is on the assumption that the 
masonry does not crush. This method of failure will now be oon- 




" Stodr of Bnorrolr Walla," Uahao'i tmulMloii, p. 68. 



ovGoQi^lc 




ABT. 1.J I fflABIUTT OF QBATITT DAMS. 331 

Bidered. WheD the reserroir is empty, the preeBiire tending to 
produce cruBhing is the weight of the dam alone, which pressure is 
dietribnted uniformly over the horizontal area of the wkII. When 
the reservoir is full, the thruat of the water modifies the distribntioa 
of the pressare, increasing the pressure at the front of the wall and 
decreasing it at the back. We will now determine the law of the 
Tariation of the pressure. 

Let A B, Fig. 7L, represent the base of a vertical section of the 
dam ; or A S may represent the rect- 
aognlar base (whose width is a unit) of 
any two bodies which are pressed against <t ^^-ir>H?j-j->;-yi 
each other by any forces whatever. ^* ■ ' ^^ 

M — the resulting moment (about A'^ of 

alt the external forces. In the 

caae of a dam, M= M, — M„— 

equations (7) and (11). nan. 

W = the total normal pressure on A B. 

In the case of a dam, W = the weight of the masomy. 
P = the maximum pressure, per unit of area, at A. 
p = the change in unit pressure, per anlt of distance, from A 

towards B. 
X = any distance from A towards B. 
P ~p X = the pressure per unit at a distance x from A. 
Y ~ti general expression for a vertical force. 

The remainder of the nomenclature is as in g 484, page 812. 
Taking moments about A gives 

M- Wx + ^(P-px)dx.x = (ii. . . (13) 

M-Wx + iPt'-ipr = 0. (14) 

For eqnilibrinm, the sum of the forces normal io A B most also 
be equal to aero ; or 

sr = -w+(*{P-px)dx=so, . . . (18) 

from which 

pr^iPl-iW. (16) 



ovGoQi^lc 



iZi VASONBT DAXB. [CQAP. XnL 

602. IfwImiiBi FxMsnrt. Combining (16) iritli (U) and i« 
daoing, 

y=*-f _l^+.5^. (1,) 

[f the stability agaioBt overtaming be determined algebmicall;, t, e., 
by equation (12), then J/'and x are known, and P can be compnted 
by eqnatioD (17). 

If the wall is Bymmetrical x = il, and (17) becomes 



^ = -T+-r (") 

Equation (18) is the same as equation (1), page 209, except that 
the latter is applicable to any form of horizontal section, vhile the 
former is applicable to only a reotangolar cross section. 

In equation (16), notice that -j is the noiform presenre onAB 
due to the .weight of the wall ; also that —jj- is the increase of pres- 
sure at A due to the tendency to overturn, and that consequently 
the uniform preBSure at £ Is decreased a like amount. 

503. The maximum pressure may be found also in another way. 
Assume that N, Fig. 71, is the center of pressore. Let p, (= B L) 
represent the pressure at B, and p,(= J £') that at A ; and any 
intermediate ordinate of the trapezoid .4 ^£f will represent the 
pressure at the corresponding point. Then, since the forces acting 
on ^ .5 must be in equilibrium for translation, the area of the 
trapezoid will represent the entire pressure on the base A B. Stated 
algebraically, this is 

f^l=W. (19) 

Also, since the forces acting qti AB must be in equilibrium for 
rotation, the moment of the pressure to the right of N must be 
equal to that to the left ; that is to say, the center of gravity of -the 
trapezoid ABLK must lie in the line NJ. By the principles of 
analytioal meohauioi, the ordinate AN \a the center of gravi^ 
ABLK'ta 



jvGooi^le 



ABT. 1.] BTABIUTT OF QBATUT DAMB, SS8 

SolTlng (19) and (20) gives 

P» = —i f-- (21) 

If the wall ia a right-angled triangle with the right angle at A, 
x = il, which, Babatittited in the above expression, shows that the 

preesare at A is —t-, »nd al") that the preasare a.t B is zero, — all 

of which is as it should be. Equation (21) is a perfectly general 
expression for thepressure betmten any two plane turfacet pressed 
together by norvial forces. Notice that equation (31) ia identical 
with the first two terms of the right-hand side of equation (17). 

The form of (31) can be changed by Bnbetitnting for x its valne 
il-d; then 

A = ^+^. m 

Equation (S2) gives the pressure at A due to the weight of the 
wall ; bnt it will also give the maximum pressure on the base due 
to both tho vertical and the horizontal forces, provided d be taken 
as the distance from the middle of the base to the poiut in which 
the resultant of all the forces cuts the base. Therefore we may 
write 

p=-5^+^. m 

&04. Equation (33) is the equivalent of equation (17), page 333. 
It is well to notice that equation (33) is limited to rectangular kori- 
zontttl cross-sections, since it was assumed that the pressure on the 
section varies as the distance back from the toe. If the stability 
against overturning ia determined algebraically, as by equation (12), 
then eqnation (17) is the more convenient ; but if the stability is 
determined graphically, as in Fig. 70, then equation (33) is the 

simpler. Notice that it d = ^l, P = —j—, which is in accordance 

with what is known in the theory of arches as the principle of the 
middle third ; that is, as long as the center of pressure lies within 
the middle third of the joint, the maximum pressure is not more 
than twice the mean, and there is no tension in any part of the 
joint. 



ovGoQi^lc 



JUSONET DAMS. [CHAP. SIII. 



ity of the load. It ie immaterial whether the deviation d is cansed 
by the form of the wall or by forces tending to prodnce overtnm- 
ing. 

505. Tension on the Xaaonry. By an analyeia similar to that 
above, it can be shown that the decrease in pressure at B, dne to 
the overturning moment, is equal to the increase at ^. ltd ^ \ I, 
then by equation (23) the increase at A and decrease at B is W, 
that is to say, the pressure at ^ is 2 ^ and that at B is zero. 
Thereforej if the center of pressure departs more than 4 1 from the 
center of the base, there will be a minus pressure, i. e. tension, at 
B. Under this condition, the triangle A YK', in Fig. 71, page 
321, represents the total presetire, and the triangle £ TX' the total 
tension on the masonry, — A IC being the maximum pressure at A, 
and BL' the maximum tension at B. 

If a good quality of cement mortar is used, it is not unreason' 
able to count upon a little resistance from tension. As a general 
rule, it is more economical to increase the quantity of stone than the 
quality of the mortar ; but in dams it is necessary to nse a good 
mortar to prevent (I) leakage, (2) disintegration on the water side, 
and (3) crushing. If the resistance due to tendon is not included 
in the computation, it is an increment to the computed margin of 
safety. 

506. If the masonry he considered as incapable of resisting by 
tension, then when d in equation (23) exceeds \l the total pres- 
sure will be borne on A Y, Fig. 71. In this case A W (the distance 
from A to the point where the resultant pierces the base) will be 
less than \l. It A K" represents the maximum pressure F, then 
the area of the triangle A YK" will represent the total weight W. 
The area at A YK" = \AK"xAY=\PyZAN', Hence 
\Px^AN' = W,OT 

P-AJL 2ff_ 

^~3^JV'-3(ii-d) ^""^ 

To illastrate the difierence between equations (23) and (24), 



ovGoQi^lc 



ABT. 1.] eXABILITT OF ORA.TITT DAHS. oSS 

assume that the distacce from the resultant to the center of the ban 
is one quarter of the length of the base, i. e., assume that d = \l. 
Then, by equation (23), the maximum pressure at A is 

P = ^+2^ = 2i^, M 

and by equation (24) it is 

^ = 3Tfr?iO = ^T- M 

That is to say, it the masonry is capable of resisting tension, equa- 
tion (25) shows that the maximam pressure isS^ times the pressnrg 
due to the weight alone-; and if the masonry is incapable of resist- 
ing tension, equation (26) shows that the maximum pressure is 3j 
times the pressure due to the weight alone. 

Notice that equation (24) is not applicable when d is lesa than 
^l; in that case, eqoation (23) must be used. 

607. Limiting Pressure, As a preliminary to the actual design- 
ing of the section, it is necessary to fix upon the maximum pressure 
per square foot to which it is proposed to subject the masonry. Of 
course, the allowable pressure depends upon the quality of the 
masonry, and also upon the conditioDs assumed in making the com- 
putations. It appears to be the custom, in practical computations, 
to neglect the vertical pressure on the inside face of the dam, i. e,, 
to assume that M, , equation (II), page 319, is zero ; this assumption 
ie always on the safe side, and makes the maximum pressare on the 
outside toe appear greater than it really is. Computed fn this way, 
the maximum pressure on rubble masonry in cement mortar in 
some of the great dams of the world is from 11 to 14 tons per sq. 
ft. The proposed Qnaker Bridge dam is designed for a maximum 
pressure of 16.6 tons per sq. ft on massiTO rubble in Portland 
cement mortar. 

For data on the strength of stone and brick masoniy, see gg 
221-23 and §§ 246-48, respectively. 

608. The actual pressure at the toe will probably be less than 
that computed as above. It was assumed that the weight of the 
wall was uniformly distributed over the base ; but if the batter ia 
considerable, it is probable that the pressure due to the weight of 
the wall will not vary uniformly from one side of the base to the 



ovGoQi^lc 



826 VASONST DAHS. [cHAP. XIIX. 

other, bnt will be greater on the central portions. The actual 
mazitnnm will, therefore, probably occnr at eome distance back 
from the toe. Neither the actoal maximom nor the point at which 
it oconra can be determined. 

Frofeaeor Rankine claims that the limiting preasnre tor the out- 
side toe shonld be leas than for the inside toe. Notice that the 
preceding method determines the maximum vertical presents 
When the maximum pressure on the inside toe occurs, the only 
force acting is the vertical pressure ; bnt when the maximum on 
the outside occurs, the thrust of the water also is acting, and there- 
lore the actual pressure ie the resultant of the two. With the pres- 
ent state of our knowledge, we can not determine the effect of a 
horizontal component upon the vertical resistance of a block of stone, 
bnt it must weaken it somewhat 

Abt. %. Odtlines or the Design. 

SOS. Width oh Top. As far as the forces already considered 
are concerned, the widtlj of the wall at the top might be nothing, 
since at this point there is neither a pressure of water nor any 
weight of DiaBOnry. But in practice we mast consider the shock of 
waves and ice, which in certain cases may acquire great force and 
prove very destructive to the upper portion of the dam. This force 
can not be computed, and hence the width on top mnst be assumed. 
This width depends to a certain extent upon the height and length 
of the dam. The top of large dauia may be used as a roadway. 
Krantz * says that it is " scarcely possible to reduce the top width 
below 3 metres (6.5 ft.) for small pouds, nor necessary to make it 
more than 6 metres (16.4 ft.) for the largest." 

Fig. 73, page .328, gives the width on top of Krantz's profile type, 
and also of the. profile recommended by the engineers of the 
Aqueduct Commission for the proposed Quaker Bridge dam. 

510. The FBOTILE. In designing the vertical cross section of a 
gravity dam to resist still water, it is necessary to fulfill three con- 
ditions: (1) To prevent sliding forward, equation (4), page 315, 
must be satisfied; (3) to resist overturning, equation (li), page H19, 
must he satisfied ; and (3) to resist crushing, equation (23), page 
323, or (24), page 334, must be satisfied. As these equations really 

• " Stndf of RewTToIr WkIIb," Hahu'B tnnalaUoii, p. 36. 



ovGoQi^lc 



ABT. %.] 0TTTLINE6 OF THE DESIQK. 837 

lUTolTe only three Tariablee, viz.: hi in and b', — the height of the 
dam and the batter ot the two faces, — they can be satisfied exactly. 
It has been shovn that there is no danger ot the dam'e eliding for- 
ward even it the width on top is zero ; and hence there are practi- 
■ cally but two conditions to be fulfilled and two variables to be 
; determined. To prevent overtnrning when the reservoir ia fnll, 
equation (12) mast be satisfied ; and to prevent crashing, equation 
(23)— or (34)— must be satisfied for the points (Figs. 69, 70, etc.) 
^!ien the reaervoir is full, and for B when the reservoir is empty. 

Although it is possible to satisfy these conditions exactly, the 
theoretical profile can be obtained only by successive approxima- 
tions. This is done by dividing the profile into elementary hori- 
zontal layers, beginning at the top, and determining the dimension 
of the base of each layer separately. The theoretical width at the 
top being zero and the actual width being considerable, a portion of 
the section at the top of the dam will be rectangular. A layer being 
given, and the profile of the portion above it being known, certain 
dimensions are assumed for the lower base of the layer ; and the 
stability against overturning is then determined by applying equa- 
tion (13), or by the method of Fig. 70 (page 330). ^Tfae maximum 
pressure at A is then found by applying equation (17) or (;'3), after 
which the maximum pressure at B when the reservoir is empty 
must be deteimined by applying equation (23). If the first dimen- 
sions do .not give results in accordance with the limiting conditions, 
others must be assumed and the compntations repeated. A third 
approximation will probably rarely be needed. 

It is not necessary to attempt to satisfy these equations precisely, 
since there are a number of unknown and unknowable factors; as the 
weight of the stone, the quality of the mortar, the character 
of the foundation, the quality of the masonry, the hydrostatic 
pressure under the mass, the amount of elastic yielding, the 
force of the waves and of the ice, etc, which have more to do 
with the ultimate stability of a dam than the mathematically exact 
profile. It is therefore sufficient to assume a trial profile, being 
guided in this by the matters referred to in g 511 and § 513, and 
test it at a few points by applying the preceding equations ; a few 
modifications to more nearly satisfy the mathematical conditions cr 
to simplify the profile is aa far as it ia wise to carry the theoretical 
determination of the profile. 



ovGoo^^lc 



3S8 UASONBX DAHB. [CHAP. XIII. 

511. Erantz'a Stndy of Eeeervoir Walls, translated from the 
French by Gapt. F. A. Mahan, U. S. A., gives the theoretical pro- 
files for dams from 16.40 ft (5 metrea) to 164 ft (50 metres) high. 
The faces are arcs of circles. The mathematical work of determin- 
ing the profiles is not given ; but it ia evident that the polygonal 
profile was deduced as above described, and that an arc of a cirele 
vras then drawn to average the irregularitiee. The largest of these 
profiles is shown in Fig. 73 by the broken line. The others are 
simply the npper portion of the largest, with the tbickness and the 
height of the portion above the water decreased somewhat and the 
radios of the faces modified correspondingly. 




The lu-ger profile of Fig. 73 is that recommended by the eof^- 
neers of the Aqnednct Commission for the proposed Quaker Bridge 
dam. The profiles of most of the high masonry dams of the world 



ovGoQi^lc 



ABT. 2.] OUTLIITBB OF THE DEBIQIT. 33d 

are exceedingly extraTngant, and hence it is not wortli while to give 
examples. 

612. Prol. Wm. Cain haa shown * that the eqaatione of condi- 
tion are nearly satisfied by a cross section composed of two tra- 
pezoids, the lower and larger of which is the lower part of a triangle 
having its base on the fonndation of the dam and its apex at the 
surface of the water, and the upper trapezoid having for its top the 
predetermined width of the dam on top (g 509), and for its sides 
nearly vertical lines which intersect the sides of the lower trapezoid. 
The width of the dam at the bottom is obtained by applying the 
equations of condition as above. The relative batter of the np- 
stream and down-stream faces depends upon the relative factors 
of safety for crashing and overturning. This section gives a 
factor of Btrfety which increases from bottom to top, — an important 
feature. 

613. Tbs Puv. If the wall is to bs one side of a rectangular 
reservoir, all the vertical sections will be alike ; and therefore the 
heel, the toe, and the crest will all be straight. If the wall is to be 
a dam across a narrow valley, the height of the masonry, and cons» 
qnently its thickness at the bottom, will be greater at the center 
than at the sides. Id this case the several vertical cross sections 
may be placed so that (1) the crest will be straight, or (2) so that 
the heel will be straight in plan, or (3) so that the toe will be 
straight in plan. Since the up-stream face of the theoretical pro- 
file is nearly vertical (see Fig. 72), there will be very little difference 
in the form of the dam whether the several cross sections are 
placed in the first or the second position as above. If the creat ig 
straight, the heel, in plan, will be nearly so ; if the crest is straight, 
the toe, in plan, will be the arc of a circle such that the middle 
ordinate to a chord equal to the span (length of the crest) will be 
equal to the maximuni thickness of the dam ; and if the toe ia 
made straight, the crest will become a circle of the same radius. 
This shows that strictly speaking it is Impossible to have a straight 
gravity dam across a valley, since either the crest or toe must be 
curved. The question then arises as to the relative merits of these 
two forms. 

S14. Straight Crest vs. Straight Toe. The amount of masonry 

■ SaginteHng JThm, vol. xlx. pp. BU-Ul 



ovGoQi^lc 



830 hajbonbt daks. [chap. xiu. 

in the two forma is the same, gince the vertical sections at all points 
are alike io both.* 

The stabilitj of the two lorniB, considered only as gravity dams, 
is the same, since the crosB sections at like distances from the center 
are the same. 

The form with a carved crest and straight toe will have a slight 
advantage due to its possible action as an arch. However, it is not 
necessary to disonas further the relative advantages of these two 
types, since it will presently be shown that both the toe and the crest 
of a gravity dam should be curved. 

616. Gravity n. Aroh Dams. A dam of the pure gravity type 
is one in which the sole reliance for stability ia tha weight of the 
masonry. A dam of the pnre arch type is one relying solely upon 
the arched form for stability. With the arched dam, the pressure 
of the water is transmitted laterally thtough the horizontal sections 
to the abutments (side hills). The thickness of the masonry is so 
small that the resultant of the horizontal pressure of the water and 
the weight of the masonry passes outside of the toe ; and hence, 
considered only as a gravity dam, is in a state of unstable equilib- 
rium. If such a dam fails, it will probably be by the crushing 
of the masonry at the ends of the horizontal arches. In the 
present state of our knowledge concerning the elastic yielding of 
masonry, we can not determine, with any considerable degree of 
accuracy, the distribution oC the pressure over the cross section of 
the arch (see Art. I, Chap. XVIII). 

If it were not for the incompleteness of onr knowledge of the 
laws governing the stability of masonry arches, the arch dam would 
doubtless be the best type form, since it reqaires less masonry for 
any particular case than the pure gravity form. The best infor- 
mation we have in regard to the stability of masonry arches is de- 
rived from experience. The largest vertical masonry arch in the 
world has a span of only 220 feet. There are but two dams of the 
pnre arch type in the world, viz. : the Zola f in France and the 

* IF the TftUsf acroM which thedMU U built haa anr considerable longltadlnkl slope. 
Be It iuuall]r will have, there will be a sllgbt difference according to the relattre posi- 
tion of the two forms. If two ends remain at the same place, the straight toe throws 
tbe dam (artber ap tbe valler, makes the base higher, and conaeqnently allghtly de- 
oreaaes the amount of maaom?. 

t For description, see Repeat on Qnaker Brld^ Dam, Oigtiuwine Mwt, toL xlx. 



ovGoQi^lc 



ABT. 3.] OUTLINES OF THE DESIGN. 331 

Bear Valley* in Sonthem California. The length of the former is ' 
205 feet on top, height 122 feet, and radiua 158 feet; the length 
of the latter ib 230 feet on top, height 64 feet, radius of tup 
335 feet and of the bottom 226 feet. The experience with larffs 
arches is bo limited (see Table 63, page 502), as to render it un< 
wise to make the stability of a dam depend wholly upon itB action 
as an arch, except under the moBt favorable conditions as to rigid 
side-hills and also under the most unfavorable conditions as to cost 
of masonry. Kotioe that with a dam of the pure arch typo, the 
failure of one part is liable to cause the failure of the whole ; while 
with a gravity section, there ia much less danger of thia. Further, 
since the average pressure on the end arch stones increases with the - 
span, the arch form is most suitable for short dams. 

ftlB. Curved Gravity Dami. Although it is not generally wise 
to make the stability of a dam depend entirely upon its action as 
an arch, a gravity dam should be built in the form of an arch, /. «., 
with' both crest and toe curved, and thus secure some of the advan- 
tages of the arch type. The vertical cross section should be so pro- 
portioned as to resist the water pressure by the weight of the 
masonry alone, and then any arch-like action will give an addi- 
tional margin for safety. If the section is proportioned to resist 
by its weight alone, arch action can take place ouly by the elastic 
yielding of the masonry under the water pressure ; but it is known 
that masonry will yield somewhat, and that therefore there will be 
some arch action in a curved gravity dam. Since but little is known 
about the elasticity of stone, brick, and mortar (see § 16), and noth- 
ing at all about the elasticity of actual masonry, it is impoesiblo 
to determioe the amount of arch action, t. e., the amount of pres- 
sure ih&t is transmitted latemlly to the abutments (side-hills). 

That it is possible for a dam to act as an arch and a gravity dam 
at the same time is shown as follows : " Conceive a dam of the 
pnre arch type, of thin rectangular cross section so as to have no 
appraciable gravity stability. Conceive the dam to be made up of 
successive horizoutal arches with key-stones vertically over each 
other. The thrust in each arch will increase with the depth, but 
the spans will, under the ordinary practical conditions, decrease 
with the depth, so that the tendency to 'settle at the crown' (move 
horizontally) will be approximately equal in each. If now we adopt 

* For descTlpttOD, eee Englnarmg Xau, vol. xlz. pp. 618-16. 



ovGoQi^lc 



332 HA80NBT DAKS. [OHAP. XIII, 

a triangular in place of a rectangalar crose section, we increase the 
areas and decrease the unit pressureB from arch-thmst as we go 
down, and hence decrease compression and consequent horizontal 
' settlement ' of the arches ; in other words, we introduce a tendency 
in the water face of the dam to rotate about its lower edge. But 
this is precisely the tendency which results from the elastic action ' 
of the mass in respect to gravity stability, which latter we hare at 
the same time introduced by adopting the gravity section. Hence 
the two act in perfect harmony, and there will be a certain size of 
triangular section (thecretically, — practically it could not be exact) 
at which precisely half the stability will be due to arch action and 
half to gravity action, each acting without any appreciable conflict 
or interference with the other." • 

617. In addition to the increased stability of a curved gravity 
dam due to arch action, the curved form has another advantage. 
The pressure of the water on the back of the arch is everywhere 
perpendicular to the up-stream face, and can be decomposed' into 
two components — one perpendicular to the chord (the span) of the 
arch, and the other parallel to the chord, of the arc. The first 
component is resisted by the gravity and arch stability of the dam, 
and the second throws the entire up-stream face into compression. 
The aggr^ate of this lateral pressure is equal to the water pressure 
on the projection of the np-stream face on a vertical plane perpeu' 
dicular to the span of the dam. This pressure has a tendency to 
close all vertical cracks and to consolidate the masonry traiiBversely, 
— which effect is very desirable, as the vertical joints are always \ea» 
perfectly filled than the horizontal ones. This pressure also pre- 
pares the dam to act as an arch earlier than it would otherwise do, 
and hence makes available a larger amount of stability due to arch 
action. 

The compression due to these lateral components Is entirely in* 
dependent of the arch action of the dam, since the arch action 
would take place if the pressure on the dam were everywhere per- 
pendicular to the chord of the arch. Further, it in no way weaken.'; 
the dam, since considered as a gravity dam the effect of the thrust 
of the water is to relieve the pressure on the back face, and con- 
sidered as an arch the maximum pressure occurs at the sides of tha 
down -stream face. 

* Editorial In Av<>Mcrtiv J^HM, ToL xlx. p. 97S. 



ovGoQi^lc 



AST. 2.] OCTLIlTEa OF THE DB9I0K. 833 

The curred dam is a little longer than a straight one, and heoce 
would cost a little more. The difference in length between a chord 
and its arc is given, to a close degree of approximation, by the formula 

in which a = the length of the arc, c = the length of the chord, 
and r = the radias. This shows that the increase in length due to 
the arched form is comparatively slight ¥or example, if the chord 
is eqiiaitotheradins, the arch is only fy, or 4 per cent, longer than 
the chord. Furthermore, the additional cost is leas, proportionally, 
than the additional quantity of maeonry ; for example, 10 per cent 
additional masonry will add less than 10 per cent to the cost. 

518. Of the twenty-five moat important masonry dams in the 
world, two are of the pure arch type, fifteen are of the curved 
gravity type, and eight are of the straight gravity type. The eight 
highest dams are of the curved gravity type,* 

619. QVAUTT OF THE KASOITBT. It is a, well settled principle 
that any masonry structure which sustains a vertical load should 
have no coDtinuona vertical jointa. Dams support both a horizontal 
and a vertical pressure, and hence neither the vertical nor the hori- 
zontal joints should be continuous. This requires that the masonry 
shall be broken ashhtr (Fig. 39, page 136) or random sqnared-atone 
masonry (Fig, 44, p^e 137), or uncoursed rubble (Fig. 45, page 137). 
The laat ia generally employed, particularly for large dams. The 
joints on the foces should be as thin as possible, to diminish the 
effect of the weather on the mortar and also the coat of repointing. 
In ordinary walls much more care is given to filling completely 
the horizontal than the vertical ones ; but in dams and reservoir 
walls it is important that the vertical joints also shall be completely 
filled. 

To prevent leakage, it is rery important that all spaces between 
the stones should be filled completely with good mortar, or better, 
with mortar impervious to water (see § 141). If the stone itself is 
not impervious, the wall may be made water tight by the ap- 
plication of Sylvester'e washes (§ 263) to the inside Ukx of the 
dam. 



* For Mnrce ot InfonnaUon oonoeming Umm dame, Me | tOO—BOilioefhj o( 



ovGoQi^lc 



33i KABOKBT DAMS. [OHA.F. Xm, 

680. BnuoeKAFET or HAaoKKX SaKL Deaiffn and Conttrue- 
tion of Masonry Dams, "R&niLiae, (MiscellnaeoDB Scientific Papen, 
pp. 550-61.) iSudy of Reservoir Walla, Krantz, (tniulstod from 
the French hj Capt F. A. Afohan, U. S. A.) Frofles of High 
Masonry Dams, McMaster, (published io Van Xostrand's Engineer- 
ing Hagaziae and also aa TSo. 6 of Van Noetrand's Science Series. ) 
Strains in High Masonry Dams, £. Sherman Qonld, (Van 
NoBtraad'B Engineering Magazine, vol. 30, p. 265 tt seg,). Hirtori' 
col and Descriptive Review of Earth and Masonry Dams, teilh 
Plans, David Gravel, (Scientific American Supplement, No. 595 
(Ma; S8, 1S87), pp. 9496-9500.) Wegmann's Design and Ooit- 
struction of Masonry Dams gives an account of methoda em- 
ployed in determining the profile of the propoeed Quaker Bridge 
dam, and also contains illnBtrations of the high masonry dams 
of the world. For a general disonssion of high masonry dams, 
inclnding a consideration of the best form for the horizontal 
cross section, a full description of the proposed Quaker Bridge 
dam and a comparison of it with other great dams, and many 
valuable points' concerning practical details, see numerous re- 
ports, correspondence, and editorials in Engineering News, Jan- 
uary to June, ld8S (vol. 19). The above articles contain many 
references to the literature, mostly French, of high masonry dams. 

Akt. 3. Rock Fill Daiis. 

521. There are three well-known types of dams, which have 
been in nae from time immemorial : earth bank, timber crib-work, 
and masonry. Recent engineering practice on the Pacific coast has 
introduced another type, viz.: the Rock Fill Dam, which is of too 
much importance to pass by without a mention here, although 
strictly it can not be classed as masonry constrnction, 

A rock fill dam consists of an embankment of irre^lar stones 
thrown in loosely, except that sometimes the faces are laid by hand. 
If the overflow is to discharge over the crest, the largest stones 
should be placed on the down-stream slope. The dam may be made 
praoticolly water tight (1) by filling the voids with smaller stones, 
gravel, sand, and earth, or (2) by placing any desired thickneas of 
earth and pnddle on the up-stream &ce, or (3) by covering the 
water slope with one or more thicknesses of planking, which is calked 
ind sometimee also pointed. Either the first or second method 



ovGoQi^lc 



ABX. 3.] ODTLUTES OS THE DBSI6H. 33S 

TOnId make a dam ptactically water tight from the beginning, and 
it would grow tighter with age ; the third method, if carefully exe- 
outed, would make the dam absolutely water tight at the beginning, 
bat would decay, aioce the upper part of the sheeting would ordi- 
narily be alternately wet and dry. 

A great number of rock fill dams hare been built on the Pacific 
slope in the past few years, for mining and irrigating parposea. A 
dam of this character has recently been completed on the Hassa- 
yampa River in Arizona, of the following dimeneions : " Height, 
110 ft.; base, 136 ft.; top width, 10 ft.; length on top, 400 ft.; 
water filopo, SK) ft. horizontal to 4? ft. vertical (^ to 1); back slopes, 
70 ft. horizontal to 180ft. vertical (f to 1); contents, 46,000 en. jda.; 
coat, by contract, 12.40 per ca. yd." * It is proposed to build a dam 
of this character in California 3fiO feet high, which is about 80 feet 
higher than any existing masonry dam, and practically ia nearly the 
eame amount higher than the proposed Quaker Bridge dam 
(Fig. 78, page 328). 

038. "Earth dama ara good and useful when only atill water not 
running over the crest is to be dealt with. Counting reservoir walls 
as dams, which they are, earth dams aro vastly mora used than any 
other. They must be made with the greatcat care, and, if of any 
considerable height, an inner wall of puddle is necessary to their 
integrity. They must be carried down to firm and impervious sub- 
soil of some kind, or they are worthless. Any considerable leak is 
at once fatal to them ; and they ara also subject to serious injury 
from muskrats, crabs, etc. Kevertbeleas, many earth dams of 
great age and great height exist, and bid fair to exist for ages, 
showiug that it is entirely possible to make them secura." 

Stone-filled timber cribs have been very much used for dams ; 
bat such structures ara sure to rot in time, since the timber can not 
always be kept wet. It seeme probable that in moat instances where 
cribs have been used a rock-fill dam would have been better, 
cheaper, and more dnrabfe. 

Masonry dams of all sizes, proportions, and agee exist in great 
abundance, and the entire snitability of masonry for the construction 
of dama is well established. This class of dams is to be preferred 
where large quantities of stone are not near at hand, or where leak- 
age ia undesirable because of loss of water or of injar; to land be- 
Kim, T<^ XX. p. MS. 



ovGoQi^lc 



836 VASONRT DAJfS. [CHA.P. XIII. 

low, or where space is valaable, or where the aurrouudingB require 
a dam of good appearance. 

533. " These three types afford an adequate choice for nearly all 
requirements, but it is obvioDS that they aire open to certain com- 
tnon objections from which the fourth type — a rock-fill dam — is 
free. They are all comparatively costly ; they require a good deal 
of labor, and much of it skilled and faithful labor, for their con- 
struction ; they can only with great inconvenience be constructed 
with water around them, which for the most part must be kept away 
by costly coSer-dams or diversions of channels ; above all, a leak is 
always a source of danger, and is apt to be destructive. They are 
all of them, ae it were, during all their existence, in unstable 
eqnilibriam — all right so long ae the balance of forces remains un- 
disturbed, and seriously endangered by a variety of causes which 
may disturb it. On the other hand a rock-fill dam is by the very 
process of its construction, if conducted with reasonable judgment, 
a structure which tends to improve with time, and which can not 
be injured but may be benefited by causes which threaten the 
other and more artificial types ; in other words, it is a structure 
which may not be very tight, but which is in stable equilibrium as 
respects all disturbing causes, being improved and never injured by 
them. 

" A rock-fill dam is appropriate where the bed on which it rests 
is either rock, hard-pan, stiff clay, or some other impervious and 
almost unwashable material. The bed may be more or less over- 
laid with gravel or loose material without barm, if it be possible 
to remove the loose material in advance, and if there be current 
enough to remove it from under the foot of the dam, as the work 
of construction progresses, it will not even involve extra expense or 
delay, and the dam may be begun on top of the stratum without 
apparent regard to it ; but whenever there is any considerable 
stratum of loose material, a rock-fill dam can only be built byback- 
ing it with earth or puddle as a timber dam would be, and the 
necessity of providing a proper apron to receive the overflow may 
make a timber or crib dam the more economical. It is obvioua 
that the place of all places for the proper use of such a rock-fill 
dam is where leakage is of no importance, either from the loss ot 
water or from injury to land below ; where skilled labor is scarce 
Aod costly, and simplicity of work rather than aggregate quantitiea 



ovGoQi^lc 



AET. 2.] OUTLINEB OP THE DE3I0K. 337 

the important coneideration ; where good materials for masonry are 
scarce or absent ; and where the surroundings do not demand at- 
tention to the question of appearance." • 

The greatest economy in this form of dam occnrs when the fill 
ie made in water ; and it is particularly advantageous In the canali- 
sation of riyers, i. «,, in forming pools in rivers for the benefit of 
navigation. It has been proposed to use rock-fill dams ezcluBively 
in the coustruction of the Klcaragua canal. 

524. In California the cost of this class of dams varies from tft 
to t3 per oubic yard, including all accessories, which is said to be 
«bout 50 per cent, cheaper than for earth dams of equal area of 
transverse cross section. 

" EdlEOrlal Id Sngineeri^f JTnw, voL sz. p. 7U. 



D.qitizeabyG00l^lc 



CHAPTER XrV. 
RETAINING WALLS. 

625. DeFUIIIOWI. lietaining wall ie a wall of masonry toi- 
fnstaining the pressure of earth deposited behind it after it is bnilt. 
A retaiDing wall is Bometimes called a snstairtirig wall. 

Face wail, or nlope wall, is a species of reiainiog wall built 
against the face of earth iu its undisturbed and natural position. 
Obviously it is mnch less important and involves Less diSSculties 
than a true retaining wall. 

Buttresses are projections in the front of the wall to strengthen 
it. They are not often used, on account of their unaightliness, ex- 
cept as a remedy when a wall is seen to be failing. 

Counterforts are projections at the rear of the wall to increase 
its strength. They are of doubtful economy, and were much more 
frequently used formerly than now. 

Land-ties are long iron rods which connect the face of the wall 
with a mass of masonry, a large iron plate, or a large wooden post 
bedded in the earth behind the wall, to give additional resistance to 
overturning. 

Surcharge. If the material to be supported slopes up and back 
from the top of the wall, the earth above the top is called the sur- 
charge. 

Retaining walla are frequently employed in railroad work, on 
canals, about harbors, etc.; and the principles involved in their 
construction have more or less direct application in arches, in tun- 
neling and mining, in timbering of shafts, and in the excavation of 
deep trenches for sewers, etc., and in military engineering. 

fi2e. Method of FAUintE. A retaining wall may fail (1) by 
revolving about the front of any horizontal joint, or (2) by sliding 
on the plane of any horizontal joint, or (3) by the bnlging of the 
body of the masonry. The first is mnch the most frequent mode of 



ovGoQi^lc 



DIFFICULTIES. 339 

failure, and the second ie the least frequent. The wall can not fail 
by the center's bulging out, unless some force acta to keep the top 
from moving fomard, — as in a cellar wall, the abutments of arches, 
etc. 

637. DITFICUITIEB. In the discussion of the stability of dams, 
it was shown that in order to completely determine the effect of the 
thrust of the water against the wall, it is necessary to know (1) 
the amount of the pressure, (3) its point of application, and (3) 
the direction, of its line of action. Similarly, to determine the 
eSeot of the thmst of a bank of earth against a wall, it is necessary 
to know (1) the amonnt of the pressure, (2) its point of application, 
-and (3) its line of action. The determination of these three^quan- 
tities requires three equations. The resistance of the wall both to 
sliding and to overturning' can be found with sufficient accuracy, as 
has already been explained in Chapter XIII — Dama; — but the 
othei elements of the problem are, in the present state of our 
knowledge, indeterminate. 

The origin of the difficnlties may be esplained briefly as follows. 
A B represents a retaining wall ; AD v& the 

■ ,'/ 

break away and come down some line as CD. The a/ .» 




A 



face of the ground. The earth has a tendency to 
break away and come down some line as CD. The 
force tending to bring the earth down is its weight ; Ar^ J 
the forces tending to keep it from coming down are 
the friction and cohesion along the line CD. The_ 
pressure against the wall depends upon the form of 
the line CD. If the constants of weight, friction, Fio. 78. 

and cohesion of anj- particular ground were known, the form of CD 
and also the amount of the thrust on th6> wall could be determined. 
Kot with standing the fact that since the earliest ages constructors 
have known by practical experience that a mass of earthwork 
will eiert a severe lateral pressure upon a wall or other retaining 
structure, there is probably no other subject connected with the 
constmctor'a art in which there exists the same lack of exact ex- 
perimental data. This lack is doubtless due, in part at least, to a 
reliance upon theoretical investigations. Of coarse, mathematical 
investigstions unsupported by experiments or experience are a very 
uncertain guide. 

This subject will be discussed further under the heads (1) 
Theoretical Formulas, and (2) Empirical Rales. 



ovGoQi^lc 



BETAININO WALL3, [CHAP. IIV. 



Aet. 1. Theoretical Foemulas. 

628. A great variety of theories have been presented, bat all rest 
upon an nncertain fonndation of assumption, and all are more or 
less defective and self- contradictory. All theories of the stability 
of retaining walls involve the three following assumptions : 

S29. FIBST AssuiIPTIOH, All theories assume that the surface 
of rupture, C D, Fig. "3-, is a plane. This is equivalent to asaum- 
ing that the soil ia devoid of cohesion, and is inelastic and homo- 
geneous, and also that if a mass of such material be sustained by a 
wall, there is a certain plane, called the plane of rupture, along 
which the particles are m equilibrium, i. e., are just on the point of 
moving. This assumption would be nearly correct in the case of 
clean, sharp sand, but would be considerably in error with a tough, 
tenacious soil. 

This assumption gives the data by which the amount of the 
thrust of the earth can be computed; that is to say, this assumption 
furnishes the conditions from which one of the equations may be 
established. 

930. fiEOOini ABBUitPTlOH. A second assumption which is always 
made is that the point of application of the lateral pressure of the 
earth is one third of the height of the wall from the bottom. The 
total pressure on the wall varies as some function of the height ; 
And it is assumed to vary ae the square of the height, and that 
therefore the center of pressure is at a point two thirds of the 
depth below the top. This is equivalent to assuming that the varia- 
tion of the pressure in a mass of earth is the same as in a liqnid, 
4. «., that the material is devoid of internal friction. 

This assumption furnishes the second of the equations required 
-to determine the effect of the thrust of earth against a retaining 
-wall. 

K31. TEIBD ASBOKPnoir. The third equation is obtamed by 
assuming the direction of the pressure. There are different theories 
ibaeed on different assumptions as to this direction. 

The theories of the stability of retaining walls in most frequent 
use will now be stated, and ' the underlying assumptions and the 
defects of each will be pointed out. 



ovGoQi^lc 



ABT. 1.] TQEOBETICA.L FOBUULAS. 341 

632. CotrtOMB's Theokt. The theory advanced bj Conlomb in 
1784 was the first to eves approximate the actual conditioDs, and 
bis method is the basis of Dearly all formulas used by engineera at 
the present time. It haa been taken up and followed out to its 
consequences by Prony (1802), Mayniel (1808), Fran<jaiBe (1830), 
Na?ier (1826), Audoy and Poncelet (1840), Hagen (1853), Scheffler 
(1857), and Moseley, as well as a host of others, in recent times. 

Coulomb assumea (1) that the line D C, Fig. 73 (page 339), is 
a straight line, down which the prism A CD tends to slide; (2) that 
the resultant pressure is applied at a point two thirds of the depth 
below the top; and (3) that the pressure exerted by this mass oa the 
wall is normal to its back face, which is equivalent to neglecting the 
friction of the earth against the back of the wall. He decomposed the 
weight, ff, of the prisma CD, Pig. 74, and the 
reaction, R, of the wall into two components 
respectively, parallel and perpendicular to the 
surface of rupture, D C. The difference of 
these parallel components, P^~ P,, he placed 
eqnal to the prism's resistance to sliding; and 
assumed the latter to be equal to n N^, in which 
H is the co-efficient of friction. There is some prism, A CD, the 
presBure of which against the wall is just sufficient to cause sliding. 
The amount of this pressure will depend upon the weight, w, of a 
anit of volume of the backing; upon the height, h, of the wall; 
upon the co-efficient of friction, /i, of earth on earth; and upon the 
distance A D, which call x. 

Under the conditions assumed, it is possible to state a value of 
R in terms of h, w, ft, and x. Coulomb assumed R to vary as x,. 
and differentiated the value of ^ to find the position of the snrtace- 
of rupture, D C, for a Maximum pressure on the wall. This Jeads^ 
to the simple conclusion that the lateral pressure exertttd by a bant 
of earth with a horizmital top is simply that due to the wedge-shaped' 
mass included between the vertical back of the wall and a line bi- 
secting the angle between the vertical and the slope of repose of the^ 
material;* that is, the pressure of the earth against the wall A B, 



* For OD algebraic demonatntlon, see Hoselej's Mechanios of EhiglDeeriDK (Sd 
Amar. Ed.), pp. 11S-16; for a graphical demonstntloii. Bee Van Nostrand's 
log Magazine, vol. tx. p. 302, and vol. izll. p. 387. 




jvGooi^le 






342 BBTA.ININQ WALLS. [CHAP. ZIT. 

Fig. 74, IB equal to the prcBeure of the prism A CE eliding along a 
perfectly smooth plane C E, which bisects the angle of repose, A CD. 
No satisfactory proof has been given of the correctneBB of this 
procedure by either Coulomb or any one else; and no defense has 
ever been made against a number of serious objections to 
^ it which have been raised. Experiments show that the 
lateral pressure of the prism ABC, Fig. 75, between two 
boards A B and A C, against A B, " is quite as much when 
the board ^ 6* is at the slope of repose, 1^ to 1, as when it 
is at half the angle; and there was hardly any difference 
whether the board was horizontal, or at a slope of ) to 1, or at 
any intermediate slope."* 

533. By this theory the pressure of the wedge ACT) (Fig. 
74) is 

P = JwA'tan'i^CD, (1) 

in which to ie the weight of a unit of the material to be supported, 
aad h is the height of the wall. This thrust is assumed to act two 
thirds of A C, Fig. 74, below A. Or, in other words, the thrust of 
the prism is equivalent to the pressure of a liquid whose weight per 
unit of volume is w tan' \ A CD. 

Equating the moment of the overturning force and the moments 
of resistance in terms of the unknown thickness, and solving the 
«quation, gives the thickness which the wall must have to be on the 
point of overturning. For example, iissume that it is desired to 
determine the thickness, t, of a vertical rectangular wall. Kepre- 
aent the weight of a cubic foot of the masonry by IF. Then placing 
the moment of the wall equal to the amount of the thrust of the 
earth, gives 

W1il.iit = P.^h (2) 

Solving equations (1) and (2) gives 

t - h i»nh ACD\ 



• Benj. Baker, an eminent Engllab engineer, In a yety IntereaHng and Inatnictlve 
article on " The Actual Lateral Pressure of Earthwork," reprinted in Van Noetrand's 
EngtoeeriDg Magulne, vol. iiv. pp. 388-12, 35&-n, and 492-609, from Proc. ol tlia 
lost, ot C, E., vol. Iiv. pp. 140-Ml, 



ovGoQi^lc 



ABT. I.] TBBORETICAL POBMrLAS. 343 

KumerooB tables have been computed which give, to a great 
number of decimal places, the thickness of a rectangular wall in 
terms of its height, the arguments being the ratio of the weights of 
a unit of volume of the wall and backing, and the angle of repose. 
Snch tables are of but Httle practical value, as will appear presently. 

S34. Surcharged Walls. The rule that the plane of rupture 
bisects the angle between the natural slope of the earth and the back 
■of the wall, holds good only when the top surface of the bank is 
horizontal and the back of the wall vertical. The formula for a 
surcharged wall, or for the case in which the back is not vertical, 
■or for both combined, may be deduced* in the same general way as 
above; bat the results for each case are too complicated for ordinary 
use, and each is subject to the same errors as the formula for a ver- 
tical wall and level top surface. There are a number of exceedingly 
Ingenious graphical solutiotui of the resulting equations, f 

fi36. Beliability of Coulomb's Theory. It is, generally conceded 
that the results obtained by this method have but little practical 
valne. " Experiments and practical experience show that walls, 
which according to this theory are on the point of overturning, 
possess on the average a factor of safety of about tjco." I One of 
the author's students experimented with fine shot, which appear to 
falfill the fundamental assumptions of this theory, and found that 
the observed resistance was 1.97 times that computed by Coulomb's 
formuIa.§ The uncertainties of the fondamental assumptions and 
the qneBtionablenesB of some of the mathematical processes are 
eafficient explanation of the difierence between the theory and 
practice. 

636. WSTBAirOH'B THEOBT. This is the latest one, having bnen j 
proposed in 1878. It was first brought to the attention of American 
engineers by Professor J. A. Du Bois's translations of Winkler's 
•' Neue Theorie dea Erddruckes," and Weyraiich's paper on retain- 
ing walla published in "Zeitschrift fiir Baukunde," 1878, Band i. 
Heft 2, which translation was published in the Journal of the Frank- 

■ See Moseley'B Mechaolcs ot En^ueerinK, pp. 4S4--3G. 

i See Van Nostraud's Eai^neeriDg Magazine, vol. U. p. SD4 : aod do., voL ixv. 
p. 3SS. For releranoea to elaborale gmpbical treatises od retaining walls, see Dn. 
Bola's Graphical Slatlca. pp. Ir-M of Introdnctton. 

I Beoj. Baker \a " The Actual Lateral Pressure ot Earthwork." See loot-note on 
psffeSia. 

S See M. Fargnaeon'B Bachelor's Thesla, University of Illinois. 



ovGoQi^lc 



344 



BETAINIirO WALLS. 



[chap. XIT. 



lin Institute, vol. eviii, pp. 361-87. The following preseotation ot 
tbiB theory is drawo mainly from that article. 

This theory assumes (1) that the enrface of rupture ia a pIaDe> 
(2) that the point of application of the resultant of the latenj. 
pressare of the eM-th, is at a point one third of the height of the 
wall from the bottom, and (3) that there is no friction between the 
earth and the back of the widl. It ie claimed that these three are, 
the only assumptions involved in this theory, and that the direction, 
of the resultant pressure is deduced from the fundameutal rela- 
tions necessary for equilibrium under the conditions assumed. 

The analysis to establish the equations fobthe^ount and direc- 
tion of the thrust of the earth is too long and too complicated to be 
attempted here; couBequently, only the final equations will be 
given. 

F Let S = the thrust of earth against 
the wall. 
w = the weight of a nnit of the 

earth. 
h = the height of the walL 
a = the angle the back of wall 

makes with the vertical. 
6 = the angle which E makes 
with the normal to the 
Fib. Tt. back of the wall. 

6 = the angle of the upper surface with the horizontal. 
fi = the angle of the plane of rapture with the vertical. 
= the angle of repoee with the horizontal. 
637. General Pormulaa. For a plane earth-surface, horizontal 
or eloping up at any angle, and the back of the wall vertical or 
leaning forward at any angle, the general relations are * 




p- r coB(0-«) ~l' 



2 COS (a + Sy 



(*> 



_ i/ aiP (0 + tf) sin (0— ~e) 
■* ' cos (« + <J) COS (or — €)' 



(«) 



jvGooi^le 



T. 1.J THBOfiETICAL POBHULAS. 34j>> 

The valne oi S required in (5) can be deduced from 

- flin (3 or — e) — Kim 2 (a — e) ,„. 

in which 

__ COfl 6 — V COB* £ — cob' . , 

cos' ~ ' '■ 

538. Horiiontal Earth-wfiiice. If the upper snr&oe iit the 
earth is horizontal, then £ = 0, and 

P^ tana Vw 

^-flin(a + tf)*^^' (^> 

and S can be found from 

, - sin sin 8 a: „^ 

,tan S = r^--7 =— (9> 

1 — Bin cos 2 « • ' \'r 

If the back of the wall is Tertical, a = 0; aod eqaation (9)< 
gives S = 0. Therefore 



= tan* 46' 



E: 



639. Bnroha^e at the Hatnral Slope. If the upper surface of 
earth has the natural slope, 6 = 0; and therefore 

. fcos (0 - a)-V h' w 

'L cos« j2coe(a + tf)* ■ • ' • ^"' 

and 6 is determined from 

" tan tf - sin COS (0 - 2 «) 

**°*-l-Bin08in{0-2«) <*^> 

If the back of the wall is vertical, rt = 0, and 5=0, which 
ahowB that E acts parallel to the top Borface of the earth. In this 
case 

J? = iCOB0A'«. (13) 

* Coippaie with eqwUlon (1), pige 848. 



ovGoQi^lc 



346 HETAIMING WALL3, [CHAP. XIV. 

640. The general eqaationa for Weyrsuch'a theory, viz., equa- 
tions (4), (5), (6), and (7), have not been solved for any special 
•caae, except for e = 0, and e = 0. The redaction is very long and 
tedious. 

641. The formulas for each of the above cases may be solved 
graphically,* but the explanations ar£ too long to be given ber«. 

542. Heliability of Weyrauoh'B Theory. On behalf of this 
theory it is claimed f that the only errors in it are those due to the 
neglect of the cohesion of the backing, and to assuming that the 
surface of niptare is a plane ; and also that " it is free from all the 
objections which may be urged against all others, and can be used 
Vith confidence." These claims are not supported by the facts. 

Weyrauch's theory is unquestionably subject to any errors Thich 
may be involved in the aBsamptions that the surface of rupture is a 
plane (see § 529), and that the point of application of the resultant 
pressure of the earth is at two thirds of the height of the wall from 
the top {see § 530). Second, the analysis purports to be perfectly 
general ; I but it is evidently ina|^licable to a wall inclined toward 
the earth to be supported, since the formulas make the thrust of 
the earth increase with the backward inclination of the wall. Ir 
fact the theory makes no difference between a wall leaning forward 
and one leaning backward. For a wall inclining at the angle of 
repose, it gives a very great lateral pressure — see eqs. (S) and (9). 
Third, the mathematical process of determining the position of the 
earface of rupture is at least questionable. Fourth, the theory errs 
on the safe side, becauee it neglects a vertical component of the 
«arth pressure which is independent of friction. § 

Weyrauch's theory differs from Coulomb's only in the form of 
the results and in the manner of deducing them ; | and hence is of 
no pi-actical value. 

643. Weyrauch's method of deducing the direction of the earth 

*S«e Jonr. Frank. Inst., vol. cvltl. pp. 880-85; Vui Noatntnd'E EDglneerfng 
Hag&zine, vol xiU. pp. 366-78 ; Howe's ReUUnliig W&lla for EBith, pp. 7-l-i. 

t Bf IIS antbor, 'Prot. Werraucli, and also by Uie tmoBlator, Prol. Du BoIb,— ms 
Jonr. Frank, InjW., vol. cvliL pp, 48IS-87. 

t See Jonr. Frank. Inst., vol. cvUi. p. 3T7 ; and also Howe's Retaining Walls for 
Garth, p. a. 

i In proof that aach a component exists, see experiments tiy Staler In Annalei dm 
fbnliil C/iatma, reprinted in Selentiflc Amirlatn Supplentaii. vol. xxlv. pp. 9I31-2S. 

I Van NoBtroDd's Engineering Magazine, vol nil, pp. SCC-TT. 



ovGoQi^lc 



ABT. t.J THEOEETICAL F0BMULA8. 347 

pressure assamoB that there is no friction betveen the earth and the 
back of the vail, or, in other worda, that the angle, d, vhich the 
thrnst of the earth makes with the back of the wall, does not de- 
pend upon the structure of the wall for its value. The formula in 
this form faila to agree with ordinary experience ; and hence it 
hoe been proposed * to modify the general formula by considering 
that the angle between the resultant pressure of the earth and the 
back of the wall is never less than the angle of friction between the 
earth and the wall. The method of doing this ie as follows : 

If 0* represents the co-efBcient of friction between the earth 
and the wall, then the direction of E must make an angle with the 
normal to the back face of the wall eqnal at least to 0'. To intro- 
doce 0' into Professor Weyrauch'a theory, it is necessary to find the 
valae of <J aa given by his formula, and see if it is greater or less than 
4^. If it is loss, use the value of 0' to determine the direction of 
£; if greater, nse the value of S and omit 0' altogether. The 
value of 0' can not be determined accurately ; but unless the back 
of the wall is exceedingly smooth, 0' will be greater than 0. If 
the back of the wall is rough rubble {§ 213) or is stepped, 0* will be 
considerably larger than 0. If the friction between the earth and 
the wall be neglected, »', c, if it is assumed that 0' = 0, then when 
rough rubble i;ctaining walls are proportioned according to Wey- 
raueh'a theory, they will have a factor of safety considerably larger 
than appears from the computations. 

This modification is inconsistent with the general the try, since 
the fundamental equations were established for that value of 6 which 
would produce equilibrium, and the corresponding value of the 
thrust was deduced accordingly. It is certainly incorrect to use one 
direction in determining the value of the thrust and another in 
applying it. Further, it is not reasonable to believe that the thrust 
ever makes an angle with the normal to the back of the wall 
greater than the angle of friction, since one of the fundamental 
conditions of statics is that if the resultant pressure makes an angle 
with the normal greater than the angle of repose, motion takes 
place. This modification of Weyranch's theory purports to give the 
relations forastateof eqnilibrinm, and yet violates the fundamental 
condition necessary for equilibrium. Neither the original theory 
aor the above modification of it are of any practical value. 

* Br Prof. Cain Id Van NoatnDd'B Engineering Magasine, vol. iiv. p. 93. 



ovGoQi^lc 



348 BETAINIKO WA13LS. [CHAP. IIV. 

544. SAMEDn'fl ThzobT. There is a.iiother class of theories, 
irhich, in addition to the asenrnptiona of § 530 and § 531, aeantne! 
that the thrust of the earth makes an angle with the hack of the. 
wall eqaal to the angle of repose of the earth. Different writera 
arrive at their resntts in different ways, hut most of them proceed 
Irom a consideration of the conditions of equiUbrium of the earth 
particles, and arrive at their resnlts by integration. Of the formnlns 
dednced in the latter way, Bankine's * are the best known. All tlie 
theories of this claes have eaaentially the same limitations and de- 
fects as Coulomb's and Weyranch's, 

545. ApmCABUlTT OF THXOKSnoAL TobihtlUl It la generally- 
conceded that the ordinary theories — Coulomb's, Weyranch's, and 
Sankine's, — types of the only ones for which there is any consider- 
able show of reasonableness, — are safe ; but " to assnme upon theo- 
retical grounds a lateral thrust which practice shows to be excessive,, 
and then compensate for it by giving no foctor of safety to the wall,, 
although the common way, is not a scientific mode of procedure." 
Thb is col; another reason for the statement, already made, that 
theoretical investigationa are of but little value in designing re- 
taining walls. The problem of the retaining wall is not one that- 
admits of an esact mathematical solution; the conditions can not b& 
expressed in algebraic formulas. Something must be assumed in 
any event, and it is for more simple and direct to assume the thick- 
ness of the wall at once than to derive the latter from equations 
based upon a namber of uncertain assumptions. 

Bear in mind that none of the above formulas apply if the back 
of the wall inclines towards the earth to be supported, or if the 
wall has a curved pro^e, or if the upper surface is irregular. It 
seems to be conceded that in these cases the surface of rupture i& 
not a plane, and hence no theory yet proposed will apply. 

In this connection it seems necessary to warn the student that • 
not all theories for retaining walls are as nearly correct as those 
referred to above. Some of them, although having all the prestige 
of antiquity and offering the advantages of extended table^for their 
application, are totally valueless, being based upon unwarranted 
assumptions, and violating the fundamental principles of mechanics. 

548. Theoretical investigations of many engineering problems 
which in every-day practice need not be solved with extreme scuu- 



ovGoQi^lc 



.A£J. 2.] £UPIBICAL BULES. 349 

Tocy, tire nsefal in detormiDiDg the relations of the varions elementa 
involved, and thus serve as a skeleton about which to group the 
reaulta of experience ; but the preceding discuBsion shows that the 
present theories of the stability of retaining walls are not sufficiently 
exact to serve even as a guide for future investigations. Further- 
more, the stability of a retaining wall is not a purely mathematical 
problem. Often the wall is designed and built before the nature of 
the backing is known; and the vifriation of the backing, due to rain, 
frost, shock, extraneous loads, etc., can not be included in any 
formnla. 

Abt. 3. Empimcal Rules. 

547. XVOUBH EUUS. The eminent English engineer Benjamin 
Baker, who has bad large experience in this line in the coustrnc- 
tioQ of the underground railroads of London, says, "Experienca 
has shown that a wall [to sustain earth having a level top surface], 
whose thickness is one fourth of its height, and which batters 1 or 
2 inches per foot on the face, possesses sufBcient stability when the 
backing and foundation are both favorable. This allows a factor o{ 
.safety of about two to cover contingencies. It has also been proved 
by experience that under no ordinary conditions of surcharge oi 
lieavy backing is it necessary to make a retaining wall on a solid 
foundation more than double the above, or one half of the height in 
thickness. Within these limits the engineer must vary the strength 
according to the conditions affecting the particular case. Outside 
of these limits, the structure ceases to be a retaining wall in the 
ordinary acceptation of the term. As a result of his own experi- 
ence, the author [Benj. Baker] makes the thickness of retaining 
walls In ground of an average character equal to one third of the 
height from the top of the footings. 

"'Hundreds of revetments have been built by royal engineer 
officers in accordance with Gen. Fanahawe's rule of some fifty years 
ago, which was to make the thickness of a rectangular brick wall, 
retaining ordinary material, 34 per cent, of the height for a batter 
of i, 25 per cent for ^, 26 per cent, for J, 27 per cent, for ^, 28 per 
oent for ^, 30 per cent, for ^, and 33 per cent, for a vertical wall.' " • 

548. TBAUTWnrs'B Bule. Trautwinef recommends that "the 

*Tmi NoBtMDd'H Engtneerlng Magazine, vol zxv. p. B70, tram Froc. Inst of 
t Engineer's Pockct-Book {Ed. 188S), p. 683. 



ovGoQi^lc 



350 BXIAISISQ WALLS. [OHAP. XXT. 

thickness oq the top of the footing coarse of a vertical or nearly 
vertical wall which is to eustaiii a backing of sand, gravel, or e^th, 
level top surface, when the backing Ib deposited looeelj (as when, 
dumped from cars, carts, etc.), for railroad practice, should not be 
less Ihan the following : 
Wall of cul-stoDe, or of flrst-Glasa large-ranged rubble in mortar, S5 per cent. 

" " good common' Hcabbled mortsr- rubble, or brick. 40 percent. 

" " well sea bbled dry rubble , SOperceoL 

When the backing ie somewhat consolidated in horizontal layers, 
each of these thicknesses may be reduced; but no rule can be given 
for this. Since sand or gravel has no cohesion, the full dimensions 
06 above should be used, even though the backing be deposited in 
layers. A mixture of sand, or earth with pebbles, paving etoaes, 
bowlders, etc., will exert a greater pressure gainst the wall than 
the materials ordinarily used for backing; and hence when Buch 
jbacking has to be used, the above thicknesseB should be increased, 
say, about ^ to | part." 

549. Details of ConrsiroTioir. The arrangement of the foun- 
dation of a retaining wall is an important matter, but has already 
been sufiicientlj discussed (Bee Part III, and also g§ 491 and 551). 
It is univereally admitted that a large majority — by some put at 
nine out of ten, and by others at ninety-nine out of a hundred — of 
foilures of retaining walls are due to defects in the foundation. 

Retaining walls are constructed of ashlar or brick, or of either 
ashlar or brick backed with rabble, or of rubble either with mortar 
or dry. Aa. the pressure at each bed-joint is concentrated towards 
the face of the wall, the larger and most regular stones should be 
placed on the front. Occasional stones or even courses should 
project beyond the back of the wall, so thiit the backing can rest 
upon them, thus increasing the rusistauce of the wall to overturn- 
ing. This object ib also promoted by building the back in steps. 
The coping should consist of large flat stones extending clear across 
the wall. 

Aa a rule, the greatest thrnst comes a^inst retaining walls when 
the mortar is green and least able to resist it, which is a reason for 
preferring cement to lime mortar. If the backing is to be filled in 
before tbe mortar hardens, it shonld be deposited in thin, horizon- 
tfil layers, or the wall should be supported temporarily by shores. 

650. Drainage. Kext to a faulty foundation, water behind the 



ovGoQi^lc 



ABT. 2.] EHPIBICAL RULES. 3S1 

wall is the most frequent cauBe of the failure of retaining walls. 
The water not only adds to the weight of the backing material, hnt 
also softens the material and changes the angle of repose so as to 
greatly increase its lateral thrust. With clayey soil, or any material 
resting upon a gtratum of clay, this action becomes of the greatest 
importance. To guard against the possibility of the backing's be- 
coming satnrated with water, holes, called weepers, are left through 
the wall. One weep-hole, three or four inches wide and the depth 
of a course' of masonry, is generally sufficient for every three or 
four square yards of front of the wall. When the backing is clean 
sand, the weep-holes will allow all the water to escape ; but if the 
backing is retentive of water, a vortical layer of stones or coarse 
gravel should be placed next to the wall to act as a drain. An 
ordinary drain at the back of the wall is often useful. 

Wnen the backing is liable to be reduced to quicksand or mud 
by saturation with water, and when this liability can not be removed 
by cffic-ent drainage, one way of making provision to resist the 
additional pressure which may arise from such saturation is to cal- 
culate the requisite thickness of wall as if the earth were a fluid. 
A pnddle-wftll is sometimes bnilt against the back of dock-walls to 
keep out the water. 

The resistance of the wall to sliding is materially increased by 
laying the lower courses of masonry with an inclination inward. 
An objection to inclining the joints, particularly in dry masonry, 
is that the water will enter them and be carried to the backing. 
This objection is sometimes met by building the face with horizon- 
tal courses, and inclining the courses in the back of the wall. The 
back of the wall for 2 or 3 feet from the top should have a batter 
of at least 1 inch in 1 foot, in order that the frost may lift the 
earth and not break the joints of the masonry. 

Walls are sometimes built with both faces inclined toward the 
material to be supported, and sometimes with a curved profile ; but 
it is generally considered unwise to do either, owing to the extra 
expense and trouble in construction. 

661. Land Ties. Retaining walls may have their stability in- 
creased by being tied or anchored by iron rods to vertical plates of 
iron or blocks of stones imbedded in a firm stratum of earth' at a 
distance behind the wall. " The holding power, per foot of breadth, 
of a rectangular vertical anchoring plate, the depth of whose upper 



ovGoQi^lc 



"8o3 BETAINING WALLS. [CHAP. XW. 

aad lower edges below the surface are respectively a;, and a:,, may 
be approximately calculated from the following formula : 

„ V ~ «,' 4 sin ,_ , , 

^ = "'-'-l-^Sr3^ M 

in which H ia the holding-power of the plate in pounds per foot of 
breadth, to is the weight in pounds of a cubic foot of the earth, 

■ and its angle of repose. The center of pressure of the plate ie 
about two thirds of its height below its upper edge, — at which point 
the tie-rod should be attached. 

"If the retaining wall depends on the tie-rods alone for its 
eecurity against sliding forward, they should be fastened to plates 
on the face of the wall in the line of the resultant pressure of the 
earth behind the wall, that is, at one third [see 1 530] of the height 
of the wall above its base. But if the resistance to sliding forward 
ia to be distributed between the foundation and the tie-rods, the 
latter should be placed at a higher level. For example, if half the 
horizontal thrust is to be borne by the foundation and half by the 
tie-rods, the latter should be fixed to the wall at two thirds of its 
height above the base."* 

662. BelieTing Arohei. In extreme caaes, the pressure of the 
■•earth may be sustained by relieving-arches. These consist of a row 
of arches having their axes and the faces of their 
piers at right angles to the face of a bank of earth. 
There may be either a single row of them or several 
tiers; and their front ends may be closed by a ver- 
tical wall, — which then presents the appearance ol 
a retaining wall, although the length of the arch- 
ways is such as to prevent the earth from abutting 
Tia. 73. against it. Fig. 77 represents a vertical transverse 

■ section of such a wall, with two tiers of relieving arches be- 
iiind it. 

To determine the conditions of stability of such a strueture aa a 
whole, the horizontal pressure against the vertical plane OD may be 
determined, and compounded with the weight of the combined 
mass of masonry and earth OAED, to find the resultant pressare 
•on the foundation. 

p. 4U. 



ovGoQi^lc 



CHAPTER XV. 

BRIDGE ABUTMENTS. 

MS. OlHXKlL Foxm. There are four formB of abutments in 
morfk or less general nae, 1. A plain wall parallel to the current,' 
shown in elevation at Pig, 78, with or without the wings J DFaaA 
BEG. The slopes may be finished with an inclined coping, as 
A D, or offset at each course, as B E — usually the latter. This form 
may appropriately be called the straight aiutmeni. 2. The wings 
may be swang around into the bank at any angle, as shown (in plan) 
in Fig. 79. The angle is usually about 30°. This form is known 



.^^^^.=^yA 



SB the wing abutment. 3. When (f> of Fig. 79 becomes 90°, we have 
Fig. 80, which is caUed the U abutment. 4. If the wings of Fig. 
SO are moved to the center of the head-wall, we get Fig. 81, vhich 
is known as the T abutment. 

' The abutment of an ordinary bridge has two offices to perform, 
-viz., (1) to support one end of the bridge, and (2) to keep the earth 
embankment from sliding into the water. In Fig. 78, the portion 
D E Q F serves both these purposes, while the wings A D F and 
B E act only as retaining walls. In Figs. 79 and 80, the portion 
D E perf ormB both offices, while the wings .4 D and B E are merely 
retaining walla. In Fig, 81 the " head " D E supports the bridge, 
and the " tail," or " stem," A B carries the train; hence the whole 
Htructnre acts as a retaining wall and also supports the load. The 
abutment proper may fail (1) by sliding forward, (2) by bulging, or 
(3) by criishing; however, it is improbable that it will fail by sliding 
forward. Its dimensions are to be determined as for a retaining 
-wall (Chap. XIV); but the mathematical theory of the lateral 



ovGoQi^lc 



864 BEIDQE ABUTKENTS. [CHAP. XV.. 

pressnre of earth is a much less perfect guide for designing bridge- 
abutmentB than it iB for simple retaining walls. The weight of the 
bridge helps the abutment to reeiat the thrust of the earth; bnt, on 
the other hand, the weight of the train on the embankment in- 
' creases the lateral pressure against the abatment. 

854. The form of the abutment to be adopted for any particular 
case will depend upon the locality, — whether the banks are low and 
flat, or steep and rocky; whether the current is swift or slow; and 
also upon the relative cost of earthwork and masonry. If the shore 
ie flat, and not liable to be cut away by the current, an abutment 
like Fig. 78 will be sufficient and most economical. However, this 
form is seldom used, owing to the danger of the water's flowing^ 
along immediately behind the wall. 

The form of Fig. 79 may be adopted when there is a contraction 
of the waterway at the bridge site, since deflecting the wing walla, 
above and below, slightly increases the amount of water that can- 
pass. This advantage can be obtained, to some degree, with the 
straight abutment (Fig. 78) by thinning the wings on the front and 
leaving the back of the wings and abutments in one straight line. 
There is not only no hydraulic advantage, bat there is a positive 
disadvantage, in increasing the deflection of the wings beyond, say, 
10° or 15°. The more the wing departs from the face line as it 
swings round into the embankment, the greater it« length and also- 
the greater is the thrust upon it. The wings are not nsually ex- 
tended to the toe, B, of the embankment slope, but stop at a height, 
depending upon the angle of deflection and the elope, such that the- 
earth flowing around the end of the wall will not get into the chan- 
nel of the stream. It can be shown mathematically that, if the toe 
of the earth which flows around the end of the wing is to be kept 
three or four feet back from the straight line through the &ce of 
the abutment, an angle of 35° to 35° is best for economy of the 
material in the wing -wallB. This angle varies slightly with the pro- 
portions adopted for the wing wall and with the details of the 
masonry. This form of construction is objectionable, since the 
foot of the slope in front of the wing is liable to be washed away ;. 
but this could be remedied somewhat by riprapping the slope, or, 
better, by making the wings longer. 

Fig. 78 is one extreme of Fig. 79, and Fig. 80 is the other. Aa 
the wing swings back into the embankment the thrust upon it in- 



ovGoQi^lc 



WING ABCTMEJJT. Wf 

I, reaching ita iiiaximam at an angle of about 45°; whea tb» 
wiDg is thrown farther back the oatward thrust decreaBes, owing to 
the filling np of the slope in front of the wing. Bringing the winga 
perpendicnlar to the face of the abatment, as in Fig, 80, also de- 
creases the lateral pressure of the earth, owing to the intersection of 
the surfaces of ruptui-e for the two sides, which is equivalent to re- 
moving part of the "prisni of maiiraum thrust." If the banks of 
the stream are steep, the base of the wing walls of Fig. 80 may b© 
stepped to fit the ground, thereby saving masonry. Under these 
conditions, also the wing abutment, Fig. 79, can be treated in the 
same way; but the saving is considerably less. When the masonry 
is stepped off in this way, the angle thus formed becomes the weak- 
est part of the masonry; but, as the masonry has a large excess of 
strength, there is not much probability of danger from this cause, 
provided the work is executed with reasonable care. 

566. Fig. 81 is the most common form of abutment. For equal 
amounts of masonry, wing abutments give better protection to the 
embankments than T abutments. The latter are more stable, be- 
cause the center of gravity of the masonry Is farther back from the 
line of the face of the abutment, about which line the abutment 
must turn or along which it will first crush. The amount of m*- 
sonry in tall T abutments can be decreased by building the tail wall 
hollow, or by introducing arches under it. The more massive the 
masonry, the cheaper it can be constructed; and, for this reason, it 
is probable that the simple T abatment is cheaper than the U abut- 
ment, although the latter may have less masonry in it. On the other 
hand, the opportunities for inspecting the masonry during construc- 
tion are better with the U than with the T abutment, and hence the 
former is usually better built than the latter. This la an important 
item, since it is somewhat common tor railroad masonry to fail by 
being shaken to pieces by the passage of trains. 

666. WWO ABtrtWEXT. Fig, 82 shows a common form of the 
wing abutment. This one is finished with stone pedestal blocks — 
marked B in plan, A in elevation, and C in section, — which is not 
always done. The thickness of pedestal blocks and the thickness. 
of the coping under the pedestal blocks vary slightly with the span 
(see g 558). The height of the parapet wall, or dirt wall (the wall 
which keeps back the top of the embankment, marked /* H' in. 
section), will vary with the style of the bridge, but should not have 



ovGoQi^lc 



366 BBIDQE ABLTMENXa [CUAP, XV. 

a thickness less than four tenths of its height (see g§ 547 and 548). 
The bridge often rests directly upon the coping. The top dimen- 
sions of the abutment will depend somewhat upon the size and 
form of bridge ; but for railroad bridges it will usuatlf not be less 
than 5 ft. vide by 20 ft. long, nor more than 6 ft. by 22 ft. 




II 



The nsnal batter is 1 in 13; sometimes 1 in 2i. For heights 
tiader about 30 ft., the top dimensions and the batter determine the 
thlckneBS at the bottom. For greater heights, the quite nniform 



ovGoQi^lc 



WING ABUTMENT. 



TABLE 87. 

QoAKTRT or Hasonbt in Wins Abutments of the Qeneeai. Fobk 

8BOWN IN Fio. 83. See g 557. 



AaiA or Lowrar 



p 


p 


^< 


/«(. 


/«(. 




































































































































sa.B 



/«(. 


/«(. 


























































































































































































108 


M 1 








































































75. a 







IflS.S 
301. c 



• DJmeDsioll iit«n« In tvopntestat blnckg = Mcu. feBt. 

-' cuplng of one Bbutmeut = SS4 " " 

Total dimeniioD stone in '■ " =889 " " 

rale ie to make the thickness four tenths of the height. The amoant 
of masonry in the abutment is computed in accordance with this 
rule, although the actual quantity is usually more than that required 
by it >jince there is no objection to the wall's being rough, no 



ovGoQi^lc 



368 BRIDGE ABCTMENTS, [CDAP, XT. 

stones are cnt out to eecnre the specified thickness, and hence the 
actual quantity of masonry nsually exceeds tlie amount required. 
The spread of the footiug courses and foundation will depend, of 
course, upon the location. 

The wings should be proportioned according to the rules for 
retaining walla (see g§ 547 and 548). The wings are not always pro- 
longed until their outer ends intersect the foot of tlie embankment 
slope; but aro frequently stopped with an end height of 3 to 5 feet 
above the footing. The thickness of the wing wall decreases from 
the body of the abutment toward the tail in proportion to the height. 
For appearance, the top of the wing is usually made uniform from 
head to tail, being uHually from 2i to SJ feet, according to the size 
of the structure. The steps should be capped with stones, not less 
than 1 foot thick, covering the entire step and extending under the 
step above not less than 1 foot. 

657. Content* of Wing Abutments. The table on page 357 
gives the quantities of masonry in wing abutments of the form 
shown in Fig. 82. Since the outlines of such structures are not 
simple geometrical figures, it is necessary to make more or less ap- 
proximations in computing the cubical contents. For eianple, fh 
Fig. 82 the wings are stepped ofi to fit the slope of the emba jkment 
as shown; and hence the comer of each course projects bej ond the 
earthwork. The amount of masonry in these projecting corners 
varies as the thickness of the courses, and for any j,-irticular abut- 
ment it could be found accurately; but, in computing a table of 
general results, it is necessary to assume some thickness for the 
courses. In this case the courses were assumed to be 1 foot thick. 
Thebackof the "head "was assumed to conform strictly to the batter 
line, although in construction itwould be stepped. The dimensions 
of the parapet wall will vjiry with the thickness of the pedestal 
blocks used, and also with the style of tlie bridge. The contents 
of the parapet as given in the table are for the dimensions shown in 
Fig. 83. . _ _ 

Footing courses were not included in the table, since they vary 
with the nature of the foundation. The area o£ the lowest course 
of masonry is given, from which the areas of the footing courses and 
of the foundation pit may be determined. The thickness at the 
top and the batter, as in Fig. 82, give, for any height found in the 
lable, a thickness of wall at the bottom of at least four tenths of its 



ovGoQi^lc 



V ABUTMENT. 369 

heigbt (see §548); for heights greater than in the tab)e, the back 
-of the wall must be stepped to keep the thickneae four tenths of the 
height.* 

668. U AbutmehT. Fig. 83 shows the standard plans of the 
Atchison, Topeka and Santa Fe R. R.f for U abutments. Tbia is 
the only form of bridge abutment used on this road, except in 
-special cases. The T abutment was once the standard, but was 
abandoned about fifteen years ago.^ 

The specifications under whicli these abntmenta are bnilt, require 
as follows : " 1. Bed-plate pedestal blocks to be 3 feet thick, and 
placed symmetrically with regard to the plates. 2. Coping under 
pedestal blocks to be 18 inches thick for all spans exceeding 100 
feet, 16 inches for 90 feet, and 14 inches for spans under 90 
feet, — said coping to be through stones, and spaced alike from both , 
sides of abutment. 3. Distances from front of dirt wall to front 
of bridge seat, and from grade line to top of bridge seat, and 
thickness of- dirt wall, to vary for different styles and lengths of 
bridges. 4. Front walls to l>e 23 feet wide under bridge seat for 
all spans of 100 to 160 feet inclusive. 5. Total width of bridge 
seat to be 5J feet, for all spans. 6. Steps on back of walls to 
be used only when necessary to keep thickness -fg of the height. 
'7. In case piling is not used, footing courses may be added to give 
secure foundation. 8. Length of wing walls to be determined by a 
-slope of 1-^ to 1 at the back end ot the walla — as shown by dotted 
line in front elevation, — thence by a slope of 1 to 1 down the outside 
— as shown on side elevation — to the intersection of the ground line 
with face of abutment. This rule may be modified in special cases. 
9. Dimeusious not given on the drawing are determined by the 
style and length of bridge, and are to be found on special sheet." 

669. Although this road is noted for the excellency of its 
masonry, this design could be improved by leaving a weep hole in 
the side walls, 2 or 3 inches wide and the depth of a course of 



• In computing the conlenta ol masonry Btructares, tt U uecesMUT to remember 
that the volame of any mass which 1b made up of priBma. wedg:es, and pyramids — or 
conee—mnst be determined by the prlsmoldal fonnala : but It the mass la compoeed 
wholly ot prisms and wedges, the contenta can be correctly toand b; osliig the aver- 



ovGoQi^lc 



BKtDQE ABtTTHENTS. [CBAP. XT. 




Fib. eiL-U ABimaiiT.-A. T. & B. F. B. B 



ovGoQi^lc 



n ABCTUENT. 



r Mabonut in U Abotmentb of thk Qenebal Fork 
anowN IN Pio. 83. See § 060. 



Sri 




OF TH« 


^ 


s;»™?^r- 






ll 
11 


BOTTOKO 


TBI 












KT. 


S 


CUPINO,* 














w 


"'«—"'« or TBI MrrsoD or 
OSWa THl Tuu. 


^1 






h 




S 






\i 






i 


1 


■ 




EC 






t5 


a 


H 






/lel. 


/«(. 


/«' 


/«(, 


/«( 


«./!. 


ou./t- 


1 




1 


23.2 


5.1 


118 


3,1 


Ill 


6.0 


1 '^' 


! : : ! : : ! I J ! ! i 


3 


22.3 


5.2 


115 


8.3 




13.8 


li 




8 


33-5 


5.2 


IIH 


8 3 


843 


18-8 


t-B»r. %,oma-»m t. 


4 


22,7 


"5 


130 


3.8 


463 


25,2 


mn Himm i 


5 


32-8 


5,4 


134 


3-4 


684 


83.0 


1 . 


at rf of 


6 




5.5 


126 


s.r. 


709 


89,0 




7 


38.3 


5.6 


129 


3-6 


887 


46.0 


i i 






8 


23. E 


5.7 


133 


3.7 


068 


58.3 








9 


38-5 


5.8 


135 


3.8 


1,101 


60.8 


1 1 








10 


33 7 


5.8 


138 


4.0 


1.338 










11 33 8 


5.9 


141 


4.4^ 


1.377 


76!8 


1 




: ins \ 1 




12 ■ 24-0 


6,0 


14* 


4.8| 


1,530 


86.0 








18 


34.3 


6,1 


147 


5.31 


1,665 


96.0 


i i 
1 1 




i 


14 


348 6.3 


150 


5.8; 


1.814 


106.8 


; ixx == = «;-;-. ^ 


15 


24-5 6.2 


153 


6.0 


1,966 


118.4 


■ ss "-"f^ --ti i 


16 


34,7 


6.3 


156 


6.4 


3,130 


130.8 




g «^ II n "sig;, V 


17 


24.8 


6,8 


139 


6.8 


3.288 


144-0 






la 


35.0 


7.8 


180 


7.2 


2,478 


158.0 




19 


2S.2 


7,6 


191 


7.6 


3;688 


172.8 


S 


^ ■-- £~Jui;^Z " 


ao 


35-3 


8,0 


20:( 


8.0 


2,9-20 


188.4 


t '^', 


; is g |:++ f 


21 


35.5 


8.4 


214 


8,4 


8.1741 204,8 


i -S ^ I^M 1 


23 


25.7 


8.8 


326 


8,8 


3.449 223,0 


i : 


28 


35.8 


».2 


338 


9.2 


3,746 340,0 


^ :S 


SflfiH 


31 


26.0 


0.6 


350 


9 6 


4,066 258.8 




2^ 


26.2 


10.0 


203 


10. 


4.408 278. 4 


1 " 


20 


36.3 


!0.4 


374 


10.4 


4.772; 298.8 


l?f:|f 


37 


26.5 


10.8 


386 


10-8 


5,160 


330.0 


s 


38 


26.7 


n.3 


299 


11.3 


5,570 


342.0 




2» 


26.8 


11-6 


311 


11.6 


8.oo;t 


364-8 




il--'sl 


30 


27.0 13.0 


834 


12-0 


6.460 


388-4 


l^' 


r" t 


SI 


37,3 il3.4 


337 


13.4 


6,941 


413.8 


33 


37.8 il2.6 


sm . 


12.8 


7,445 


438-0 


. J gSa 


33 


87.5 13-2 


863 


13.3 


7.97:l 


464.0 




84 


37.7 ,13.6 


37« 


13.6 


asafl 490,8 




85 


27.8 |14-0 


390 


14.0 


9,103 518.4 


|i|. 


. . , , . , , , 


















• F"r riin.pnBlon« 


of oopltitr snd pedesul blocks. 








hot\m. "^ 


* 





jvGooi^le 



'362 BBIDOE ABCTHENTS. [CSAP. XT. 

maaonry, for each 4 or 5 square yards of wing wall. Cinders, or 
Band and gravel are Bometimes used to fill in between the wing walls 
to give a better drain^e, and also to decrease the lateral thrust of 
the earth. 

560. Content! of U Abntments. The table on page 361 gives 
the contents of U abutments of the form shown in Fig. 83. The 




_ J ij li LI Lf U) 
1 n n " n n n r» 
V v y I) ^,' U 1^ 




T ABUTMENT 



quantities were computed on the basis that the thickness of the 
walls was four tenths the height, except that no wall was taken of a 
loss thickness than that given by the thickness at the top and the 
batter as in the drawing. 

661. T ABnmitT. Fig. 81 shows the ordinary form of T abnt- 



ovGoQi^lc 



T ABUTMENT. 



QnAsmT aw Habonsy in T Abuthrhts of thb Oehxral Fobh 
BBOWN IN Pio. 84. Sbb §563. 





^S 


:r?H°/H's,. 


(Wtitt or Hawhrt. 












|e 


_ 






■s 




s 
1 










'^ 








1 


1 






'R 






tc 


^ 




/•«'. 


/«.(. 


fttt 


/«( 


cu.ft. 


Cll.fl. 


CH.fl. 


|°SS!S=IEIS|SIE 1 




22.8 


S.8 


183 


607 


13,6 


60 I 


e 


28.0 


6.0 


188 


743 


18.0 


73 




7 


23.3 


8.3 


148 


888 


34.5 


84 


ft, «*^^*S 
1 Sffiffig 








:I 


8 


23 3 


6.3 


148 


1,039 


33.0 


96 






^•a-> 


S 


23.5 


6.5 


168 


1.179 


40.5 


108 








;i 


:o 


38.7 


a. 7 


168 


1.BS4 


50,0 


130 








11 


28.8 


6.8 


168 


1,496 


60.5 


m 


1 l«; 






m 




12 


24.0 


7.0 


188 


1,660 


73,0 


144 1 






: i 


13 


34.2 


7,2 


173 


1,831 


84,5 


166 ! 


T .^oos 






22a 


14 


24.8 


7.8 


178 


2,006 


96,0 


188 


4m 






; b 


15 


34.5 


7.5 


184 


2.188 


112.5 


180 ■ 






Soe 


18 


24.7 


7.7 


189 


3.374 ■ 138.0 


193 


tiii 


£ 




II II ff 






34.8 


7.8 


195 


2,566 144.6 


304 1 


!ff';|, 






18 


25.0 


8.0 , 300 


3,768 163.0 


318 


■^S. - «;d !» 


IB 


35.3 


8.3 206 


2,968 180.5 


238 


»l« sli.8=-;3 
sis; i,5|-;""xx 


20 

■21 
23 


35.8 
25.5 
35.7 


8.3 311 

8.6 217 

8.7 323 


3.174 
3,8H8 
3,608 


300,0 
220,6 
343.0 


340 1 
352 1 
264 ; 


28 


35.8 


8.8 336 


8,833 


364.6 


376 


84 


26. 


9-0 284 


4,064 


288,0 


268 


BiZi:,%-,t.-i..". 


25 


26.3 


9,3 , 340 


4,301 


313,5 


800 


nii.i-fi'j'i' 




H6.3 


9.8 246 


4,544 


888.0 


312 


27 


36.5 


9.5 


2.53 


4,7931 364,6 


334 i 


:ae 


26.7 


9.7 


258 


5,047' 893 


386 


29 


36.8 


9.8 


264 


5,308 430.5 


348 


m'MR 


30 


27.0 


10.0 


270 


5.675 ■ 450,0 


360 


81 


37.3 


10.2 


276 


5,H48i 480.5 


372 


3-2 


87.3 


10.8 


382 


6,127| 512.0 


884 


83 


37.6 


10.6 1 289 


6,418, 544,6 


396 


34 


27.7 


10.7 395 


6.706' 578.0 


408 


3S 


37.8 


10,8 ! 801 


7.008 613,5 

1 


430 


lii=. =.==,==,, 










Ar«orcopln( 


on a wing., per 


t.ofl«nBth=. R 


•q.f(. 


fill 


^..-..^otcopfn* 


ODbrHge-Mkt 


= 188 







jvGooi^le 



864 BRIDOB ABUTMEHTS. LCHAP. Xt- 

ment. For railroad bridges the head is nsnally Dot lesB than 5 ft. 
X 20 ft., nor more than 6 ft, X 22 ft., under the coping, according 
to the size of the bridge. The tail wall is UBually 10 or 13 ft. wide, 
and of such length that the foot of the ^lope of the embankment 
will juBt reach to the back of the head wall. The batter on the 
head wall is 1 to 13 or 1 to 24 all around. The tail wall is generally 
built vertical on the sides and the end. Notice the batter at the 
top of the free end of the tail wall. This is known as the " frost 
batter," and is to prevent the frost from dislocating the corner of 
the masonry. The drainage of the ballast pocket should be pro- 
vided for by leaving a space between the ends of two atones. 
Formerly the tail wall was sometimes only 7 or 8 feet wide, in which 
caae the ties were laid directly upon the masonry without the inter- 
vention of ballast ; but this practice has been abandoned, as being 
very destructive of both rolling stock and roasonry. 

According to the common theories for retaining walls, T abut- 
ments with dimensions as above have very large factors of stability 
against sliding, and overturning, and crushing. 

562, Content! of T Abatments. The table on page 163 gives the 
contents of the abutments of the form shown in Fig. 84. The 
height of the tail above the under aide of the bridge-seat coping will 
vary with the thickness of the pedestal blocks, and with the style of 
the bridge ; and hence the table gives the quantities in the abutment 
below the bridge-seat coping und above the footing. The quantity 
of masonry above this line will vary also with the amount of ballast 
used. The term "wedge" in the table is used to designate tliat 
part of the tail included between the head and a vertical plane 
through the lower edge of the back face of the head, 

663. FOUHDATIOB. Usnally but little difficulty is encountered 
in securing a foundation for bridge abutments. Frequently the 
foundation is shallow, und can be put down without a coffer-dam, 
or at most with only u light curb (see g§ 316-;'0). Where the ground 
is soft or liable to scour, a pile foundation and grillage is generally 
employed. For the method of doing this, see Art. 3, Chapter XI ; 
and for examples of this kind of foundntion, see Fig. 84 (page 362), 
Fig. 86 (page 380), and Fig. 90 (page 386). 

Where there is no danger of undei' washing, and where the fonu- 
dation will at all times be under water, the masonry may be started 
upon a timber platform consisting of timbers from, say, 8 to 13 



ovGoQi^lc 



QCALITT OF MABONBT. 365 

incheB thick, laid side by side upon sills, and covered by one or 
more layers of timbers or thick planks, according to the depth of 
the foundation and the magnitude of the structure. For an exam- 
ple of a foundation of this class, see Plate II. For a discussion of 
the method of failure by sliding on the fonndation, see § 491. 

564. auAUTT or HASontX.— Bridge abutments are built of 
first-class masonry (g 307) or of second-class (gg 209 and 213), ac- 
cording to the importance of the structure. See also the specifica- 
tions tor bridge pier masonry (^ 591-600). The coping should be 
composed of as large stones as practicable — not less than 13 inches 
thick, and 15 or 18 inches thick is better and more frequently used. 

Sometimes, the bed plates of the bridge rest directly upon the 
coping, but usually upon a stone pedestal block (see Figs. 82 and 
S3), in which case small pedestals, upon which the rail stringers 
rest (see Fig. 90, page 386), are also generally need. 

666. Con. For data on the cost of masonry, see g§ 233-38. 



ovGoQi^lc 



CHAPTER XVL 
BRIDQE PIE»S. 

566. The eelection of the site of the bridge and the arrangemeDt 
of the sptuiB, although important in themselves, do not properly be- 
long to the part of the problem here considered ; therefore they 
■will be discussed only briefly. The location of the bridge is nsually 
a compromiBe between the intereste of the railroad or highway, and 
of the river. On navigable streams, the location of a bridgf, itfr 
keight, position of piers, etc., are subject to the approval of engi- 
neers appointed for the porpose by the United States Government. 
The law requires that the bridge shall cross the main channel nearly- 
at right angles, and that the abutments shall not contract nor the 
piers obstruct the water way. For the regulations governing tho- 
varioDS streams, and also reports made on special cases, see the 
various annnal reports of the Chief of Engineers, XS. S. A., partlcQ- 
larly Appendix X,, of the Report for 1878. 

The arrangement of the spans is determined mainly by the rela- 
tive expense for foundations, and the increased expense per linear 
foot of long spans. Where the piers are low and foundations eoeily 
secured, with a correspondingly light cost, short spans and an in- 
creased number of piers are generally economical, provided the piers 
do not dangerously obstruct the current or the stream is not navi- 
gable. On the other hand, where the cost of securing proper foun- 
dations is great and much difficulty is likely to be encountered, long 
spans and the minimum number of piers is best. Sound judgment 
and large experience are required in comparing and deciding upon 
the plan best adapted to the varying local conditions. 

Within a few years it has become necessary to build bridge pier» 
of very great height, and for economical considerations iron has 
been substituted for stone. The determination of the stability of 
Buch piers is wholly a question of finding the stress in fmme struc- 
tures, — the consideration of which is foi-eign to our subject. 



ovGoQi^lc 



ABT. 1.] THEOBT 0? STABILITT. 



Art, 1, Theory of Stability. 

667. ItETEOD OF Fauurz. A bridge pier may fail in any on»- 
of three waya : (1) by alidiiig on any section on account of the ac- 
tion of the wind against the train, bridge, and exposed part of the 
pier, and of the current of the stream against the immersed part of 
the pier ; or (2) by overturning at any section when the moment of 
the horizontal forces above the sectioD exceeds the moment of the 
weight on the section ; or (3) by crushing at any section nnder tho- 
combined weight of the pier, the bridge, and the train. Tha^ 
dimensions of piers are seldom determined by the preceding condi- 
tions ; the dimensions required at the top (§ 584) for the bridge. 
seat, together with a alight batter for appearance, generally give 
sufBcient stability a^inst sliding, overturning, and crushiug. How- 
ever, the method of determining the stability will be briefly out- 
lined and illnstrated by an example. 

568. Stabiutt AOAivbt SuBIiie. Effect of the Wind. The- 
pressure of the wind against the truss alone is usually taken at 50- 
Ibs. per sq. ft. against twice the vertical projection of oue tmss^ 
which for well-proportioned iron trusses will average about 10 sq. ft. 
per linear foot of span. The pressure of the wind against the truss 
and train together is usually taken at 30 Iba per sq. ft. of truss and 
train. The train exposes about 10 sq. ft, of surface, per linear foot. 
The pressure of the wind against any other than a flat surface is. 
not known with any certainty ; for a cylinder, it is usually assumed 
that the pressure is two thirds of that against its vertical projection. 

569. Effect of Current For the pressure of the current of" 
water against an obstruction, Weisbach's Mechanics of Engineering 
(page 1,030 of Coxe'a edition) gives the formula. 



(!)■ 



in which P is the pressure in pounds, s the exposed surface in 
sq. ft., A a co-efiBcient depending npon the ratio of width to length 
of the pier, w the weight of a cubic foot of water, v the velocity in 
ft per sec., and g the acceleration of gravity. For piers with 
rectangular cross section, k varies between 1.47 and 1.33, the first 
being for square piers and the latter for those 3 times as long as. 



ovGoQi^lc 



368 BBIDOE FIEBS. [CUAP. ZVI. 

Wide ; for cyliaders, i = about 0.T3. The law of the Tarution of 
the yelocity with depth U not certainly known; bnt it is probable 
that the velocity raries as the ordioates of an ellipee, the greatest 
velocity being a little below the enrface. Of coarse, the water haa 
ita maiimam effect when at its highest stage. 

S70. Effect of Im. The pier is also liable to a horisontal press- 
ure dne to floating ice. The formnlaa for impact are not applica- 
ble to this case. The assumption is sometimes made that the field 
of ice which may rest against the pier, will simply increase the sar- 
face exposed to the pressure of the carrent The greatest pressare 
possible will occnrwhen a field of ice, so large that it is not stopped 
by the impact, strikes the pier and plows past, crashing a channel 
throngh it eqnal to the greatest width of the pier. The resulting 
horizontal pressure is eqnal to the area crushed multiplied by the 
cmsbiag strength of the ice. I'he latter varies with the tempera- 
ture; bnt since ice will more down stream ih fields only when 
melting, we desire its minimum strength. The crushing strength 
of floating ice is sometimes put at 20 tons per sq. ft (300 lbs. per 
eq. inch); bnt in computing the stability of the piers of the St. 
Lonis steel-arch bridge, it was taken at 600 lbs. per sq. inch (43 
tons per sq. ft.). According to experiments made under the 
author's direction,* the crushing strength of ice at 23° P., varies 
between 370 and 760 lbs. per sq. in. 

Occasionally a gorge of ice may form between the piers, and 
dam the water back. The resulting horizontal pressure on a pier 
will then be eqnal to the hydrostatic pressure on the width of the 
pier and half the span on either side, dne to the difference between 
the level of the water immediately above and below the bridge 
opening. A pier is also liable to blows from rafts, boats, etc ; but 
as these can not occur simnltaneously with a field of ice, and will 
probably be smaller than that, it will not generally be necessary to 
consider them. 

A lateral pressure on the pier is possible, due to the earth's be- 
ing washed away from one side and not from the opposite. It will 
be on the safe side, and near enough for this purpose, to assume 
that this effect is equal to the pressure of a liquid whose density is 
the difference between that of the water and the saturated soil dis- 
placed. Under these conditions, the actual tendency to slide is 
• The TacHHoaRlPH, Ualvenlty of IHinols, No. 9 (1894-96). pp. 88-48. 



ovGoQi^lc 



ART. 1.] THEORY OF STABILITT. 369 

«qnal to the square root of the sum of the squares of the down-stream 
lorces and the lateral thrnst. However, this refinement is unnecee- 
-saty, particularly since a pier which is reasonably safe against over* 
taminET ond crushing will be amply safe against eliding. 

671. Seiuting Force*. The resisting force is the friction due to 
the combined weight of the train, bridge, and the part of the pier 
above the section considered. For the greatest i-efinement, it would 
be necessary to compute the forces tending to slide the pier for two 
conditions : viz., (1) with a wind of 50 lbs. per sq. ft. on truss and 
pier, in which case th<! weight of the train should be omitted from 
the resisting forces ; and {'i) with a wind of 30 lbs. per sq. ft on 
truss, train, and pier, in which case the weight of a train of empty 
■box cars should be included in the resisting forces. For a table of 
we^hts of maeonry, see page 200. If the water can find its way 
under the foundation in thin sheets, the weight of the part of the 
pier that is immersed in the water will be diminished by 62^ lbs. 
per cu. ft. by buoyancy ; but if it finds its way under any section 
by absorption only, then no allowance need be made for buoyancy. 

The resisting force is equal to the product of the total weight 
and the co-eflBcient of friction. For values of the co-eflBcient of 
friction, see the table on page 315. The tenacity of the mortar is 
usually n^lected, although it is a \evj considerable element of 
strength (see § 137). 

672. SluniTT AQADTBT OviBTtrxitnro. The forces which tend 
to produce eliding also tend to produce overturning, and the forces 
■which resist sliding also resist overturning ; hence, there remains to 
determine only their points of application. The stability can be 
■determined either by moments or by resolution, as was explained for 
dams ; but in this case, it is easier by moments, since there are sev- 
eral horizontal forces, and it requires considerable work to find their 
resultant as demanded by the method by resolution of forces. 

673. A. By Xoments. By this method, it is necessary to find 
the arm of the forces, i.e., the perpendicular distance from the line 
of action of the forces to a point about which the pier tends to turn. 
This is the same method as that used in §§ 493-98, which see. 

The center of pressure of the wind on the truss is practically at 
the middle of its height ; that of the wind on the train is 7 to 9 
feet above the top of the rail ; and that of the wind on the pier ie 
»t the middle of the exposed part The arm for the pressure of the 



ovGoQi^lc 



370 BRIDGE PIEBS. [CHAP. XTU 

ice abould be measured from high water. The center of pressnre- 
of the ciirreDt ia not easily determiaed, since the law of the Taria- 
tion of the velocity with the depth is not Icuovn ; but it will proba- 
bly be safe to take it at one thinl the depth. Finally, the downward 
forces will usnally act vertically throagh the center of the pier. 

From these data the overturning and resisting moments am. 
easily be computed. For equilibrium, the summation of the former 
must be less than the latter. The above principles will be further 
elucidated in g§ 579-8l> by an example. 

674. B. By BeiolDtlon of Foroet. This is the method explained 
in g 499 (page 3^0). In that case the problem was ^ery sim- 
ple, since there were but two forces ; but in the present case there 
are several horizontal forces and also several vertical ones. The first 
step is to find a single force which is equivalent in every respect to 
the combined effect of all the horizontal forces ; the second is to 
find an xijuiralont for all of the vertical forces ; and the third is to 
find the resultant of these two forces. 

The horizontal distance, x, of the point of application of the re- 
snltant of all the vertical forces, back from the toe of the pier, is 
found by the equation, 

_ gum of the momenls of the vertical forces . , 

~ sum of the vertical forces 

The weight of the train and bridge act vertically through the center 
of the pier ; and if the pier is symmetrical, as it usually is, the 
weight of the pier will also act through its center. Therefore, x in 
equation (3) will usually be half the length of the pier. 

The vertical distance, y, of the point of application of the re- 
sultant of all the horizontal foi'cos above any horizontal joint is 
found by the equation, 

_ sum of the moments of the fiomoiifttl forces ,„. 

sum of the horizontal forces ' " ^ ' 

Having found x and y, as above, draw a vertical line at a distance 
* back from the down stream end of the pier ; on this line lay off a 
distance y above the horizontal joint under consideration. The 
point thns determined corresponds to « of Fig. TO (page 3'iO). Con- 
struct the parallelogram of forces by laying off, to any convenient- 



ovGoQi^lc 



ART. i-l THEORY OF STABILITY. 371 

scale, (1) a. horizontal line equal to the sum of all the horizontal 
forces acting on the pier, and (2) ii vertical line equal to the sum of 
all the vertical forcfs ; and complete the diagram by drawing the 
reBultaot. The stability of the pier is determined by the ratio of 
A C to iV C, Fig. 70. 

S7&. StaBILITT AGAIKBT CBITSHIItO. HepreBont the maximum 
pressure by P, the total weight on the section by W, thearea of the 
section by S, the moment of inertia of the section by /, the length 
of the section by I, and the overturning moment by M; then from 
equation (1), page 305, we have 

For the particular case in which the pier has a rectangular horizon- 
tal cross Ecctlon, the above formula becomes the same b& equation 
(18), (page aJ2,) as deduced for an element of a masonry d&m. 

The method of applying the above equation vrill be explained in 
§ 581 by an example. 

576. ExAXFLE OF Method or Cokfutiito 8tabilitt. Fig. 85 
shows the dimensions of the channel pier of the Illinois Central R. 
R. bridge over the Ohio River at Cairo, 111, This pier stands be- 
tween two 533-foot spans. Its stability will now be tested by the 
preceding principles. 

677. Stability against Sliding. Wo will examine the stabil- 
ity against sliding on the top footing course. The wind surface of 
the truss = 10 sq. ft. X 523 = 5,230 sq. ft. The wind pressure 
(^inst the truss at 30 lbs. per sq. ft. = 30 lbs. X 5,230 =156,900 
lbs. -= 78 tons ; and the wind pressure on the truss at 50 lbs, = 
50 lbs. X 5,230 = 261,500 lbs. = 131 tons. 

The wind pressure on train at 30 lbs. per sq. ft. ,= 30 lbs. X 
533 X 10 = 156,900 lbs = 78 tons. 

The pressure of the wind against a section of the pier 52 fL 
long, is 20 lbs. x 52 X 14 = 14,560 lbs. = 7 tons. 

The pressure doe to the ice is found as follows: Assume the 
thickness to be 1 foot , and also assume the crushing strength of 
ice to be 200 lbs. per sq. in. =, say, 15 tons per sq. ft. The pier is 
16 ft. wide at the high-water line. Hence the resistance required in 
the pier to crash its way through a field of ice is IS tone X 16 X 1 
^ 240 tons. 



ovGoQi^lc 



BBIDOE FIEBS. [CHAF. STI, 




ovGoQi^lc 



ART. 1.] THBOBT OP STABILm. 873 

The preesore due to the current is fooiid as follows: From 



19 ft. = 1,330 sq. ft., which valno is equivalent to asanming that 
the riTer may scour to the top of the footing conrsea. * represents 
a co-efficient, which, if the pier were rectangular, would be about 
1.4, and if the pier were cylindrical would equal about 0.73. We 
will asBume it to be 1.1, — a trifle more than the mean of tbese two 
Talues. to = 62.5 lbs. per cu. ft. The surface velocity at the 
bridge site was measured* " when the Mississippi and the Ohio 
were at about the same stage," and found to be 4 miles per hour 
(= 6 ft per second); but as high water may occur in the Ohio at 
the time of moderately low water in the Mississippi, the possible 
mazimnin velocity is greater than the above, and bence we will as- 
sume that it is 10 ft per second. The velocity of the water at 
various depths below the surface of a stream varies as the ordinate 
of an ellipse; but the effect of the mean velocity is approximated 
with sufficient accuracy for this purpose by assuming that the mean 
pressure is half of that due to the surface velocity. Substituting 
these numbers, the above equation becomes P = 1,330 x 1.1 X 
6S.5 X Vtf = ''0-5 tons = 70 tons with sufficient accuracy. Divid- 
ing this by 2 to get the pressure corresponding to the mean velocity* 
we have the pressure of the current equal to 35 tons. 
<Jollecting the preceding results, we have: 

Wind Ml tnuB, 78 tons. 

" " Inia, 78 " 

" pier, 7 " 

Pressure of ice, 8*) " 

" " water 86 " 

Total force tendlcg to slide the pier od the foot- 
ing = 488t<»w. 

578. The weight of the bridge will be assumed at 3 tons per 
lineal foot; and hence the total weight is 2 tons X 533 = 1,046 
tons. 

The weight of a train of empty cars is about 0.5 ton per lineal 

• Thlid Atmnal Report of the Uhnoii Sodet; of Eagliiaeia, p. 78. 



ovGoQi^lc 



374 BBIDOE PIEBS. [CHAP. SVt. 

foot; and hence the total weight of the train ie 0.5 tons x 523 = 
261 tons. 

The amonnt of masonry below the high-water line = 67,946 ca. 
ft.; the amount above the high water lino = 24,534 en. ft.; and 
hence the total masonry = 93,480 cu. ft. We will assume the 
weight of the masonry to be 150 lbs. per cubic foot. Then the 
weight of the masonry is 150 lbs. X 93,480 = 6,936 tons. 

OoUecting these results, we have: 

Weight of the bridge 1,046 tona. 

" " " train of empty cars, 361 " 

" " " masonry, 6,986 " 

Total weight to resUt sliding = S,248 toua. 

Sliding cannot take place, if the co-eEBcient of friction ie eqnal 
to or greater than 438 -i- 8,243 = 0.053. For values of the co-ef- 
ficients of friction, see the table on page 315. In the above ex- 
ample, .the factor of safety against sliding is at least 12 to 15. 

679. Stability against Overtaming. We will coneider the 
stability against overturning about the top of the upper footing 
course. The wind on the trass = 78 tons; the arm of this force = 
height of the pier {Vid it) + half the depth of the trass {ZO ft.) = 
153 ft.; and therefore the moment of this force = 78 tons X 153 
ft. = 11,934 foot-tons. 

The pressure of the wind on the train = 78 tons; and the arm 
of this pressure = distance from footing tn top of pier (123 ft.) -|- 
dislance from top of pier to top of rail (8 ft.) -\- distance from top 
af rail to center of train {8 ft.) = 139 ft. Therefore the moment 
at this pressure is 78 tons X 139 ft. = 10,842 foot-tons. 

The pressure of the wind against the pier is 7 tons (§ 577); the 
arm of this force = i (302 + 150) — 79 = 97 ft. ; and the moment of 
this force = 679 foot-tons. 

The pressure of the ice is 340 tons, the arm is 70 ft., and the 
moment is 16,800 foot-tons. 

The pressure of the water is 35 tons. The center of pressure 
lies somewhere between one third and one half of the depth from 
the top; and-as the increased area at the base of the pier compen- 
sates in part for the deerease of velocity with the depth, we will as- 
sume that it is at half the depth. The arm then is 36 ft., and the 
moment is 35 tons x 36 ft. = 1,260 foot-tons. 



ovGoQi^lc 



ABT. 1.] THEOKT OF STABILITY. 

Collecting theBe resnlta, tee have: 
it of the wind on the trun. 



" " pressure of the ice, 



Total OTertumIng momeot . 



11,884 fc- 


Dt-tons. 


. 10,842 




679 




. 16,800 




, 1,260 




= 41,015 foot-b>D>. 



580. The total weight aboTO the joint considered is (§ 678) 
S,343 tons. This force acts vertically down through the center of 
the pier; hence the ami iB 3I.fi ft., aad the total moment resisting 
OTertuming is 8,343 X 31.5 = 359,654 foot-tons. The factor of 
safety against overturning about the top of the upper footing 
course ia 259,654 -f- 41,515 = 6.3. 

Assuming the traiu to be o9 the bridge, and that the vind 
pressure on the tmsB is 50 lbs. per sq. ft., and following the method 
pursned above, it ie found that the factor of safety against over- 
turning under thnse conditions is 6.4. 

581. Stability against Cnuhing. The maximum pressure on 
the section will occur when the loaded train is on the bridge and 
all the horizontal forces are acting with their full intensity. The 
load when an evipty train ia on the bridge is (§ 578) 8,243 tons. 
Assuming that a loaded train will weigh li tons per lineal foot, we 
must add (0.75 tons x 533 =) 392 tons to the above for the 
-difference between a loaded and an unloaded train. Then the total 
■direct pressure is 8,343 + 392 = 8,635 tons. The area of the sec- 
tion at the top of the footing course is 1,160 sq. ft. Hence, the 
maximum direct pressure is 8,635 -;- 1,160 = 7.4 tons per sq, ft. 

The moment to overturn, M, = 41,515 foot-tous. The greatest 
length of the section = 63 ft. The moment of inertia of the sec- 
tion about an axis through its center and perpendicular to ita 
length = 287,917 (ft.). Prom § 575, the maximum pressure 

s ^si 

Substitating the sbore qaaotities in this equation givee 

•^ = '■* + Iw^sn = '•* + ■'■5 = llil '!»» pel •!■ ft- 
Since it is highly improbable that all the forces will act at the 
«ame time with the intensity assumed in the preceding compnto- 



ovGoQi^lc 



3?t! BBIDQB PIEBS. [CHAP. ZVt^ 

tioDB, we may eonolade that tbe presenre will never exceed 11.9 
tons per Bq. ft. A compariBon of tbia with the Taluee of the com- 
pressive strength of masonry as given in g 233 (p^^ l^^) shows 
that this presBure is entirely safe. 

Since this is ao unusually high pier under an nnnsnally long 
span, and since the overturning and resisting moments and also the 
top dimensions of the pier vary with the span, we may draw the 
conclusion that any pier which has sufficient room on top for the 
bridge seat (§ 584) and which has a batter of 1 in 13, or 1 in 34, it 
safe against any mode of failure. 

683. FreniiTe on the Bed of the Foondation. The caisson, 
is 70 feet long, 30 feet wide, and 50 feet high. The load 
on the base is eqnal to the weight on the top of the footing ptttt- 
the weight of the footings plus the weight of the caisson. 
The weight above the footing = 8,636 tons (§581). The weight- 
of the footings = 1,300 sq. ft. X 4 ft. X 150 lbs. = 390 tons. Tbe. 
weight of the caisson = 70 ft. X 30 ft. X 50 ft. X 100 lbs. = 6,360 
tons. The total weight on the bed = 8,635 + 390 + 6,250 = 14,- 
275 tons. Tbe area = 70 ft. x 30 ft. = 3,100 sq. ft. The direct 
pressure per nuit of area = 14,375 -i- 3,100 = 6.8 tons per sq. ft. 

The overturning moment, M, is eqnal to the moment about tb&; 
top of the footing (g 661) plus the product of tbe sum of the bori- 
tontal forces and the distance from the footing to tbe base of the 
caisson; or, tbe moment about the base = 41,515 foot-tons -)- 438 
tons X 54 ft. =: 65,167 foot-tons. The moment of inertia, I, = 
^30(70)' =857,500 (ft.), i = 70 ft. The concentrated preaaare 
caosed by the tendency to overtam is 

Ml 65,167 X 70 „„^ 
3T = 2^857:500 =*-^'«^ 

The caisson was sunk all the way through, and rests, on sand ; 
consequently the water will find its way freely under the entire 
foundation, thus causing bnoyaucy to act with its full force. This 
upward force of the water will be equal to the volume of the im- 
mersed masonry multiplied by tbe weight of a cubic foot of water;, 
or tbe buoyancy = (67,946 + 5,300 + 105,000) X 62.4 = 5,558 tons. 
The lifting effect of buoyancy is (5,558 -r- 2,100 =) 2.62 tons per 
sq. fi 

Therefore, the total pressure is not greater than 6.8 + 8.7 — 3. ft 
, = 6.9 tons per sq. ft. 



ovGoQi^lc 



AET, 3.] DETAILS OP COHSTBUCTION. 877" 

The preaanre voaM oever bo so mach, for the following reaeona : 
1. There is no probability th&t both spAas will be covered hy s train 
of msximnm weight at the eame time that the mssitnum effects of 
the wind, of the current, and of the ice occnr. Z. The friction oa 
the sides of the caisson will sustain part of the load. A friction of 
600 lbs. per sq. ft., which was about the amount experienced in 
sinking these piers (see g 455), wonld decrease this pressure about. 
1^ tons per sq. ft. 

Therefore, we conclude that the pressure on the sand will be at. 
least as tnuch as 6.8 — 1.5 — 2.6 = 2.7 tons per sq. ft.; and that it. 
may possibly, but not probably, amount to 6.8 + 3.7 — 3.6 — 1.5 = 
5.4 tons per sq. ft. The larger value was taken at the greatest pos- 
sible one for the sake of establishing the conclusion stated in th» 
last paragraph of g 581. 

683. Other Examplta. At the St. Louis steel-arch bridge- 
tfae greatest pressure possible on the deepest foundation (bed- 
rock) is 19 tons per aq. ft. The pressure at the base of the 
New York tower of the East River suspension bridge is about. 
7^ tons per sq. ft. upon a stratum of sand 2 feet thick overlying^ 
bed-rock ; and at the base of the masonry the pressure is about 11^ 
tons per sq. ft.* The corresponding qnantities for the Brooklyn 
tower were a little over a ton less in each case. At the Plattamonth 
bridge f the maximum pressure caused by the weight of train, bridge, 
and pier ia 3 tons per sq. ft, At the Bismarck bridge f the pressure- 
due to the direct weight is 3 tons per sq. ft. on clay. 

Aet. 2. Details of CoHSTECcnoN. 

684. TOI DncIHBIOni The dimensions on the top will depend 
somewhat upon the form of the cross section of the pier, and also 
upon the style and span of the bridge; but, in a general way, it may 
be stated that, for trussed spans of 100 ft. or over, the dimenaiona 
under the coping will not bo less than 5 ft. X 20 ft. ; for 250-ft. 
spans, 8 ft. X 30 ft.; and for 500-ft spans, 10 ft. X 40 ft. Appar- 
ently 6 ft. X 22 ft. nnder the coping is the favorite size for spans of 
100 to 200 ft 

* 7. ColltDffwood, Mristant engineer, In Tan Noitrand'i Engfn'K >U8-> voL xvL 
p. «S1. 

t Beport of Oea B. Koricon, chief engineer. 



ovGoQi^lc 



3*8 BBIDOE PIEES. [CHAP. XVI, 

5S5. BOTTOX SuiEltslolts. Theoretically the dimeusioDsat the 
bottom are determined by the area necessary for stability; bnt the 
top dimensions required for the bridge seat, together with a slight 
batter for the sake of appearance, gives sufficient stability (g 581). 
Only high piers for short spans — a combination not likely to occur 
in practice — ^are liable to fail by overturning or crushing. 

686. Batteh. The nsnal batter is 1 inch to a foot, althoagh l 
-an inch to a foot is very common. In high piers it is customary to 
use a batter of 1 to 24, and offset the masonry and introduce a water- 
table at the high-water Hue, so as to give an average batter of about 
1 to 13. This construction very much improves the appearance, 
«nd does not add materially to the cost. 

A corbel course, or "belt course," is sometimes introduced im- 
mediately under the coping for appearance's sake. For an exam- 
ple, see Fig. 85 (p^e 372), Fig. 87 (pa^ 383), and Fig. 88 (page 
384). 

6S7. Cbobb Sectimi. The up-stream end of a pier, and to a 
considerable extent the down-stream end also, should be rounded 
or pointed to serve as a cut-water to turn the current aside and to 
prevent the formation of whirls which act upon the bed of the 
«tream around the foundation, and also to prevent shock from ice, 
logs, boats, etc. In some respects the semi-ellipse is the best form 
for the ends ; but as it is more expensive to form, the ends are 
usually finished to intersecting arcs of circles (see Figs. 85, 87, and 
89 — pages 372, 383, and 385, respectively), or with semi-circulai 
ends. Above the high-wat«r line a rectangular cross section is as 
good as a curved outline, except possibly for appearance. 

A cheaper, but not quite as efficient, construction is to form the 
two ends, called starlings, of two inclined planes. As seen in 
plan, the sides of the starlings usually make an angle of about 45° 
with the sides of the pier (see Fig. 90, page 386). A still cheaper 
construction, and the one most common for the smaller piers, is to 
finish the up-stream end, below the high-water line, with two in- 
clined planes which intersect each other in a line having a batter of 
from 3 to 9 inches per foot, and build the other three sides and the 
part of the up-stream face above tJie high-water line with a batter 
cf 1 in 12 or 1 in 24. Of course the simplest construction is to 
make the pier rectangular in horizontal cross sections and give it the 
■same batter on ill faces. 



ovGoQi^lc 



ART. 2.] DXTAIL8 OF COKSTBUCTION. 379 

Occasionally, tor economy, piere, particularly pivot piers, are 
built hollow — sometimea with and sometimes without interior cross 
walls (see Fig. 86, page 380). The piers of the bridge across the 
' Missouri River at Glasgow, Mo., are solid ap to the high-water line, 
and above that each pier consists of two stone columua. The piers 
of the bridge over the Missouri at St. Charles, Mo., have a somewhat 
similar construction, except that the secondary piers are connected 
by a comparatively thin wall. 

With piers subjected to a severe pressure from ice, it is customary 
to protect the edge of the noee with an angle-iron or a railroad rail. 

689. Pivot Fixbs. These differ from the ordinary piers only 
in that they are circular, are larger on top, and have plumb sides. 
Pivot piers are about 25 to 30 feet in diameter, under the coping, 
for spans of 250 to 360 feet, respectively. 

Fig. 86 shows the pivot pier for the Northern Pacific R. R 
bridge over the Bed River at Grand Forks, Dakota. The specifica- 
tions for the grillage were as follows: "Fasten the first course of 
timbers together with f-inch x 20-inch drift bolts, 18 inches apart; 
iasten second course to first course with drift bolts of same size at 
every other intersection. Timbers to be laid with broken joints. 
Pat on top course of 4-inch X 12-inch plank, nailed every 2 feet 
with jV'i^'^'i X S-inch boat spikes. The last course is to be thor- 
oughly calked with oakum." 

Pivot piers are protected from the pressnre of ice and from 
shock by boats, ejc, by an ice breaker which is entirely distinct 
from the pier. The ice breaker is usually constructed by driving a 
group of 60 or 70 piles in the form of a V (the sharp end np stream), 
at a short distance above the pier. On and above these piles a 
strong timber crib-work is framed so as to form an inclined ridge 
up which the cakes of ice sjide and break in two of their own weight 
Between the ice breaker and the pier two rows of piles are driven, 
on which a comparatively light crib is constrncted for the greater 
security of the pier and also for the protection of the river craft. 

68S. QlFALITT Ot HABOITBT. Bridge piers are usually quarry- 
foced ashlar, i. e. , first-class masonry (see § 207) backed with rubble. 
Good concrete, if made with reasonable care, is equally aa good as 
ordinary rubble masonry, and is sometimes cheaper, — since it affords 
an opportunity to use up the refuse from the quarry. 



ovGoQi^lc 



BBIDOB PIERS. [CHAP. Z7I. 



7crib;G00l^lc 



ABT. 2.] DETAILS OF CONSTRL-CTIOS. 381 

For an iUuBtrated description of the method of building concreta 
bridge piera, see Engineering News, vol. xix. pp. 443-44. 

690. BPlcincATlolls. The follnwing specificationa for the ma- 
sonry of the railroad bridge over the ^f iseouri River near Sibley, Mo., 
^Octave Channte, engineer) may be taken as an example of the best 
practice.* 

S81. 0«a«rftl aaqnlnmuit*. " The stone to be used Id tbeee pieia must be 
of what is kuowD M the best qualltj of Cottonwood llmeHtoiie, or otber stoue 
wbicb. Id tbe opiuioa of the eugineer, Is of equally good quality and in every 
way suil&ble for the purpose for which it is to be used. It must be a oucd and 
durable, free from all drys, Bhake s, or flaws of any kind »h»teser,-end must 
be of such a character as will, in the opTnTou of the engineer, withstand the 
ac tion o f the w eather. No alone o( an ipferior quality will be accepted or 
«veD permitted to be delivered upon tlie ground. The masonry in the bridge 
piers must be of tbe Ijest ^nd larg est atones that the qunrry will afford, and 
muat l>e quarried in time to aeoaoii again a t froat before being used. 

" The face stonescompoaing the starling, and the ends and aides of the river 
piera from the neat line alxiut low water up for a distance of twelve (18j feet, 
Uid also the pedestal blocks of the main piera will be of Uinneaota granite, 
or a granite of equal quality approved by the engineer. ' 

" All masonry of the main piers shall be regular coursed ashlar of the best 
ieacrlptioD, and muat be laid in mortar of the proportions of sand and cement 
hereinafter specified. 

" All stones must tie so shaped that the bearing beds abail be parallel to tbe 
natural beds, and be prepared by dressing and bammering before they are 
brought on tbe walls, as tooling and hammering will not be allowed after the 
Stones are in place. They are to be laid to a firm bearing on their natural beds 
In a full bed of mortar, without the uae of chips, pinnera, or leveiers. No 
shelving projections will tw allowed to extend beyond the under bed on either 
side. The stone and work are to be kept free from all dirt that will interfere 
with the adhesion of mortar. Stones must be sprinkled with water before 
being placed in poaiiton on ihe wall. In laying stone in mortar, their beds are 
to be so prepared that when settled down they may rest close and full on tbe 
monar. In handling tbe stones care must be used not to iujure the joints of 
those already laid; and in case a stone is moved after being set and tbe joint 
broken. It must be taken out, tbe mortar thoroughly cleaned from tbe beds, 
and then reset, 

" Wherever the engineer shall so require, atones shall have one or two 1^- 
tnch iron dowels passing through them and iuto tbe stones below. The holes 
for tbe dowels shall be drilled through such stones Ijefore they are put In 
position on the walls. After the stones are in place the boles shall be con- 
tinued down into the under stones at least six (6| inches ; tbe dowel pins will 
then be set In and tbe holes filled with neat cement grout. Cramps binding 

♦ywpeeifleaaotMforflrst-clawmBSOPrr, seejao?; see also Appendix! 



ovGoQi^lc 



BSIDOE FIEBS. [cHAP. XTI. 



the seTeral atones of a course togeiber may be inserted when required by the 
eDgineer ; !d such caae tbey will be couater-suck into tbe stones wtiicb they 
fastCD together. 

S0S. 7km ftonai. " Tbe face stones must be accurately squared, jointed, 
and dreaacd uii their beds and builds ; and tbe Joints must be dressed back at 
least twelve inches (13j from tbe face. FacestoueBare to be brought to s Joint, 
wheo laid, of not more IhaQ three quarters (}) of an inch nor less than one 
half (ii inch. The courses shall not'be leas than eighteen (18) inches In thick- 
ness, decreasing from bottom lo top of the wall. Courses to be well bonded. 
The face stones shall break joints at least twelve (12) inches. The face stones 
may be left rough, except the stones forming the starting, which must be care- 
fully dressed to a uniform surface. The eAgts of face stones shall be pitched 
true and full to line, and on corners of all piers a chisel draft one and a half 
(li) inches must be carried up from base to tbe under side of the coping. No 
projection of more than three (3) inclii^s from the edge of face stones will be 
allowed. No s-toue with a hollow face will be allowed in the work. 

69S. Stntoher*. "Xacb stretcher shall have at least twenty (SO) Incfaeb 
width of bed for all course? of from eighteen (IB) to twenty (20) inches rise, 
and for all thicker courses at least as much bed as rise ; and shall have an 
averse length of ai least three and one half (8^) feet, and no stretcher aliall be 
less than three (8) feet in length. 

SH. Hsadan. "Each header shall bavc a width of not lees than eighteen 
Inches (IS) and shall hold, back into the heart of the wnll, tbe size that It shows 
on tlie face. The headers bLhU occupy at least one fifth 1^^ of the whole face 
of the wall, and be, as nearly as practicable, evenly distributed over it, and be 
80 placed that tbe headers in each course shall divide equally, or nearly so, the 
■paces Iwlween the headers in the course directly below. In walls over six 
feet vO)iu thickness, the headers shall in no case be leas than three and out' lialf 
feet (3J) long; and in walls over nine (9) feet thick, the headers shall be equal 
in length to one third the thickness of the wall, except when this length of 
bender exceeds six (6) feet.^uo header over e\x (6) feet long being required. 

S9B. Ba«ldng. " Tbe headers must alternate front and back, and their 
binding effect be carried through the wall by inleniiediale stones— not less in 
length and thickness than the headers of tbe same course^laid crosswise !□ 
the interior of the wall. Tbe slrelcbers and all stones In tbe heart of the wall 
shall be ot the same general dimensious and proportions us the face stones, 
aud shall have equally good bed and bond, but may have less nice vertical 
Joiola,— although no space greater than Ave (5) inches in width shall be left be- 
tween stones. All stones in the backing must be well fitted to their plAces, 
and carry the course evenly quite through the wall. 

5M. Coping. " The tops of the bridge piers, cap stones of tbe pedestals, 
and such other ports of the masonry as the engineer shall direct, shall be cov- 
ered with coping of such dimensions as prescribed. A.1I coping stones shall 
be neatly bush-hammer dressed on the face, bed, top, and Joints; and shall be 
well and carefully set on the walls, brought to one quarter (i) inch joints, and. 



ovGoQi^lc 



ABT, a.] DETAILS OF CONSTRUCTION. 



Tia. ffl. — Sbore Pnn. Buir Bridoi. 



jvGooi^le 



SSi BBIDGE FIEBS. [chap. ZTT. 

If required, be doweled, the dowels being well secured in twd to the coping 
iritlt grout. No coping Btone sLall be Icbs (ban nine (9| equare feet on top, 

097. FolDtimg. " All maBOnr)' is to be pobited bo as to fill the Joints solid, 
fbc surface of the wall is to be scraped clean and the ^ints freed of all loose 
raortar aad refilled solid by using proper ramming tools. Joints must be well 
wet before being pointed. Hortar used in pointing is to be composed of one 
part Portland cement and one part sand. 

SOS. Osnient. " llie cement used iti the work shall be equal in quality to 
the best brands of Milwaukee or Louisville cement, and shall be ground so 
tbnt at least 90 per cent. In weight will pass a standatd sieve of 3,500 mesbes 
(c the square Inch, iind shall have a lensile strength— afler being exposed one 
hour, or until set, in air, and the balance of the twenty- four hours in water not 




Tia. 88.— Top or Piut, Hi 



TwIowW" F— of at least 40 pounds per square inch; and, after being exposed 
-one day in air and six days in water, from 60 to 100 pounds per square inch. 

" All cemenia shall be furnished by llie contractor subject to approval by 
the engineer. The contractor shsU provide a suitable building for storing the 
cement, in which the same must be placed before being tested. The engineer 
shall be notified of tbe receipt of cement at least three days before it is required 
for use. and the inspector may take a sample from each package for testing. 

600. Xortar. "The mortar shall be composed of tbe above cement and 
clean, dry. sharp sand Id the proportion of one part cement to two parts of 
sand by weight.* The sand and cement shall be Ihoroughly mixed dry, and, 
after adding sufficient water to render Ibe mass plastic, shall be mixed and 
worked until of uniform consistency throughout. 

" Hortar remaining unused so long as to have taken an initial set shall not 
be used in the work. 



ovGoQi^lc 



a-] 



DETAILS OF COKSTRUCTION. 



eoo. FtdMtel KaMmTj. " The pedestals shall be founded upon t, bed of 
'Concrete or upon piles, ae may be directed by the engineer. The masonry In 
the pedestals shall be of the best de- 
scription of couiwd aahhu- composed of 
the ItmeHtooe and the mortar described 
above, the stimea U> be not less than 
twelve (13) inches thick, and to have 
horizontal beds and vertical Joints on 
the face. When the walls do not ex- 
ceed three and one half (81) feet in 
thtclcnesB, the headers shall run entirely J 
through, or a single stone — square and 
of the proper thichuen— may be used. 
In walls over three and one half (8^) 
feet in thickness, and not over seven 
<7) feet in thiobnesi, headers and 
^stretchers shall alternate, and there 
mtuil be aa many headera aa stretchers. 
The space in the Interior of the walls 
■hall be fliled with a single stone 
to fit such space, and said stone shall 
be of the same heiglit as the headers 
»ndatretchenof thecourse. In all the 
masonry of these pedestals the slope 
must be carried up by steps and in a 
-cordance with the plans of theengineer. 
All the quoins must have hammer- 
dressed beds, builds, and joints, and 
draft comets." 



601. •gfSlfPT.M ffg 

Puss. Fig. 8S (page 372) shows 
the channel pier of the IHinois 
Central B. B. bridge over the 
Ohio at Cairo, IlL 

Fig. 86 (page 380) shows the 
pivot pier of the Northern Pacific 
B. R. bridge over the Red River 
■at Grand Forks, Dakota. 

Fig. 87 (page 383) shows one 
of the two shore piers of the 
bridge over the Missouri River, 
near Blair, Neb.* This pier stands between two 




)-ft. spans. 



• From tlie Report of Geo. 8. MorUon, chief engineer of the bridge. 



ovGoQi^lc 



BSIDQE PlEBa [CHAP. XT^ 



"aK. 't'<.'— -I 



Fn. (0.— Pin or Bi. Cwuz Kitib Bi 



jvGooi^le 



ABT. 2.] DETAILS OP CONBTEUCTION. 387 

" The Tertical joints are shown as they actually are in the stmot- 
nre." The masonry ia 145 It from top to bottom. 

Fig. 88 (page 384) shows the top of the pier between two 525-ft 
channel spans of the LouisTille and Nashville B. R bridge across 
the Ohio River at Henderson, Ey. 

Fig. 69 (page 385) shows the actual arrangement of the stones in 
one of the courses of one of the channel piers (Fig. 86) of the Blinois 
Central R. R. bridge over the Oliio River, at Cairo, JR. 

80S. Fig. 90 (page 386) shows the river pier of the Chicago, Bur- 
lington and Northern R. R. bridge across the St. Croix River. This 
pier stands between a draw of 370 feet and a fixed span of 153 foet. 
The thickness of the courses is as follows, in order from the bottom 
ap : Two conrses, including the footing, 28 inches; two 26 inches ; 
one each 34, 22, 21, 19, and 17 inches; two 15 inches; four 14 
inches ; one 13 inches ; one 13 inches ; and the coping 18 inches. 

The following table givestheqnantityof masonry in the pier and 
illnstrates the manner of computing the contents of such stractnies. 
Notice tha* the order in the table is the same as that in the pier ■ 
1.8., the top line of the table relates to the appermost masonry, etc. 

TABLE 40. 
CoNTBSTS or THE PiER sEowH EH FiQ. W (page 8S6). 



Btringer ResU. 
Bridge Seaia. . . 
Coping. 



ax3.75'X8.(r XS-lff... 
8 X 3.75' X B-ff Xl.«'.. 
7.6' X ai-C X 1.5' 



Ice Bnaker. . . . 
Footing Course. 



f(3xM' + 8.6)88' + cax8.6' + «.5')ae.l'l~-. 

(3x8.a' + 7.1')8.8'xi^ 

, [8.6'x8.«)(— -t-l.c) 

i(8.6X8.«'+4.8'x4.8-)iaiT 

B,6- X 89.4' X 3.88' 

" 4.8' X 8.88' 

Total = 880.89 cubic yards = 



608. Iron Tubular Pien. For a description of an iron tubular 
pier, see g 415 ; and for a description of a pier founded upon screw 
piles, see Engineering Newa, vol. xiii. pp. 210-12. 



ovGoQi^lc 



388 BBIDGK PIEBS. [CHAP. XTI. 

604. Timber Barrel Piers. The Chicago, Burlington and Qnincy 
E. R. has constructed a few "barrel piers" as an experiment, the 
object being to reduce the cost of foundations, and also to find some 
cheap anbstitate foi masonry. The barrels are cjUndrioal, 8 feet ia 
diameter, and 30 to 30 feet in length. The stares are 10 inches 
thick, 8 inches wide on the outside, and are dressed to fit together 
to form a cylinder. Tlic staves are bolted at the top and bottom 
to two inside rings made of I-beams, aud are further held in place 
by strong outside hoops of iron. These caissons or barrels are sunk 
by excavating the soil from the inside. The bottom and top por- 
tions of the caisson are filled with concrete, and the intermediate 
portion with saud. On top of the wooden barrel, an iron frame is 
placed, upon which the truss rests. Two barrels constitute a pier. 
The advantages claimed for the wooden caissons are that they can 
be put in witho.ut interfering with traffic, or without loss of time in 
sinking by the passive of trains. The objection to them is that they 
Are not durable. 

605. COHTSirrB or B&IDOX FlESa. The table on page 389 gives 
the quantity of masonry in bridge piera having rectangular cross 
sections and such dimensions on top and hatters as occur most 
frequently (see g§ 584-67). The quantities in the first four columns 
cover most of the eases for highway and single track railway bridges; 
and the quantities in the last two columns are applicable to double 
track railway bridges. Since that portion of the pier below the 
water should have more or less pointed ends, and since there is 
likely to be an offset in the profile — particularly of high piers, — 
the quantities in the table (being for a rectangular cross section) are 
mainly useful in milking preliminary estimates. 

Thecontentsof piersof other dimensioni than those in the table 
may be computed by the following formula : • 

confenfs = thl + b(l + Ijh' + lil^h?, 
in which / = the length on top under the coping, 

t=" thickness on top under the coping, 
h = " height to the under side of the coping, 
5 = " batter — i. e., b = -^ or ^, 
The length on the bottom = I + 2bh; and the thickness on tho 
bottom = t-\-Zbh. To illustrato the method of applying this for- 



ovGoQi^lc 



ART. 2.] 



DETAILS OF CONSTKCCTION. 



I BkCTIOS AMD 



'S%- 




DnDDiwoN or TBI Pub oi 


«TOPU«.« 


IHSCOFDra. 




Fmotwo 


Btt. X 


-B. 


arc.. 


8»(C. 


8 ft.. 


Mft. 


noPIBB. 


BUUr 1 : IE 


Bauerl^iM 


B..^rni. 


B«HePl;M 


BMIerliU 


Batl«rl:M 


rati. 


CB-Vd.. 


cu.v<i<. 


c».Vd». 


™-lrfi. 


CU.IKI". 


». V<I>. 


s 


80. « 


19.49 


36.64 


35.58 


40.90 


89.88 





25.07 


23.68 


83.51 


80.90 


49.84 


47.67 


7 


20.68 


37.85 


88.57 


86.86 


59.06 


65.98 


8 


34.74 


83.18 


44.81 


41.90 


68.53 


64.48 


9 


89.84 


88.62 


51.24 


47.55 


78 28 


78.0* 


10 


46-09 


40.97 


67.86 


68 38 


88.33 


81.8ft 


11 


80.53 


45.61 


64.67 


69.09 


98.50 


90.88 


12 


66.14 


60.14 


71.69 


65.03 


1U9.03 


99.68 


18 


61.98 


54.85 


78.89 


71.03 


119.88 


108.83 


14 


67.01 


69.64 


86.81 


77.18 


180.90 


118.09 


15 


74.07 


64, S3 


93.93 


88.83 


142-25 


137.49 


16 


80,40 


69.48 


101.73 


89.61 


158-88 


187.08 


17 


86.98 


74.58 


109.75 


96.00 


165,79 


146.69 


18 


93.65 


79.66 


117.98 


103.49 


177.99 


186.49 


1» 


100.58 


S4.87 


136,43 


109.06 


190.45 


166.40 


ao 


ior.6« 


90.18 


186,07 


115.72 


208.22 


178.47 


31 


114.96 


95,57 


148 94 


123.49 


216.28 


188.67 


32 


133.46 


101 06 


158.01 


139.36 


229.80 


198.98 


38 


130,15 


106.83 


162,38 


136.34 


243.24 


207.46 


34 


138,04 


112.27 


171.84 


148.89 


257.17 


318.05 


25 


146,14 


118.03 


181.66 


150.58 


271-39 




36 


154,45 


128.86 


191.63 


157.79 


380.91 


239! 85 


27 


163,96 


139.79 


aoi,74 


165-17 


. 300.74 


2.W.67 


38 


171.69 


185,81 


212-16 


173 68 


[ 815,87 


381.62 


29 


180,62 


141.93 


332.79 


180.18 


881.27 


273.09 


80 


189.77 


148.13 


283.68 


187.85 


347,01 


284.51 


S3 


308-73 


160.81 


256 19 


208.47 


879.43 


807.78 


S4 


328.64 


178.86 


279.58 


319,53 


413.06 


881.69 


86 


848,36 


187.30 


303.98 


285,98 


447.99 


855.99 


38 


270.91 


301.13 


329.86 


2.*i2.84 


484.17 


480.92 


40 


293.47 


2I.i.33 


■855.74 


370.13 


■ 531.86 


406.43 


13 


316.98 


239.98 


388.17 


287.88 


560.61 


432.57 


41 


841.46 


344.91 


411.59 


306.03 


600.64 


469.32 


4e 


866.00 


260.29 


441.05 


324,60 


643.15 


486.47 


•«' 


8B!1.86 


3T6.09 


471.66 


843.66 


1 684,99 


S14.S3 


80 


420.83 


293.29 


503.83 


368.13 


739 34 


542-78 


53 


449.33 


308.90 


686.07 




774,88 


671.80 


M 


478.86 


335.93 


569.96 


403.45 


831.98 


801.47 


56 


509.45 


348.38 


604.96 


424.39 


870.45 


681.71 


58 


541.13 


361.34 


641.11 


445.57 


930.41 


663.37 


SO 


678.85 


879.62 


878.48 


467.43 


1 971.78 


694.02 



jvGooi^le 



390 BBIDOB PIBBS. [CHAP. XVL 

mala, aasume that it is required to find the coutenta of a pier 4 leet 
thick, 20 feet long on top, and 30 feet high, having a batter on 
all fonr faces of 1 inch per foot. Then I = 20, t = i, b = -^j and 
the preceding formula becomes 

eonienta = 4 X 20 X 30 + V. (20 + *) (30)' + | x yir X (30)' 
=4,450 cubic feet. 

606. Co§T. For a general discussion of the cost of masonry, see 
gg 226-38 (pp. 153-60) ; and for data on the cost of bridge pier 
maaonij, see § 235 (p. 157). 



ovGoQi^lc 



CHAPTER XVa 



Aet. 1. Watee Wat BEQUlBElk 

607. The determinstioD of the amount of water wayreqaired in 
007 given ca^ is a problem that does not admit of an exact matlie- 
matical Bolacion. Although the proportioniog of culverts is in a 
measure indeterminate, it demands an inbelliguit treatment. If 
the culvert is too small, it is liable to cauee a washout, entailing 
possibly loss of life, intermptions of traffic, and cost of repairs. 
On the other band, if the culvert is made unnecessaril; large, the 
■cost of construction is needlessly increased. Any one can make a 
culvert large enough ; but it is the province of the engineer tc 
design one of sufficient bat not extravagant size. 

BOB. The Faotobb. The area of water way required depends 
npon (I) the rate of rain-fall, (2) the kind and condition of the 
soil, (3) the character and inclination of the surface, (4) the condi- 
tion and inclination of the bed of the stream, (6) the shape of the 
area to be drained and the position of the branches of the stream, 
(6) the form of the mouth and the incliQation of the bed of the 
-culvert, and (7) whether it la permissible to back the water up above 
the culvert, thereby causing it to discharge nnder a head. 

1. It is the maximum rate of rain-fall during the severest stomu 
which is required in this connection. This certainly varies greatly 
in different aectiona ; but there are almost no data to show what it ia 
for any particular locality, since records generally give the amount 
per day, and rarely per hour, while the duration of the storm 
is seldom recorded. Further, probably the longer the series of 
observations, the larger will be the maximum rate recorded, since 
the heavier the storm the less frequent its occurrence ; and bence a 
record tor a short period, however complete, is of but little value 
in this connection. Further, the severest rain-falls are of company 
lively limited extent, and hence the smaller the area, the larger the 

Ml 



ovGoQi^lc 



898 CULVERTS. [chap, xvii^ 

possible maiimnm precipitation. Finally, the effect of the rain-fall 
in melting snow would have to be considered in determining the- 
mtuimnm amount of water for a. gi^en area, 

2. The amount of water to be drained off will depend upon the. 
permeability of the surface of the ground, which will vary greatly 
with the kind of soil, the degree of saturation, the condition of 
oultiTation, the amount of vegetation, etc. 

3. The rapidity with which the water will reach the water 
courses depends upon whether the surface is rough or smooth, steep 
or flat, barren or covered with vegetation, etc. 

4. The rapidity with which the water will reach the culvert, 
uepends upon whether there is a well-dedned and unobstructed 
channel, or whether the water finds its way in a broad thin sheet. 
If the water course is unobstructed and has a considerable inclina- 

' tion, the water may arrive at the culvert nearly as rapidly as it 
falls ; but if the channel is obstructed, the water may be much, 
longer in paesitig the culvert than in falling. 

5. Of course, the water way depends upon the amount of area- 
to be drained: but in many cases the shape of this area and the 
position of the branches of the stream are of more importance than 
the amount of the territory. For example, if the area is long and 
narrow, the water from the lower portion may pass through the 
culvert before that from the upper end arrives ; or, on the other 
hand, if the upper end of the area is steeper than the lower, the 
water from the former may arrive eimultaneouBly with that from 
the latter. Again, if the lower part of the area is better supplied 
with branches than the upper portion, the water from the former 
will be carried past the culvert before the arrival of that from the- 
latter; or, on the other hand, if the upper portion is better supplied 
with branch water courses than the lower, the water from the 
whole area may arrive at the culvert at nearly the same time. In 
lai^e areas the shape of the area and the position of the water 
■lourses are very important considerations. 

6. The efQciency of a culvert may be materially increased by so 
lOranging the upper end that the water may enter it without being 
retarded (see § 639). The discharging capacity of a culvert can 
also be increased by increasing the inclination of ite bed, provided 
the channel below will allow the water to flow away freely after 



ovGoQi^lc 



ART. 1.] WATER WAY REQUIRED. 393 

hariog passed the culTert. The last, although very important, ia 
frequeatly overlooked. 

7. The discharging capacity of a culvert can be greatly increased 
by allowing the water to dam up above it. A culvert will discharge 
twice as much under a head of 4 feet as undei'a head of 1 foot. 
This can only safely be done with a well-conetructed culvert. 

009. FOBKHLAB. The determination of the values of the difFer- 
ent factors entering into the problem is almost wholly a matter of 
judgment. An estimate for any one of the above factors is liable 
to be in error from 100 to 300 per cent., or even more, and of 
course any result deduced from such data must be very uncertain. 
Fortunately, mathematical exactness is not required by the problem 
nor warranted by the data. The question is not one of 10 or 30 
per cent, of increase ; for if a 2-foot pipe is insufficient, a 3-foot 
pipe will probably be the next size — an increase of 335 per cent., — 
and if a 6-foot arch culvert is too small, an S-foot will be used — 
an increase of 180 per cent. The real' question is whether a 3-foot 
pipe or an 8-foot arch culvert is needed. 

Numerous empirical formulas have been proposed for this and 
similar problems ; * but at best they are all only approximate, since 
no formula can give accurate results with inaccurate data. The 
several formulas, when applied to the same problem, give very 
discordant results, owing (1) to the sources of error already re- 
ferred to and (3) to the formulas' having been deduced for localities 
differing widely in the essential characteristics npon which the 
results depend. For example, a formula deduced for a dry climate, 
as India, b wholly inapplicable to a humid and swampy region, as 
Florida; and a formula deduced from an agricultural region is 
inapplicable in a city. 

However, an approximate formala, if simple and easily applied, 
may be valuable as a nucleus about which to group the results of 
personal experience. Such a formula is to be employed more as a 
guide to the judgment than as a working rule ; and its form, and 
also the value of the constants in it, should be changed as subse- 
quent experience seems to indicate. With this use in view, a few 
formulas will be referred to briefly. 

There are two classes of these formulas, one of which purports 

* For a general note ou empirical lonmilaa, «ee | m. 



ovGoQi^lc 



394 CDLVERTS. [chap. XVII. 

to give the quantitj of water to be diecharged per nnit of drainage 
area and the other the area of the water way in terms of the area of 
the territory to be drained. I'he former giTes the amoant of water 
sappoBed to reach the culvert; and the area, elope, form, etc., of 
the culvert mast be adjusted to allow this amount of water to pass. 
There are do reliable data by which to determine the discharging 
capacity of a culvert of any given form, and hence the use of the 
formulas of the first class adds complicatioa without securing any 
-compensating reliability. Most of the formulas in common use for 
proportioning water ways belong to this class. Such formulas will 
not be considered here. 

The two following formulas belong to the second class. 

610 Hysr'i Formala. Of the formulas giving a relation be- 
tween the area of water way and the area to be drained, Myer's is 
the one most frequently need. It is 

Areatf water way, in tgy^M'e feet ■:= C ^Drainage area, in aeret, 
in which C is a variable co-efficient to be assigned, for slightly 
rolling prairie, is usually taken at 1; for billy ground at 1.5; and 
for mountainous and rocky ground at 4. For most localities, at 
least, this formula gives too large results for small drainage areas. 
For example, according to the formula, a culvert having a watei 
way of one square foot will carry the water from a single acre only. 
Further, if the preponderance of the testimony of the formulas for 
the quantity of water reaching the culvert from a given area can 
be relied upon, the area of water way increases more rapidly than 
the square root of the drainage area as required by this formula. 
Hence, it appears that neither the constants nor the form of this 
formula were correctly chosen; and, consequently, for small drainage 
areas it gives the area of waterway too great, and for large drain- 
age areas too small. 

611. Talbot's Pormnla. Prof. A. N. Talbot proposed the fol- 
lowing formula, "more as a guide to the judgment than as a work- 
ing rule :" * 

Area itf water wag, in»guaTefeet= C •^(Drainage area, in aertt)*, 
in which (7 is a variable co-efficient. Data from various States gave 
values for C as follows: "For steep and rocky ground, Cvariea 



ovGoQi^lc 



AST. 1.] WATER WAT BEQUIBED. 395 

from f to 1. For rolling agiicQlttiral country subject to floods at 
times of melting of edow, and with the length of valley three or 
four times its width, C is about \; and if the stream ia longer in 
proportion to the area, decrease C. In districts not affected by 
accumulated snow, and where the length of the valley is several 
times the width, \ or \, or even less, may be need, ^should be 
increased for steep side slopes, especially if the upper part of the 
valley has a much greater fall than the channel at the culvert." 

The author has tested the above formula by numerous culverts 
and small bridges in a small city and also by culverts nnder high- 
ways in the country (ail slightly rolling prairie), and finds that 
it agrees fairly well with the experience of fifteen to twenty years. 
In these teats, it was found that water ways proportioned by this 
formula will probably be slightly flooded, and consequently be com- 
pelled to discharge under a small head, once every four or five 
jears. 

612. In both of the preceding formulas it will be noticed that 
the large range of the " constant " C affords ample opportunity for 
ihe exercise of good judgment, and maJtes the results obtained by 
the formulas almost wholly a matter of opinion. 

613, Practical Hethod. Valuable data on the proper size of 
«ny particular culvert may be obtained (1) by observing the existing 
openings on the same stream, (S) by measuring — preferably at time 
-of high water — a cross section of the stream at some narrow place, 
■and (3) by determining the height of high water as indicated by 
iirift and the evidence of the inhabitants of the neighborhood. 
With these data and a careful consideration of the various matters 
referred to in g 608, it is possible to determine the proper area of 
water way with a reasonable degree of accuracy. 

Ordinarily it is wise to take into account a probable increase of 
flow as the country becomes better improved. However, in con- 
structing any structure, it is not wise to make it absolutely safe 
^against every possible contingency that may arise, for the expen- 
diture necessitated by such a course wonld be a ruinous and un- 
justifiable extravagance. Washouts can not be prevented altogether, 
oor their liability reduced to a minimum, without an unreasonable 
expenditure. It has been said — and within reasonable limits it ia 
^e — that if some of a number of calverte are not carried awa^ 



ovGoQi^lc 



896 CUL7EBTB. [cHAP. XTII. 

each year, they are not well deBigned; that is to say, it is only a 
qneetion of time when a properly proportioned colTert will periBlL 
in some excessiTe flood. It is easy to make a culvert lai^ enoqgii 
to be safe under all circumBtances, but tlie difCerenoe in cost be- 
tween such a structure and one that would be reasonably safe would 
probably much more than overbalance the losses from the washing 
out of an occasional cnlrert. It is seldom justifiable to provide for 
all that may possibly happen in the course of fifty or one hundred 
years. One dollar at 5 per cent, compound interest will amount to 
tll.47 in 50 years and to $131.50 in 100 years. Of course, the 
question is not purely one of finance, but also one of safety to human 
life; but even then it logically follows that, unlese the engineer is 
prepared to spend ti:U.50 to avoid a given danger now, he is not 
justified in spending $1 to avoid a similar danger 100 years hence. 
This phase of the problem is very important, but is foreign to the 
subject of this volame. 

614. In the construction of a new railroad, considerations of 
first cost, time, and a lack of knowledge of the amonnt of fnture 
traffic as well as ignorance of the physical features of the country, 
usually require that temporary structures be first put in, to be re- 
placed by permanent ones later. In the mean time an incidental 
hut very 'mportant duty of the engineer is to make a careful study 
of the requirement of the permanent structures which will ulti- 
mately replace the temporary ones. The high-water mark of streams 
and the effect of floods, even in water courses ordinarily dry, should 
be recorded. With these data the proper proportioning of the 
water way of the permanent Btnictures becomes a comparatively easy 
task. Upon the judgment and ability displayed in this depends 
most of the economical value of the improvements; for, aa the road 
will have fixed or standard plans for culverts, abutments, piers, 
etc., the supervision of the construction will not be diflScult, 

Art. 3. Box and Pipe CcLvEirra. 
616. Stofe Box Cvitiet. This culvert consists of vertical side 
walls of masonry with flag stones on top from one wall to the other. 
Masonry box culverts were constructed much more frequently for- 
merly than at the present time. The lack of suitable stone in many 
parts of the West led to the adoption of vitrified pipes (§ 637) and 
iron pipes (§ 631) instead of masonry box culverts. However, in 



ovGoQi^lc 



ART. 2.] BTONB BOX CDLTBETS. 397 

many localities they are built frequently enoagh to warrant a brief 



616. Fonndatioit A common foandstion for masonry box ciil< 
Terts is a stone pavement (§ 319) under tbe entire culvert, npon 
which the side walls rest (see Fig. 91a). This is not good practice ; 
for, sinca the paving is liable to be washed oat, it endangers tho 




wall. The tendency of the pavement to undermine may be dimin- 
ished (1) by driving sheet piling or by setting deep curb-stones at 
both ends, or (2) by extending the paving to a considerable distance 
beyond bo^-h ends. The first is the better method ; bnt nsnally 
these devices only postpone, and do not prevent,'final failure. The 
water is nearly certain to carry the soil away from ander the pave- 
ment, even if the curb-stones or sheet piles remain intact. 

Sometimes culvert foundations are paved by laying large stones 
flatwise. This practice is no better than ordinary st^ne paving, un- 
ices the flags are large enough to extend under both walls ; but 
atones large enongh for this can seldom be obtained. 

A much better method is to give each side wall an independent 
ioundation and to pave between the walls only (see Fig. Qlb). Am 
important advantage of this method is that each wall can be placed 
separately, which facilitates the keeping of the water away from the 
foundation pit. Indeed, if the foundations are deep, or if there is 
not much current, the paving may be entirely omitted. If the cur- 
rent is only moderate, it is sufficient to build in, at each end of the 
culvert, between the ends of the side walls with solid masonry up 
to the bed of the stream ; but if the culvert is long, it is wise to 
build one or more intermediate cross walls also. If the current is 
strong, the cross walls at the ends should be carried down deep, 
and the space between the side walls should be paved with large 
stones closely set and deeply bedded. The beet job possible is se- 
cured by setting the paving in cement mortar. In this connection, 
see Figs. 94, 95, and 96 (pages 403, iOi, and 406). 



ovGoQi^lc 



[CHAP, xra^ 

alle^jp&rticnlarly at the end of the 
tions b^riow the effect of frost. 
f box culT6>rts are nsoally finished 
iular to the axViB of the culvert as 
7 stepping the eVids off as shown 
B to become clogget^l and to have 
, and probably its oWl^esistence 
i its upper end. ThiB a*«Miger la 




ig the Bide walla at the upper end] 
)4 (page 403). If the month of' 
L with drift, the open top is a well 
In this way the full discharging- 
atained. The lower end niay be 

linner at the outer end, thus pro^ I 
effect as is obtained in splaying \ 






,. 96 (page 406). 

) a relatiouBhip between the thiolc- | 

ad to be supported, let t 



\ 



ibove the top of the culyert ; 
7ounds per square inch ; 
strength (§ 15) ; 
over the cover stone, in pounds. 



N 



jvGooi^le 



AKT. 8.] 8T0NB BOX CULTEBTS. SSft"- 

For simplicity, consider a section of the cnlTert only a foot long. 

The coTer atones are in the condition of a beam supported at the 
ends and loaded uniformly. By the principles of the resistance of 
materials, one eighth of the uniform load multiplied by the span ig 
equal to one sixth of the continued product of the modulus of rup- 
ture, the breadth, and the square of the thickness. Expressing, 
this in symbols as above, and reducing, gives 

^^ (i>. 



= / 



Ordinarily, earth weighs from 80 to 100 lbs. per en. ft, bnt for 
convenience ve wiU assnme it at 100 lbs. per cu. ft, which is on 
the safe side ; then W = 100 S S. The maximnm moving load for^ 
railroad bridges may be taken at, say, 2 tons per foot of track. 
This is distributed over at least 8 square feet ; and hence the liva- 
load is eqn&l to one quarter of a ton, or 500 poands, per square foot* 
i. e. the live load is equal to an embankment & feet high. Therefore., 
the maximum live load — a locomotive — ia provided for by adding 5 
■ feet to the actual height of the embankment The table on pag&- 
13 shows that for limestone R = 1,500. Substituting these values, 
in equation (1), above, gives for limestone 

T=0.Z0SVH+5, (2). 

By substituting the corresponding value of R from the table on 
page 12, we have for Bandtt&ne 

r=0.355Vir+6, (8> 

For hightaat/s, it is sufficiently exact to drop the 6 under the 
radical, i. e., to neglect the live l(»d ; and equation (1) then becomes, 
for limestone 

T=0:208VH,' (4), 

and for sandatone 

I' = 0.25S/ff. (5> 

The preceding formulas give the thickness which a stone of 
average quality must have to be on the point of breaking; and henc«- 



ovGoQi^lc 



400 CCLVEBTS. [CHAP. XVIL. 

in applying them it will be necessary to allow a margin for safety, 
cither by selectiog the stone or by increasing the computed thick- 
ness. If reasonable care is used in selecting the stones, it is probably 
safe to double the thickness found as above. To allow for any given 
fiictor of safety, multiply the thickness found by applying the above 
formulas by the square root of the factor of safety. Thus, to allow 
for a factor of 4, multiply the thickness found as above by 2 ; for a 
factor of 6, mnltipty by 2J ; and for a factor of 9, multiply by 3. 

619. The thickness of the cover stones does not, however, de- 
pend alone upon the depth of the earth, the live load, and the span. 

In the first place, the pressure on the cover stone does not vary 
directly as the depth of the earth above it. {a) The earth itself 
acts more or less as a beam to support part, at least, of the weight 
over the opening. That earth may act thus is proven by the fact 
that an excavation can be carried Iiorizontally into an embankment 
<ir side hill without supporting the roof. ' The beam strength of tht 
earth increases with the compactness and the tenacity of the soil 
and with the square of the height of the embankment above the 
roof. This effect would be zero with clean sand ; but, owing to the 
nature of that material, it would seldom be employed for filling over 
A culvert. Hence, under ordinary conditions, part of the load is 
BUpportfld by the beam strength of the earth itself. Therefore, a 
low embankment may produce a greater strain in the cover than a 
much higher one. (5) The prism of earth directly over the culvert 
will be partially supported by the adjacent soil ; that is to say, the 
particles of earth directly above the culvert will act more or less as 
arches resting upon the earth at the sides of the culvert, thus par- 
tially relieving the c<Jver stones. This effect would be greater with 
sharp eand than with clay, but would be entirely destroyed by shock, 
aa of passing trains, (c) The stones at the center of the culvert 
would be relieved of part of their load by an action similar to that 
mentioned above, whereby the weight over the center of the culvert 
is transferred towards its ends. However, the relief caused by this 
action is but slight. 

In the second place, the pressure due to the live load is trans- 
mitted downward in diverging lines, thus distributing the weight 
over a considerably larger area than that assumed in deducing eqoar- 
tioDs (3) and (3) above. 

In the third place, the cover most be thick enough to resist the 



ovGoQi^lc 



-ABT. 2.] 8T0KE BOX CULVERTS. 401 

■eftbct of frost, as well as to support the earth and live load above it. 
The freezing, and consequent ezpausiOD, of the earth is a force 
tending directly to break the cover stones. That this is an impor- 
tant consideration is proved by the fact that tliese stones break 
near the ends of culverts as frequently as near the middle, although 
the weight to be supported is greater at the latter place. 

620. It is impossible to compute, even approximately, the effect 
of the preceding factors ; but experience shows that the thickness 
is independent of the height of the embankment, provided there is 
sufBcient earth over the cover stones to prevent serious shock, — say 
3 feet for railroads and 1 to 2 feet for highways. 

The thickness employed on the railroads in States along the 
fortieth parallel of latitude is generally about as foUovfs, irrespec- 
tive of the height of the bank or of whether the cover is limestone 
■or sandstone : ' 

8PAH or CDI.TMBI. Thickness or Corn, 

a feet 10 Inches. 

S feet 12 laches. 

A feet 15 inclies. 

On the Canadian Pacific R. R., the minimum thickness of cover 
atones for spanB of 3 feet is 16 inches, and under 3 feet, 14 inches. 

621. Quality of Masonry. Box culverts are usually built of 
rubble masonry (§ 313) laid in cement mortar. Formerly they were 
■often built of dry rubble, except for 3 or t feet at each end, which 
was laid in mortar. It is now generally hold that box culverts 
■should be so built that they may discharge under a head without 
damage. It is usually specified that the cover stones must have a 
solid, well-leveled bearing on the side walls of not leas than 15 
inches. The most careful constructors close the joints between the 
cover stones by bedding spalls m mortar over them. 

63S. Speeifieatiotu.' A11 atose box culverta shall bavea wnterwnjat l«sat 
2iX.S feet. The side walls sbiOl not be less than two feet <2') thick, and 
shall be built of aound, durable stouca not less thaa six inches (6") thick, laid 
In cement mortar [usually 1 part Rosendale cement to 3 parts sand]. Tbe 
walla must be laid in true horizontHi courses, but in case the thicknCBs of the 
course Is greater than 12 Intbea (IS"), occasionatly two stones may be uied to 
make up the thickness. The walls must oe laid so as to be thoroughly bonded, 
and at least one fourth of the area of each course must be headers going en- 

*F«nnsrlvBi)ia Railroad. 



ovGoQi^lc 



4ff4 COLTEET3. [CSAP. STTL 

tlrely through the wall. The top course must have one half its area of througb. 
(rlones, and ibe remaioder of thla course must consist of sloue go\ng at least 
one half of the way across the wall from tbc inside face. Tbe face stones ot 
each course must tie dressed to a straight edge, and pitched off to a true line. 
All of the coping stones of head walls must lie througlis, and must have the 
upper surface hammer-dressed to a straight edge, and the face pitched off 
to a line with margin draft. Cbver stones shall have a thickness of at least 
twelve inches <13 "J for opening of three feet (3'), and at leaslUinches (14") for 
opening of four feet (4') ; and must be carefuUj selected, and must be of such 
length as to have a tiearing of at least one foot (1) on either wall. 

The beds and vertical joints of the face stones for a distance of six 
Inches (6") from the face of the wall shall be so dressed as to require a mortar 
Joint not thiclier than tbree fourths of an inch ({"). Joints between the cov- 
ering stones must be not wider than three fourths of an inch <{"), and the 
bearing surface of cover stones upon side walls must be ao dressed as to 
require not more than a one-inch (1") mortar joint. 

The paving sliiill consist of flat stones, set on edge, at right angles with the 
line of the culvert, not less than twelve inches (12 ') deep, and shall be laid in 
cement mortar and grouted. , 

623. ExampleB. The box colvert shown in Fig. 94 (page 403), 

JB presented as being on .the whole the beet (see § 617). The table 
sccompanjing the diagram gives the various dimensions of, and also 
quantities of masonry in, bos culveii^s for different openings. The 
former data and the diagrams are ample (or the construction of any 
box culvert ; while the latter data will be useful in making esti- 
mates of cost (§ 636). In the headings of the colums under " Size 
of the Openings," the first number is the span of the culvert, and 
the second is the clear height of water way. The quantities of 
masonry in the table were computed for a cross wall at each end of 
the culverts, of the section s'lowu in Fig. 94 ; but in many cases, 
this should be 3 feet deep instead of 2, as shown. In using the 
table this correction is easily iipplied. 

824. The box culvert shown in Fig. 95 is the one employed in 
the construction of the !'Weet Slioro R. E." — New York City to 
Buffalo. The data in the table accompanying tlie dii^ram give the 
dimensions and qnantities of masonry of various sizes. In the head- 
ings under "Size of the Openings," the first number is the span of 
the opening and the second is its'height. 

Box culverte of the general form show* in Fig. 95 are sometimes 
built double; i.e., two culverts are built side by side in such a 
maimer as to have o^e side wall in common. The following table 



ovGoQi^lc 



LBT. 2.] 



STONE BOX CULVBHIB. 



p: 

> 









E-i 



g ^ 



£>%&&&^ s S ^ 









5-1 

ill 






irf.|| 



jvGooi^le 



CCLVEBIB. 



[chap. im. 




EC 




p: 








g 


£ 


( i 


"= 


X 




CQ 


D 


E-i 


s 






^ 




'1 


1 


|fI 


' s 


f^ DC 


K 


1^1 




I ^• 


1 


|r "1 


1 


•^ X 


1 X 


O 




m 


i 










ir> 






^ 


s d 


s 


1 E 








i 


£ 




8SS 


I 

s 

s 


d 

7 

T 






1 


3- 


^!-i.i.S 


^iS 




d 

s 




!!| 



JS2 
Iff 

-h 



m 



^1 , 



15! 

lis I 

si. ^'8 ,til 

SI- -la fill 



jvGooi^le 



ART. 2.] STONE BOX CULVBRTS, 405 

gives the dimeiiBionB and qaantities for such box culverts. The 
dimensions not given in the following table ore the same as in the 
table accompanying Fig. 95. 

TABLE 44. 
DniKNBionB aus ConTBiiTfl of Doijbib Box Ctn-vxRii. 





Sum Of THi OPENtm. 




'&? 


v.* 


lOi." 


'£! 


i^^ 


I>M™«™! 


wv 
w 

IS.1S 
0.8M 

0.407 


rv 

U.1S 
0.W1 
0.444 


yo" 

Sl.DO 

i.m 

0.481 


»'0" 

at. IS 

l.JW 

o.ut 








OtannmK 

Muoiii7lntw<i<iDilwa1I«.liieiLJd«. 

a—oarr la tniok, per toot of lenctb from bi- 
■Jde to Ipslde DfeDd walJi, in oD. rda. . . 


9.S» 







The standard double box culvert employed in the eotiBtmctioa 
of the Canadian Pacific B. S. differed from the form described 
abore in having (1) shorter end walls, and wings at an angle of 30*^ 
■with the axis of the colTwt, and (2) a triangular cat-water at the 
upper end of the dlTision wall. 

625. The culvert shown in Pig. 96 is the standard on the Inter- 
colonial Bailway of Canada, and is very substantially constrncted. 

626. Cost. With the data accompanying Figs. 94 and 95 (pages 
\ 403 and 40i), and the table of cost of masonry on page 160, it is 

an easy matter to make an estimate of the .cost of a box culverts 
For example, assume that it is proposed to build a culvert 30 feet 
long — out to out of culvert proper — having a water way 3 feet wide 
and 4 feet high, and that estimates of the cost of the general forms 
shown in Fig. 94 and ateo of that of Fig. 95 are desired. 

StUmaUifor o 8 X tjt. Box Culvert qfOie Q«neral Form ifuwn in S\g. M. 

Huonryln «eTid wbIIr— IB.ffi cu. fda ....(^ $S.GOpeTCU. ;d. = tU.OB 

" EG feet of trunk (1444 >»=)Se.I0car(li.O*^B0 " " ■• US.» 

PartDR " 85 " " " (0.111 ■»=) S.JS ■• " at!.«l •' '■ - B.W 

Total cHHt SlKl.ie 

BttimaUtfor aiy.Aft. Box OultartqfAe Oerteral Firrm Aoan in Fig. OSl 



7 in Sendiralli 



i_S4.aOcu.Td» a »S SO per OIL yd, = $84.70 

-UDk(»< 1.14B=)3T.Wim.]'d>.$WiO " " => 8e.4S 
" (81 X O.S70=) 8.88 " ■■ ® tt.00 ■' '■ = 1179 



ovGoQi^lc 



CULTEETB. [CHAP. XTII. 






'I 

i 
i 



jvGooi^le 



4RT. 3.] VITRIFIED PIPE CCLVERTS. 407 

If the price for the masonry does not include the expense for 
the neceseKrj ezcaTation, the above estimates should be increased 
by the coat of excavation, which will vary with the situation of the 
culvert. 

To make a comparison of the relative coat of the two types 
of culverts just mentioned, we may proceed as follows : The coat 
per foot of the trunk of a 3 X 4 culvert of the form shown in 
Fig. 94 is (1.444 cu. yds. of masonry @ 13.50 plus 0.111 cu. yds. 
of paving ® 13.00) 15.28; and the corresponding cost for Fig. 
95 is (1.148 CO. yds. of masonry @ t3.50 plus 0.370 cu. yds. 
of paving @ )2.00) $4.76. The difference, in cost per foot 
is {$5.28- J4. 76) $0.52 in favor of Fig. 95. The cost of the 
cnd.wallsforFig. 94 is (16.88 cu. yds. @ $3.50) $59.08; and the 
corresponding cost for Fig. 95 is (34.30 cu. yds. @ $3.50) $84.70. 
The difference is $25.62 in favor of Fig. 94. Since in the former 
the cross wall extends but 3 feet below the floor of the culvert, 
irhile in the latter the end walls extend 3 feet, the difference tn cost 
shonld be decreased by the cost of the difference of the f oundationa. 
If the cross walls of Fig. 94 be carried down another foot, the 
amount of masonry will be increased 3 cu. yds. and the cost $7.00; 
and the difference in coat of the end walls will be ($35.63 — $7.00) 
$18.62 in favor of Fig. 94. Under these conditions, for a culvert 
'40 feet long, the two types will cost the same; for lengths less than 
■40 feet Fig. 94 is the cheaper, and for lengths greater than 40 feet 
Fig. 95 is the cheaper. If the end walls of Fig. 95 are carried 
down only 2 feet, the amount of masonry will he decreased by 3. 4 
en, yds. and the cost by $11.90; and then the difference of cost will 
l>e($25.63 — $11.90) $13.73. Under this condition, for a culvert 
30 feet long, the two types will cost the same; for lengths less than 
30 feet Fig. 94 is the cheaper, and for lengths greater than 30 feet 
Fig. 95 is the cheaper. We may conclude, therefore, that for 
lengths under 35 or 40 feet the type shown in Fig. 94 ia a little 
cheaper, while for greater lengths than 35 or 40 feet that in Fig. 
95 is slightly cheaper. For the smallest size the length of equal 
cost is abont 10 feet. 

There is no material difference in the first coat of the two types; 
but the culvert shown in Fig. 94 is the more efScient. 

627. TnaiTIED Pipe CtllTZBTt. During the past lew years 
vitrified sewer pipes have been extensively employed for small col- 



ovGoQi^lc 



408 CULVEETS. [chap, xvir 

verts tmder both highvaye and railroads. The pipe generally- 
employed for this pnrpoae is that known to the trade as culrert 
pipe or "ertra heavy" or "double strength" sewer pipe, which is 
20 to 40 per cent, (varying with the maker and the size) heavier than 
the quality ordinarily employed for sewers. 

Apparently the heavier pipe is need on the supposition that the 
lighter is not strong enough for cnlverts. In moet cases, at least, 
this is an erroneous assumption. 1. With the same depth of earth 
over the pipe, there is but little more pressure on the pipe when 
jused as a culvert than when employed in a sewer. At moet, the- 
difference of pressure is that due to the live load, which can not 
exceed the weight of an additional 5 feet of earth (see g 618), and. 
will generally he much less (see the second paragraph of g 619). 
8. Experience demonstrates that the lighter pipes are not deficient 
in strength when used in sewers, however deep they are laid. 
According to experiments made by bedding the lower half of the 
pipe in sand and applying a pressure along a comparatively ftarrov^ 
area, the average cmahing strength of ordinary sewer pipe was- 
;i,400 Iba. per sq. ft. of horizontal section, and (or culvert pipe 
12,000 lbs. per sq. fi* . If the pressure had been applied more 
nearly as in actual practice, the pipes would have borne consider- 
ably more. The firat of the above results is equal to the weight of 
24 feet of earth, and the second to that of 130 feet, although actual 
embankments of these heights would not give anything like ih&- 
abovti pressures (see g 619). 

There is a little difference between culverts and aewers in the- 
exposure to frost; but no danger need be apprehended from this. 
cause, provided the culverts are so constructed that the water is 
carried away from the lower end, since ordinary soft drain tile are 
not in the least injured by the expansion of the frost in the earth, 
around them. 

62S, Constmction. In laying the pipe, the bottom of the trench. . 
should be rounded out to fit the lower half of the body of the pipe,. 
with proper depressions for the sockets. If the ground is soft or 
sandy, the earth should bo rammed carefully, but solidly, in and 
around the lower part of the pipe. On railways, three feet of earthi 
between the top of the pipe and the bottom of the tie has been 
found anfScient. On highways pipes have stood from 10 to Ifi 
years under heavy loads with only 8 to 13 inches of earth over- 
* For addJtloDol data, see Note 7, pt^e 617. 



jvGooi^le 



ART. 3.] VITEiriED PIPE CULVBKTB. Wftt 

them; but as a rule it is not wise to lay them vith less than 12 to- 
18 inches of earth covering. 

In many cases — perhaps in most — the joints are not calked. If 
this is not done, there is liability of the water's being forced out at 
the joints and washing away the soil from aronnd the pipe. Even 
if the danger is not very imminent, the joints of the larger pipes, 
at least, should be calked with hydraulic cement, since the coat is. 
very Bmall compared with the insurance of safety thereby secured.. 
Sometimes the joints are calked with clay. Every cnlvert should 
be built so that it can discharge water under a head without damage- 
to iteelf. 

The end sections should be protected with a timbet or masonry- 
bnlkhead, althongh it is often omitted. Of course a parapet wait 
of rubble masonry or brick-work laid in cement is best (see Fig. 97)^ 




Bio. 7T. Fio. 98. 

The fonndation of the bnlkhead should be deep enough not to ho- 
disturbed by frost In constructing the end wall, it is well to in- 
crease the ^11 near the outlet to allow for a poesible settlement of 
the interior sections. When stone and brick abntments are too 
expensive, a £iir aubetitute can be made by setting poets in the 
ground and spiking plank on as shown in Fig. 98. When planka 
are used, it is best to set them with considerable inclination towards 
the road bed to prevent their being crowded outward by the pressure 
of the embankment. The upper end of the cnlvert should be so 
protected that the water will not readily find its way along the out- 
side of the pipes. In case the mouth of the culvert shonld become 
submerged. 

The freezing o! water in the pipe, particularly if more than 
half full, is liable to burst it; eonse<|uently the pipe should have a. 
sofBcient fall to drain itself, and the outlet should be so low that. 



ovGoQi^lc 



410 CCliVERTS. [chap. XTir. 

there is no danger of back-water's reaching the pipe. If properly 
drained, there is no danger from froet. 

When the capacity of one pipe is not sufficient, two or more 
may be laid side by side. Although two Bmall pipes do not hare 
tm much discharging capacity as a single large one of eqoal cross 
section, yet there is an advantage in laying two small ones side by 
side, since then the water need not rise so high to utilize the full 
capacity of the two pipes as would be uecessary to discharge itself 
throngh a single one of larger size, 

629. Examplea. Fig. 99 (page 411) shows the standard yitri- 
£ed pipe culrerts employed on the Kansas City and Omaha B. B. 
This construction gives a strong, durable cuhert which passes water 
freely. The dimensions of the masonry end walls and of the con- 
crete bed for the intermediate sizes are nearly proportional to those 
«hown in Fig. 99, Table 46 (page 411) shows the quantities of 
masonry required for the principal sizes. 

630. Cost. Prices of vitrified pipe vary greatly with the con- 
ditions of trade, and with competition and freight. Current (1888), 
non-competitive prices for ordinary sewer pipe, in car-load lots 
/. 0. b. at the factory, are about as in the table below. 

TABLE «. 
Cost ahd Weight of Vitbitibd Seweb Pipb. 





PucinsFooT. 


Abu. 


WCIOHT PCB 


'SJZS.* 


12 lucbea. 


15 ceots. 


.78 aq. fl. 


45 1be. 


500 feet 




28 " 




56 " 


400 ■' 


18 " 


80 " 


1.40 •' '■ 


63 " 


860 ■' 


18 '• 


88 ■' 


1.76 ■• '■ 


75 " 


800 " 


20 " 


63 " 


2.18 " " 


90 •' 


260 " 


83 " 


67 " 


2.64 •■ ■■ 


110 ■■ 


280 " 


34 " 


87 " 


8.14 ■• •' 


140 " 


200 •' 



Culvett pipe costs about 20 to 25 per cent, more than as above, 
and second quality sewer pipe about 30 to 25 per cent, less. The 
latter difiers from first quality in being less perfectly glazed, less 
perfectly burned, or not perfectly round, or in having fire cracks in 
the glazing, blisters on either surface, eicrescences or pimples on 
the inside, or a piece broken out of the end. Frequently such 
pipe ie as good for calverts as first quality seWer pipe. 



ovGoQi^lc 



ABT. ».J TITBIPIED PIPE CULVEKIB. 




TiTBirtm Fin CCLvnti.— K. C. & 0, K. B. 



U&soHRT Rbqdibed fob YmuriKD Pipe Cultkbtb or thb Qsuxjui, 

TOBM HHOWM ABOVE. 



ITM*. 




DUKCTKB 


ofPtpi. 




14 IncbM. 


IB inches. 


a) Inches 


Mlnobta. 


n«h. t« 


If 




ir 


"til?- 






Total Uauorr 


S.4T 


BIB 


7fl5 


G.H 


Concrete, per lineal too».. 


O.CCT 


o.ioa 


0,,» 


...» 



jvGooi^le 



413 CULVERTS. [chap. XVII. 

631. iBOir Pips Cultebtb. In recent yeara, iron pipes have 
been much need for culvert*. In tnanj localities good stone is not 
available, and hence stone box culverts (|§ 615-26) can not be need. 
In such localities vitrified stoneware pipes are nsed ; but as thej 
are not made larger than 3 feet in diameter, iron or stone is the only 
material available for permanent culverts requiring a greater water 
way than that obtained by using one or two of the largest vitrified 
pipes. Apparently, stone culverts if well built should last forever; 
but, as constructed in the paat, they have been found to lest rela- 
tively only a short time. Hence, with the increasing cheapness of 
iron, there has been an increasing tendency to use iron pipe for even 
large cnlverta, Caet-iron pipes from 12 to 48 inches in diameter 
and 1'4 feet long are in common use by all of the prominent roada 
of the Mississippi Valley. Some of the roads cast their own, while 
others buy ordinary water pipe. The lightest water pipes made, or 
even such as have been rejected, are sufficiently strong for use in 
culverts. The dimensions used on the Chici^, Milwaukee and St. 
Paul R. B. are about as follows : 

TABLE 47, 

DiMSNBIONH OF CAST-IaOH Cni.TSBT PiFE. 



iKODEDuMCm. 


w„„™^. 


r.„„. 


Wbidbt pn LnmiL Four 






A iDch. 




16 




88 " 


88 •■ 


ao 




118 •■ 




69 ■' 


u 




175 •' 




66 •' 


80 






" 




86 




820 " 


i ■■ 


« " 


4a 




400 " 


\-- 


42 " 


48 " 


510 " 


41 " 



632. Cosfltrnotioii. In constructing a culvert with cast iron, 
the points requiring particular attention are (1) tamping the soil 
tightly around the pipe to prevent the water from forming a clian- 
nel along the outside, and (2) protecting the ends by suitable head 
walls and, when necessary, laying riprap at the lower end. The 
amount of masonry required for the end walls depends upon the 
relative width of the embankment and the number of eections of 
pipe used. For example, if the embankment is, say, 40 feet wide 
at the base, the culvert may consist of three 12-foot lengths of 



ovGoQi^lc 



JUff. S.] IKOlf PIPE CULVEUT8. 413 

pipe and a light end wall near the toe of the bank ; bnt if the 
embankment ia, aay, 33 feet wide, the culvert may consist of two 
12-foot lengths of pipe and a eomparatiyelj heavy end wall well 
back from the toe of the bank. The smaller sizes of pipe usoally 
jome in 13-foot lengths, but sometimes a few G-foot lengths are 
included for use in adjuetlug the length of culvert to the width of 
bank. The larger sizes are generally 6 feet long. 

Fig. 100 (page 414) shows the method employed on the Atchi- 
son, Topeka and Santa F^ B. R. in putting in coet-iron pipe 
^ulvertB. Table 48 (page 414) gives the dimensions for the end 
walls for the various sizes. The length of pipe is determined by 
taking the multiple of 6 feet next larger than the length given by 
the position slope as in Fig. 100. To allow for settling, the pipe is 
laid to a vertical curve having a crown at the center of 1 inch for 
-each 5 feet in vertic^ height from bottom of pipe to profile grade. 

Where the soil is treacherous, it would be wise to lay the pipes 
■on a bed of broken stone to prevent undue settling. In this con- 
nection, see Figs. 96 and 99 (pages 406 and 411). 

633. Fig. 101 (page 415) shows the method employed on the 
Chicago, Burlington and Quincy R. H. of putting in cast-iron pipe 
'Culverts. This conatmction has given entire liatisf action. 

The same road has recently commenced the use of iron for cul- 
verte up to 12 feet in diameter. For diameters greater than 4 feet, 
the pipes are cast in quadrants 3, 4, 6, and 8 feet long, which are 
afterwards bolted together, through outside flanges, to form a 
cylinder of any desired length. The different segments are so com- 
bined as to break joints around and also along the pipe. The body 
-of the pipe was formerly 1| inches thick ; but is now IJ, etiSened on 
the outside by ribs. The sections are put together without any chip- 
ping, drilling; or other skilled labor. Between the different sec- 
tions is a recess in which a tarred rope smeared with neat cement 
mortar is placed before bolting the segments together, which makes 
tne joints tight.* 

634. Cost. The cost of cast-iron pipe varies greatly with com- 
petition and the conditiona of trade. The price ranges from #36 to 
$36 per ton for firat quality water pipes,/, o. b. at the foundry; or 
approximately, say, IJ cents per pound. 

' For innatntloD of deUUs, ue ^Uroad Qaatu, voL zlx. pp. 123-34. 



ovGoQi^lc 



[chap. XVII. 



i 


\ 



a I, 




1 

s 

I 


i 


if:::: 


1 




i 




1 




! 


SVSBS 




\ 





jvGooi^le 



ART. 3.] IROS PTPB CCLVERTB. 



n*. 101,— Cmt-Uom Pira CuLrna.^C, B. A ^ & & 



ovGoQi^lc 



■416 CCLVERTS. [chap. XYU, 

Table 47 (ptige 412) shows that' the average weight of the pipe 
per foot per square foot of water way ia about 60 pounds ; ana 
hence the cost of the trunk of a cast-iron pipe, exclusive of trans- 
portation and labor, is about (60 X li) 90 cents per lineal foot per 

■ sq. ft. of area. The cost of sewer pipes ia, from Table 46 (page 
411), about 22 cents per foot per square foot of water way ; and for 
culvert pipe about 30 cents. 

Assuming the cost of rubble masonry to be $3.50 per cubic yard 
and of paving to be $^.00 per cubic yard, the averse cost of the 
masonry in the trunk of the box culvert ehown in Fig. 95 (page 
404) is 40 cents per lineal foot for each square foot of water way ; 
and the corresponding cost for the culvert of Fig. 94 (page 403) is 
46 cents. The end walls required for these different forms of cul- 
verts are essentially the sanift; and hence the above comparison 

■ shows approximately t .e relative cost of the different forms of cul- 
verts. According to this showing, cast-iron pipe is the most ex- 
pensive ; but this difference is partly neutralized by the greater 

■ease with which the iron pipe can be put into place either in new 
work or in replacing a wooden box-Kiulvert, 

635. The following figures give the cost of a 7-foot cast-iron 
^iulvert of the form referred to in § 633, which see. 

43 ((. body ® 136.55 per foot (I.M ceots per pouod) ILIU-Sa 

8 It. speciala @ ta9.48 " " ' aas.BS 

Bolts and washers 29.81 

Unloading IT..™ 

Putting (n place I4H.95 

Stone for end walls, 70cu. yds., ® tl.QO 105.00 

Stone forriprap fouudation, 60cu. yda., ® tlOO 60.00 

BflmoTiDg temporary bridge 235.63 

Total 11.947.15 

Excluding the cost of removing the temporary bridge — which 
is not a part of the culvert proper, — and of the riprap foundation — 
which the anusnal conditions required, — the cost of the culvert was 
433.03 per foot, or 83 cents per lineal foot for each square foot of 
water way. 

636. TmBB Box Ctltebts. Timber box calvertB should be 
used only where more substantial material is not attainable at a 
reasonable cost. Many cnlverts are constructed of timber an^ 



ovGoQi^lc 



AET. 2.] BOX AND PIPE CULVERTS. 417 

periodically renewed with the same material, and many are con- 
strncted of wood and replaced with stone, or sewer or iron pipe. 

The latter is an example of what may be called the standard 
' practice in American railroad building ; i. e., constructing the road 
as quickly and cheaply as possible, asiug temporary BtmctuTes, and 
-completing with permanent ones later as the finances of the company 
will allow and as the requirements of the situation become better 
understood. After the line is open, the permanent atructures can 
be built in a more leisurely manner, at appropriate seasons, and 
thus insure the maximum durability at a minimum cost. 

There is a great variety of timber box culverte in common use, 
but probably there are none more durable and efficient than those 
tised on the Chicago, Milwaukee and St. Paul H. R., — shown in 
Pig. 102 (page 418).* On this road, it is the custom to replace the 
wooden boxes with iron pipes before the timber has seriously de- 
cayed. If experience has shown the size of the wooden box to be 
about right, the timbers are cut out a little and an iron pipe is 
placed inside of the box without disturbing the earth. 

For timber bos culverts of sizes larger than can be made of 
plank, the Atchison, Topeka and Santa 'Ek R R. employs bridge- 
tie box cnlTerte. These are made by laying 6x8 inch abwed 
bridge ties flatwise, in contact, to form a floor. These ties are 
gained at the ends so as to leave a shoulder 1 inch deep against 
which the inside of the side walls bears. Upon this floor, vertical 
side walls are constructed by laying ties flatwise, one on top of the 
other ; the lowest timber in each side wall is fastened to each tie in 
the floor by a drift-bolt 12 inches long, and each timber in the side 
wall is fastened to the one below it by a 12-inch drift-bolt every 3 
feet. The lengths of the ties employed in the side walls are so ad- 
justed as to make the exposed ends conform closely to the slope of 
the embankment. Thereof consists of 6- x 8-inch ties set edgewise, 
in close contact, with a shoulder 1 inch deep on the inside, both ' 
ends of each piece being also drift-bolted to the side wall. 

637. TIHBEB BA£S£L CmTXSTB. For a number of years past 
the Chicago, Burlington and Qnincy R. R. has found it desirable, 
in view of the absence or poor quality of the stone along its lines, to 
use a timber "barrel-culvert" when the opening is too large for a 

* From SaOrvad CkuMt. 



ovGoQi^lc 



CULVERTS. [chap. xtii:. 



m 



iim 



"^f 

$ 



vm 



xS 



iGooi^le 



IXt. 3.] ARCH CULTSBIS. 419 

timber box-culvert. The staves are 10 or 12 inches thick, accord- 
ing to the size of the culvert, and 8 inches wide on the outside, 
dressed to form a circle 4^ or 6 feet in diameter. Iron rings — made 
of old rails — spaced about 10 feet apart, are used as a form upon 
which to construct the culvert aud also to give it strength. The 
staves break joints and are drift-bolted (g 381) together. M soon 
as the timber is thoroughly seasoned, the culverts are lined with a 
single ring of brick, and concrete or stoue parapet walla are built. 
If, at any time, the timber fails, it is the intention to put iron pipe 
through the present opening. 

The timber costs about $13 per thousand feet, board measure, 
at the MisaiBBippi Biver ; and the cost of dressing at the company's 
shops is about tl.50 per thousand. 

Abt. 3. Abce Culvebts. 

638. In this' article will be discussed what may be called ths 
theory of the arch culvert in contradistinction to the theory of the 
arch. The latter will be considered in the next chapter. 

By the theory of the arch culvert is meant an exposition of the 
method of disposing a given quantity of masonry so as to secure (1) 
maximum discharging capacity, (2) minimum liability of being 
choked by drift, and (3) maximum strength. Attention to a few 
points, which are often neglected in the design of culverts, will se- 
cure these ends without additional cost. 

639. OiaXBAL FOBII OF CVLTBKT. Splay of Wingfl. There 
are three common ways of disposing the wing walls for finishing 
the ends of arch culverts. 1. The culvert is finished with a straight 
wall at right angles to the axis of the culvert (see Fig. 103). 3. The 



JLM 



wings are placed at an angle of 30° with the axis of the culvert 
(see Fig. 104). 8. The wing walls are built parallel to the 
axis of the culvert, the back of the wing and the abutment 
being in a straight line aud the only splay being derived from thia- 



ovGoQi^lc 



420 CULVERTS. [CHAP. XVll. 

ning the wings at their outer ends (see Fig. 105). The first method 
ia shown on a larger scale in Plate II, the third in Plate III, and 
the second in Plate IV. 

The quantity of masonry required for these three forms of wings 
does not differ materially, Fig. 105 requiring the least and Fig. 103 
the most The most economical angle for the wings of Fig. 104 is 
abont 30° with the axis. 

The position of the wings shown in Fig. 104 is much the most 
common and is better than either of the others. Fig. 103 is ob- 
jectionable for hydraulic considerations which will be considered 
in the nest section, and also because it is more liable to become 
choked than either of the others. Fig. 105 does not have splay 
enough to admit the natural width of the stream at high water, 
and does not gire snflScient protection to the toe of the embankment. 
610. Junction of Wings and Body. With a calvert of the 
general form outlined in Fig. 104, 
there are two methods of joining 
the wings to the body of the cul- 
vert. The more common method 
is shown in Figa. 106 and 108; and 
. the better, but less common, one is 
^ shown in Figs. 107 and 109. 
""''"' ""■""■ ^^^ j^^^ shown in Figs. 106 

and 108 is very objectionable because (1) the comers reduce the 
o»pacity of the culvert, and (2) add to its cost. 



/ \/ \ 



,', I I 



^ 



Fio. :<». Fid. 10ft. 

1. The sharp angles of Fig. 106 materially decrease the amount 
oi vater which can enter under a given head and also the amount 



ovGoQi^lc 



ART. 3.] ARCH CCLVEHTS. 421 

which can be discharged. It is a well-eBtabliahed fact in hydraulics 
that the discharging capacity of a pipe can be increased 200, or 
eren 300, per cent, simply by giving the inlet and outlet forms some- 
what similar to Fig. 107. Although nothing hie this increase can 
be obtained with a culvert, one finished at both the apper and the 
lower end like Fig. 107 will discharge considerably more water than 
one like Fig. 106. The capacity of Fig. 107 decreasee as the angle 
between the wing and the axis increases ; hence, the less splay the 
better, provided the outer ends of the wings are far enough apart 
to accommodate the natural width of the stream at high water. 
Also the less the splay, the less the probability of the culvert's being 
choked with drift. Fig. 106 is very bad for both the admission and 
the discharge of water, and also on account of the great liability 
that drift and rolling stones will catch in the angles between the 
wings and the end walls. In this latter respect Fig. 108 is slightly 
better than Fig. 106. 

2. Every angle adds materially to the cost of the masonry. In 
a culvert like Fig. 106, there are fonr unnecessary comers. This 
form probably owes its prevalence to the desire to have a uniform 
batter on the face of the wing, and to have the face of the wing 
wall intersect the end wall back of the arch stones. Satisfying both 
of these conditions gives a culvert in ground plan like Fig. 106; 
and satisfying the second one only, gives Fig. 108. Practically 
there is but little difference between these two forma — both are 
objectionable, as already explained. If the wing of Fig. 108 is 
moved inward, and the corner of the wing, which would other- 
wise project into the water way, is rounded off to a gentle curve, 
Fig. 109 is obtained. This form is simple, efficient, aud, on the 
whole, the best. 

Plate III shows another method of joining the wing to the 
end wall without having an unnecessary angle. In this case, the 
face of the wing up to the springing line of the arch is a warped 
surface, which is in some respects undesirable, although it saves 
a little masonry. However, the face of the wing wall could be 
built vertical up to the springing line and then battered; or 
the wing could be moved forward and the comer be rounded off 
as in Fig. io9. 

641. 8«mi-ciroalar vs. Segmental Arohei. There are two 
classes of arches employed for culverts, viz., the semi-circular and 



ovGoQi^lc 



488 CDLTBRTS. [CHAP. IVII. 

the eegmental. The first ia by far the more common; but neverthe- 
lesB the latter is, on the whole, much the better. 

1. For the same span, the segmental arch reqairea a shorter iD- 
tradoB (the inside cnrve of a section of the arch perpendicular to 
its axis). For example, the culverts shown in Plates IV and V 
have the same span, hat the intrados of the semi-circular arch is 
15.71 ft., while that of the segmental arch is 10. 73 ft. ; that is, the 
intrados of the segmental is only 68 per cent, of the intrados of the 
semi-circular arch. ■ This difference depends upon the degree of 
fiatucss of the segmental arch. The above example is an extreme 
case, since the segmental arch is unnsiially flat, the central angle 
being only 73° 44'. (The rise is one sixth of the span.) With a 
central angle of, 120°, the intrados of the segment is 77 per cent, of 
the semi -circle. 

Or, to state the above comparison in another and better form, for 
the same length of intrados the segmental arch gives the greater 
span. For example, a segmental arch on the same general plan as 
that of Plate V, but having an intrados equal to that of Plate lY, 
would have a span of 14.64 ft., which is 46 per cent, greater than 
the span of the semi-circnlar arch shown in Plate IV. A segmental 
arch with a central angle of 130° has a span 33 per cent, greater 
than a semi-circular arch having the same length of intrados. This 
difference constitutes an important advantage in favor of the seg- 
mental arch cnlvert, since the wider the span the lees the danger of 
the culvert's being choked by obstrnctions, and because it will pass 
considerably more water for the same de[)th. 

2. For the same length of intrados, the segmental arch gives the 
greater water way. The water way of the culvert shown in Plate 
IV is 87.6 square feet ; hut the same length of intrados in a seg- 
mental arch culvert having 73° 44' central angle (the same as Plato 
V) would have a water way of 98.3 square feet; and with a central 
angle of 120° would have a water way of 99.5 square feet. In both 
examples the increase ia one eighth. 

3. On the other hand, the segmental culvert will require a 
thicker arch. It will he shown in the next chapter that arches 
can not be proportioned strictly in accordance with mathematical 
formulas ; and hence the exact difference in thickness of arch which 
should exist between a semi-circular and a segmental arch can not 
be computed. According to estahlished rules of practice, small 



ovGoQi^lc 



ART. 3.j ARCH CULVERTS. 423 

"Segmental arches are from 10 to 25 per cent, thicker than semi- 
oircalar ones. Thia difference is not very great, and its effect upon 
the 006L of the cnlvert is, proportionally, still less, since the coat 
per yard of arch masonry is leas for the thiciier arches. Then, we 
may couclade that, aince for the same apan the intrados of seg- 
mental arches is from 20 to 40 per cent, shorter than the semi- 
circle, the segmental arch requires a less volume of arch masonry 
than the semi- circular, and also costs less per cubic yard. The arch 
masonry per foot of length of the segmental arch culvert shown in 
Plate \' is only 71 per cent, of that in the semi-ciroalar one shown 
in Plate IV. The dimensions and contents of arch culverts of the 
general forms shown in Plates IV and V are given in Tables 51 
and 53 (pp. 430 and 431 respectively), from which it appears that 
the segmental arch contains only from 60 to 76 per cent, ae much 
masonry as the semi-circular, the average for the six spans being 
almost exactly 70 per cent. The coat of these two classes is shown 
in Tables 56 and 57 (pages 437 and 438), from which it appears that 
the average coat of segmental culverts 20 feet long and of different 
spans is only 59 per cent, of the cost of aemt-circnlar ones of the 
same length and span; and the average cost of an additional foot 
in length for the segmental is only 86 per cent, of that for a circnlar 
one. The water ways of the semi-circular culverts are a little the 
greater, and hence the difference in cost per square foot of water 
way is not as great as above; but, on the other hand, the form of 
water way of the segmental culvert is the more efficient, and hence 
the above comparison is about correct. 

4. Will the segmental, i. e., the flatter, arch require heavier abut- 
ments (side walls)? Unquestionably the flatter the arch the 
greater the thrust upon the abutment; but the abutment not only 
resists the thrust of the arch which tends to turn it over outwards, 
but also the thrust of the embankment, which tends to push it in- 
wards. It is impossible to compute, with any degree of accuracy, 
■either the thrust of the arch or of the embankment; and hence it is 
impossible to determine either the relative value of these forces or the 
thickness which the two abutments should have. Experience seems 
to indicate that the thrust of the earth is greater than that of th« 
arch, as is shown by the fact that nearly all semi-circular culverts 
have abutments of much greater thickness than are required to n- 
.eist the thnut of the arch; and hence we may conclude that ezp»- 



ovGoQi^lc 



424 CL'LTEBTS. [CHAP. ZVU. 

rienoe has shovii tliat tbe thrust of the earth neceBBitates a heavier 
ftbutmeDt than does the thrust of the arch. If this be true, then 
the abutment for aegmeotal arches may be thinner than thoae for 
semi-circular ones; for, since the thrust of the former is greater- 
thsD the latter, it ezerts a greater force ontvard, which counter- 
balances a larger part of the inward thrust of the embankment, and 
thus leaves a less proportion of the latt«r to be resisted by the mass 
of the abatment. Segmental arch culverts are not often built; and 
designers appear to have overlooked the thrust of the earth, since 
the side walls of segmental arches are generally thicker than for' 
semi-circular ones (compare Plates IV" and V). • 

The conclnsions may, therefore, be drawn that segmental arch 
culverts are both cheaper and more efficient than semi-circular ones. 

642. As built, many semi-circnlar arches are practically seg- 
. mental; that is, the side walls are built so high, or the backing is 

made so heavy, that practically the abutments are less than 130° 
apart, and hence the two lower ends of the arch are really only a. 
part of the side wall, and should be built square. 

Further, it is shown in §§ 681-82 that a trne arch of more thaa 
about 90 to 130 degrees is impossible. 

643. EXUCPUES. Under this head will be ^ven a brief descrip- 
tion of four series of arch cu