Skip to main content

Full text of "Concrete Engineers' Handbook"

See other formats


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 








J^9^^/ ,> ! p , 


THl.« On* 



"Ms QrawJ/ii/Book (n fn& 


Coal Age ^ Electric Railway Journal 
Electrical Wjrld '^ Engineering News-Record 
American Machinist ^ The Contractor 
Engineering 8 Mining Journal ^ Po we "^ 
Metallurgical 6 Chemical Engineering 
Electrical Merchandising 






pRorsasoR ov btbuoturajj BwaunnBiNa, thb uxiversitt 









Harvey Whipple, Adei^bert P. Mills, 
Walter S. Edge, A. 6. Hillberg 
AND Leslie H. Allen 

F1B8T Edition 
Second Impression 




6 A 8 BOUVERIE ST., E. C. 


Copyright, 1918, by thb 
McGraw-Hill. Book Company, Inc. 

Fvni prinHngf May, 1918 
Second printing^ November, 1918 



This handbook has been prepared to make available in concise form the best of present day 
knowledge concerning concrete and reinforced concrete and to present complete data and 
details, as well as numerous tables and diagrams, for the design and construction of the principal 
types of concrete structures. Although intended as a working manual for the engineer, the 
first few sections of the book may be read with profit by any one engaged in concrete work. 
In these sections an effort has been made to present the latest authoritative knowledge in 
regard to the making and placing of concrete in such form that it may be applied in the field 
to the betterment of construction. * 

In preparing this book the authors have been ably assisted by Mr. S. C. Hollister, Research 
Engineer of the Corrugated Bar Company. Special credit is due Mr. Hollister for his applica- 
tion of the slope-defiection method to the development of formulas for rigid frame structures; 
for material relating to flexure of annular sections and to restraint of standpipe sides by con- 
nection with the base, this material being published here through the courtesy of the Corrugated 
Bar Company. 

The authors also are greatly indebted to Messrs. Harvey Whipple, Adelbert P. Mills, 
Walter S. Edge, A. G. Hillberg, and Leslie H. Allen for the important chapters which they have 
prepared. These men are specialists; and their contributions will prove of great value to the 
engineering profession. ' 

The chapter on dams, written by Mr. A. G. Hillberg, has been made brief, but a sufficiently 
extensive presentation is given to enable the reader to decide intelligently what type of dam 
to adopt. It is desired to call attention to the paragraphs dealing with siphonic spillways 
as there is practically no text-book information on that subject. 

In writing this book the authors have drawn from the three volumes of HooPs '^ Reinforced 
Concrete Cons.truction " only where the preparation of new material would have been sub- 
stantial duplication. 

The authors are under obligation to Messrs. Wm. J. Fuller and Frank C. Thiessen for many 
helpful suggestions and for assistance in making calculations for some of the tables and diagrams. 
Acknowledgments are also due to Mr. C. M. Chapman and others for suggestions and criticisms 
of manuscript; and to a large number of engineers who have supplied data and details, and have 
generously given their views in regard to both theory and practice. 

The authors are indebted to Mr. Clifford E. Ives for his excellent work in preparing all 
drawings made expressly for this handbook. 

G. A. H. 
April, 1918. N. C. J. 





Section 1. Materials 


\bt. Paqs 

1. Classificationi composition, and uses 

of the principal cementing 

materials 1 

a. Gypsum plasters 1 

5. Common lime 1 

c. Hydraulic lime 2 

d. Puzzolan or slag cement 2 

c. Natural cement 3 

/. Portland cement 4 

2. Portland and natural cements com- 

pared 4 

3. Constitution of Portland cement 5 

4. Setting and hardening of Portland 

cement 5 

5. Manufacture of Portland cement .... 6 
a. Raw materials 6 

6. Proportioning the raw materials. 7 

c. Grinding and mixing 7 

d. Burning the cement mixture 7 

e. Treatment of the clinker 7 

6. Manufacture of natural cement 8 

a. Raw material 8 

6. Process of manufacture 8 

7. Testing of cement 8 

a. Sampling 8 

b. Uniformity in cement t-osting .... 8 

c. The personal factor 8 

' . Kinds of tests 8 

e. Fineness 8 

/. Normal consistency 9 

g. Time of setting 9 

h. Tensile strength 10 

i. Relation between tensile and 

compressive strength 10 

j. Compressive strength 10 

k. Soundness 10 

I. Specific gravity 1 

m. Chemical analysis 1 

8. Specifications for cement 1 

9. Containers for cement 1 

10. Storing of cement 1 

11. Seasoning of cement 12 

Abt. Pagk 

12. Use of bulk cement 12 

13. Weight of cement 12 


14. Definitions 12 

15. General requirements 12 

16. Classification — Coarse and fine 12 

17. Qualities of fine aggregates — General 13 

18. Qualities of coarse aggregates — Gen- 

eral 13 

19. Materials suitable for coarse aggre- 

gates 13 

20. Igneous rocks 13 

a. Granite 13 

b. Trap-rock of diabase 14 

21. Sedimentary rocks 14 

o. Sandstone 15 

6. Limestone 15 

22. Metamorphic rocks 16 

23. Gravel 16 

24. Blast-furnace slag 17 

25. Cinders 17 

26. Materials suitable for fine aggregates 17 

a. Special characteristics of sand. . . 18 

b. Crushed stone and screenings. ... 19 

c. Sea sand 19 

d. Standard sand 19 

27. Requirements of fine aggregate as to 

shape and size of particles 19 

28. The selection of sand 20 

29. Requirements of coarse aggregate as 

to shape and size of particles 20 

30. Impurities in aggregates 21 

31. Size and gradation of aggregate par- 

ticles 22 

a. Grading of mixtures 22 

6. Grading, density, and strength. . . 22 

c. Money value of grading 23 

32. Mechanical analysis of aggregates ... 23 

33. Specific gravity of aggregates 25 

34. Voids in aggregates 25 

a. Percentage of voids 25 



Abt. Paob 

&. General laws 25 

c. Effect of moisture on voids in sand 

and screenings 26 

d. Percentage of voids determined 

by weight 26 

36. Tests of aggregates 27 

36. Notes on the selection and testing of 

aggregates 30 

37. Specifications for aggregates 30 


38. General requirements 31 

30. Examination of water 31 

40. Functions of water 31 

41. Influence of quantity of water on 

strength of concrete 32 

42. Influence of quantity of water on flux- 

ing of cement 32 

43. Influence of quantity of water on 

lubrication of concrete mixture. . 33 

44. Influence of quantity of water on 

space occupied in resulting con- 
crete 33 

45. Harmful effects of voids caused by 

excess water 34 

46. Excess water the cause of "day's 

work planes" 34 

47. Excess water the cause of large lait- 

ance deposits 31 

48. Excess water and waterproof concrete 35 

49. Excess water causes unsatisfactory 

concrete floor surfaces 35 

50. Excess water prevents bonding new 

concrete to old 35 

51. Excess water and concreting in cold 

• weather 35 

52. Suggested procedures to guard 

against use of excess water 36 


53. Types of reinforcement 36 

54. Surface of reinforcement 37 

55. Quality of steel 37 

56. Working stresses 37 

57. Coefficient of expansion 37 

Art. Paob 

58. Modulus of elasticity 37 

59. Steel specifications 37 

a. For bars rolled from billets 37 

6. For rerolled bars 40 

60. Factors affecting cost of reinforcing 

bars 42 

61. Deformed bars 42 

a. Diamond bar 42 

b. Corrugated bars 43 

c. Havermeyer bars 43 

d. Rib bar 44 

e. Inland bar 44 

/. American bars 44 

62. Wire fabric 45 

a. Welded wire fabric 46 

5. Triangle-mesh wire fabric 47 

c. Unit wire fabric 48 

d. Lock-woven steel fabric 49 

c. Wisco reinforcing mesh 50 

63. Expanded metal 50 

a. Streelcrete 51 

h, Kahn mesh 52 

c. Corr-X-metal 52 

d. Econo 53 

c. G. F. expanded metal 53 

64. Rib metal 55 

65. Self-centering fabrics 55 

a. Hy-rib 56 

6. Corr-mesh 56 

c. Self-centering 56 

d. Chanelath 56 

c, Ribplex 57 

/. Dovetailed corrugated sheets .... 57 

66. Reinforcing systems for beams, gird- 

ers and columns 57 

a. Kahn system 57 

6. Cummings system 58 

c. Unit system 58 

d, Corr system 60 

c. Hennebique system 60 

/. Pin-connected system 60 

g, Luten truss 60 

h. Xpantruss system 60 

i. Shop fabricated reinforcement 

system 60 

Section 2. General Methods of Construction 



1. Properties of concrete, dependent 
upon properties and relative pro- 
porUons of constituent materials 63 

2. Theory of proportioning 

3. The strength elements of concrete.. . . 

4. Proportioning for high strength con- 






Aht. Paos 

5. Weakness due to poor proportioning . 64 

6. Unit of proportioning 65 

7. Arbitrary proportions 65 

8. Proportioning by void determinations 65 

9. Proportioning by mechanical analy- 

sis 68 

10. Proportioning by maximum density 

tests 68 

11. Checking materials on the job 69 

12. Proportions and the measurement of 

materials 70 

13. Proportioning bank-run gravel 70 

14. Proportioning crusher-run stone 71 

15. Proportioning blast-furnace slag and 

cinders 71 

16. Proportioning water 71 

17. Success in proportioning 72 

Mixing, Transportinq and Placing Con- 
crete ■ 

18. Mixing concrete 72 

19. Amount of water to be used in mixing 

concrete 72 

20. Transporting concrete 72 

21. Depositing concrete in forms 73 

22. Continuous and even depositing in 

forms 73 

23. Continuous depositing to avoid stop- 

page planes 74 

24. Bonding set and new concrete 75 

25. Removal of entrained air 75 

26. Spading, puddling and tamping 75 

27. Depositing concrete through water. . 76 

28. Remixed and retempered concrete. . . 76 

29. Concreting in hot weather and in 

cold weather 76 

a. Pre-heating aggregates and water 77 

b. Means for heating aggregates. ... 77 
c Enclosure and heating of forms. . 77 

d. Protection against frost 78 

6. Freezing of concrete 78 

/. Use of anti-freezing mixtures .... 78 

g. Protection against heat 78 

Field Tests of Concrete 

30. Object of field tests 78 

31. Limitations inherent in field tests. . . 78 

32. Comparative tests on field-molded 

and structural specimens 79 

Abt. Paqe 

33. Value of tests on field-molded test 

specimens 79 

34. Transverse tests on beam specimens. . 79 

35. Core drill test specimens from actual 

structures 79 

36. Suggested methods for making and 

testing field specimens of con- 
crete 79 

37. Pre-use tests of materials 82 

Waterproofing Concrete 

38. Meaning of "waterproof" 82 

39. Resistance of concretes to water ac- 

tion 82 

40. Resistance of concretes to water pene- 

tration 83 

41. Degree of impermeability attainable. 83 

42. Porosity of commercial concretes .... 83 

43. Excess water as a cause of porosity. . 84 

44. Shrinkage cracks 84 

a. Types of shrinkage cracks 84 

b. Shrinkage cracks and porosity ... 85 

c. Prevention of shrinkage cracks. . . 85 

45. Pervious concretes and laitance 86 

46. Effect of temperature and atmos- 

pheric effects on water-tightness 86 

47. Integral waterproofing compounds . . 86 

a. Integral waterproofing classifica- 

tion 87 

b. Value of integral waterproofing 

compounds 87 

c. Rendering defective structures 

impervious 87 

48. Waterproofing by cement grouting . . 88 

49. Membranous waterproofings 88 

a. Application of membranous 

waterproofing 88 

b. Continuity of membrane 89 

c. Protection of waterproofing 89 

50. Rules for making concrete impervious 89 

Finishing Concrete Surfaces 

51. Character of surface finish desired. . . 90 

52. Removing form marks 90 

a. Tooling 90 

b. Rubbing 91 

c. Brushing 92 

d. Sand-blasting 92 

53. Use of colored aggregates 92 




Ajit. Page 

64« Addition of colon to concrete 93 

66. Use of white cement 93 

66. Plaster 'finishes 93 

67. Surfacing concrete floors 93 

68. Specification for the production of a 

rubbed surface 93 


69. General requirements 93 

60. Economical considerations 94 

61. Lumber for forms 94 

62. Removal of forms 95 

63. Number of sets of forms in Iniihiing 

work 96 

64. Examples of form design 96 

a. Column forms 96 

5. Beam and girder forms 103 

c. Slab forms 104 

d. Column heads 108 

c. Wall and pier forms Ill 

66. Design of forms 120 

a. Values to use in design 121 

6. Drafting room methods 123 

66. Tables and diagrams for designing 

forms 124 

a. Notation 124 

b. Fiber stresses allowed 124 

c. Formulas used 125 

67. Systematizing form work for build- 

ings 131 

a. Saw mill and yard 131 

6. Shop procedure 131 

C. Cleat spacing 134 

d. Planning of field work 134 

c. Stripping of forms 134 

68. Steel forms 135 

69. (Construction nott»s 138 

Hknuixi} and Placing Kkinfoiukmknt 

70. Chei*king, assorting and storing t^tocl 139 

71. Bending of reinforcement 139 

a. Types of bends 139 

6. Hand devices 130 

r. Power-operated benders — . 141 

d. Care to bo exercised in bendinc . Ml 
f. Bending of slab reinforcement. 1 12 

72. PLicing of reinforcement U2 

73. IVviccs for supporting reinforcing 

bars 113 

Manupacturb and Use op Co.nx'Rete 
Stone, Block and Brick 

Abt. Page 

74. Development of the industry 146 

75. Two main lines of work 147 

76. Methods of manufacture 147 

a. Dry tamp method 147 

h. Pressure methotl 148 

c. Wet-cast method 14S 

77. Consistency 148 

78. Commercial molds 149 

79. Operation of machines 1.50 

a. Tamping 150 

80. Gang molds for wet-cast products. . . 151 

81. Materials 151 

a. Cement (storage and conveying) 151 

h. Aggregates (kind and quality). . . 152 

82. Mijdng 1.52 

a. Mixers (general type) 153 

h. Mixing dry and mixing wet 153 

c. Agitation subsequent to mixing 

in wet-cast work 153 

d. Mixing facing materials 153 

83. Placing 154 

a. Buckets and hoppers 1.54 

h. Wheelbarrows 15.5 

c. Pallets . 155 

d. Bankers 15(5 

84. Curing 150 

a. Natural curing 156 

h. Steam curing 157 

85. Special molds 159 

a. Wood molds 159 

6. Plast<}r molds 150 

c. Glue molds 100 

d. Combination molds 161 

c. Waste molds 161 

86. Sand molds and casting in sand 161 

87. Surfaces 162 

a. Face design in standard unit.s. . . . 162 

h. Facing materials 163 

c. Colors 164 

d. Spraj'ing 165 

e. Brushing 165 

/. Rubbing \m 

g. Tooling 16(5 

h. Mosaics 167 

i. Efflorescence 167 

j. ^Vir bubbles . 167 

k. Crazing 1 67 

88. Specifications of the Amcritau ('<m- 

crete Institute l(>s 




Section 8. Constniction Plant 

Preparation of Concvrete AudREOATSs 
Art. Pagb 

1. Preparation of crushecl-Htono aggre- 

gate 171 

a. Preparation of site for quarrying. 171 

6. Quarrying 171 

c. Drills 171 

d. Stone crushers 172 

e. Screening and grading of crushed 

stone 172 

/. Washing crushed stone 172 

g. Crushed limestone 172 

2. Screening of sand and gravel 172 

3. Washing of sand and gravel 173 

Handling and Storage of Materials 

4. General considerations 173 

5. Storage and care of stone 174 

6. Shoveling materials directly from cars 

to ground 174 

7. Storage and care of sand 174 

8. Conveyance economics 174 

9. Unloading economies 175 

10. Proper size and type of shovel 175 

11. Clam-shell buckets 176 

12. Bucket unloaders and conveyors. . . 177 

13. Belt conveyors 177 

14. Storage and handling of sack cement 177 

15. Bundling and storage of empty 

cement sacks 178 

16. Storage and handling of water 179 

17. A typical installation 179 

Concretino Plant 

18. Plant economics 181 

a. First cost 181 

h. Cost of installation 181 

A«T. . Paqb 

c. Cost of operation 181 

d. Cost of maintenance 181 

e. Cost of removal 181 

/. Salvage 181 

19. Balancing the plant 181 

20. Typical plants 181 

21. Machine vs. hand mixing 186 

22. Types of mixers 187 

a. Drum mixers 187 

h. Trough mixers 187 

c. Gravity mixers 188 

d. Pneumatic mixers 188 

23. Machine mixing 189 

a. Time of mixer operations 189 

6. Time of mixing 189 

c. Drum speeds 190 

d. Loading the mixer 190 

Charging hoppers 190 

Power loaders 191 

Low charging mixers 191 

e. Measuring materials 192 

/. Discharge of the mixer 192 

24. Transporting and placing concrete . . 193 

a. Barrows 193 

h. Concrete carts 193 

c. Buckets 194 

d, Cableways and buckets 195 

c. Spouts or chutes 195 

/. Sections used in spouting 196 

g, Hoiste 200 

25. Spouting plants 203 

a. Boom plants 203 

6. Guy line plants 203 

c. Tower plants 203 

d. Combinations of spouting systems 203 

e. licgulating flow of concrete in 

spouting plants 203 

Section 4. Concrete Floors and Floor Surfaces, Sidewalks, and Roadways 

Concrete Floors and Floor Surfaces 

1. The concrete floor problem 205 

2. "Dusting" of concrete floors 206 

3. Making good concrete floors and floor 

surfaces 206 

4. Special surface finishes 206 

a. Surface grinding 206 

6. Integral pigments 206 

r. Finish produced by removal cf 

water from surface 207 

fi. Integral hardeners and surface 

compounds 207 

.*). Causes of common defects in concrete 

floors 207 

r>. Remedial measuren 209 

a. Retopping 209 

b. Chemical hardeners 210 


Ajn. Paos 

c. Use of oils 210 

d. Floor coatings and paints 210 

Concrete Sidewalks 

7. Structural functions 210 

8. Essential qualities 210 

9. The making of concrete sidewalks. . . 211 

a. Porous subbase 211 

h. Concrete base 211 

c. Top of wearing surface 211 

d. Surface finishing 211 

e. Surface protection and curing 211 

/. Protecting sidewalks in hot weather 212 

g. Special surface finishes 212 

10. Vault light pavements 212 

AxT. Facub 

11. Concrete curbing 212 

12. Summary 212 

Concrete Roadways 

13. Structural functions 213 

14. Essential qualities 213 

15. One-course and two-course pave- 

ments 213 

16. The making of concrete roadways. . . 213 

a. Porous subbase 213 

b. Proportioning and selecting of 
materials 213 

c. Joints 214 

d. Curing 214 

e. Consistency 214 

Section 6. Properties of Cement Mortar and Plain Concrete 

Strength op Cement Mortar and Plain 


1. Strength in general 215 

2. Laboratory tests, their use and signifi- 

cance 215 

3. Neat, mortar, and concrete strength 

compared 217 

4. Aggregates of mortar and concrete ... 218 

5. Effect of mineral character of aggre- 

gates 219 

6. Effect of shapes and size of aggre- 

gates 221 

7. Relation between density and 

strength 222 

8. Effect of mica, clay, and loam in ag- 

gregates 224 

9. Effect of consistency 225 

10. Compressive and tensile strengths 

compared 227 

11. Strength of plain concrete columns. . 229 

12. Effect of method of mixing 230 

13. Effect of method of placing 231 

14. Effect of regaging 232 

15. Effect of curing conditions 234 

16. Effect of freezing 236 

17. Effect of salts 236 

18. Effect of hydrated lime and water- 

proofing compounds 238 

19. Effect of sea water used in gaging. . . 238 

20. Effect of oils used in gaging 240 

21. Effect of laitance 241 

22. Rate of increase in mortar strength, 

retrogression 241 

23. Transverse strength 242 

24. Shearing strength 243 

25. Adhesive strength 246 

a. Adhesion to concrete previously 

placed 246 

h. Adhesion or bond to steel (see 

AH. 2, Sect. 6) 247 

26. Strength of natural cement mortar 

and concrete 247 

27. Strength of cinder concrete 248 

28. Working stresses (see Appendix B) . . 250 

Elastic Properties of Cement Mortar 

AND Concrete 

29. Stress-strain curves for mortars and 

concretes 250 

30. Yield point 251 

31. Modulus of elasticity 251 

Contraction and Expansion of Cembnt 
Mortar and Concrete 

32. Coefficient of expansion 252 

33. Moisture changes 253 

Durability of Cement Mobtab and Con> 


34. Fire-resistant properties 254 

35. Weathering qualities 255 

36. Abrasive resistance 255 

37. Action of sea water 256 

38. Action of alkali 257 

39. Action of acids, oils, and sewage 257 

40. Electrolysis in concrete 258 




41. Effect of manure. 


Miscellaneous Properties of Cement 
Mortar and Concrete 

Abt. Paob 

43. Porosity 261 

44. Permeability and absorptive proper- 

ties 262 

45. Protection of embedded steel from 

corrosion 262 

46. Weight of mortar and concrete 263 

42. Rise of temperature in setting 259 

Section 6. General Properties of Reinforced Concrete 

1. Advantages of combining concrete 

and steel 265 

2. Bond between concrete and steel 265 

Pull-out teste 266 

Relation of bond stress to slip of 

bar as load increases 266 

Bond resistance in terms of com- 
pressive strength of concrete. . . 267 
Distribution of bond stress along 

a bar 267 

Variation of bond resistance with 
size, shape, and condition of sur- 
face of bar 267 

Anchoring of reinforcing bars 268 

Influence of method of curing 
concrete 268 

Influence of freezing of concrete . . . 268 

Influence of age and mix of con- 
crete 268 

Effect of continued and repeated 
load 268 

Effect of concrete setting under 

pressure 269 

Beam teste 269 

3. Length of embedment of reinforcing 

bars to provide for bond.. ..... 270 

4. Ratio of the moduli of elasticity 270 

5. Behavior of reinforced concrete under 

tension 271 

6. Shrinkage and temperature stresses . . 271 

7. Weight of reinforced concrete 272 

Section 7. Beams and Slabs 

Rectangular Beams and Slabs 

1. Forces to be resisted 273 

2. Distribution of stress in homogeneous 

beams 273 

3. Assumptions in theory of flexure for 

homogeneous beams 275 

4. Plain concrete beams 275 

5. Purpose and location of steel rein- 

forcement 275 

6. Tensile stress lines in reinforced con- 

crete beams 275 

7. Flexure formulas for reinforced con- 

crete beams 276 

8. Assumptions in flexure calculations . . 276 

9. Flexure formulas for working loads — 

straight line theory 276 

10. Flexure formulas for ultimate loads . . 279 

11. Flexure formulas for working loads 

and for ultimate loads compared 280 

12. Lengths of simply supported beams . . 280 

13. Shearing stresses 280 

14. Methods of strengthening beams 

againstfailureindiagonaltcnsion 281 

15. Moment and diagonal-tension teste — 

General 281 

16. Bond stress 284 

17. Web reinforcement in general 285 

18. Region where no web reinforcement 

is required 286 

19. Vertical stirrups 289 

20. Method of placing stirrups from the 

moment diagram 291 

21. Bent rods and vertical stirrups for 

web reinforcement 296 

22. Pointe where horizontal reinforce- 

ment may be bent 297 

23. Transverse spacing of reinforcement 300 

24. Depth of concrete below rods 300 

25. Ratio of len^h to depth of beam for 

equal strength in moment and 
shear 301 

26. Economical proportions of rectangu- 

lar beams 302 

27. Rectangular beams with steel in top 

and bottom 302 

a. Formulas for determining per- 
centages of steel in double-rein- 
forced rectangular beams 304 

28. Deflection of rectangular beams 304 

a. Maney's method 305 



Art. Paqb 

h. Tumeaure and Maurer's method 305 

29. Slabs 306 

a. Moments in continuous slabs 306 

h. Provision for negative moment in 

continuous slabs 306 

r. Floor slabs supported along four 

sides 307 

d. Cross reinforcement in slabs 307 


30. T-beams in floor construction 307 

31. Tests of T-beams 307 

32. Flange width 308 

33. Bonding of web and flange 308 

34. Flexure formulas 308 

a. Formulas for determining dimen- 
sions and steel ratio for given 
working stresses 309 

35. Designing for shear 310 

36. General proportions of T-beams . 310 

37. Economical considerations 310 

38. Ck)nditions met with in design of T- 

beams 310 

39. Design of a continuous T-beam at the 

supports 311 

40. T-beams with steel in top and bottom 313 
a. Formulas for determining per- 
centages of steel in double-rein- 
forced T-beams 313 

41. Deflection of T-beams 313 

Special Beams 
42. Wedge-shaped beams 


Abt. Page 

43. Beams of any complex or irregular 

section 314 

a. Analytical method 314 

h. Graphical method 316 

Shear and Moment in Restrained and 
Continuous Beams 

44. Span length for beams and slabs 318 

45. Recommendations of Joint Com- 

mittee as to positive and negative 
moments 318 

46. Theorem of three moments 318 

47. Uniform load over all spans 323 

48. Fixed and moving concentrated 

loads 324 

a. Influence lines 324 

49. Moving uniform loads 329 

50. Maximum moments from uniform 

loads 330 

51. Beam concentrations 331 

52. Negative moment at the ends of con- 

tinuous beams 334 

53. Bending up of bars and provision for 

negative moment 335 

54. Continuous beams with varying 

moment of inertia 339 

Desionino Tables and Diagrams for 
Beams and Slabs 

65. Illustrative problems (12 in all) 341 

56. Leffler's comprehensive beam chart. 348 

57. Beard and Schuler's comprehensive 

charts 350 

Section 8. Columns 

1. Column types 371 

2. Plain concrete columns or piers 371 

3. Columns with longitudinal reinforce- 

ment 37 1 

4. Columns with hooped and longitudi- 

nal reinforcement 372 

5. Columns reinforced with structural- 

steel shapes 372 

0. Working stresses 373 

7. Recommendations of the Joint Com- 

miUee 373 

8. Tables and diagrams 374 

9. Reduction formula for long columns. . 381 
10. Columns supporting bracket loads. . . 381 

Section 0. Bending and Direct Stress 

1. Theory in general 385 

2. Analytical determination of stresses in 

rectangular sections 387 

a. Case I. — Compression over the 
whole section — steel top and 

bottom 387 

Case II. — Tension over part of 
section — stet^l top and bottom. . 394 

c. Case III. — Tension over part of 

section — steel in tension face only 40.*J 
3. Graphical determination of stresses . . 40<> 

a. Rectangular sections 40i> 

h. Hollow-circular sections 407 

c. Solid circular and other sections . . . 409 



Section 10. Moments in 

1. Importance of the subject 411 

2. Method of analysis 411 

3. Application of method of analysis to 

simple cases 414 

4. Conception of rigidity of building 

frames 415 

5. Moments at interior columns in 

beam-and-girder construction . . 416 

a. All terminals hinged 416 

6. All terminals fixed 417 

c. Columns hinged. Outer girders 

with constant moment 418 

d. Columns fixed. Outer girders 

with constant moment 418 

e. Point of inflection at center of 

columns. Outer girders hinged . 418 
/. Point of inflection at center of 

columns. Outer girders fixed. . . 418 
g. Point of inflection at center of 

columns. Outer girders with 

constant moment 419 

Rigid Building Frames 

Abt. ^^^" 

h. Point of inflection at center of 

upper columns. Lower columns 
fixed. Outer girders with con- 
stant moment 419 

i. Point of inflection at center of 
upper columns. Lower columns 
fixed. Outer girders hinged. . . 419 

j. Point of inflection at center of 
upper columns. Lower columns 
fixed. Outer girders fixed 420 

6. Moments at exterior columns in 

beam-and-girder construction. . . 420 

o. Inner end of girder hinged 425 

h. Inner end of girder fixed 425 

7. Moments in columns in flat-slab 

construction 425 

8. Criteria for maximum combined 

stresses in columns 426 

a. Interior columns 426 

b. Exterior columns 426 

9. Wind stresses in building frames 427 

10. Roof frames 428 

11. L-frames 429 

Section 11. Buildings 

Floors — General Data 


1. General types of concrete floors. . . 

2. Floor loads 431 

3. Economic considerations 432 

4. Floor surfaces 432 

5. Small floor openings 433 

6. Provision for the attachment of 

shaft-hangers and sprinkler pipes 434 

7. Bedding machinery 437 

8. Waterproof floors 438 

9. Tests 438 

10. Basement floors 438 

Monolithic Beam and Girder Con- 

11. Ordinary type of beam and girder 

construction 439 

12. Hollow tile construction 447 

Flat Slab Construction 

13. General description 467 

14. Advantages over the beam-and-girder 

type 467 

15. Classes of buildings to which adapted 458 

16. Remarks regarding design 459 

17. Sjrstems 461 

o. Barton Spider Web system 461 

h. Cantilever Flat-slab construction 463 

c. Simplex system 466 

d. Mushroom sjrstem 465 

e. Watson system 466 

/. Akme system 467 

g. Corr-plate floors 470 

h. S-M-I system 471 

t. Three-way system 476 

18. Patents ^9 

19. Loading tests 480 

20. Methods of design and problems 487 

a. Computations for Akme system.. 488 
6. Computations for Corr-plate 

floors 489 

c. Computations based on Pitts- 

burgh Ruling 492 

d. Computations based on Chicago 

Ruling 494 

e. Computations based on Ruling of 

American Concrete Institute. . . 495 



Art. Paqs 

/. Roof design 496 

g. Beams in flat-filab floors 497 

h. Columns 497 

i. Brick exterior wall supports 498 

21. Tables 498 

22. Ck>nBtruction methods and safeguards 506 

Unit Constbuction 

23. Method of construction in general. . . . 508 

24. Advantages of the unit method 508 

25. " UnU-biU" system 508 

26. Ransome unit system 509 

Stbel Frame (Construction with Con- 
crete Slabs 

27. Types of construction 511 

28. Wrapping of I-beams 511 

29. Types illustrated 511 


30. Structural design 512 

31. Loading 512 

32. Prevention of condensation on con- 

crete roof slabs 513 

33. Concrete roof surfaces ;. . 516 

34. Separate roof coverings 517 

35. Drainage 518 

36. Parapet walls 521 

37. Saw-tooth construction 526 


38. Trainshed of UnU-biU construction 




39. Details of design 527 

40. Loading 530 

41. Column brackets 530 

Walls and Partitions 

42. Bearing walls 532 

43. Curtain walls 533 

44. Brick and other veneer 539 

45. Window openings 540 

46. Door openings 543 

47. Basement walls 545 

48. Partitions 545 


49. General design 549 

50. Methods of supporting stairs 550 

51. Stair details 551 

Elevator Shafts 

52. Elevator-shaft pits 552 

53. Pent houses 553 

Provisions for Contraction and Expan- 

54. Methods employed 554 

Section 12. Foundations 

1. Bearing capacity of soils 557 

2. Pressure on the soil 558 

3. Plain concrete footings 558 

4. Advantages in using reinforced con- 

crete for foundations 558 

5. Wall footings 559 

0*. Types of column footings 559 

7. Single column footings 559 

8. Combined column footings 565 

9. Cantilever footings 568 

10. Raft foundations 570 

11. Examples of column footings 571 

12. Concrete piles 572 

a. Piles molded in place 572 

h. Piles molded before driving 574 

Section 13. Retaining Walls 

1. Earth pressure 575 

a. Rankine's formula for resultant 

active earth pressure 576 

h. Coulomb's wedge of maximum 

pressure 577 

c. Comparison of Coulomb and Ran- 

kine results 580 

d. Useful interpretation of results of 

earth pressure theories 580 

2. Live load on top of fill — ^Equivalent 

surcharge 580 

3. Live load on top of fiU — ^Pressure dis^ 

tribuUon ^1 

4. Stability of a retaining wall 581 



Art. Paob 

a. So-called "Factor of Safety". ... 584 
h. Factor of limitation 584 

5. TVp^ of retaining walls 584 

6. Design of plain concrete walls 584 

a. Formulas and diagrams for the 

two principal types 584 

h. Trautwine's table 586 

c Selection of preliminary section. . . 587 

7. Design of cantilever or T-walls of 

reinforced concrete 587 

a. To determine approximate base 

width 587 

6. Stem 589 

c Base slab 590 

Art. Paqe 

d. Expansion joints 592 

3. Design of counterforted walls 592 

a. Thickness and spacing of counter- 

forts 592 

b. Vertical or face wall 593 

c. Back floor slab 593 

d. Cantilever toe slab 596 

e. Methods of reinforcing counter- 

forted wall 596 

9. Specifd types of reinforced-concrete 

walls 600 

10. Construction of retaining walls 601 

a. Back-filling and drainage 601 

6. Forms 602 

Section 14. Slab and Girder Bridges 

Slab Bridges 

1. Slabs under concentrated loading. . . 603 

a. Illinois tests 603 

ft. Ohio tests 604 

c. Tests by Goldbeck 605 

2. Slab bridges of single span 605 

3. Slab bridges of multiple spans 607 

a. Concrete pile trestles 607 

6. Pier trestles 610 

c. Trestles with framed bents 610 

d. Cantilever flat-slab construction.. 612 

Simple Girder Bridges 

4. Deck girders 613 

5. Through girders 617 

Continuous Girder Bridges 
(Monolithic construction) 

6. Expansion joints 622 

7. Examples of typical bridges of the 

continuous girder type 623 

8. Analysis of stresses in rigid viaduct 

structures 625 

a. Viaduct frames 628 

h. Four-span viaduct frame with 

rigidly connected column tie 629 

c. Three-span viaduct frame with 

rigidly connected column tie 631 

d. Two-span viaduct frame with 

rigidly connected column tie . . . 632 

e. One-span viaduct frame with 

rigidly connected column tic . . . 633 

/. Four-span viaducl frame 633 

g. Three-span viaduct frame 634 

h. Two-span viaduct frame 634 

i. One-span frame, unequal columns 636 

j. Temperature stresses 636 

k. Effect of fixed bases 638 

I, Viviuct bent 638 

Cantilever Bridges 

9. Theory of design 639 

10. Examples of cantilever bridges 639 

Section IS. Concrete Floors and Abutments for Steel Bridges 

1. Concrete floors on steel bridges 643 

2. Abutments for steel bridges 645 

3. Types of abutments 645 

4. Pier abutments of plain concrete 645 

5. Pier abutments of reinforced concrete 646 

6. Wing abutments 646 

7. Cellular abutments 647 

8. U-abutments 647 

9. T-abutments 647 

10. Buried-pier abutments 648 

11. Skeleton and arched abutments 648 

12. Care in constructing abutments 649 



Section 16. Afches 

General Data 

Art. Paoe 

1. Definitions 651 

2. Curve of the intrados 651 

o. Three-centered curve 652 

6. Semi-eUipse 652 

c. Parabola 653 

3. Arrangement of spandrels 653 

4. Piers and abutments 653 

5. Depth of filling at crown 654 

6. Loads 654 

7. Empirical rules for thickness of arch 

ring 656 

8. Approximate formula for best shape 

of arch axis 658 

9. Proper thickness of arch ring in the 

haunch for given thicknesses at 
crown and springing 658 

10. Dead loads and their action lines 658 

11. Approximate method of testing trial 

arch 658 

12. Use of temporary hinges in arch 

erection 659 

13. Use of reinforcement in concrete 

arches 660 

14. Classification of arch rings 660 

Analysis of the Arch by the Elastic 


15. Deflection of curved beams 660 

16. General procedure in arch analysis. . 661 

17. Notation 662 

18. Formulas for thrust, shear, and 

moment .•. 663 

19. Division of arch ring for constant j . . 663 

20. Loadings to use in computations 664 

21. Use of influence lines 664 

22. Internal temperature investigations . 664 

23. Shrinkage stresses 664 

24. Deflection at any point 664 

25. Method of procedure in arch-ring 

design 666 

26. Uncertainty as to fixedness of ends of 

arch 665 

27. Skew arches 665 

28. Unsymmetrical arches 665 

Art. . . Page 

a. Origin of coordinates between 

divisional lengths 665 

6. Origin of coordinates at crown. . . 666 
c. Origin of coordinates at left 

springing 667 

29. Arch structure of two spans with 

elastic pier 667 

Cochrane's Formulas and Diagrams for 

Use in the Design op Symmetrical 

Arches in Accordance with 

THE Elastic Theory 

30. Accuracy of formulas and diagrams . . 669 

3 1 . Difficulties and uncertainties involved 

in applying the elastic theory . . 669 

32. Best shape of arch axis 669 

33. Variation in thickness of arch ribs. . . . 670 

34. Influence-line diagrams 673 

35. Diagrams for moments, thrusts, and 

average stresses 681 

36. Approximate method of correcting 

maximum moments when actual 
arch axis deviates from assumed 
axis r)8(> 

Details of Arch Bridges 

37. Spandrel details in earth-filled 

bridges 691 

38. Spandrel details in open-spandrel 

bridges 694 

39. Piers and abutments 702 

40. Railing and ornamental details 702 

Construction of Arches 

41. Arch-ring construction 702 

42. Centering 704 

o. Timber centers 704 

h. Steel centers 712 

Three-hinged Arches 

43. General discussion 715 

44. Methods of analysis 715 

45. Common type of hinges 717 

46. Methods of construction 717 

47. Details of design 719 



Section 17. Hydraulic Structures 





1. Preliminary studies 723 

a. Locating 723 

6. Geological investigations 723 

c. Selecting a suitable type of dam.. 724 

d. Height of structure 724 

e. Hydrographic investigations 725 

/. Capacity of reservoir 726 

2. Design of foundation 728 

a. Grouting 728 

b. CulK)ff walls 728 

c. Caissons 728 

d. PiUngs 728 

e. Sheet piling 728 

3. Design of dams of gravity section . . . 729 

o. Hydrostatic pressure 729 

h. Profiles of dams 729 

c. Uplift 731 

d. Wind pressure 732 

e. Ice thrust 732 

/. Initial stress 732 

g. Temperature stresses. . . . .\ 732 

h. Stresses in masonry and on foun- 
dation 733 

t. Shearing stresses 734 

j. Final calculation 734 

4. Design of arched dams 736 

a. Constant radius dams 736 

h. Constant angle dams 738 

5. Design of reinforced-concrete dams. . 739 

a. Cut-off walls 740 

b. Foundation mattress 740 

c. Buttresses 740 

d. Bracing 741 

e. Apron 741 

/. Multiple-arch dams 743 

6. Earthem dams with concrete core 

wall 745 

7. Passing the discharge 748 

a. Form of spillway 748 

6. Discharge capacity 749 

c. Profiles of spillways 749 

d. Overflow dams 753 

e. Sluices 754 

/. Siphonic spillways 755 

8. Movable dams 768 

• a. Requiring operating machinery . . 759 

6. Operating under hydrostatic pres- 
sure differences 759 

Abt. Page 

c. Automatically operating 759 

9. Pish ladders 759 


iO. General types 760 

11. Quality of concrete for reservoir 

masonry 760 

12. Open basins with embankment walls.. 761 

13. Concrete floors for reservoirs 761 

14. Groined and flat floors 762 

15. Concrete walls for open reservoirs. . . . 763 

16. Partition and outside walls 764 

17. Provision for ice 764 

18. Covers or roofs for reservoirs and 

basins 764 

19. Groined arch construction 764 

20. Construction details of columns and 

roof 764 

Standpipes and Small Tanks 

21. Analysis of stresses in standpipes . . . . 765 

22. Restraint at base 765 

23. Shear at base 767 

24. Small tanks 768 

25. ConBtruction details of tanks and 

standpipes 769 

26. Precautions in construction 771 

Elevated Tanks 

27. Analysis of stresses 771 

28. Supporting tower 774 



General considerations 775 

Factors in culvert design 776 

a. Culvert efficiency 776 

b. Waterway required 776 

c. Length of culverts 776 

d. Design of ends 777 

31. Pipe culverts 777 

a. Pressure in trenches 778 

6. Strength of pipe 781 

c. Circular culverts cast in place.. . . 782 

Box culverts 783 

a. Forms of box culverts 785 




Art. Page 

h. Loading 785 

c. Design of crossnsection 786 

Type I. The closed frame 787 

Type II. Open frame with fixed 

walls 788 

Type III. Open frame hinged 

at the base 789 

Diagrams 794 

d. Construction 794 

33. Arch culverts 796 

a. Design of cross-section 796 

Abt. Paos 

6. Forms 797 

Conduits and Sewers 

34. Stresses due to internal pressure 799 

35. External earth pressure on circular 

pipe 799 

36. Large conduits and sewers not circu- 

lar 799 

37. Construction 802 

38. Longitudinal reinforcement 802 

39. Examples of reinforced-concrete pipe. 803 

40. Forms for sewers 803 

Section 18. Miscellaneous Structures 

Deep Grain Bins or Silos 

1. Action of grain in deep bins 805 

2. Janssen's formulas for pressure in 

deep bins 805 

3. Conclusions from tests 806 

4. Design of walls 808 

a. Vertical load carried by walls. . 808 
h. Wind stresses on a horizontal sec- 
tion 808 

c. Thickness of walls 808 

d. Horizontal reinforcement — circu- 

lar sections 808 

e. Rectangular sections 808 

/. Hexagonal bins 809 

6. Construction 809 

Shallow Bins 

6. Sloped sides— level full (Case I) 811 

7. Partly vertical sides — level full (Case 

II) 811 

8. Sloped, sides — fill heaped to angle of 

repose (Case III) 811 

9. Partly vertical sides — fill heaped 

(Case IV) 812 

10. Thrust due to P* 812 

11. Data for bin design 812 

12. Submerged storage for coal 813 

13. Dock pockets 815 


14. Dead load stresses 816 

15. Stresses on annular sections in flexure 816 

16. Wind stresses in chimneys of rein- 

forced concrete 818 

17. Chimney with no vertical reinforce- 

ment 818 

18. Longitudinal shear in chimneys 819 

19. Temperature stresses in chimnesrs. ... 819 

20. Chimney construction 820 

21. Bases for chimneys 820 

Section 19. Estimating 

Estimating Unit Costs 

1. Division of the work 823 

2. Estimating unit cost of concrete 823 

a. Materials 823 

h. Labor 824 

c. Plant 825 

d. Summary 825 

3. Estimating unit cost of foniiH 826 

n. Considerations involved 826 

6. Materials 827 

c. I^abor 827 

d. Summary 828 

4. Estimating unit cost of steel rein- 

forcement 828 

5. Estimating imit cost of surface finish 829 

Estimating QuANTiTifis 

6. Systematic procedure advisable 830 

7. Rules for measurement of concrete 

work 830 

8. Estimatingamountof form work. . 831 

9. Estimating amount of steel 832 

10. Estimating amount of surf ace finish . . 832 



Appendix A 


Standard specifications and tests for Port- 
land cement 833 

Appendix B 

Worlcing stresses 845 

Appendix C 

Rulings Pertaining to Flat-Slab Design 

Ruling on the design of cantilever flat- 
slab construction in the city of 
Pittsburgh 847 

Ruling covering design of flat-slab con- 
struction in the city of Chicago. . . 849 

C'hicago reinforced-concrete flat-slab rul- 
ing amended '. . 851 


Final report of special committee of the 
American Society of Civil Engi- 
neers — part pertaining to flat-slab 
design 854 

Standard building regulations for the use 
of reinforced concrete. American 
Concrete Institute, 1917 — part 
pertaining to flat-slab floors 858 

Appendix D 

Standard notation 


Appendix E 


Concrete barges and ships — extract from 
report of the Joint Committee of 
the American Concrete Institute 
and Portland Cement Association 863 






a by 


± he residue 

water this 
ste of lime 

1 water, as 

d clay 


lar a^ 

Sloped &. 
Partly vt 


Division of \ 
Estimating m. 

a. Materials. . 

b. Labor 

c. Plant 

d. Summary.. 
Estimating uiu 
n, Gonsidepr' 

'» Mateo 





1. CUssification, Composition, and Uses of the Principal Cementing Materials, — Cement- 
ing materials used in structural work may be divided into two main classes — non-hydraulic 
and hydraulic. Non-hydraulic cements, as the name implies, will not set and harden under 
water; while hydraulic cements will harden in either water or air. Following is a list of the 
structural cements of commercial importance: 

--. . J ,. f Gypsum pi 
Non-hydrauhc | ^^^^^ ,; 


Hydraulic ' 

Hydraulic lime 

(Grappier cemenlj a by-product) 
Puzzolan cement 
Natural cement 
Portland cement 

(Adulterated or modified Portland cement) 

la. Gypsum Plasters. — Gypsum plasters are made by partial or complete de- 
hydration of relatively pure or impure natural gypsum. The setting of these plasters is a 
recrystallization from a solution formed by admixture of the partially or totally dehydrated 
material with water, reforming the original substance. [Pure gypsum is a hydrous crystaUine 
calcium sulphate (CaSOf + 2H2O) ; and in its raw uncalcined state is used as an adulterant 
to retard the setting of Portland and natural cements. Plaster of Paris (CaSO* H- HHjO) 
is also used for the same purpose and is a refined plaster made from pure gypsum by 

Gypsum plasters, of one variety or another, are used principally on interior walls and floors. 
They are also used in the form of molded hollow blocks and tiles for fireproof interior partition 
walls, and as one variety of '' stucco '^ for the architectural adornment of buildings. 

16. Common Lime. — Common lime is made by burning limestone (CaCOt) 
at a temperature of about 900®C. until its carbon dioxide (COi) is driven off as gas. The residue 
is common lime (CaO), known commercially as "quicklime." On addition of water this 
product slakes with evolution of heat and much increase of volume, forming a paste of lime 
hydrate, or calcium hydroxide (Ca[0H]2) known as "lime putty" or, on dilution with water, as 
"cream of lime." 

Even in the purest Umestone to be found in nature some impurities are present. Generally 
a part of the lime (CaO) is found replaced by a certain percentage of magnesia (MgO), and clay 



[Sec. 1-lc 

is also present to some extent. [Clay is composed chiefly of silica (SiOz) and alumina (AI2OS), 
and usuaUy contains some iron oxide (FetOs).] In the manufacture of quicklimei magnesia 
acts in much the same manner and may be considered the equivalent of limB, which makes it 
possible to use limestone which is high in magnesia. Quicklimes are divided into four main 
types according to the relative content of calcium oxide (CaO) and magnesium oxide (MgO). 
These are: 

1. High-calcium; quicklime containing 90% or over of calcium oxide. 

2. Calcium; quicklime containing not less than 85% and not more than 90% of calcium 

3. Magnesian; quicklime containing between 10 and 25% of magnesium oxide. 

4. Dolomitic; quicklime containing over 25% of magnesium oxide. 

Following is an average analysis of ten high-calcium quicklimes and two dolomitic quick- 

! siOt 

1 '^ 





CaO MgO 

% % 

High'Calcium . . . 81 

Dolomitic 0.87 




94.98 1.39 
60.13 1 36.12 

[Analysis also shows carbon dioxide (COi) and water (HfO) to be present in small amounts.] 

Practically all lime used in construction is made into mortar by adding sand to the paste 
of lime hydrate, as sand is not only cheaper than lime but diminishes the great shrinkage which 
accompanies the setting and hardening of lime putty. This hardening is due mainly to crys- 
tallization, but in addition some of the water in the hydroxide is gradually replaced by carbon 
dioxide from the atmosphere, causing a small part of the hydroxide to revert to the original cal- 
cium carbonate (CaCOs). 

Although common lime is used chiefly in combination with sand as a mortar in laying 
ordinary brick and stone masonry, it is also used extensively as an interior wall plaster and 
for gaging hydraulic cement mortars, either to make them easier to work or to induce their 

HydrcUed lime is quicklime slaked at the place of manufacture. Its market form is that of 
a dry powder, and as such it can be mixed with sand more easily than can lime paste made by 
slaking ordinary quicklime on the work. 

Ic. Hydraulic Lime. — Hydraulic lime is made by burning argillaceous or silicious 
limestone at a temperature not less than lOOO^C. When showered with water the product 
slakes completely or partially without sensibly increasing in volume, and possesses hydraulic 
properties due to the combination of calcium with silica contained in the limestone as an 
impurity, forming calcium silicate. It is the universal practice to slake the lime at the place 
of manufacture on account of the better results obtained. 

Grappier cement is a by-product in the manufacture of hydraulic lime, produced by grind- 
ing the lumps of underbumed and overbumed material which do not slake. As might he 
inferred, grappier cement possesses properties similar to those of hydraulic lime. 

Hydraulic lime is not manufactured in the United States on account of the abundance 
of raw materials suitable for the manufacture of Portland cement, with which hydraulic lime 
cannot compete as a structural material. A number of hydraulic limes and grappier cements 
are marketed as "non-staining cements'' — ^that is, they do not stain masonry For this reason 
a considerable amount of this cementing material is annually imported from Europe for purposes 
of architectural decoration. 

Id. Pazzolan or Slag Cement — Puzzolan cement is made by incorporating 
hydrated lime with a silicious material, such as granulated blast-furnace slag, of suitable fine- 
ness and chemical composition. In Europe a natural puzzolanic material, such as volcanic ash. 

Sec. l-lc] 


is used at some plants in place of the blast-furnace slag. Silica, when finely enough divided, is 
soluble in water and chemically active. For this reason the materials are finely pulverized and 
intimately mixed by grindmg, but are not calcined, the formation of calcium silicate taking 
place slowly and at ordinary temperatures. 

Although this t3rpe of cement possesses hydraulic properties, it should not be confused with 
slag Portland cement (sometimes called steel Portland cement) which is produced by calcining 
finely divided slag and lime in a kiln and pulverizing the resulting clinker. Anal3n9is of the small 
number of puzzolan or slag cements manufactured in this country shows approximately the 
following range in composition: 


AlsOi + FesOi 
+ FeO 








COs + HtO 


27.2 to 31.0 

11.1 to 14.2 

60.3 to 51.8 


0.15 to 1.42 


They are normally slower in setting than Portland cements and on ^is account are usually 
't reated with materials which will hasten the set — such as burned clay, high-alumina slags, caus- 
tic soda, sodium chloride, or potash. Puzzolan cements made from slag may be distinguished 
by their light lilac color, absence of grit, and low specific gravity (2.60 to 2.85). They also are 
high in sulphides, which render them liable to disintegration in air, nor are they suited for use in 
sea water, where there is always an excess of sulphates. 

Puzzolan cement is not as strong or reliable as either natural or Portland cement and should 
be used only in unimportant structures or in unexposed work, such as foundations, where weight 
and bulk are more important than strength. 

le. Natural Cement. — Natural cement, as its name implies, is made from rock as 
it occurs in nature. This rock is an argillaceous (clayey) limestone, or other suitable natural 
rock, and it is burned at a temperature of from 900** to 1300°C., the clinker being then finely 
pulverized. The product does not slake, but possesses strong hydraulic properties, calcium 
silicate being formed and acquiring strength and rigidity through crystallization. 

Unfortunately, the composition and characteristics of natural cement are subject to consider- 
able variation. This is to be expected, since the composition of the rock from which it is made 
not only varies in different localities but is further subject to variation to some extent at least, 
even in the same deposit. Portland cement, on the contrary, is an artificial mixture, subject 
to control. This fact, together with its slow setting, is mainly responsible for the decrease in the 
use of natural cement and the adoption of Portland cement in all important structures. 

In spite of these disadvantages, however, it is a significant fact that natural cements do hot 
show disintegration with passage of time, while Portland cements frequently are most erratic 
in behavior. On comparing anal3rses of typical natural and Portland cements, it is at once 
noticed that natural cements have a higher percentage of silica, about the same percentage of 
alumina and a lower percentage of lime than have Portland cements. Excess lime, so generally 
prevalent in hydrated Portland cements, and not necessarily resultant on "free lime," is fre- 
quently the cause of much trouble. There is a distinct field of usefulness for natural cement 
which is largely overlooked by engineers at the present time. In many cases, perplexing prob- 
lems could be effectively solved by its employment. 

The following summary shows the range in composition of an average analysis of six well- 
known American natural cements: 












5.2 to8.8 


31.0 to 57.6 



[Sec. 1-1/ 

[Analysis also shows varying small amounts of alkalies (K20 and NasO), anhydrous sulphuric* 
acid or sulphur trioxide (SOs), carbon dioxide (COs), and water (H2O). Magnesia (MgO) is 
usually regarded as equivalent to lime in its action. The specific gravity of natural cements 
range from 2.7 to 3.1, with an average of 2.85.] 

Natural cement is adapted to many uses, but its relatively low strength and slow hardening 
limit its field to structures where high stresses will not be imposed for several months after plac- 
ing the concrete, as in large or massive structures where weight and mass are more essential than 
early strength — ^that is, in such structures as dams, abutments, foundations, and many under- 
ground structures. Mortar made with natural cement (either alone or mixed with lime mor- 
tar) is excellent for la3ring ordinary brick and stone masonry. 

1/. Portland Cement. — Portland cement is made by finely pulverizing the clinker 
produced by burning a definite artificial mixture of siHcious (containing silica), argillaceous 
(containing alumina), and calcareous (containing lime) materials to a point somewhat beyond 
where they begin to fuse or melt. The product is one that does not slake and possesses strong 
hydraulic properties. The essential components of Portland cement — ^namely: silica, alumina, 
and lime — are obtained from many different sources, but the proportions used of the raw mate- 
rials are always such that the chemical composition of the different Portland cements is constant 
within narrow limits. The percentages of the principal components range about as follows: 











19 to 25 1 5 to 9 

2 to 4 

60 to 64 


[Small amounts of alkalies (KiO and NasO) and sulphur trioxide (SOs) are also present. Mag- 
nesia (MgO) is considered by some as an impurity, while other investigators claim it is equiva- 
lent to lime (CaO) in its action. Alumina (AlsOa) and iron oxide (FesOi) do not act entirely 
alike but are usually considered to have the same functions.] The specific gravity of Portland 
cements range from 3.1 to 3.20, with an average of 3.15. 

Portland cement is by far the most important cementing material used in modem engineer- 
ing construction. It is adapted for use in concrete and mortar for all types of structures where 
strength is of special importance, or in structures exposed to wear or to the elements. It 
should invariably be employed in reinforced-concrete construction because of its high early 
strength and generally uniform quality. 

A number of special cements employing Portland cement as a base are made by grinding in 
adulterating materials after calcination. These adulterants include clay, slaked lime, sand, 
slag, natural cement, limestones, and natural puzzolanic material or tufa. The action of these 
materials is essentially to promote combination between lime from the cement and silica from 
the adulterant, with formation of silicate of lime. In some cases these silicious adulterants 
improve the quality of concrete made from such cements, but this result cannot be expected 
from all forms of adulteration. 

Sand and puzzolanic material have perhaps been used the most extensively and successfully 
of any of the adulterants, producing products known as sand cemenl and tufa cement respectively. 
These cements have been used principally on large work where freight rates are high and lung 
wagon hauls combine to make the cost of undiluted Portland cement excessive. Cement 
specifications in common use are of a character to exclude any grinding in of materials after 
calcination, presumably on the ground that specifications permitting any adulteration would 
be subject to abuse so that the results obtained would be uncertain. 

2. Portland and Natural Cements Compared. — The distinguishing properties of natural and 
Portland cements and the chief differences in manufacture may be summarized as follows: 

Sec 1-3] 


Natural cement 

Portland cemeut 

Raw material 

Natural rock 

Mostly the vertical stationary 
Low, but variable 
Variable, not under control 

Yellow to brown 
Relatively rapid 
Low, especially at early age 
Usually rather coarse 
Will not usually stand steam 

Artificial mixture 

Slanting cylindrical revolving 

Relatively high 
Controllable within narrow 


Bluish or steel gray 
3.1 to 3.2 . 
Relatively slow 
Relatively high 
Relatively fine 
Required to stand steam test 

Type of kiln used 

Calcination temperature 

Chemical composition 


Specific Kravitv 

Rate of settins 


Decree of Krindine 


In structures where either natural or Portland cement may be used, and where economy 
is the governing consideration, the choice of cement should be based on a comparison of the 
costs per cubic yard of the required mortar or concrete mixtures. Decisions usually are 
desired either between a 1 : 2 natural cement mortar and a 1 : 3 Portland cement mortar, or 
between a 1 : 2: 4 natural cement concrete and a 1 : 4: 8 Portland cement concrete. 

3. Constitution of Portland Cement.' — The latest optical and microscopical examinations 
of Portland-cement clinker, and of all the substances which were formally considered as likely 
to be formed in manufacture, show Portland cement to be made up largely of the three com- 
pounds 3CaO-SiOa, 2CaO'SiOj, and 3CaO-AlaOj. The tri-calcium silicate appears the best 
cementing compound and it is probable that the higher its percentage, the better the cement. 
The small amounts present of FeaOj, MgO, alkalies, etc. have but little effect on the three 
major compounds but their presence aids materially in manufacture by promoting the combina- 
tion of CaO with AI2O3 and SiO,. 

A perfectly burned cement clinker consists of about 36% of tri-calcium silicate, 3CaO-Si02; 
33% of di-calcium silicate, 2CaO-Si02; 21% of tri-calcium aluminate, 3CaO-Al20a; and 10% 
of minor constituents. The principal cementing compound, tri-calcium silicate, 3CaOSiOj, 
is the last constituent to form completely in Portland-cement manufacture; and this compound 
is formed by the combination of CaO with 2CaO-Si02. When cement clinker is not perfectly 
burned there is evidently less 3CaO-Si02 formed and more 2CaO-Si02. There is also a certain 
percentage of free lime (CaO) present, the amount depending upon the degree of burning. 

4. Setting and Hardening of Portland Cement.' — The setting and hardening of Portland 
cement is caused principally by hydration in the order named of the three major constituents — 
SCaOAUOj, 3CaO-Si02, and 2CaO SiOa. When water is added to Portland cement, these 
constituents form first amorphous and later both crystalline and amorphous hydrated materials 
which act much as does ordinary glue, except that since they are of mineral origin and largely 
insoluble, hardening progresses even under water. 

Of these hydration products, the compound tri-calcium aluminate (3CaO-Al20a) when 
mixed with water sets and hardens very quickly; tri-oalcium silicate (3Ca0-Si0«) sets and 
hardens somewhat less rapidly; and di-calcium silicate (2CaOSi02) reacts slowly. Hardening 
occurs only after the lapse of a long period of time. The initial set of cement is due undoubtedly 
to the hydration of 3Ca0'Al203; the early hardness and cohesive strength is due to this hydration 

1 From paper before Am. Cone. Inst., Feb., 1016, by G. A. Rankin, Geophysical Laboratory, Carnegie Inst, 
of Wash. 

* See Klein and Pbillips: Tech. Paprr^ 43, U. S. Bureau of Standards. 
Bates and Ki.ein: Tech. Paper , 78, U. S. Bureau of Standards. 



[See. 1-5 

and to that of the 3CaO*SiOt; while the gradual increase in strength is due to the further hydra- 
tion of these two compounds together with the hydration of the 2CaO-SiOt. 

The compound 3CaO'SiOx appears to be the best cementing constituent of this group, as 
it is the only one of the three which when mixed with water will set and harden within a reason- 
able time to form a mass which is comparable in hardness and strength to Portland cement. 
Although dCaO-AljOt sets and hardens rapidly, it is rather soluble in water and is not particu- 
larly durable or strong. The compound 2CaO'SiOi, however, requires too long a time to harden 
to be in itself a valuable cementing material. 

6. Manufacture of Portland Cement. 

6a. Raw Materials. — Silica, alumina, and lime — ^the essential components of 
Portland cement — occur as ingredients in a large number of natural materials of widely vary- 
ing character. In none of these, however, do the three components occur in the exact pro- 
portions required in Portland-cement manufacture so that an artificial mixture of several 
materials has to be resorted to. The following combinations of raw materials are used in differ- 
ent cement plants in this country : 

1. Cement rock and limestone. 

2. Marl and clay (or shale). 

3. limestone and clay (or shale). 

4. Blast-furnace slag and .limestone. 

5. Chalk and clay. 

Cement rock is an argillaceous limestone containing about 68 or 72% of lime carbonate, 
18 to 27% of clayey matter, and not over 5% of magnesium carbonate. It is a dark slatey 
Umestone, rather soft in texture, and is almost ideal for cement making due to the fact that 
it is easy to quarry and grind, and is usually so well balanced in composition that but a small 
amount of comparatively pure limestone needs to be added. This rock is found in many parts 
of the country but so far has been used in the manufacture of Portland cement only in the 
Lehigh district of eastern Pennsylvania and western New Jersey, a district producing nearly 
one-third the entire output of the United States. 

Limestone suitable for cement manufacture is composed principally of calcium carbonate 
together with more or less impurities. The following shows the approximate range of com- 
position of such limestones: 




AlsOs + FetOi 




0.3 toS.O 

0.2 to 2.1 


Sulphur as SOt and various alkalies may also be present in small percentages. 

Af arZ is almost pure calcium carbonate. It Lb a soft, wet earth found in the basins of dried- 
up lakes and in swamp regions, deposited either by chemical agencies or through the phys- 
ico-chemical agencies of certain forms of vegetable and animal life. 

Clays and shales are of the same general composition, differing only in degree of solidifi- 
cation. Clays result from the decay of shales, and like their parent rock, are composed chiefly of 
silica (SiOs) and alumina (AUOj), and usually iron oxide (FejOi). The proportion of silica in 
clay suitable for cement manufacture should not be less than 55 to 65%; and the combined 
amount of alumina and iron oxide should be between one-third and one-half the amount of 
silica. A clay with these proportions of principal constituents is highly siliceous, produeinK 
a cement clinker which is comparatively easy to grind. Clay containing a greater portion of 
alumina produces a hard clinker and a quick-eetting cement which is more severely attacked by 
sea water. 

Blast^ftirnace slag is a compound formed from impurities in tl;e iron ore and the limcst^mc 

Sec. 1-56] 


used aa a flux in the blast furnace. Following is a typical analysis of slag used in the manu- 
facture of cements: 



AhOi + FesOs 







12.60 i 49.98 


Slag, when allowed to cool quickly, becomes a hard glassy mass, very tough and durable. In 
order to use slag economically in cement manufacture the molten slag is nm into large cisterns, 
and there converted into a granular substance by directing against it innumerable little streams 
of air and water. This disintegrated slag has the appearance of coarse brown sugar; and in this 
form it is used as a raw material for Portland cement. 

Chalk is a soft earthy variety of calcium carbonate, formed from the remains of minute 
organisms. It also sometimes contains small amounts of silica, alumina, and magnesia. 
Its use as a material for cement manufacture is limited. 

56. Proportioning the Raw Materials. — ^Two rules are in use for proportioning 
the raw materials used in the manufacture of Portland cement. Newberry's rule is as follows: 

Max. lime = 2.8 (%SiO,) + 1.1 (%A1,0,) 

Eckel's rule (called the '^ cementation index"), which is really a modification of the above rule, 
takes the magnesia and iron oxide into account : 

2.8 (%SiOJ + 1.1 (%AhO,) -f 0.7 (%Fe«0,) 
%CaO -h 1.4 (%MgO) 

= 1 

A value of the cementation index below 1 means an excess of lime or magnesia in the cement 
which will cause expansion, or imsoundness. In practice it is customary to reduce by about 
10% the proportion of lime found by the above rule to avoid any chance of obtaining an unsound 
cement. Although the above rules are not based on the most recent investigations of the 
constitution of Portland cement there is no immediate prospect of any change being made in 
practice in the methods of proportioning because the present rules are known by experience to 
produce excellent results. 

6c. Grinding and Ifizing. — The admixture and grinding of the raw materials 
before calcination is accomplished by either a wet or a dry process. In the wet process, prin- 
cipally for plants using marl, the raw materials are ground and fed into rotary kilns in the 
form of a slurry containing sufficient water to make it of a fluid consistency. In the dry process 
raw materials are ground and mixed in the dry state. The larger portion of Portland cement 
manufactured in the United States at the present time is made by plants using the dry process. 

hd. Burning the Cement Mixture. — Rotary kilns are used in almost all American 
Portland-cement plants. These kilns are slightly inclined to the horizontal and revolve at 
about the rate of one revolution per minute. The ground raw materials are fed in at the upper 
end and are carried forward and tumbled over and over by the slant and revolution of the kiln. 
As the materials advance they are reduced by the hot gases from the burning fuel which is fed 
in at the lower end. The clinkers formed vary in size from y^ in. up to about IH ^' ^^ 
diameter. It takes about an hour for a particle of raw material to traverse the entire dis- 
tance from the feed to the outlet. 

he. Treatment of the Clinker. — After cement clinker is cooled it is crushed and 
passed through preliminary grinding mills. Then gypsum is added and the clinker ground to 
a fine powder. If the clinker was used without the addition of gypsum, it would take an 
almost immediate set. Approximately 2 lb. of gypsum (CaS04) (or pLaster of Paris) is used to 
every 100 lb. of clinker. 


6. Manufacture of Natural Cement 

6a. Raw Material. — The raw material used in the manufacture of natural cement 
is a natural argillaceous limestone containing from 13 to 35% of clayey material. About 15% 
of the clayey material is silica, the balance being alumina and iron oxide. The kind of limestone 
generally used contains a considerable proportion of magnesium carbonate in place of calcium 
carbonate. When the rock varies greatly in composition, materials from different strata are 
mixed together to give as uniform a product as possible. The wide range allowed in the com- 
position of natural cement, however, does not warrant great refinement in the analysis of the rock. 

66. Process of Manufacture. — Natural cement is usually manufactured in ver- 
tical kilns lined with firebrick. These kilns are of the mixed-feed type, the rock (not crushed) 
and fuel being charged in alternate layers. The temperature required in burning is considerably 
below that required in Portland-cement manufacture because temperatures higher than 1300^0. 
would fuse the material to a slag having no hydraulic properties. The temperature employed, 
however, is sufficient to cause the formation of silica compounds with the lime and magnesia. 
The clinker is taken out at the bottom of the kiln as it is burned, and then it is crushed and 
ground. Grinding of the clinker is not usually carried as far as that of Portland although some 
of the newer mills use grinding machinery similar to that in Portland-cement plants. 

7. Testing of Cement — For standard methods of cement testing, see Appendix A. 

la. Sampling. — Tests should be conducted only on representative samples. 
For method of sampling, see Appendix A, page 834. 

76. Uniformity in Cement Testing. — ^In order to obtain results in cement testing 
which will be of the greatest value, definite and uniform methods should be used. Results 
depend not only on the quality of the cement but also on the temperature and percentage of 
water used in mixing, the method of mixing and molding test specimens, the temperature and 
humidity of the air, the character of the sand used, and the type of apparatus employed. 

7c. The Personal Factor. — The personal factor has considerable effect on results 
obtained in cement testing and, on this account, only experienced, well-qualified men should be 
employed in making tests. Results by untrained or careless operators are really worse than 
nothing and may be positively misleading. The comparative results, however, by any one 
experienced observer are generally consistent and are of value. It is usually advisable to have 
the testing done at some well-established and properly equipped cement-testing laboratory. 

Id, Kinds of Tests. — The following cement tests made regularly are recom- 
mended for construction work of importance and also in all cases where the cement to be used 
does not work satisfactorily: 
Time of setting. 
Tensile strength of standard mortar. (Compressive strength of standard mortar the 

best criterion.) 
On unimportant construction it is generally safe to use a well-known brand of Portland cement 
without testing, or to make simply the test for soundness. 

7e. Fineness. — Fine grinding has a great influence on the properties of cement. 
It increases the ability of the cement to react readily with water and enables the cement 
particles to coat the sand grains more thoroughly. In other words, the finer the cement, all 
other conditions being the same, the stronger will be the mortar produced with a given sand. 
The fineness of cement is measured by determining the percentage by weight which will 
be retained on a standard 20O-mesh sieve. ^ Standard specifications' require that the residue 
shall not exceed 22%. Most mills are now equipped to grind cement to such a fineness that 
even less than 10% is retained. 

It has long been generally recognised that the coarser particles in cement are practically 

* For description of stancUrd tieve, ace Appendix A, p. 83G. 
*For ■tandard specifications, see Appendix it, p. 833. 

Sec. 1-7/1 MATERIALS 

inert and that at least the earlier cementing value is due chiefly to the grains that will pass the 
No. 200 sieve. Because of this fact the present standard method of testing for fineness is unsat- 
isfactory, as no attempt is made therein to determine the further fineness of the greater and 
more valuable part of the cement. To remedy this defect an air-analyzer' has recently been 
perfected at the Bureau of Standards which makes it possible to further divide a cement which 
passes a 200-me8h sieve into four definite sizes. This apparatus had been standardized and will 
undoubtedly come into extensive use. 

Increased fineness has the efifect of making a cement quicker in setting and hardening, 
the high-alumina cements being the most afifected. Fine grinding also affords additional 
opportunity for seasoning and thus indirectly improves the soundness of cement. 

If, Normal Consistency. — Tests for time of setting, strength, and soundness 
are greatly influenced by the quantity of water used in mixing. In order to have all results 
comparable with one another, a determination is made in each case of the quantity of water 
necessary to be added to a given weight of cement to give a standard or normal consistency. 

A simple method of finding normal consistency is to mix a quantity of cement paste and 
make up from the paste a ball about 2 in. in diameter. The ball is then dropped upon the 
testing table from a height of 2 ft. The paste is of normal consistency when the ball does not 
crack and does not flatten more than one-half of its original diameter. The finer the cement, 
the more water is required for normal consistency. For this test the room and the mixing 
water should be kept at standard temperature. 

Another method of finding normal consistency which is more commonly used and gives 
more concordant results is by the use of the Vicat needle apparatus (see Fig. 2, Appendix A, 
page S37). The manner of making this test is explained in Appendix A, page 838. 

7g. Time of Setting. — The time of setting of a cement may vary within wide 
limits and is no certain criterion of quality, but it is important in that it indicates whether or 
not the cement can be used advantageously in ordinary construction. A cement may set so 
quickly that it is worthless for use as a building material (since handling cement after it com- 
mences to set weakens it and causes it to disintegrate), or it may set so slowly that it will 
greatly delay the progress of the work. 

Age of cement has a great effect upon the setting time, and tests should preferably be made 
after delivery of the cement on the work. Most cements absorb moisture from the air and lose 
some of their hydraulic property on storage. It also occasionally happens that the gypsum 
added in manufacture loses its effectiveness in a short time, and in consequence the cement 
becomes quick setting. The cause of this loss of effectiveness of the gypsum is due usually to 
the composition of the cement and may be remedied by increasing the lime content. 

Aside from the consideration of age, the conditions which accelerate setting are: finely 
ground and lightly burned material; dry atmosphere; small amount of water used in gaging; 
and high temperature of both water and air. Since the time of set is influenced by so many 
factors, tests should always be made with extreme care under standardized conditions. 

There are two distinct stages in setting: (1) the initial set; and (2) the hard or final set. 
The best cements should be slow in taking the initial set but after that should harden rapidly. 
Portland cement should acquire the initial set in not less than 45 min. when the Vicat needle 
is used (see Appendix A, page 837), and hard set in not more than 10 hr.' The time of initial 
set is controlled largely by the amount of sulphate (gjrpsum or plaster of Paris) which is added 
in making the cement. 

A cement has taken its initial set when it wiU not thoroughly reunite along the surfaces of 
a break. It has taken its final set when it begins to have appreciable strength and hardness. 

There are two methods in common use for finding the time of setting. The method usually 
preferred is by the use of the Vicat needle apparatus explained in Appendix A, page 838. The 

> Copies of Tech. Paper 48, the publication upon this subject, may be obtained, free of charge, upon appli- 
cation to the Bureau of Standards, Washington, D. C. 
* See standard specifications in Appendiz A, p. 833. 


other method is by the use of the standardized Gillmore needles described in Appendix A, 
page 841. 

7^. Tensile Strength. — The testing of cement in tension is to obtain some meas- 
ure of the strength of the material in actual construction. In other words, tests of tensDe 
strength are made primarily to determine whether the cement will be likely to have a continued 
and uniform hardening in the work, and whether it will have such strength when placed in 
mortar or concrete that it can be depended upon to withstand the strain placed upon it. 

The small shapes made for testing are called briquettes (see details of standard test piece in 
Appendix A, page 842) and have a minimum cross-sectional area of 1 sq. in. — that is, at the 
place where they will break when tested. Standard mortar used in testing is composed of 1 
part cement to 3 parts of standard sand^ from Ottawa, HI. 

It is customary to store the briquettes, immediately after making, in a damp atmosphere 
for 24 hr. They are then immersed in water until they are tested. This is done to secure 
uniformity of setting, and to prevent the drying out too quickly of the cement, thereby prevent- 
ing shrinkage cracks which greatly reduce the strength. 

Specifications for tensile strength of cement usually stipulate that the material must pass a 
minimum strength requirement at 7 and 28 days. This is required in order to determine the 
gain in strength between dififerent dates of testing so that some idea may be obtained of the 
ultimate strength which the cement will attain. A first-class cement, when tested, should 
give the values for tensile strength stated in the standard specifications (see Appendix A, 
page 833). 

7i. Relation between Tensile and Compressive Strength. — Since cements are 
rarely depended upon to withstand tensile stresses, the test for tensile strength has undoubtedly 
become standard on accoimt of the popular belief that there exists a more or less definite and 
constant relation between the tensile and compressive strengths. It can be shown, however, 
that the ratio of compressive to tensile strength of cement mixtures is by no means constant 
at all ages and varies greatly with different cements and with different mixtures. Thus the 
tensile strength cannot usually be regarded as any more than a very approximate indication 
of the probable compressive strength of the same cement. 

7j. Compressive Strength. — Compressive strength of cement mortar is undoubt- 
edly a better criterion by which to judge the suitability of a cement for use in construction. 
The American Society for Testing Materials has tentative specifications and methods of tests 
for compressive strength of Portland-cement mortar* which, when adopted as standard by the 
Society, will be inserted in and made a part of the American Specifications and Methods of 
Tests for Portland Cement. A foreign standard specification is as follows: 

''Slowly setting Portland cement shall show a compressive strength of at least 120 kg. per 
sq. cm. (1710 lb. per sq. in.) when tested with 3 parts by weight of standard sand, after 7 days' 
hardening, 1 day in moist air and 6 days under water; after further hardening of 21 days in the 
air at room temperature (15** to 20°C.) the compressive strength shall be at least 250 kg. per 
sq. cm. (3570 lb. per sq. in.). In cases of controversies, only the test after 28 days is decisive. 

"Portland cement which is intended for use under water shall show a compressive strength 
of at least 200 kg. per sq. cm. (2850 lb. per sq. in.) after 28 days' hardening, 1 day in moist air 
and 27 da3r8 in water." 

7k, Soundness. — A cement to be of value must be perfectly sound; that is, 
it must remain constant in volume and not swell, disintegrate, or crumble. Excess of either 
lime, magnesia, or sulphates may cause unsoundness. The usual method of testing is to form 
a small pat of neat cement about 3 in. in diameter, }i in. thick at the center, and tapering to a 
thin edge. This pat should remain 24 hr. in moist air and 5 hr. in an atmosphere of steam at a 
temperature between 98 and lOO^C. upon a suitable support 1 in. above boiling water. To 
pass the soundness test satisfactorily, the pat should remain firm and hard, and show no signs 

t See Appendix A, p. 841. 

« See Proe. of the Society, vol. zvi (1916). pert I (pp. 590-593). 

Sec. 1-7/] MA TE RIALS 1 1 

of crackingi difltortion, checking or disintegration. The steam test is what is called an acceUr- 
nted test and is for the purpose of developing in a short time (5 hr.) those qualities which tend to 
destroy the strength and durability of a cement.^ 

11, Specific Gravity. — A test for finding the specific gravity of Portland cement 
was originally considered to be of value in detecting adulteration and underburning, but is no 
longer thought to be of much importance in view of the fact that other tests lead to more definite 
conclusions. One trouble has been that specific gravity is not alone lowered by the above 
causes. Seasoning of cither cement or cement clinker, for instance, although known to be 
desirable and in some cases absolutely necessary, lowers the specific gravity materially. On the 
other hand, many underbumed cements show a specific gravity much higher than that set by 
standard specifications. These considerations, together with the fact that the principal adul- 
terants have a specific gravity very near that of Portland cement, make it difficult in the specific 
gravity test to obtain results from which accurate conclusions can be drawn. The test in any 
case is without value unless every precaution is taken to have accurate results, as otherwise 
only very large amounts of adulterated material could be discovered. When the specific 
gravity of a cement falls below 3.10, standard specifications' allow a second test to be made 
upon an ignited sample — ^the idea being that ignition will lower the specific gravity of adulter- 
ated cement. This second test, however, is usually of little value as the ignition loss of most 
adulterants is low and as the specific gravity of an ignited sample of cement is invariably higher 
than that of the original sample. 

7m. Chemical Analysis.* — If the tests of a cement for time of setting, strength, 
and soundness seem to indicate adulteration, resort may be had to chemical analysis. Such 
analysis is not usually made in routine commercial testing. Chemical analysis not only serves 
as a valuable means of detecting adulteration but shows the amounts of magnesia (MgO) and 
sulphuric anhydride (SOO contained in the cement. Specifications usually limit the amount of 
MgO to about 5% and SOs to about 2% because of fear that more of these materials may make 
the cement unsound. 

8. Specifications for Cement. — Standard si)ecifications are given in Appendix A. 

9. Containers for Cement. — Cement may be obtained in cloth or paper bags, in bulk, and 
in barrels. 

Cloth bags are the containers most generally used since manufacturers will refund the 
extra charge for the bags when returned in good condition. The consumer, however, must 
prepay the freight when returning the empty bags to the mill. The cloth bag will stand trans- 
portation, and its size and shape make it convenient to handle. If properly cared for, it may be 
used over and over again. Paper bags are more delicate and have no return value. Wooden 
barrels are advisable when the work is in a damp location, as in marine construction. Bulk 
cement requires special preparations for handling and storage. 

10. Storing of Cement — Cement either in containers or in bulk should be stored within 
a tight, weather-proof building, at least 8 in. away from the ground and an equal distance from 
any wall, so that free circulation of air may be obtained. In case the fioor of a storage building 
is laid directly above the ground, it would be well to give the cement an additional 8-in. eleva- 
tion by means of a false floor, so as to insure ventilation underneath. The cement should 
further be stored in such a manner as to permit easy access for proper inspection and identifi- 
cation or removal of each shipment. When cement is not mill-tested, a proper period before 
cement is needed should be allowed by the contractor for inspection and tests, this period being 
determined by the provisions of the specifications governing his contract. 

Where cement in bags is stored in high piles for long periods, there is often a slight tendency 
in the lower layers to harden, caused by the pressure above; this is known as warehatue set. 

1 The Lackawanna Railroad Co. requires that Portland cement uaed in its Btruetures shall remain sound after 
being subjected to boiling under a 20-atmosphere pressure. This is called the AtUoelavt TeMt. 

* See Appendix A, p. 833. 

* For method to be followed in making a chemical analysis, see Appendix A, p. 834. 

12 ffrfkh7h hS^^IS thty HA S :jb*PfK "Sec- 1-1 1 

*a^ti «^it ^f<^ ">» » *r>i>i <.irfv^ '^ ,r^ '•2H£4 tc^ ^^1=^11'. ''ic.: 

,'^,prfrf^9^ u^, '\mX .*/ 'M 't^, *^rfj^^\. fna^ ^^^nexit <»CLtaiiki fir.^r mnhiwntt at bxe or loosvHr 

l(^*tH( 't^. ^r ,^j.t^. Irt wux-Y* it at -jm^L l>iim^ tim taut of 

f^4t V/ f< y^fikU; Mid ti'.^A V/ eaHxnttti^ of hxne viurii da 

,< ti0Mtf0i^ «t t^<<« m;..4 hefofvt nhippiiki^ 1»U v^ith the best milfe. tke slock 

»//v m ^rt;rvyii ^/f ni«b t^*At * eh^mx vill be taken on fresh 

^f«#^r^/ife< mAf be himpy, b^n t^je lamps aie casilj broken iq>. If. bomrvr, the cement 

^j^^tn «'ibj^9^'t^ U/ ^^xtifamk'r^ daropoefliy or has been vet, faonps w3I be formed which sie hnrd 

ft/»d diffi^jlt to frrifih, A dMtmetKn shoold be mjM&e, so that the latter wiQ not be used with- 

ff$i wifiun^ uti4 TP^ysHiffD of hardened portions. 

IS* Us# of Mfc C«m«rt« — Within the past few jean eonsideiabfe eement has been shipped 
Iff ^/tjlk Uf iyKfttt^iUpfiAtyrt faetr^ries and to cfmgtmeiioa jobs adjacent to raifaoad tracks. Eron- 
fmty kaff, in tb#9ie instaoeeM, resulted from the saving in labor, and from theeliminatiaa of pack- 
MK^ I//MM and expense. There iieems to be no difficulty in diipping balk cement in tight box car?. 

i$. Weight of Cement — ^A barrel of Portland cement weighs 376 Ib^ not including the 
t^rrel, and a ^mg of Portland cement wei^is 9IIb.; in other words there are 4 bags to a barrel. 

A barrel of natural r;ement varies in weight according to the locality in which it is manu- 
fa/'t4jred< A barrel (A Western conent tisuaOy weighs 265 lb. and a barrel of Eastern cement 
Htlt) lb, A ^iag (ff natural <;ement is usuaUy one-third of a barrri. 

A barrijl of puz^olan cement is usually assumed to contain 330 lb. net, and there are 4 
fmgs tf> the barrel, 

A r>ernerjt barrel wf;fghs abr>ut 20 lb. on an average. 


14* Deflnitlooe. — ''Aggregates'^ is a general classifying term applied to those inert {i.e., 
t'hnmiv.nlly ina^:tive} materials, both fine and coarse, which, when bound together by cement, 
form the sulMitance known as concrete. Fine aggregates are materials such as natural sand or 
fiH'.k fwriMmings. Coarw*, or large aggregates, or ballast are materials such as natural gravel, 
cnjshod HK'k, or by-prfxluct materials such as cinders or crushed blast-furnace slag. 

10* General Requirements. — ^Aggregates, fine and coarse, compose approximately 90% 
or mom of tho sulMtance of concrete. From this it follows that the properties of aggregates 
in list rorresp<jnd and \ye at least equal to the properties desired in the concrete. 

Thn usual Nc*rvicc requirements are that aggregate shall be dense, hard, durable, structur- 
ally strong and, for aggrogaten in concretes exposed to water action, insoluble. Further, since 
ronrrrtn \n formed by l)on(iing of aggregates with cement, they must permit by their physical 
rliAract4triNtics (such as roughnosM) the adhesion of cement; and always all particles must be 
climn, NO that a surface* coat of one kind or another may not prevent physical contact with oe- 
uitMit, or destroy its properties through chemical action. 

16* Claieiflcatlon of Aggregates. — The usual cUssification of aggregates is into two divi- 
sions, based upon sixes 

rofjr«c OQurrytUen an* hII particles of gravel, crushed stone, or other materials 
above ^4 in. diameter. 

Fine agQftgtiitn arc all particles below }^ in. in sixe. Particles of such sise are 
further divided by defining '*sand" as all mineral particles from 2 mm. (^ in.) to 0.5 mm. in 
diameter; "silt," all particles from 0.5 mm. to 0.005 mm. in diameter; ''clay," all particles 
having a diameter less than 0.005 mm.; and "loam" as a mixture of any of the above finer 
varii^t ira with organ io mat tor — i.s., of vcKetable or animal origin. It is particularly such organic 
matter rather than sise of particle which rcndcra loam unfit for concrete work, as through some 

Sec 1-171 MATERIALS 13 

chemical action notyet fully understood, possibly through fonnation of anorganic acid, it injures 
OT inhibits the proper action of cement. 

17. Qualities of Fine Aggregates. — General. — When it ia remembered that the finer natural 
materials are derived from rocks by disintegration and by "weathering," or breaking down 
through frost action, water and wind erosion, or kindred agencies, the differences in quality 
sooft«nfoundinsanddeposits, wi th possibly the presence of foreign materials, are not surprising. 
Rirther, sands necessarily partake of the qualities of the rock from which they are derived. 
Silicious quartz sands are best for concrete work, but crushed sands from any durable rock will 
answer, if natural sand of proper quality cannot be obtained. 

18. Qualities of Coarse Aggregates. — General. — For coarse aggregates, any crushed rock of 
durable character, or any clean, hard, natural gravel not subject to ready disintegration may 
properly be used. In general, the better the stone or gravel, the better the resulting concrete. 
For this reason, granite, trap, or hard limestone are preferred for large aggregates, but any rock 
will serve which is sound, which has adequate strength and does not contain objectionable 
mineral inclusions liable to decompose, such as 

iron pyrites, FeSi, which may form sulphuric 
acid by oxidation (see Fig. I). 

Since the properties of any concrete are so 
closely related to the properties of its compo- 
nents, it is essential to an understanding of the 
value of any stone as an aggregate that something 
be known of the origin, nature, and properties 
of the .varieties in common use. 

19. Materials Suitable for Coarse Aggr«- 
gates. — Roughly, rocks suitable for use as aggre- 
gat«fall into three groups. These are: (1} Gran- 
ite and other igneous rocks; (2) sandstones and 
other sedimentary rocks; (3) limestones and re- 
lated rocks. A fourth division comprises slates 
and shales, but as these weather rapidly with for- 
mation of clay, they ore unsuited for use in con- 

firetti. Yia. 1.— Iron pyrilca (Fe6i) in soncntc aurcHBO^ 

The physical character of a rock depends ^|«njs^^ *l«t"yins Bdi«™nt mmtrii. (M.«iii- 
upon two things — its mineral constituents and 

its structure. If the mineral constituents are themselves durable, but massed together in a 
manner structurally weak, rapid weathering, with formation of sand through liberation of 
mineral grains, is to be expected. Such a rock would make a poor concrete. On the other 
hand, a dense structure with like mineral constituents would make an excellent aggregate. 
A dense structure and weak mineral constituents are sometimes associated, but Nature has 
generally cared for such rocks by bringing about their decomposition, so that they exist only an 

SO. IgneouB Rocks. — Igneous rock is a general term descriptive of all rocks formed from 
molten matter which has consolidated either into mineral, or glass, or both. Among such 
rocks are granite and trap rock. Many classifications of a more-or-less satisfactory nature 
have been devised; but all sorts of gradations exist between the various types, rendering their 
descriptive identification difficult, 

30a. Granite. — Granite is well-known by its characteristic appearance (sec 
Fig. 2). In structure, it is a blend of quarti (crystallized silica dioxide), orthoclose, and mica, 
though this latter may he replaced by hornblende. It ia exceedingly den!*e, hard, and durnhle, 
consisting entirely of minerals wilh no glass or uncrystallized material Iwtwccn its constiturnl 

Granites possess the strength and durability desirable in an aggregntr, but they are of low 


toughness. Inaddition, if used inconcret«s exposed to more than ordinary heat, asin chimneys, 
there is a decided tendency to disintegrate, due to unequal mineral grain expansions. Granites 
are not often used as aggregate, their ornamental value precluding less profitable use. 

206. Trap Rock or Diabase. — Trap rock (Fig. 3) and fine-grained basic and vol- 
canic rocks are generally hard, of high abrasive value adhering well to cement. These rocks 

(NatuTBl BUe.) (Magnified 00 dianu.) 

Fia. 2.— Granite. 

have a closely interlaced mineral structure and generally good resistance to stress. Care should 
be taken not to choose a trap rock having a considerable percentAge of iron present in low oxide 
form, as this may absorb oxygen, forming a higher oxide, with expansion and probably rupture. 
In general, trap rock (and rock of similar character, in which class are included many of tho 
"green-9t«nes") makes a very excellent aggregate although in some respects its cxcellcnco 

8ut{ac« appauance. Intenial atnictuie. 

(Natunfun.) (Magnified eO diaru.) 

Fia. 3.— Trap idcIi. , 

has been exaggerated. It baa, however, a very high compressive strcnglli and, as this quality 
is very desirable in concretes, its use has become widespread. It is not always procurable 
without excessive coat but, where price is not prohibitive, its use is advantageous. 

SI. S«dimentai7 Rocks. — To the sedimentary series of rocks belong all those solidified 
depoHits which have accumulated at the bottom of bodies of water. Originally, these materials 
were derived from the land surface and transported to the sea or lakes, either by mechanical 
carriage, or by solution in water. Many of the minerals contained in sedimentary rocks were 


derived directly from the decay of igneous or volcanic rocks. Although additional chemical 
changes supplementing this more-or-less complete decomposition of the original minerala may 
have resulted in the formation of new minerals found in the sedimentary series. With passage 
of time and the action of various chemical and mechanical t^encies, these sedimentary deposits 
solidified into the stratified rocks of one kind or another found throughout the entire surface 
of the earth. 

21a. Sandstone. — One of the moat important of the sedimentary rocks is sand- 
stone (F^g. 4). In structure, it is natural concrete, composed of finely divided mineral parti- 
cles, cemented together in more-or-less close relation by iron or alumina or by calcium com- 
pounds. The character of any sandstone depends, therefore, on the mineral charact«r of ita 
component grains; oa the size and shape of these grains; on their arrangement within the rock; 
and on the nature of the material cementing them together. 

Quartz particles form by far the greatest percentage and the most desirable constituent of 
sandstone. Feldspar is also frequently present, and occasionally, hornblende, chlorite, garnet, 
magnetite, and calcite. In the best sandstones, the grains are arranged uniformly through the 

mass, although frequently coarser and finer particles are arranged in layers, giving a stratifieil 
appearance to the stone. 

So far as its use in concrete is concerned, the most important feature of a sandstone is the 
nature of the cementing material combining its constituent grains. ArgiUacanis gandstoTtes 
in which the cementing material is Ume (usually lime carbonate) may be crushed with compara- 
tive ease, but they disintegrate rapidly on exposure to weatliering agents, such as water or air. 
Such stone may be readily identified by its effervescence when treated with a drop of hydro- 
chloric acid. Sandtlones cemented by oxide of iron are generally red in color, the shade being a 
rough indication of the amount of iron present. Many of these sandstones disintegrate very 
rapidly on exposure to the weather, forming the so-called "rotten stones" so often found in 

8and»lone8 cemented solely by clay should never be used in concrete, as the simple pene- 
tration of moisture is sufficient to disintegrate them, rendering them practically valueless as 
aggregate. A good accelerated test is to boil }^-in. fragments of the stone in water. Rapid 
disintegration indicates a weak stone, with a tendency to weather rapidly, and unsuited for use 
as aggregate. 

ai6. Limestone. — Limestone is carbonate of lime deposited on the floors of 
bodies of water and subsequently hardened into rock. This precipitation of lime may have 
be«i effected from the water, or through the agency of animal or vegetable life. That is to say, 


some limestonea lire chemicsil preoipit&teg, while others are fonned [ram the shells and other 
hard parts of animals, aa well as from hardened tissues of certain plants. Such plant and animal 
forms fossilized are often seen in limestone fragments {see Fig. 10, page 19), 

Compact limestone (Fig. 5) varies in texture from coarse to exceedingly fine. (It is prac- 
tically impossible to obtain a good photograph of the surface of limestone on account of its 
dark color and uniform texture.) It is only occasionally pure carbonate of lime, usually contain- 
ing greater or less percentages of magnesia. Either magnesian hmestone, or pure calcic lime- 
stone is very well suited for use as a concrete aggregate. Any considerable percentage of clay 
in limestone, however, is very undesirable, aa it softens the rock and renders it very liable to 
disintegration. Limestone is found in many colors. White, gray, yellow, blue, and green are 
those of most frequent occurrence. 

In general, limestone makes a very good coarse aggregate for concrete. When crushed to 

the finer sizes, it has a flaky fracture which renders it somewhat unsuitable for use as sand unless 

it is rerolled. Natural limestone sands are of infrequent occurrence, as limestone is soluble 

to as high a percentage as 00%, so that the usual 

weathering processes result in solution, rather than 

fragmentary disintegration. 

23. HetamorpbicRocks.^Rocksofeitherigneous 
or sedimentary origin have often been subjected to 
such severe treatment in the long course of geologic 
history, that their ordinary character is much altered. 
Crushing of the earth's crust, the weight of overlying 
material, and contact with hot molten rock from the 
interior are among the causes contributing to the 
change. Such rocks are classed as "metomorphic." 
There are many metamorphic rocks, the whole 
group constituting a very high percentage of the sur- 
face of the earth's crust. Some of them are of value 
as aggregates in concrete; while others, notably the 
slates and shales, have a weak stratified structure 
and weather so rapidly that their value in concrete is 
almost nothing. 

The above cla8si6cation gives a general indics- 
Fio. 6.— iBMnal untian of limexoDe. *■'"" *>' *■■"«* "^^ *'"'=*' "*'*'■' cniahed are of value 
(Mscnifinl 20 dikiH.) as concrete aggregates. The first caution to be ob- 

served in selecting them is to be sure that they do nnt 
contain objectionable impurities in their substance; and the second important caution is to be 
sure they are clean. 

2S. Gravel. — Gravel of good quality (Fig. 6) makes excellent concrete (see Tech. Paper 58, 
Bureau of Standards, Washington, D. C). Gravel is nothing more nor less than natural rock, 
broken away from parent ledges and worn round by the rolling of streams. Its natural proper- 
ties, therefore, are identical with the rock of which it once formed a part. Provided it has not 
decayed through being in relatively small masses, the properties natural to this parent rock 
are to be expected of a gravel. The surface of gravel is usually very rough; and from consider- 
ntions of character of surface presented for adhesion of cement, it should produce as good as, and 
even better, concrete than crushed stone. Certainly, there is no reason against its use, provided 
it is clean and of good mineral quaUty. 

It is desired to emphaaiie in this connection that not the least Important of all the qualitie» 
of stone or gravel is its cleanneas. The percentage of concretes in which cement and aggregates 
have little or no adhesion due to a coating of dirt (a coating of "matter out-of-plnce ") is sur- 
prisingly large; and the careless acquiescence of engineers in the use of such materials is result- 
ing in a general inferiority of concrete structures. "Dirt"'in such cases may be visible {is 

Sec 1-241 MATERIALS 17 

when the coating is clay, or of tenacious dust due to crushing) or it may be quite inviaible (such 
as a coating of colloidal, transparent, organic matter) requiring chemic^ procedure for its 
detection. A coating of any character is not to be disregarded, when first-quality concretes 
are desired. At best it is a detriment and oftentimes proves a serious defect, greatly weakening 
the concrete. 

84. BUst-funiace SUg.— Slag from blast furnaces, crushed to proper size, has much to 
recommend it for mass construction. Slag is a hard though very porous material, of high 
compressive strength; and in certain localities is relatively 
cheap as compared to stone of good qualities. Offering 
a rough, pitted surface for the adhesion of cement, it 
produces a very strong concrete, but care should be taken 
that its sulphur content is low, else passage of time may 
bring about disintegration of the concrete. Some steel 
companies exercise great care in the preparation uf slag 
for aggregate, weathering it in thin layers for 2 or 3 years 
before marketing, but the advisability of its use in con- 
cretes exposed to dampness and especially in thin sec- 
tions is yet in controversy (see Proc. Am. Soc. Test 
Mat., 1913). 

SS. Cinders. — Furnace cindera as an aggregate are 
used only in inferior grades of mass concrete, or for 
fireproofing. Cinders have low structural strength, high 

porosity, and oftentimes as an added objection, high sul- -t,vS liiomn^'lTiudiMm* ^ii£mrfl^' 
phur content. In more than one instance, sulphuric acid ^ diinu.) 
resulting from sulphur decomposition in cinder concrete 

floors has eaten away conduits and piping, and has even attacked reinforcing and structural 
steel. Cinder concrete is of value chiefly because of its cheapness and low specific gravity^ 
but discrimination is required in its use. 

36. Haterials Suitable for Fine Aggregates. — All fine aggregates are essentially rock 
fragments, crushed to varying degrees of fineness, either by the natural processes of weathering, 
disintegration, or glacial action, or by man with his machines. Sand deposits arc masses of 
weathered rock minerals, transported, collected, and 
sorted by the age-long action of streams (see Fig. 7). 
From the earliest ages the formation of sand, 
silt, and clay has been going on through the break- 
ing down of rocks. The changes involved in these 
processes are part physical and part chemical. All 
changes produced at or near the surface by atmos- 
pheric agents, which result in more or less complete 
disintegration and decomposition, are classed under 
the general term of "weathering." The action of 
physical agents alone, which result* in the rock break- 
ing down into smaller particles without destroying 
its identity, is termed "disintegration" (see Fig. 8). 
rio. 7— CoocretmMn^ ^^'"'" """""■« On the other hand, the action of chemical agents de- 
stroys the identity of many of the minerals by the 
formation of new compounds, and this latter process is known as "decomposition." Silt 
and clay generally result from decomposition; and, as such chemical change has altered the 
character of the material (usually to its detriment so far as concrete purposes are concerned), 
that is one reason, but not the only reason, against permitting their presence in concrete sand. 
Since coarse sands arc of a size to retain and partake of the nature and properties of the 
parent rock, the structure of at least larger particles should be identical with the structures of 
1 Fiom "EnfliiiMriDC OeoloiT," by Rim uid Wataon. 

18 rO.WHKTK t.SdlSKEHU- HASOBOOK |S«^ l-26o 

Huch rocki; and their strcngih nnd fllnesa for use in runrrrte may he judged with mon tkr leas 
(trcurscy from a runaiili- ration nf Ihf Mrurturv and xlivngth of thoMC rorki kaown to be miitable 
(or concrete work. Tlicrr hit few rook* which do not contain silica in greater or leaser 

BocniiHC nf iu hanltifm and n-HiHtiincc to clicniical nRcnta, quart! or silica is, therefore, the 

cunimoncHt mineral in Hnml, Other tnitieraln siii'h an feldspar, mica, etc., though origjnally 

prcHcnl, JM-cailHc of Iheir lexNcr n'HiHlaiii'c, liikve l>n'n niciir rrndilr dcrompoHed by the action 

of the eleiiienls; and by rciuuin of their mmpli'tc iliMiiileKration with rivultant fine stat« of gub- 

divi^'icin, hiive Ikmti n'nicivcd by wind and water. Quarti cry»- 

liiK Ihc-n-fon-. reiiiiiiu a.i (he most evi.lent Burvivors of the 

|)im'nt r<n-k aricl (heir survival is evidence of their desirable 

S6<i. Special Charmcteristio ol Sand. — A simple 
and illiiNtriitive rxiinipio of n rock from which quartt oand may 
bi- .ieriveil h Kiind-time. This Ktime is built up almost wholl> 
ijf ciutirtu Kr'ii'i", cemented tiigelher by iron oxide, calcium car- 
liuniilr-, or rlny, (irid on the nnture of the cementing material 
Ki... H HUI^ .l..[r.,.ii Ivhif lin- deiKTiili t1i<- HlreriKlli imd hardness of the :<tone. The struc- 
iriirjii!.'r"iMVt'!)m"ii.'«i'r,(Ti„'r''' *""' "^ " '""''' f">'i<l-(one l^ shown in Fig. 9. In this atone the 
cetiietiliiiKn>iitiTiiil is iron oxiae, 

I'niler I'-Hrtiii '■ Ii'iiiim of iim' ii frniii'ii'iit of HiiinUtotie, such ivh would be represented by a 

liirice .fioil Krr hinUi lie inilll lis ii iiiiitcriiil for i-oiicn>1e. Such H i-onditiim would be repro- 

m;,,,-l l.v ..ih,.'ri,MK .■oi,..|i-l., cohliiliiiuK HwU .siiiui to evln-me of hei.t. iia in a fire. Under 

like I'oiiililii'iiH ii w'liil'l I \|»-i'ti'i| tliiit II liirni- rriiKiiicut of (he siime stone would break up, or 

"s|.iill," itml II 1. iiiliipillv foiiiiil lliril HONK' Hiiiiii Kriiiiis ri-|H-iit in iiiiiiiiituri' the behavior of the 
liirj(rr iHcirn iif iiloiii-. Hiirh It I'oii.hlioii of linil is. of course, iinusiiiil and extreme, and would 
not iircjiKlicc till' iim- of nii'li sriml for most |uir|iiisrs. 

If lliii I'l-ini-iiliim iiiiili'iiiil of II sniiilstoiii' siiiiil Ki"in be nilrium I'lirbonatc, it may bedia- 

ceiiii-iit ni mind in ciisy of di-ti-i'lion bv addiiiK a drop of 
niurialii: ..rid mid noliiiic e(rerves,>i.n<r, or the lurk of it. 

<;iiiy ceiiii'iit in II Hiiiidslone is iiiiili; undesirable. It 
Ih freigiK-nlly llio riuui that a siind IIioukIi) to lie qnite per- 
fect tor uwi in c<iii'n>te, by rcitsun of ils whili-ni'^'* and nood 
griulinit in nhi:, is in reiility (luile diiUKerons. Cluy U not a 
HtronK cement, and a mmd of whiidi the piirticli's are built 

up with thin Rementinn material is readily crii-lird. Tliis ^'^i- fl— Haul Mndaione. yuiitt. 
, ,, . .^.11 I't 1 I UTiiiiia cpzncnte<] byiroD oiJdi?. (.Nat- 

lit wtpcrially tnic m concrete, for the clay reiidily absorlw uralBJic.) 
water and becomes a xoft paste, leaving the nimponcnt 

uand grains loose and without coiitncl with th(! I'ortliind cement save at the outer surface of the 
outside particles. This, of course, weakens the concrete seriously if all the sand is of this same 
general character. 

Not all sanda, however, are composed of mineral grains, os are those previously mentioned. 
It ia not infrequently the coJie that the rock from which they came has been formed by the 
fossiliiing, or partial foH.silizing, of minute prehistoric, shells. A section of limestone, built up 
in this way, is shown in Fig. 10. Other rocks contain like fossil materials in combination with 
quartz grains and cementing material, as in fossiliferous sandstone. In many sands derived 

See. 1-2661 MATERIALS 19 

from such rocks the structure of the shell ia so perfectly preserved that the fossils retain 
their hollow structure and, further, are easily decomposed by agents which either reach them 
before their incorporation in the concrete, or afterward, by dissolviag out the softer portions. 
The use of such a aand, therefore, is not advisable in concrete which is inteoded to be irapervious 
to water, or to possess a high strength. 

Mb. Crushed Stone and Screenings. — Crushed stone screenings, when free 
fram clay, usually make excellent sand. These screenings ordinarily give a stronger mortar 
1 han natural sand but are likely to contain an undue amount of dust, especially when obtained 
from soft stone, and should be screened and washed to get rid of the finest particles before being 
used in mortar or concrete. 

Crushed limestone makes a concrete of excellent early strength, provided the crushings are 
rerolled, as limestone breaks with a flat, scaly fracture, giving particles that are structurally 
weak and that are very hard to compact in the manner necessary to 
give an impervious concrete. For work exposed to water this 
point is of great importance. f\irthermore, if there is porosity in 
such concretes, the high solubility of the limestone fragments in 
water is a further disadvantsge. In such cases, percolation pro- 
reeds Bt an increasing rate with passage of time, due to bodily 
removal of the fine aggregate by solution, leaving a honeycomb 
structure behind. 

26c. Sea Sand. — Sea sand is usually well suited 
for use as fine a^regate for concrete, so far as structure, mineral 
composition, and cleanness are concerned. It is, however, usually 

of such fineness that its use is inadvisable if undiluted by coareer p .. ii i,^ - ■- 

particles. Saline deposits on the grains, when derived from pure tume. (Msga>B«d20iii>Tng.) 
sea water, should not be of a nature detrimental to concrete. It 

is unwise to take such sands close to tide limits, as the newer sands close to water, t«em 
with minute organic life. 

39d. Standard Sand. — The standard sand used in tests of mortars is a natural 
sand obtained at Ottawa, 111., passing a screen having 20 meshesand retained on ascreen having 
30 meshes per lin. in., prepared and furnished by the Ottawa Silica Co. at a cost of 2 cts. per 
lb., f.o.b. cars, Ottawa, 111. The grains of this sand are rounded and readily compacted, the 
percentage of voids being about 37%. 

It is to be noted that standard sand gives about the lowest value attainable with sand in 
combination with cement, because of its uniform size of grain. Yet the present acceptance 
tests for a commercial sand provide only that it shall in like combination with a like cement, 
attain not less than 757d of the strength of the lowest value obtainable. This standard is 
decidedly low and permiUf the u-sc of almost any sand, even one of poor quality. 

27. Requirements of Fine Aggregate as to Sliape and Size of Particles. — It is exceedingly 
difficult in choosing a fine aggregate for concrete work to balance all considerations. Time is a 
factor of utmost importance in all construction operations. Therefore, where delays would 
be entailed by the selection of one sand, the quaUties of which are superior to those of another 
sand that is more readily obtained, it is more than probable that considerations of superior 
quality will have little weight. It is unfortunately true that regardless of all that has be<:onie 
known in regard to the importance of sands, their quality will be generally disregarded in favor 
of cheapness or convenience until engineers and ownera demand and insist upon concretes of 
proper quality and refuse payment for those not cominguptostandard. When inferior materials 
at a less price are as readily marketable when incorporated in concrete as first-grade materials, 
the contractor, as vendor, is not to be censured if he realizes every opportunity afforded htm 
to le^ie the largest possible profit. The buyer and his agents receive only what they demand. 
Usual specifications for concrete sands permit of little discrimination on the part of the 
supervising engineer, however conscientious he may be. Provision that the sand shall be "clean, 


sharp and coarse'' means nothing, as no standards are defined as comparisons and the deter- 
mination is left solely to the judgment of individuals oftentimes quite incompetent and unskilled. 

Sharpness as a quality requirement for sand is archaic. It has little or no definite meaning; 
and rarely are two individuals agreed as to how sharpness should be determined. To some it 
defines the sound given off when sand is rubbed in the hand. To others, it is measured by abra- 
sive quality, determined in the same way. To others, it indicates a certain angularity judged 
solely by the eye. If it were but remembered that all natural sands are water-borne and water- 
worn, with inevitable rounding of grains, the fallacy of '' sharpness," whatever its interpretation, 
as a standard of quality in natural sands, would be evident. 

Cleanness in sands is most important, for reasons before given. Not all dirt coatings on sand 
are detectable, short of laboratory procedures; and unfortunately, much sand is judged as to 
cleanliness by rubbing in a hand that itself is usually none too clean, the fitness of the sand being 
judged by the deposit it leaves behind. Judging a sand in this way without supplemental 
tests betokens ignorance, or carelessness, or both. Cleanness is an imperative necessity, but 
it should be judged by adequate tests, not by such haphazard methods as the foregoing. 

Coarseness in sands, as opposed to excessive fineness, is a desirable quality, but coarseness 
alone, without finer materials and especially when judged without standards, is no criterion 
of fitness for use in concrete. As before pointed out, coarse sands have less surface area than 
have fine sands, requiring less cement and being more readily coated. Such a requirement, 
if properly judged, is therefore advantageous. As an insurance against excessive clay or loam, 
the requirement of coarseness may also be of benefit. 

What really is needed is: (1) a general and thorough understanding by engineers and 
contractors alike as to the fundamental relations existing between the various materials forming 
concrete (see chapter on ''Proportioning" in Sect. 2); (2) an appreciation of the importance 
of sands in the production of good concrete; (3) their selection on a basis of quality; and (4) to 
make the foregoing of value, a rigid insistence upon conformity to standard by tests of earh 
shipment made to the job with ruthless rejections of inferior materials. A single test on a 
sample which may or may not be representative of the bulk of material is of no value whatever, 
unless it is supplemented by comparative tests on the materials actually delivered. 

28. The Selection of Sand. — The only logical procedure in the selection of a sand for con- 
crete is: 

1. Determine its granulometric analysis by screening. 

2. Determine its cleanness by washing, or by chemical tests. 

3. Determine its actual strength value in concrete by test. 

4. Check all shipments for cleanness and uniformity of grading. 

89. Requirements of Coarse Aggregate as to Shape and Size of Particles. — Since stone is 
one of the strongest, if not the strongest constituent of concrete, the greater the percentage of 
stone (i.e., the nearer concrete actually approaches natural stone in strength and density) the 
stronger is the concrete. It follows, then, other things being equal, that the larger the stone, 
the stronger will be the concrete, since each piece of stone has greater mass density than would 
its components unless compacted and united by Nature's unapproachable processes. 

There are, however, certain limitations as to size of stone imposed by certain classes of work. 
In reinforced work, the plastic concrete must fit itself closely around the reinforcing metal, so 
that 1 to IH in. is the greatest diameter of particle that experience demonstrates is advisable 
to use. 

Concrete of this character obviously requires more cement than would concrete using 
larger stone, since the stone surface to be coated is greater. In mass work, on the other hand, 
crushed stone of 2}^ to 3 in. diameter may be advantageously employed, with less cement. 
For those reasons, if for no others, richer mixtures are specified in reinforced work and leaner 
mixtures in mass work. It should Ix^ borne in mind, however, thut size of stone is not alone the 
determining factor in this regard, but that grading of stone and size and grading of sand is of 

Sec. 1-30] MATERIALS 21 

even more importance as influencing the quantity uf ccinont required, with a corresponding 
effect on the quality of concrete. 

Plunu are large stone, 5 in. or more in least diameter, thrown into plastic mass concrete, 
largely with the object of using them as cheap space-fillers. Their use is also to be commended 
for the reasons before given, provided that the plums themselves are of the proper quality of 
stone and that they are not of such size as to cut through the concrete section. A safe rule to 
follow in the use of plums is that they shall be of a maximum size such that not less than 6 
in. of concrete shall intervene between them and the forms at any point. In using very large 
plums, this thickness of intervening concrete should be materially increased. 

The shape of particle of large aggregates is of relatively little importance. Cleanliness, 
grading, and character of rock have far greater influence on the concrete than has angularity or 
roundness of particle. 

SO. Impurities in Aggregates. — In order that cement may adhere to sand grains and to 
particles of coarse aggregate, each grain or particle must bear no coating such as would prevent 
either proper chemical action between cement and water, or a proper bond between cement 
and aggregates. 

This requirement applies to both fine and coarse aggregates, but is of relatively more impor- 
tance with respect to the former, as sand in its natural state and as used in ordinary concrete 
construction is more likely to contain dirt in sufficient amount and of such kind to cause appreci- 
able injury than is coarse aggregate. 

Clay and silt are impurities of most frequent occurrence in sand and gravel. These mate- 
rials are the result of decomposition of natural rock of various kinds and it is almost inevitable 
that they should be associated with sand. 

Each of these impurities causes injury to mortar or concrete not only when it exists as a 
coating on the sand or gravel particles, but is equally undesirable when it occurs in such 
amounts, or so unequally distributed, that its extremely fine grains "ball up " and stick together 
when wetted, so as to remain in lumps in the finished mortar or concrete. If, however, the par- 
ticles of these impurities are distributed so that they do not bind together on the addition of 
water; and if they are not contaminated by organic matter, experiments have shown that with 
sand that is not too fine, no serious harm results in lean mortars and concretes from their pres- 
ence to the extent of from 10 to 15%. In fact, either clay or silt are often found beneficial as 
they increase the density by filling some of the voids, thus increasing the strength and water- 
tightness besides making the mortar or concrete work smoothly. In rich mortars and con- 
cretes the density and consequently the strength is lowered by even slight additions of clay or 
silt as the cement furnishes all the fine material that is required. 

A coating of organic matter on sand grains, such as loam, appears not only to prevent the 
cement from adhering but also to affect it chemically. In some cases a quantity of organic 
matter so small that it cannot be detected by the eye and is only slightly disclosed by chemical 
tests has prevented the mortar or concrete from reaching any appreciable strength.^ Tannic 
acid, colloidal sewage, manure, sugar, tobacco juice are instances of organic contamination 
destructive to concrete. 

Mica in sand or stone is objectionable because of its low mechanical strength and its lami- 
nated scaly structure. Even a small amount of this impurity in sand may seriously reduce the 
strength of a mortar or concrete. Mica is especially injurious in sands for concrete surface 
work as the scaly flakes cause the surface to dust and peel. 

Furthermore, it is to be noted that water does not wet the surface of mica. Necessarily, 
this precludes attachment of cement, so that not only is the material weak of itself, but it lies 
in the mass without attachment, inviting disintegration. 

Mica schist is totally unfit for use as large aggregate, both because of the foregoing reasons 
and also because of its rapid decomposition on exposure to air. 

1 A new proc«B8 for the detection of such coatings is being developed under the auspices of Committee C-0 
of the A.S.T.M., at Lewis Institute, Chicago, fll. by Prof. Abrams and Dr. Harder. See footnote on p. 29. 


Mica above 1% in concrete sands is very objectionable. 

Iran pyrites or fooVa gold — a bright, yellow substance with metallic luster — ^is chemically 
iron sulphide (FejS). This is a very common impurity in stone; and its undesirability lies in its 
ready oxidation in the presence of water with formation of sulphuric acid (H1SO4), which latter 
readily attacks the cement of concrete with disastrous consequences (see Fig. 1, page 13). 

In large fragments of stone, as in coarse aggregate, the presence of a small amount of this 
substance is not objectionable, but when fine aggregates are made from this same rock, the py- 
rites previously isolated are exposed, with resultant increase in quantity and rate of oxidation 
and with corresponding formation of acid. 

Some sand deposits also contain unoxidized iron sulphide, though such deposits are an 

Finely powdered dust present in crushed stone screenings causes approximately the same 
effect upon the strength of mortar or concrete as does the presence of silt or clay in like quanti- 
ties. It is essential for the best work that this dust be removed by screening and washing, 
in the same manner that silt is removed from sand and gravel, though it may later be used 
advantageously in known quantities by recombination. 

81. Size and Gradation of Aggregate Particles. — The weakest and most changeable element 
in any cement (plus water}-sand-stone combination is the cement matrix in which the aggre- 
gates are embedded. The strongest and least changeable elements are the sand and stone. 
It follows, therefore, that strong, enduring concrete should contain as large a percentage as 
possible of aggregates consistent with proper embedment and cohesion, together with requisite 
plasticity to permit ready placing in forms. This conclusion also is true from an economic 
point of view, since sand and stone are much cheaper, bulk for bulk, than cement. 

Further, tests of concrete show that, with the same percentage of cement to a unit volume 
of concrete, that mixture which gives the smallest volume and has, therefore, the greatest den- 
sity,' usually produces the strongest and most impermeable concrete. This rule, it should be 
said, does not strictly apply to water-tightness, as permeability is influenced by size of voids 
as well as density. 

81a. Grading of Iftixtures. — Other things being equal, the best aggregates, fine 
and coarse, for use in concrete are those which are so graded in sizes of particles that the per- 
centage of voids, or hollow spaces, in the resulting concrete is reduced to a minimum. The same 
law applies also to mortar mixtures, so that if concrete is considered as a quantity of relatively 
large stone set in a pudding, or bedding of mortar, the best sand as to size is one which, if mixed 
with the given cement in the required proportions to standard consistency, will produce the 
smallest volume of mortar; and the best concrete will be one in which the particles of stone are 
BO graded aa to permit in a given volume a maximum quantity of stone being bedded in such 
a mortar.* 

816. Grading, Density, and Strength. — ^It has been found that the densest mix- 
ture occurs with particles of graded sizes; and also that the least density occurs when the grains 
are all of the same size. Coarse sands, or fine sands alone are thus inferior to graded sands for 
concrete, but of the two extremes the coarse sand is preferable because its particles are more 
readily coated with cement particles. Further, a coarse sand has a less total grain surface in 
a unit volume than a fine sand, even when the sands considered contain the same proportion 
of solid matter and voids. Less total grain surface means less cement and less water required 
to coat the grains. Furthermore, these interactions are cumulative, for the additional amount 
of cement and water required in the case of fine sand reduces the density of the resulting mortar 
and likewise its strength as well as increasing its cost. 

81c. Money Value of Grading. — One reason is here evident, both from an 
engineering as well as from a purely doUars-and-cents standpoint, that care and attention bo 

t The density of a mortar or conerete m here referred to ia the ratio of the volume of the solid partielea to 
the total Tolume. 

* P<» tha uee of 6 to 8-iii. acsrecate, see Sng. Rec., May 1 . 1915. 

Sec. 1-32] 



given to the grading of sands for concrete. If it were possible to compute the total saving in 
the annual concrete production of the United States, both direct (by lessened quantity of cement 
required) or indirect (by increased durability and usefulness and prevention of disintegration) 
that would result from the use of proper sands, the amount would be almost beyond imagina- 
tion. It is probable that from this neglect alone not less than 20% of the total annual expendi- 
ture for concrete is unnecessary waste. It has been proven in England that on a strict 1:2:4 
basis, using different aggregates, the cement required for a quantity of concrete varied from 
100 to 130 bags — a difference of 30% in direct cement cost, without counting the variation in 
quality of the several concretes, with like variation in their durability and value. At the pres- 
ent time, a comfortable ignorance is general among engineers and contractors alike on these 
important matters, but in the not distant future, an awakened intelligence on the part of all 
will demand reform. 

82. Mechanical Analysis of Aggregates. — ^The value of an aggregate, sand or stone, with 
reference to its size may be determined by means of a sieve analysis. This analysis consists of 
sifting the material as supplied through several different sieves, and then plotting upon a dia- 
gram the percentage by weight which is passed (or retained) by each sieve — abscissse (hori- 
zontal) representing size of grain and ordinates (vertical) representing percentage of any size 
passing each sieve. ^ Such a sieve analysis may appear of little use as regards the making and 
placing of 100,000 yd. of concrete, but experiment has developed definite laws establishing the 
relation of percentages and sizes of particles to maximum density and strength of concrete so 
that such a sieve analysis may be directly translated into terms of commercial and engineering 

A typical analysis of three natural sands — a fine, a medium, and a coarse sand is given 
in Fig. 11.' Uniform grading is indicated by an approach to a straight line and the variation 
from the grading found to give best results in practice is observed without difficulty. 

A mechanical or sieve analysis is also useful in studying the size of the particles of the coarse 
aggregate.* Fig. 12 illustrates the analysis of a bank gravel and a crushed stone as it came from 
the crusher, without screening. 

The following mechanical test for sand has been used in the laboratory of the Board of 
Water Supply of the City of New York: 

Samples of the material proposed for use in mortar or concrete shall be prepared for testing by passing them 
through a No. 4 sieve. Of the material passing this sieve not more than 95 % shall pass a No. 8 sieve, not more than 
40 % a No. 50 Steve, and not more than 15 % a No. 100 sieve. 

* It is greatly to be regretted that there is no standard in the United States in matters of this kind. Such va- 
riations make for confusion and waste. A proposed standard, which has much to recommend it, varies successive 
screen openings by a constant ratio of -v/2* The following sixes of sieves are desirable for analysing sand, although 
a very useful analysis may be made with fewer sites: 

Commercial No. 

5 8 










Approximate sise of hole 
in inches. 











* Sieves are given commercial numbers, which agree approximately with the number of meahes to the linear 
inch. The actual sise of hole, however, varies with the gage of wire used by different manufacturers and every 
set of sieves should be separately calibrated. The screen with H-in. openings is generally used for separating out 
large mat.i*rial from sand. The No. 4 sieve with four meshes per linear inch is practically its equivalent. 

* A new portable instnunent for making mechanical analysis of sands quickly in the fidid is now manufactured 
by Kolesch St Co., of New York (see Sn^. Rec., June 26, 1915: also Fig. 2, Sect. 2, p. 60). 

* For ooarse aggregate analysia, the following sixes of sheet brass sieves with round holee are desirable; 3 in., 
2H in.. 2 in.. \\i in.. \\i in., 1 is., f( in., H in. and M in. As is also the ease with a like analysis of sand, a 
straight line on a niedinnical-«nalysis diagram indicates a uniform grading. 

oied, provided ther« 


widt. No, 23 <rin. 
'tie, So. 35 wire. 
I, wide, No. 40 wire. 

Id the Bt&ndard specificatiane far concrete patremeot adopt«d by the American Concrete 
Institute, the fine agKregate is required to pass, when dry, a Bcreen having ^j-in. epeniags. 
Not more than 20% is allowed to pass a sieve having 50 meshes per linear inch, and not more 

Fia. II. — Typical mecbanicsl uulyne of But. medium, and couth iuiiiIs. 

than E% is allowed to pass b. sieve having 100 meshes per linear inch. The coarse aggregate 
is specified aa such as will pass a 13^-in. round opening and will be retained on a screen having 
? j-in. openings. It is also required that natural mixed aggregate shall not be used as it cornea 
from deposits, but shall be screened and used as specified. 

The Joint Committee recommends tiiat not 
more than 30% of sand by weight should paas a 
sieve having 50 meahca per linear inch. 

According to heretofore accepted theory, sand 
for mortar or concrete should have its grains 
uniformly graded. Recent testA,' however, by 
Prof. McNeilly at Vanderbilt Univeraity seem to 
indicate that there should be a jump in the grading 
from the sieve No. 40 size particles to the sieve 
No. 10 size particles. The following conclusions 
h™ were derived from these tests: 

Plo. 12.^Ty^oal nMchiinlol kutyKO at buolt The beet uiinc ol cnini in ■ eommerololly-eieTed ■care- 

■nvel and ctiuhed etone. g^tD ii about ee (oUowe: 63% tobeMnuhtbelweMi the No. * 

udNo. I0uev«;47% finea to be puMd by tbe No. Maieve 
(thii include* (he cement). 
There ie reaeon to bdiev* ibat tbe coane accnsale in concrel 
to tb* ebore; that ia to aay. there abouid be a juiap in aiie from coar 
cnuled material. 

For the method of proportioning by mechanical analysis, as developed by Wm. B. Fuller 
and by the U. S. Bureau of Standards, see page 68. 

'. 27. 1015. (Coneluaiau derived from theae leatf have not aa yel noBted leival 

Sec. 1-33] MATERIALS 25 

Tests lor Speciic Gnfitj el A iaptc^fcs u— -Ihe qwofif gnmtr of « — Menrf k the 
ratio of the wei|^t of an absoliitelj aoHd ant T o i u aae <tf the iwhsreiifie to tbe wei^j^ of a ttnit 
volume of water. This ratio for aggre^tes may be deteiiiufied as folknrs: 

1. By pouring a grven weight of sand into a crren rolame of water and findssg tbe tneroMi' 
in volume of the liquid. (Enough matenal sbooki be used to grre sttfinettt aeeoney J 

2. By suspending pieces of eoarse aggregate by a thread froin ehemkai acaies nad DOtiiic 
weight in air and weight when hanging in ws;ter. (The diflereDee m weight is tbe weight of 
the water which the aggregate dt^ptaeeaj 

Finding specific gravity of partieks of sand and stone is preiimifiafy to one method of 
determining the pereentage of voids. Tbe apectfie gravity of aaad is praetkalfy a constant, 
with a value of 2.65. Tbe ^leeific gravity of gravel is also quite uniform, the avera^ being 
2.66. Average values for stone varies with the kind and the locality, raagiog frois 2A (iHiAd* 
stone) to 2.9 (trap). Cinders have an average apectfie gravity of 1.5. 

Before determining spedfie gravity, sand or fine stone should be dried in an oven at a 
temperature as hig^ as 212^F. until theie is no further loss of weight. If the stone is of a 
porous nature, it rikonld be moistened sufficiently to fill its pores, and then the moisture on ttte 
surface should be removed by means of Uotting paper. Bueh a proeeduie, of course, does not 
determine absohite spedfie gravity but gives a result that should be used in determimng the 
percentage of voids for proportioning eonerete mixtures. 

34. Voids m Aggregates. 

Ma. PevDestage el Voids. — ^The percentage of voids in dry sand ranges from 
28% for a eoaxse, weD-graded natural sand to 40 to 45% for a wery uniform natural or screened 
sand. The range for eoarse aggregates is from 25 to 55%. 

The percentage of voids in aggregates may be determined by two methods: 

1. By determining tbe specific gravity of the sc^ particles and then weighing a given 
volume of tbe aggregate and computing therefrom the percentage of voids. (Pores in porous 
stone should be fflled with water. See Art. 33.) 

Let 8 — Bpea6e gravity, W >» weight per cubic foot of the dry aggregate, and P <" per- 
centage of voida. Then, since water weighs 62.5 lb. per cu. ft., 

'' - ^"^ (» - ^) 

2. By ^mAin^ the amount of water required to fill the voids in a given volume of 
Let V » vohuoe of water required to fill the voids and T * total volume, or given volume 
of the aggregate. Then 

With sand or fine broken stone the percentage of voids by this method should be obtained 
by dropping the aggregate into a vessel containing water. If /C -« volume displaced by the 
aggregate and T » given volume of the aggregate, then 


^ " T 

Pouring water into fine aggregate does not give reliable results because it is physically impossible 
to drive out all tbe air. 

The percentage of voids is considerably affected by the degree of compactness of the 
aggregate. Moderate shaking of coarse aggregate, for example, will reduce the volume of 
voids by as much as 10%. Loose measurement is usually considered preferable for the coarse 
aggregate ance sand and cement separate the stones to a considerable extent in the concrete 
as placed. Voids in sand are usually determined with reference to the dry material well shaken. 

846. General Laws. — 1. A mass of spheres of any uniform sise if carefully 
piled in the most compact manner would have 26% voids. If the same mass of spherw were 



[Sec. 1-34C 

poured into a receptacle and the spheres allowed to arrange themselves, it has been found by 
experiment that 44% would be the smallest percentage of voids which could be obtained under 
the best conditions. 

2. A material having particles all of a uniform size and shape contains practically the same 
percentage of voids as a material having particles of a corresponding similar shape but of a 
different uniform size. 

3. In any material the largest percentage of voids occurs with particles all of the same size 
and the smallest percentage occurs with particles of such different sizes that the voids of each 
size are filled with the largest particles which will enter them. Thus, an aggregate consisting 
of a mixture of stones and sand has a less percentage of voids than sand alone. 

4. Materials with round particles contain less voids than materials with angular particles 
screened to the same size 

84c. Effect of Moisture on Voids in Sand and Screenings. — The percentage 
of voids in sand is greatly affected by moisture. The reason for this lies in the fact that when 
water surroimds a particle of sand it occupies space and separates this particle from grains 
adjacent to it. Since fine sand has a larger number of grains per unit volume, it is more 
affected than is coarse sand. If either loose or tamped sand is mixed with a small percentage 
of water and kept either loose or thoroughly tamped, it will be found to increase considerably 
in volume and weigh less per cubic foot. A maximum volume will be obtained with the addition 
of from 5 to 8% of water by weight. Greater percentages will give a less increase of volume 
until finally when the sand is thoroughly saturated it will have a volume slightly less than the 
original. A sand that has, say, 35% voids, may contain from 27 to 44% of voids depending 
upon the degree of compactness and the percentage of water. Natural sand as it ordinarily 
comes from the bank contains from 2 to 4% of moisture by weight. 

34d. Percentage of Voids Determined by Weight. — The specific gravity of 
gravel particles and of sand grains is usually nearly constant, varying between 2.6 and 2.7. 

Percentages of Voids in Sand and Gravel Corresponding to Different Weights 

PER Cubic Foot 

(Based on an Average Specific Gravity of 2.65) 

Percentages of 
moisture by weight 


per cubic foot of sand or gravel 
























































































Sec. 1-361 MATERIALS 27 

On account of this fact the percentage of voids in sand and gravel may be considered to vary 
inversely as the weight per cubic foot of dry material. Knowing the weight per cubic foot and 
assuming a specific gravity of 2.65, the percentage of voids in dry sand and gravel may be 
readily found as explained under Method (1) in Art. 34a. The percentage of voids in moist 
sand or gravel may be determined in the same manner as for the dry aggregate except that the 
weight per cubic foot of the moist material should be considered as decreased by the weight of 
moisture which the sand or gravel contains. The foregoing table gives percentages of 
voids for sands and gravels of different weights per cubic foot and with different percentages 
of moisture by weight. The table may be used for any aggregate with a specific gravity of 
approximately 2.65. 

36. Tests of Aggregates.' — ^Tests of an aggregate for use in mortar or concrete may be 
divided into two general classes: 

1. Tests to determine the general suitability of the aggregate. 

2. Tests to determine those characteristics of the aggregate which have an influence on 
its general suitability. 

Tests of the first class comprise those for determining the quality of the mortar or concrete 
that can be made from the given aggregate. These tests may be called Teats for Acceptance, 
Tests of the second class include those which may be made to determine the cause of any 
failure of an aggregate to pass the tests of the first class. These tests of the second class, which 
may be called Tests for Quality y are useful not only to discover the cause of failure of an aggregate 
to pass the tests for acceptance, but may be employed to determine the methods of improving 
a given aggregate and of comparing different aggregates as to special characteristics. 

No standards for acceptance tests of concrete aggregates have been established by the 
American Society for Testing Materials, although the need of such standards is now fully 
realized and will soon be satisfied. The most advanced practice in this direction is probably 
that represented by the procedure of the Materials Testing Division of the New York Public 
Service Commission whose methods have been given wide publicity. The standards quoted 
in abridged form below are derived from this source (Eng. Rec,f Jan. 8, 1016 and Eng. News, 
Feb. 4, 1915), 

Fine Aggregatet. — Complete tests of a fine aggregate comprise: 

1 . Determination of % retained on No. 4 square-hole sieve. 

2. Mechanical analysis of portion passing No. 4 square-hole sieve. 

3. Determination of silt by washing on No. 100 sieve. 

4. Determination of silt by deeantation. 

5. Compressive tests of 2*in. cubes. 

6. Microscopical examination. 

7. Weight per cubic foot. 

8. Voids. 

9. Specific gravity. 

10. Reaction to litmus. 

11. Quantitative test for organic matter as indicated by loss on ignition.* 

12. Density in mortar. 

13. Determination of insoluble silica. 

It is seldom necessary to make more than the first six of the above tests and frequently only one or two of them 
are necessary. 

1. The entire sample is screened on a No. 4 square-hole sieve and the % retained is computed upon the basis of 
the original weight of the sample. No correction is made for moisture contained. 

(A No. 4 sieve has clear openings of 0.20 in., while the M-in- sieve recommended by the Joint Committee of 
the National Engineering Societies has 0.2&-in. clear openinpi made by drilling round holes in a plate. The differ- 
ence in results obtained with the two sieves is negligible and either is satisfactory.) 

> See also the foUowing articles on sand testing: 

Clotd M. Cbapman and Nathan C. Jobnson: "The Economic Side of Sand Testing." Sng. Rtc.t June 12, 19 
and 26, 1915. 

Clotd M. Chapman: "The Testing of Sand for Use in Conerete," Eng. New, Feb. 5, 1914, and Mar. 12, 1914. 
Ralph E. Goodwin: "Standard Practice Instructions for Concrete Testing," Eng. Netpt, Feb. 4 and 11, 191iS. 
* A new colorimetric test b being developed. See footnote on p. 29. 


2. The mechanical analysis of the portion of the sample which has passed the No. 4 sieve is made by use of 
sieves Nos. 8, 16, 30, 50, and 100. About 150 grams of this material is separated by the method of quartering, and 
110 grams of this is weighed and dried beneath a gas burner. One hundred grams of the dry material is placed on 
the No. 8 sieve, the other sieves being nested below the No. 8 in the order of increasing fineness, and the entire nest of 
sieves is mechanically agitated. With the agitator used, the amount of sieving is standardised by always turning 
the hand crank 200 revolutions. At the end of the operation the material retained on each sieve, and that which 
has passed the No. 100 sieve, is weighed. The % of the whole sample which passes each sieve is now computed, 
and the mechanical anal3r8is curve is plotted (see Art. 32). 

3. Silt by washing on the No. 100 sieve is determined for a sample obtained by quartering which weighs about 
220 grams in its natural moist condition. The sample is dried slowly at temperatures not greatly above ordinary 
room temperature to avoid baking any clay or similar matter. Two hundred grams of the dry sample is now weighed 
on the No. 100 sieve, soaked in water for a few moments to soften any lumps, washed under a gentle stream of water, 
dried under a gas burner, and reweighed. The % of silt is the loss in weight multiplied by 100 and divided by 200. 

(The washing test obtains the true silt value because it removes clay which adheres to the grains as a coating 
which b not separated by sieving in a dry state.) 

4. Determination of silt by decantation is a test for field use only. About 20 c.c. quartered from a carefully 
selected sample is placed in a 100-c.c. graduated glai^s cylinder with about 30 c.c. of lukewarm water. The mixture 
is stirred with a wire for 30 sec., allowed to settle for 30 sec, and the water decanted into a second 100-c.c. cylinder. 
The sand left in the first cylinder is stirred up with a fresh portion of water and the process repeated. This is done 
4 times. After 1 hr. the volume of silt in cylinder No. 2 and the volume of clean sand in cylinder No. 1 is noted 
and recorded. The % of silt is 100 multiplied by the number of cubic centimeters of silt in cylinder No. 2 and 
divided by the sum of the volumes of silt in cylinder No. 2 and clean sand in cylinder No. 1. 

(The method of decantation is better than the method of allowing the silt to settle on top of the sand in one 
cylinder but is not as satisfactory ss the method of washing.) 

5. Compressive tests of 2-in. cubes are made by the methods recommended by the Committee on Uniform 
Tests of Cement, of the American Society of Civil Engineers (Trana. Am. Soc. C. £., vol. 75, p. 665). Fine aggre- 
gate must not be dried, but the natural moisture is determined on a separate sample, and is counted as a part of the 
water used for mixing, not as a part of the weight of sand. Proportions are 1 : 3 by weight. The consistency em- 
ployed is 60% more water than that required to make standard Ottawa sand mortar of "normal consistency." 
Ottawa sand specimens of the same consistency are made with each set of specimens from commercial sand. Tests 
are made at ages of 3, 7, and 28 days. Specimens whose weights vary more than 3 % from the average are rejected. 
Cubes are stored in water up to time of crushing. A spherical bearing block and two thicknesses of blotting paper 
above and below are used in testing. Since the wet consistency used lowers the strength at early periods, the results 
are permitted to fall below those for Ottawa sand specimens of standard consistency by the following amounts or 
lees: at 3 days — 10%; at 7 days— 5%: at 28 days— 1 %. 

(The wet consistency is used because it more nearly represents working conditions and because some sands fail 
in wet consistencies although satisfactory in "normal consistency." No tensile tests are required because "it is 
thought that compressive tests more nearly represent the conditions of the work and that modem practice is tending 
toward compressive tests.") 

6. Microscopical examination is made for the purpose of detecting the presence of a crust or film of organic 
matter on the grains which cannot be detected by other means. 

7. Weight per cubic foot is determined by using an 8 by 16-in. cylindrical concrete mold and a second cylinder 
of smaller diameter but high enough to have a cubic capacity slightly greater than that of the 8 by IG-in. cylinder. 
The smaller cylinder is placed within the larger one and filled with fine aggregate. It is now drawn out allowing 
the aggregate to flow out at the bottom into the larger cylinder. The material is now struck off level with the top of 
the measure and weighed. After weighing, the material is dried and again placed in the larger cylinder by use of 
the smaller cylinder as before. The weight of the material dry is then determined, the volume being found by 
measuring down from the top of the cylinder. 

8. Voids are determined by using the sample whose apparent volume and dry weight have been determined in 
test No. 6. It only remains to determine the absolute volume of solid matter. This can be computed from the 
weight, assuming the specific gravity 2.65, or it can be found by placing the entire sample in a receptacle containing 
water and weighing the whole, emptying out the aggregate, and reweighing the receptacle filled with water to the 
previous level. The difference in weighings minus the weight of dry aggregate is the weight of a volume of water 
equal to the absolute volume of the aggregate. This divided by the weight of water is the absolute volume oi 
aggregate. A hook gage clamped to the water receptacle is used to determine water levels accurately. Finally, 
the % of voids in aggregate 

100 (Apparent volume — Absolute volume) 

Apparent volume 

9. Specific gravity is determined by slowly and carefully introducing the aggregate into a specific gravity 
flask or a graduated glass cylinder containing water and noting the volume of water displaced by a known weight 
of material. The specific gravity of the aggregate 

Weight of aggregate in grams 

Displaced volome in cubic centimeters 

Sec. 1-351 MATERIALS 29 

10. The reaction to Utmiu domonatrates the presence of injuriouB alkalies. 

11. Organic matter as indicated by the ignition lose is determined as follows :> 

Two 25-gram samples are weighed from a 75-gram sample containing natural moisture which has been pre- 
viously selected by the method of quartering. These are placod in beakers " A " and " B " and each is covered with 
60 c.e. of water at lOO^F. The sand and water in beaker "A" is stirred briskly with a glass rod, allowed 30 sec. 
to settle, and a part of the water decanted onto a previously dried and weighed filter paper. The remaining portion 
is again stirred, allowed to settle and a further quantity of water decanted. The process is repeated until all the 
water has been filtered. An additional 30 c.c. of water at lOO^F. is added to the sand in the beaker and the process 
is repeated again. The filtrate is allowed to drain, and the washed sand is dried under a gas burner. The same 
procedure is followed with beaker " B," using a second filter paper. The filter papers with filtrate on them are now 
dried in an oven at 100^. to constant weight (a temperature of lOO^C. must not be exceeded) and weighed. The 
excess in weight over the dry weight of the filter papers is the weight of silt. The filter papers are now folded care- 
fully and ignited thoroughly in a platinum crucible. The residue is weighed, and the dry washed sand is also 
weighed. Finally, % loss on ignition 

100 (Weight of silt - Weight of crucible ash) 
Weight of silt -f Weight of dry washed sand 

(This test is seldom necessary because more direct results are obtained by tests of 2-in. cubes.) 

12. Density is determined by weighing the relative amounts of cement, sand, and water, according to the 
proportions used, mixing, placing mix in a graduated cylinder and noting the final volume of the set mortar. The 
net weight of this mortar is also determined and compared with the sum of the individual weights of cement, sand, 
and water in the mix. The weight of material left adhering to mixing slab and tools is thus ascertained. This 
loss is apportioned between the cement, sand, and water according to the relative weights of each as orifi^nally 
combined, and the corrected amount by weight of each constituent in the set mortar is thus computed. The 
corrected weights of cement and aggregate in the set mortar are now multiplied by their respective specific gravi- 
ties to obtain absolute volumes and the sum of these absolute volumes divided by the total volume of set mortar 
is the density, or solidity ratio. 

13. The determination of insoluble silica is seldom called for but when required calls for the services of an 
experienced analytical chemist. 

Coarse AggregaUB. — Complete tests of a coarse aggregate comprise: 

1. Mechanical analysis. 4. Voids. 

2. Cleanliness. 5. Specific gravity. 

3. Weight per cubic foot. 6. Crushing strength of stone. 
It is usually necessary to make only thto first two of the above tests. 

1. Mechanical analysis of coarse aggregate is made by mechanically agitating a nest of six sieves, the clear 
openings in the wire meshes of which are 2 in., IH in., 1 in., ^i in., H in., and H in. respectively. A 25-lb. sample 
obtained by quartering a larger aample is used. The material retained on each sieve after 100 bumps of the 
rocker apparatus is weighed, and the percentage passing each sieve is computed. 

2. Cleanliness is judged by inspection only. 

3. Weight per cubic foot is determined by pouring the material slomdy into an 8 by 16-in. cylinder mold from 
a height of 2 ft. above the bottom, striking off the top and weighing. If the material is very wet it is previously 
dried sufficiently to remove surplus water, but not enough to dry out the pores. 

4. Voids are determined by the method used for fine aggregate. 

5. Specific gravity is determined by weighing a number of representative particles of the stone af ttt thorough 
Atyra^ to remoire all moisture in the pores, allowing the material to cool, filling the poreA by boiling in water and 

1 A eolorimetrio test for organic impurities in sands is being developed under the auspices of committee C-9 
of the A.S.T.M. Sodium hydroxide (NaOH) is added to a sample of sand at ordinary temperature and the depth 
of color resulting has been found to furnish a measure of the effect of the impurities on the strength of mortars 
made from such sands. See Circular No. 1 of Structural Materials Research Laboratory, Lewis Institute, Chicago. 

The method for fidd tests is describcni in the circular as follows: 

*' Fill a 12-os. graduated prescription bottle to the 4M-os. mark with the sand to be tested. Add a 3 % solu- 
tion of sodium hydroxide until the volume of the sand and solution, after shaking, amounts to 7 oi. Shake thor- 
oughly and let ptand over night. Observe the color of the clear supernatant liquid. 

" In approximate field tests it is not necessary to make comparison with color standards. If the clear super- 
natant liquid is colorless, or has a light yellow color, the sand may be considered satisfactory in so far as organic 
impurities are concerned. On the other hand, if a dark-colored solution, ranging from dark reds to black is obtained 
the sand should be rejected or used only after it has been subjected to the usual mortar strength tests. 

" Field tests made in this way are not expected to give quantitative results, but will be found useful in: 

1. Prospecting for sand supplies. 

2. Checking the quality of sand received on the job. 

3. Prdiminary examination of sands in the laboratory. 

"An approximate volumetric determination of the silt in sand can be made by measuring or estimating the 
thickness of the layer of fine material which settles on top of the sand. The % of silt by volume has been found to 
vary from 1 to 2 times the % by weight." 



[Sec. l-3(> 

allowinc the material to stand in the water until cool, removing the surface water with a towel, weighing, placijig 
particles in a 300-o.c. graduate, pouring in a sufficient measured volume of water to cover the stone, noting the com* 
hined volume of water and stone, computing the volume of stone alone, and computing the specific gravity by divid- 
ing the original dry weight of the stone in grams by the displaced volume in cubic centimeters. Data are also 
afforded by this test for determining the % of absorption of the stone based on dry weight. 

6. Crushing strength of the stone has never been used as an acceptance test of aggregate. 

86. Notes on the Selection and Testing of Aggregates. — Sand and sionefor important or special 
work should be tested in some tpell-equipped laboratory andj of course, the tests should be made 
before the aggregates are purchased or the concrete mixed. The tests should be upon representative 
samples, and materials should be checked for uniformity <u delivered,. In sampling natural sands 
and gravels the greatest care must be used. 

Sand should also be regularly tested as construction work progresses. Often a sand that is 
found entirely suitable at ttie start will be found entirely different in later deliveries. Proportions 
of materials should change whenever the mechanical analysis shows a decided change in the grada- 
tion of the sand grains. 

Mechanical analyses of sands, or volumetric tests of mortars or concretes made from the given 
sands, sometimes show that a stronger mortar or concrete may be obtained by mix^ing two sizes of 
sands from different portions of the same bank, making the single requirement that a definite pet^ 
centage is to be retained on a certain sieve. Mechanical analyses and volumetric tests are also 
lis^vl in studying two or more sands to determine the one most suited for the gicen work. Pre- 
quenUy a properly proportioned mixture of sand and crushed stone screenings vHU produce a better 
sand for mortar or concrete than either one used separately. 

For the best results, sand for concrete requires more fine material than sand for mortar. 

For maximum waterAighJlness a mortar or concrete may require a slightly larger proportion of 
fine grains in the sand than for mcucimum density or strength. Gravel tends to produce a more water- 
tight concrete than broken stone. 

A high unit weight of material and a correspondingly low percentage of voids are indications of 
coarseness and good grading of particles. However, the impossibility of establishing uniformity 
ef weight and measurement due to different percentages of moisture and different methods of 
handling make these residts merely general guides thai seldom can be taken as positive indiccUions 
cf true relative values. This is especially true of the fine aggregates in which percentages of void^ 
increase and weights decrease with the addition of moisture up to 5 to 8%. 

Aggregates that contain harmful impurities may sometimes be made satisfactory for concrete 
work by washing. 

Some sands which contain impurities have been found to prevent hardening with one brand of 
cement and to give satisfactory results with another brand. 

A chemical analysis of aggregates is desirdble in many cases, and a microscopical examination 
loiU often prove of value. 

Cement adheres more readily to sand grains unth rough, unpolished surfaces. 

Usually an artificial sand or crushed stone will safety contain a greater percentage of fine 
material than natural sand. 

87. Specifications for Aggregates. — The specifications adopted by the Public Service 
Commission for quality of concrete aggregates used in New York City subway construction 
are noted below (Eng. News, Feb. 11, 1916; Eng. Bee., Jan. 8, 1916). 

Fin9 aggregatet shall conform to the following requironents: 
Mechanical Orading. — 

flise of opening square 
holes (inches) 

Commercial number 
of sieve 

Limit of fineness ( % 

passing). Not more 

may pass 

limit of coarseness ( % 

passing). Not less 

must pass 










Sec. 1-381 



Sitt.— >Not ov«r 6 % by dry weight shall pus a No. 100 sieve when screened dry. Not over 10 % , dry weight , 
shall pass a No. 100 sieve when washed on the sieve with a stream of water. 

Both of the above tests shall be made, and neither limit shall be exceeded. 
The following test is for field use only: 

Not over 10% by volume shall be tilt when the test is made by decanting from test tubes (method desoribcd 
in Art. 35). 

Strength in Mortar. — Fine aggregate shall be of such quality that mortar composed of 1 part Portland cement 
and 3 parts fine aggregate by weight will show a tensile and compreesive strength at least equal to the strength of 
1 : 3 mortar of the same consistency made with the same cement and standard Ottawa sand. Fine aggregate shall 
not be dried before being made into mortar, but shall contain natural moisture. 

Organic Matter.*— Itom on ignition shall not exceed 0. 1 % of the total dry sand by weight, nor 10 % of the silt 
obtained by decantation. 

Coaree aggregate for concrete shall conform to the following requirements: 
Me^anieal Orading — 

Sixe of opening 

square noles 


Limit of fineness (% 

passing). Not more 

may pass 

Limit of coarseness 

(% passing). Not 

less must pass 


• ■ • 
















• ■ 

Cleanlineee. — ^All broken stone aggregate must be so free from dust that samples caught as the material falls 
from the conveyor bdt at the plant will be within the limit of fineness. All gravel must be thoroughly washed at the 
plant. ' 


88. General Requirements. — ^The water used in mixing mortar or concrete should be free 
from oil, acids, alkalies, or vegetable matter, and should be of a quality fit for drinking purposes. 
The presence of oils is easily detected by the well-known iridescent surface film. Vegetable 
matter can sometimes be detected by observing floating particles, or by turbidity. Chemical 
determinations are better and more certain. 

89. Examination of Water. — Tests of water for acidity or alkalinity may be made by means 
of litmus paper, procured at any chemist's. If blue litmus remains blue on immersing in the 
water, then the property is either neutral or alkaline; if the color changes to red, then the prop- 
erty is acidic. If there is a dangerous amount of acid present, the change in color will be very 
rapid. likewise, if red litmus changes very quickly to blue, the water will be found to contain 
a dangerous amount of strong alkali. If the change of color is slow and faint in either test, 
the indication may be disregarded. A solution of phenol-phthalein is a delicate test for 

Whenever a water does not appear satisfactory, its effect upon the strength and setting 
qualities of a cement should be determined by direct test on mixtures. 
40. Functions of Water. — ^The functions of water in concrete are: 

1. Water reacts with cement to form a binding material which unites otherwise non- 
cohering sand and stone. 

2. Water operates to flux both dissolved and undissolved cementing substances over the 
surfaces of sand grains and stcme particles (or pieces of gravel), rendering possible extensive 
and close adhesion by carrying these substances into the minute and multitudinous surface 
irregularities of the particles, where they are absorbed as water is later absorbed or evaporated. 

3. Water acts as a lubricant between sand particles and stone particles so that placement 
of harsh and irregular materials in molds and forms is rendered easy. 

1 For eotorimeine test aee footnote on p. 29. 



IScc. 1--4 1 

4. Water itself occupies space in the mass. 

41. Influence of Quantity of Water on Strength of Concrete. — ^The first function of water 
cited in the preceding article is basic and essential to the manufacture of concrete. If there 
is insufficient water, obviously its reaction with cement will prematurely cease; and if there is 
too much water, it is equally obvious that the cementing products may be too dilute to develop 
proper strength since cement depends for its early strength and for a considerable part of its 
later strength upon the hardening of amorphous or glue-like substances. Undue dilution of 
these substances is readily possible, but it is accompanied by impairment of strength, just as 
glue may be a valuable adhesive when of proper consistency, while the same glue, if too dilute, 
may be useless for like purposes although later evaporation may gradually restore its cementitious 

14> 21 

Age ot time of breaking in days 
FkG. 13.-^Tinie-strenKth curves for concretes mixed with different percentaces of wftter. 

value. Cement depends further for its strength upon interlacing crystals; and crystallixation 
takes place only from saturated or supersaturated solutions. If, therefore, excess water is 
added, such strength as these crystals may confer is further impaired. The influence of water 
in greater or less quantities on the strength of concrete is shown in Fig. 13.^ 

42. Influence of Quantity of Water on Fluxing of Cement. — The second fimction of water 
cited above — ^its action as a carrier (or flux) of cementing substances — ^is obvious. In bringing 
sand, stone and cementing substances into intimate contact (Fig. 14), it acts physically in a 
manner analogous to its earlier chemical r61e. 

Inevitably, however, this desirable function of water is closely dependent upon its fourth 

* See L. N. Eowabim: Proc. Am. Soc. Test. Mat., 1917. 
D. A. Abkams: ConcreU (C. M. Edition). July. 1917. 

Fio. H.— Cement nBrticlee 
fluted over surfuH of HUuT gnini. 

Sec. 1-431 MATERIALS 33 

function — vis., its occupaacy of space. It is rmdily accii that if the minute irregularities in 
tlicsurfaceof stone are firat filled with water; and, because of initial excess of water, the cement- 
ing solution is then too weak, a strong, intimate attachment of cement will be inhibited, both 
because the irregularities are already filled and also because of excessive dilution. 

Furthermore, water, particularly when charged with gelatinous aluminates from cement, 
bos ability to occlude a very high perci^ntage of air. This air, as minute bubbles, firmly attaches 
itself to the Bond and stone.' It also remains between parti- 
cles to such an extent as oftentimes to completely isolate a large 
percentage of the materials. Given excess water , therefore, 
and a proportionate amount of occluded air, detriment to con- 
crete is sure to arise from the primary fault in an increasing 
ratio (see Hg. 15). This explains to a large degree the lower 
strengths with prolonged mixing in present-type machines 
found by some investigators. 

4S. Influence of Quantity of Water on Lubrication of 
Concrete Mixture. — The function of water as a lubricant of 
concrete is very important, but its importance can be over- 
estimated, particularly when balanced against the detrimental 
effects which may and often do result. It is not necessary 
to add great quantities oF water to concrete to make it easy- 
flowing if the concrete is sufficiently mixed. The more con- 
crete is mixed, the smoother working it becomes and the less water is superficially evident. 
Cement is continually hydrating in the mixing action ; and in process of hydration, large 
quantities of hydrated lime arc formed. This has a very pronounced effect in lubricating 
the mass, and furthermore keeps it coherent. Excess water, on the other hand, promotes 
separation of the constituent materials, offsetting the good effects of hydration and render- 
ing the concrete extremely harsh in working and 
difficult to handle. . More mixing, therefore, or 
more efficient mixing through improved mixing 
devices, should be relied upon for easy placing, 
rather than excess water. 

44. Influence of Quantity of Water on 
Space Occupied in Resulting Concrete. — The 
fourth function of water in concrete — ^that of 
occupying space in the mass — is so important 
and so varied in its manifestations that a. largo 
treatise would be too small tor adequate pre- 
sentation. A few leading considerations may, 
however, serve to stimulate individual thought 
in this regard. 

There are few substances so incompressible 
as water. Beyond question, although water is 
mobile, a given quantity occupies definite space. 
Concrete in forms iseseentially in a confined space. 
In this form space are cement, sand, stone, and 
tooter, each occupying its proportionate share of 
the total volume of concrete. So long as the form remains tight, these substances must all 
remain substantially in place. If the form leaks, which is contrary to practice, more or less 
water, with a greater or leas quantity of cement in solution, may escape. This escape may be 
before, or nfler the mass has taken either initial or final set. In this latter case, a hollow space, 

<SaO«T«*u>: "Collotilvl Chemietry," p. 70-118. 


or void of greater or lees volume remaina in the concrete where once was water in greater or lest 

quantity or a aolution too dilute to solidify (see Fig. 16). But leakage need not necesBarily 

<i«cur in this way. After forma are removed, any uncombined water or dilute solution will b<- 

fnv to escape, either by gravity, or by capillary suction aided by evaporation, leaving behind .-w 

a void the space it required in the maas. Surh 

action is evidenced in hundreds of concretes 


4fi. Harmful Effects of Voids Caused by 
Excess Water. — From the above it is evident 
that the more uncombined water, the more voids 
in the set concrete. Conversely, the more voids, 
the less the cloeeness of compacting of sand and 
Btone; and the lesa this compacting, the less the 
density, strength, durability and value of tht^ 
hardened mass. Furthermore, the proportions 
of the concrete are seriously unbalanced (see Arts. 
2 and 16, Sect. 2). 

The space loss referred to is only a small 
part of the ultimate damage. Physical stress 
due to loading may be the least intense of the 
stresses to which concrete is subjected. Physical, 
. . or chemical and physico-chemical stresaes set 

(Micnifiat IS dikin*,) Up in the mass after hardening, through dis- 

ruptive freeiing, or through percolating water 
alone or carrying chemically active agents, are of far greater intensity. Each pore, or void, 
is a potential aid to such destructive agente; and enlarged by initial attack, soon become 
an active aid and abettor. First loss, therefore, may be of minor importence. Induced 
wealuiessea, augmenting primary deficiencies, must be reckoned with to an increasing degree. 

46. Excess Water the Cause of Day's Wwk Planes. — Perhaps the commonest evidence, 
te be found on eyery hand, as to the efFecte of excess water in concrete arc "day'fl work planes." 
In the early hfe of a structure such as a buttress wall, these planes are hidden either by a smooth 
mortar surface at contact with forms, or by a later-applied wash or coat of cement plaster. 
But as months pass and the structure is subjected to water action in greater or leas degree, 
from one source or another, theseplanesaremademoreandmoreevidentby seepage along them. 
When such seepage is in quantity, it may be detected aa a film of water, or, with rapid evapora- 
tion, by crysteUine depoate. When seepage is leaa, it may be evidenced by a pateh of efflores- 
cence, but in each case the underlying cause of water passage is "laitence," which is largely 
caused by the use of excess water. 

47. Excess Water the Canoe of Large Laitance Deposits. — "Laitence," or "day's work 
planes," may be of small bulk, relative to the total mass of concrete, yet in some instances 
laitance is found to an exaggerated and oftentimes to a dangerous extent. Whenever forms 
are filled by dumping concretes continuously in one spot, with dependence upon hoeing-down, 
or natural flow for distribution of heavier materials into lower parte of forms, it is inevitable 
that water and the finer materials suspensible in water, including much of the cement, should 
separate from the heavier materials; and that they should form, when solidified above the con- 
crete a depo«t or stratum of greater or less thickness and extent, which will be entirely composeil 
of muck or "laitance." This material is chalky and of low strength. It is very absorbeni ; 
and when saturated is of little better value than so much wet, sandy clay. 

Instences of the formation of "laitence" in quantities and in situations where it is dangerous 
are found in columns poured in two or more sections. Although not approved by building codca. 
contractors, for their own convenience or to save on forms, will sometimes pour half a column, 
allowii^ it to set before continuing to the top. Inevitebly, lighter piaterials rise in the form. 

Sec 1-481 MATERIALS 35 

Necessarily the joint thus formed in the middle of the column is of inferior material; and of 
a material which cannot bond with material subsequently poured. If this procedure is again 
followed, the lighter part of this latter also rises so that a stratified column, with another 
"laitance" section at the column head will result. If after removal of forms this material 
should become wet from any cause, crushing and sliding is to be expected with possibly collapse 
of the column and its supported load. Columns should not only be poured in one section, but 
they furthermore should be poured of concrete of such consistency that "laitance** will not 
accumulate; and it would also be a desirable precaution to overflow the form to remove such 
accumulations as may rise. It is better to waste a portion of material at the top, in order to 
be sure that there may be no ''laitance" at the column head, rather than to have any question 
as to strength or security.^ 

48. Excess Water and Waterproof Concrete. — It is difficult to find a truly waterproof 
field concrete, largely because excess water is so generally used in mixing. The majority of 
structures are of such size that they cannot be poured continuously. This necessarily means 
stoppage of work for greater or less intervals. Stoppage of work with wet concretes always 
means a layer of 'Maitance;" and this inevitably prevents succeeding layers from bonding, 
entailing a chain of consequences. A radical change in such field procedures is demanded, if 
these difficulties are to be overcome. 

49* Excess Water Causes Unsatisfactory Concrete Floor Surfaces. < — ^It is difficult to 
insure that concrete floors shall be dustless. The functions of a concrete floor are to bear loads 
as well as to withstand abrasion and impact, these latter being the severest service to which 
it is subjected. It is unfortunate that the top of a concrete floor is the surface on which depend- 
ence is placed, as in possibly nine cases out of ten, this surface coat, both by virtue of its initial 
consistency and also because of water later brought to the surface through troweling, is partly 
or wholly composed of "laitance.'' 

To remedy these defects, floor hardeners of one kind or another, are added to the concrete 
-in mixing. Few of these substances should have any real place in the concrete-floor industry. 
Most are inferior to quartz sand in hardness and strength, but because of the prevalence of 
unsatisfactory concrete floors and because of the human tendency to escape consequences by 
purchasing immunities, such alleged remedies find ready sale. If, instead of bu3dng integral 
floor hardeners, less water were put into concrete floors and a good quality of graded sand with 
Portland cement used in well-mixed and properly placed concrete, there would be less need of 

60. Excess Water Prevents Bonding New Concrete to Old. — One result that can be gua- 
ranteed is the failure of effective bonding between new concrete and old. Various expedients 
from time to time have been claimed to bring about effective results in this regard, but Uttle 
has as yet been unquestionably accomplished. Washing the surface of old concrete with hydro- 
chloric acid is ineffective and wrong except so far as it may clean off surface dirt and carbonated 
deposits. Picking the surface rarely goes deep enough or covers enough surface. The inher- 
ent difficulty tmderlying all attempts at bonding, is the identical trouble that causes day's 
work planes, or that makes the wearing surface of concrete floors unsatisfactory, i.e., the exist- 
ence of a light, chalky, insecure material, substituted at the critical plane for a substance which 
should be durable and secure. 

61. Excess Water and Concreting in Cold Weather. — Concreting in cold weather is always 
attended by some lidc, even when forms remain in place until milder weather. Heating of 
aggregates is seldom adequate, and the heat transmitted through wooden forms after pouring 
is small in quantity. It should be remembered that at 40^F. the reaction between water and 
cement and the production of cementing strength is only one-fourth as rapid as at SO^'F. and 
less than one-ninth as rapid as at 70^F. Dilution by excess water of such feeble solutions 
increases the danger, as is evidenced by frequent winter failures. Furthermore, at d9**F. 

1 See CtLLMOKs: ''On liioea. HydnaUe CemenU and Mortan/' 1872; p. 242. 
* See Sect. 4, "Concrete Floon and Floor SorfaeeB. Sidewalks, and Boadwnyt." 


some subtle change occurs in water which decreases its chemical ability even before actual 
freezing;^ and at this latter point occurs expansion of 8% by volume, with exertion of some 300 
tons disruptive pressure per square inch of surface. 

The greater the quantity of water in a cold-weather concrete, therefore, the greater the 
liability to dilution, little strength, frost disruptions, and failures. The potency of excess 
water in these respects is just beginning to receive due recognition. 

62. Suggested Procedures to Guard Against Use of Excess Water. — Excess water in con- 
crete should be rigidly guarded against. To insure the use of less water, specifications must 
embody provisions giving the engineer authority for its regulation. To this end, the following 
partial specification is suggested: 

1. Concrete shall be an intimate mixture of sand, stone (or gravel), cement and water 
of the several kinds and qualities herein specified and in proportions as specified, subject to 
modification by the Engineer. 

2. The proportions and quantities of all materials, including water, shall be as directed by 
the engineer and shall be subject at all times to such change as his tests or judgment may 
dictate as advisable. 

3. All materials shall be accurately measured in measures of approved type and known 

Cement shall be measured by the standard sack or, if in bulk, by weight, 94 lb. being 
taken as an equivalent of one sack. Loose measurement of cement is prohibited. 

Sand and stone shall be measured in struck measures of a capacity and type approved by 
the engineer. Measurement in wheelbarrows of a type which do not admit of a struck measure- 
ment will not be permitted. 

Waler shall be measured at each mixer in containers adapted to ready adjustment and to 
accurate delivery of variable quantities. Supplementing the delivery of such measuring con- 
tainers by additions of water, because of slowness of discharge or for any other reason, will not 
be permitted. 

4. Concrete of a plastic consistency shall be required in all parts of the work, unless per- 
mission be given by the engineer for the use of drier and stiffer mixtures. Sloppy and overwet 
concretes are strictly prohibited. The quantity of water, therefore, will he subject to regulation 
at all times by the engineer according to the requirements of the aggregates in use at theU time. The 
rejection and removal of overwet concretes either before or after placing in forms may, at 
the engineer's discretion, be required of the contractor without compensation. 


63. Types of Reinforcement. — The reinforcing steel in reinforced-concrete construction 
is mostly in the form of rods, or bars, of roimd or square cro6S-«ection. These vary in size from 
^ to ^ in. for light floor slabs, up to Ihiio iHin. asa maximum size for heavy beams and 
columns. Both plain and deformed bars are used. With plain bars the adhesion between steel 
and concrete is depended upon to furnish the necessary bond strength. With deformed bars 
the usual adhesion is supplemented by a mechanical bond, the amount of this bond in any 
given case depending upon the shape of the bar. The adhesion of concrete to flat bars is less 
than for roimd or square bars, but the flat deformed bar possesses advantages over other forms 
when used as hooping for tanks, pipes, and sewers where the reduced thickness of the bar allows 
the concrete section to have a greater effective depth for the same total thickness of concrete. 

Wire fabric and expanded metal in various forms are used to a considerable extent in 
slabs, pipes, and conduits. These t3rpes of reinforcement are easy to place and are especially 
well adapted to resist temperature cracks and to prevent cracking of the concrete from impact 
or shock. 

> See O. D. Yak Enoklbn: Ceniury Magtuin*, April, 1917. 

Sec 1-54] MATERIALS 37 

A number of combinaiioiuB of forms are employed to a greater or less extent. Theae 
combinations are known as systems. 

64. Surface of Reinforcement — A rough surface on steel has a higher bond value for use 
in concrete than a smooth surface, consequently a thin film of rust on reinforcement should not 
cause its rejection. In fact in the case of cold-drawn wire which presents a very smooth surface, 
a slight coating of rust is a decided advantage. Loose or scaly rust, however, should never be 
allowed. Reinforcement in this state of corrosion may be used if first cleansed with a stiff 
wire brush or given a bath of hydrochloric acid solution (consisting pf 3 parts acid to 1 part 
water) and then washed in clean running water. Oiling and painting of reinforcing steel 
should not be permitted as its bonding value is greatly reduced thereby. 

65. Quality of SteeL — Authorities differ as to the quality of steel to be used for reinforce- 
ment. Mfld steel is the ordinary structural steel occurring in all structural shapes. High steel 
or steel of hard grade has a greater percentage of carbon than mild steel and is also known as 
high-carbon or high elastic-limit steel. 

Brittleness is to be feared in high steel, although this quality is not so dangerous when the 
metal is used in heavy reinforced-concrete members — for example, in heavy beams or slabs — 
as the eoncrete to a large extent absorbs the shocks and pro- 
tects the steeL All high steel should be carefully inspected 
and tested in order to prevent any brittle or cracked material 
from getting into the finished work. Steel of high elastic Via. 17.— Cold-iwkied wauan hmr, 
limit is seldom onployed where f^ain bars are used. 

Gold twisting increases the elastic limit and ultimate strength of mild-steel bars. The 
increase, however, is not definite, varying greatly with slight variations in the grade of the 
reeled SteeL A square twisted bar is sbown in Fig. 17. 

M. Wofkmg Stresses. — Hmc generally aeeq>ted working stress for mild steel is 16,000 
lb. pa- sq. in. sad 1S,000 to 39,000 lb. per sq. in. for high steel and cold-twisted steel. A stress 
not greater than 16,000 lb. per sq. in. is reeomme&ded by the Joint Conunittee for ail grades of 

ST. CocAcieat cf RipssMS M. — ^Tbe coefBeient of expansion of steel is approximately 
0.0000065 degree Fahreinheit. 

S6u Modidiis ftf Elaslkity. — ^The modulus of elasticity of all grades and kiiMis of steel is 
about the same and is usually taken as 30,000,000 lb. per sq. in. in both tension and compression. 

9iL Stsel Bpecificationfi. — The following ^leeifications are those of the Association of 
American Steel Manufacturers for concrete reinforoenkent bars rolled from billets, adopted 
Mardi 22, 1910 (revised 1912 and 1914k 



1. Manuft»etur€.—^tael nmy be made by either tbe ttpesy-hemrth or Beaaemcs' p too o a a . Ban ahall be ruUed 
from standard new billete. 

2. Chemietd and Phj/ncaJ Froperiiet. — Tbe ebemical and jrfiyaical propertiee oball ooofurfD to tbe limito as 
Bbown is tbe table on tbe foUowiac pas«- 

5. Chemical J^tierminaiione. — lo order to determine if tbe material oonCorma to tbe ebonioal limitations 
in ea mib ed io paiasiap b 2 berein. analyeie afaaU be made by tbe manufacturer from a teet ins^ taken at tbe time 
of tbe pouring of eaeb melt or blofr of ateel. and a correct copy of sueb analyais sball be f umiobed to tbe en^oeer 
or hie inapector. 

4. Yi€id Fmni. — For tbe porpoae c of tbeK apeeificatiooe. tbe yield point aball be determined by careful ob- 
•enration of tbe drop of tbe beam of tbe teating machine, or by other equally accurate method. 

6. Furm igf Specumens. — (a) Tensile and bending test specimens may be cut from tbe bars as roUed. but 
and bending test specimeas of deformed bars may be planed or turned for a leogtb of at least 9 in. If deemed 

by tbe maaolaeturer in order to obtain uniform crossHwetion. 

(6) TeoMlo and ^'^^'4'fv^ test specimens of cold-twisted bars shall be cut from tbe bare after twistingt i^od shall 
be tested in full «i*<' without furtbiT treatment, unless otherwise q>eeified as in (e), in which case the conditions 
thereon stipttlai«<i shall govern. 

(r) If it i^ deaired that tbe testing and aooeptanoe for eold«twisted bars be made upon the hot-rolled bars 
before \mo€ twisted, the bot-roUed bats aball meet tbe requirements of tbe structural-ateel «rade for plain bars 
shown in this specahcation. 



[Sec. 1-59 

Properties considered 



Hard grade 








Phosphorus maximum: Bessemer 

Ot>en-hearth. ■. * 












. 50.000 




180 deg. 
d - 2< 

180 deg. 
d - 3< 

Ultimate tensile strength, lb. ^tx sq. 

Yield point, minimum lb. per sq. in. . 

Elongation. % in 8-in. minimum .... 

Cold bend without fracture: Bars 
under H-in. diameter or thickness. 

Bars M-in. diameter or thickncBs and 

The intermediate and hard gra 

Tens. str. 

180 deg. 

d - 1( 

180 deg. 
d - It 

dee will be 

Tens. str. 

180 deg. 

d - 2( 

180 deg. 
d - 2< 

used only 

Tens. str. 

180 deg. 

d - 2< 

90 deg. 
d - 2f 

when spe 

Tens. Btr. 

180 deg. 

d - 3( 

90 deg. 


Tens. str. 

180 deg. 

d- 3< 

90 deg. 
d - 3t 

Tens. str. 

180 deg. 

d -4f 

90 deg. 
d - 4( 

6. Number of Testa. — (a) At least one tensile and one bending test shall be made from each melt of open- 
hearth steel rolled, and from each blow or lot of 10 tons of Bessemer steel rolled. In case bars dififering H in* and 
more in diameter or thickness are rolled from one melt or blow, a test shall be made from the thickest and thinnest 
material rolled. Should either of these test specimens develop flaws, or should the tensile test specimen break 
outside of the middle third of its gaged strength, it may be discarded and another test specimen substituted there- 
for. In case a tensile test specimen does not meet the specifications, an additional test may be made. 

(b) The bending test may be made by pressure or by light blows. 

7. Modifieatione in Slongaiion for Thin and Thick Material. — For bars less than ^B in. and more than H 
in. nominal diameter or thickness, the following modifications shall be made in the requirements for elongation: 

(a) For each increase of H in* in diameter or thickness above H in. a deduction of 1 shall be made from the 
specified percentage of elongation. 

(ft) For each decrease of M« in. in diameter or thickness below ^e in. a deduction of 1 shall be made from the 
specified percentage of elongation. 

U) The above modifications in elongation shall not apply to cold-twisted bars. 

8. Number of TwitU. — Cold-twisted bars shall be twisted cold with one complete twist in a length equal to 
not more than 12 times the thickness of the bar. 

9. Finiah.—Mtkieirial must be free from injurious seams, flaws or cracks, and have a workmanlike finish. 

10. Variation in Weight. — Bars for reinforcement are subject to rejection if the actual weight of any lot varies 
more than 5 % over or under the theoretical weight of that lot. 

The following specifications are those of the American Society for Testing Materials for 
roncrete reinforcement bars rolled from billets: 

Standard Specifications for Billet-btbel Concrete Reinforcement Bars 

(.\merican Society for Testing Materiab) 

1. (a) These specifications cover three classes of billet-steel concrete reinforcement barB. namely: plain, 
deformed and cold-4wisted. 

(6) Plain and deformed bars are of three grades, namely: structural-steel, intermediate and hard. 

2. (a) The structural-steel grade shall be used unless otherwise specified. 

(6) If desired, cold-twisted bars may be purchased on the basis of tests of the hot-rolled bars before twist- 
ing, in which case such tests shall govern and shall conform to the requirements specified for plain bars of structural- 
steel grade. 

Manufacture.'—^, (a) The steel may be made by the Bessemer or open-hearth process. 

(6) The bars shall be rolled from new billets. No reroUed material will be accepted. 

4. Cold twisted bars shall be twisted cold with one complete twist in a length not over 12 times the thieknes» 
of the bar. 

Chemical Froperiite and Teste. — 5. The steel shall conform to the following requirements as to rhemiral 

Sec. 1-69] 



Phoaphorua, Baaaemer not QV«r 0. 10% 

Opan-haarth not over 0.06% 

6. An analyais of eaeh melt of ateel ahall be made by the manufacturer to determine the pocentacea of car- 
bon, mansaneae, phosphorua and sulphur. This analysis shall be made from a test insot taken during the pouring 
of the melt. The chemical composition thus determined shall be reported to the purchaaer or his representative, 
and ahall conform to the requirements specified in Sect. 5. 

7. Analyses may be made by the purchaser from finished bars representing each melt of open-hearth steel, 
and each melt, or lot of 10 tons, of Bessemor steel. The phosphorus content thus determined shall not exceed that 
specified in Sect. 5 by more than 25%. 

Phyneal Properiie* and Te»t$, — 8. (a) The bars shall conform to the following requirements as to tensile 
properties : 

Properties considered 

Plain bars 

Deformed bars 






' mediate 


Tensile strength, lb. per sq. 

Yield point, min., lb. per sq. 

Elongation in 8 in. min. % > 





















Tens. str. 

Tens. str. 

Tens. str. 

Tens. str. 

Tens. str. 

Tens. str. 


> See Sect. 9. 

(&) The yield point shall be determined by the drop of the beam of the testing machine. 

0. (a) For plain and deformed bars over ^4 in. in thickness or diameter, a deduction of 1 from the percent- 
ages of elongation specified in Sect. 8(a) shall be made for each increase of H in- in thickness or diameter above 
H in. 

(6) For plain and deformed bars under lit in. in thickness or diameter, a deduction of 1 from the percentages 
of elongation specified in Sect. 8(a) shall be made for each decrease of Ms in* in thickness or diameter below 3^s in. 

10. The test specimen shall bend cold around a pin without cracking on the outside of the bent portion, as 
follows : 

Thickness or diameter of 

Plain bars 




Deformed bars 





Under Hin.. 
H in. or over 

180 deg. 

d - t 

180 deg. 

d - t 

180 deg 
d - 2< 
00 deg. 
d " 2t 

180 deg. 
d - 3< 
90 deg. 
d " 3t 

180 deg 
d - ( 

180 deg 
d " 2t 

180 deg. 
d - 3( 
90 deg. 
d ~ Zt 

180 deg. 
d - 4( 
90 deg. 
d -* 4< 

180 deg. 

d - 2< 
180 deg. 

d m zt 

d ~ diameter of pin about which the specimen is bent. 
( " thickness or diameter of specimen. 

11. (a) Tension and bend test specimens for plain and deformed bars shall be taken from the finished bars, 
and shall be of the full thickness or diameter of bars as rolled; except that the specimens for deformed bars may be 
machined for a length of at least 9 in., if deemed necessary by the manufacturer to obtain uniform cross-section. 

(6) Tension and bend test specimens for oold-twisted bars shall be taken from the finished bars, without 
further treatment; except as specified in Sect. 2(b). 

12. (o) One tension and one bend test shall be made from each melt of open-hearth steel, and from each melt, 
or lot of 10 tons, of Bessemer steel; except that if material from one melt differs H in. or more in thickness or diam- 
eter, one tension and one bend test shall be made from both the thickest and the thinnest material rolled. 

(&) If any test specimen shows defective machining or develops flaws, It may be discarded and another stieci- 
men substituted. 



[Sec. 1-59 

(fi) If the percentage of elongation of any tension test specimen is less than that specified in Sect. 8(a) and. any 
part of the fracture ia outside the middle third of the gage length, as indicated by scribe scratches marked on the 
specimen before testing, a retest shall be allowed. 

PermianbU Variations in Weight. — 13. The weight of any lot of bars shall not vary more than 5% from the 
theoretical weight of that lot. 

Finish. — 14. The finished bars shall be free from injurious defects and shall have a workmanlike finish. 

Inspection and Rejection. — 15. The inspector representing the purchaser shall have free entry, at all times 
while work on the contract of the purchaser is being performed, to all parts of the manufacturer's works which con- 
cern the manufacture of the bars ordered. The manufacturer shall afford the inspector, free of cost, all reasonable 
facilities to satisfy him that the bars are being furnished in accordance with these specifications. All tests (except 
check analyses) and inspection shall be made at the place of manufacture prior to shipment, unless otherwise speci- 
fied, and shall be so conducted as not to interfere unnecessarily with the operation of the works. 

16. (a) Unless otherwise specified, any rejection based on tests made in accordance with Sect. 7 shall be re- 
ported within 5 working days from the receipt of the samples. 

(6) Bars which show injurious defects subsequent to their acceptance at the manufacturer's works will be 
rejected, and the manufacturer shall be notified. 

17. Samples tested in accordance with Sect. 7. which represent rejected bars, shall be preserved for 2 weeks 
from the date of the test report. In case of dissatisfaction with the results of the tests, the manufacturer may make 
claim for a rehearing within that time. 

Reinforcing bars rolled from old rails are being used tc a considerable extent in reinforced- 
concrete work and seem to be giving satisfaction, especially for unimportant work such as 
footings, retaining walls, and possibly in slabs where the failure of one rod could not wreck 
the structure. The specifications for rail-steel concrete reinforcement bars adopted by the 
Association of American Steel Manufacturers April 20, 1912 (revised April 21, 1914) are as 

Manufacturers' Standard Specifications for Rail-steel Concrete Reinforcement Bars 

1. Manufacture. — All steel shall be rolled from standard section Tee rails. 

2. Physical Properties. — The physical properties shall conform to the following limits: 

3. Yield Point. — For the purposes of 
these specifications, the yield point shall 
be determined by careful obserration of 
the drop of the beam of the testing ma- 
chine, or by other equally accurate method. 

4. Form of Specimens. — (a) Tensile 
and bending test specimens may be cut 
from the bars as rolled, but tensile and 
bending test specimens of deformed bars 
may be planed or turned for a length of 
at least 9 in. if deemed necessary by the 
manufacturer in order to obtain uniform 

(fr) Tensile and bending test speci- 
mens of hot-twisted bars shall be cut from 
the bars after twisting, and shall be tested 
in full sise without further treatment, un- 
less otherwise specified. 

6. Number of Tests,-^ia) One ten- 
sile and one bending test shall be made 
from each lot of 10 tons or less of each 

Properties nmsiilcrpd 

Rail-steel grade 


and hot- 
twisted bare 

Ultimate tensile strength, minimum, lb. 

per sq. in. 
Yield point, minimum, lb. per sq. in. . . . 

Elongation, % in 8-in. minimum 

Cold bend without fracture: Bars under 
H in. diameter or thickness. 

Bars ^ in. diameter or thickness and 






Tens. str. 

180 deg. 
</ - 3< 

90 deg. 
d - 31 

Tens. str. 

180 deg. 
•f - 4r 

90 deg. 

d ^ At 

sise of bar rolled from rails varying not more than 10 lb. per yd. in nominal weight. Should a test specimen 
develop flaws, or should the tensile test specimen break outside of the middle third of its gaged length, it may 
be discarded and another test specimen substituted therefore. In case a tensile specimen does not meet the 
specifications, an additional test may be made. 

(b) The bending test may be made by presnure or by light blows. 

6. Modifications in Elongation for Thin and Thick Material. — For bars less than lis in. and more than ?i 
in. nominal diameter or thickness, the following modifications shall be made in the requirements for elongation: 

(a) For each increase of >i in. in diameter or thickness above H in., a deduction of 1 shall be made from 
the specified percentage of elongation. 

(b) For each decrease of Me in. in diameter or thickness below lis in., a deduction of 1 shall be made 
from the specified percentage of elongation. 

7. Number of Twists. — Hot-twisted bars of rail carbon steel shall be twisted with one complete twist in a 
length equal to not more than 12 times the thickness of the bar. 

Sec. 1-591 



8. Finish. — Material must be free from injurious seams, flaws or cracks, and have a workmanlike finish. 

9. Variation in Weight. — Bars for reinforcement arc subject to rejection if the actual weight of any lot 
varies more than 5% over or under the theoretical weight of that lot. 

Properties considered 

Plain bars 

and hot- 
twisted bars 

Tensile strength, lb. per sq. in 

Yield noint. lb. ner so. in 




Tens. str. 




Tens. Btr. 

Kloniration in 8 in. %* 

1 See Sect. 5. 

Thickness or diameter of bar 

Rerolled bar specifications have also been adopted by the American Society for Testing 
Materials after an extended series of tests. The specifications follow : 

Standard Specifications for Rail-steel Concrete Reinforcement Bars 

(American Society for Testing Materials^ 

1. The specifications cover three classes of rail-steel concrete reinforcement bars, namely: plain, deformed, 
and hot-twisted. 

Manufacture. — 2. The bars shall be rolled from standard section Tee rails. 

3. Hot-twisted bars shall have one complete twist in a length not over 12 times the thickness of the bar. 

Phyeieal Propertiee and Te»t9.—A. (a) The bars shall conform to the following minimum requirements a^ 
to tensile properties: 

(6) The yield point shall be deter- 
mined by the drop of the beam of the test- 
ing machine. 

5. (a) For bars over 9i in. in thick- 
ness or diameter, a deduction of 1 from the 
percentages of elongation specified in Sect. 
4(0) shall be made for each increase of }i 
in. in thickness or diameter above 94 in. 

(6) For bars underlie in. in thickness 
or diameter, a deduction of 1 from the per- 
centages of elongation specified in Sect. 4(a) 
shall be made for each decrease of Ms in. 
in thickness or diameter below ^{e in. 

6. The test specimen shall bend cold around a pin without cracking on the outside of the bent portion, as 


7. (a) Tension and bend test sped- 
mena for plain and deformed bars shall be 
taken from the finished bars, and shall be of 
the full thickness or diameter of bars as 
rolled; except that the specimens for de- 
formed bars may be machined for a length 
of at least 9 in., if deemed necessary by the 
manufacturer to obtain uniform eross- 

(6) Tension and bend test speci- 
mens for hot-twisted bars shall be taken 
from the finished bars, without ' further 

8. (a) One tension and one bend test 
shall be made from each lot of 10 tons or less of each sise of bar rolled from rails varying not more than 10 lb. per 
yd. in nominal weight. 

(6) If any test specimen shows defective machining or develops flaws, it may be discarded and another speci- 
men substituted. 

(c) If the percentage of elongation of any tension test specimen is less than that specified in Sect. 4(o) and any 
part of the fracture is outside the middle third of the gage length, as indicated by scribe scratebes marked on the 
specimen before testing, a retest shall be allowed. 

Permi**ibU Variation* in Weight. — 0. The woght of any lot of bars shall not vary more than 5% from the 
theoretical weight of that lot. 

FiniMh. — 10. The finished bars shall be free from injurious defects and shall have a workmanlike finiah. 

Inspection and Rejection. — 11. The inspector representing the purchaser shall have free entry, at all times while 
work on the oontraet of the purchaser is being performed, to all parts of the manufacturer's works which concern 
the manufacture of the bars ordered. The manufacturer shall afford the inspector', free of cost, all reasonable facili- 
ties to satisfy him that the bars are being furnished in aooordance with these specifications. All tests and inspec- 
tion shall be made at the place of manufacture prior to shipment, unless otherwise specified, and shall be so con- 
ducted as not to interfere unneoessariJy with the operation oi the works. 

12. Bars which show injurious defects subsequent to their acoeptance at the manufacturer's works will be 
rejected, and the manufacturer shall be notified. 

Plain bars 

and hot- 
twisted bars 

Under fi in. . 
14 in. or over. 

180 deg. 
d ^ 3t 
90 deg. 
d ~ 3t 

180 deg. 
d - 4< 
90 deg. 
d - 4< 


■- diameter of |nn about which the specimen is bent. 
— thickness or diameter of the specimen. 



[Sec. 1-60 

$0. Factors Affectiiig Cost of Reinforcing Bars. — In order to insure nxinimum cost and 
prompt delivery of steel reinforcing bars, the steel schedule for a reinforced-concrete structiire 
should call for bars of as few different sizes and lengths as possible. Bars of odd 16th sixes 
are seldom to be found in stock (except the Jlc-in. size which is frequently used for slab rein- 
forcement) and shipments from the mill on such sizes are likely to be very slow. Designers 
should always bear in mind this fact and arrange to use either round or square bars in )^-in. 
sizes. Wherever possible, steel lengths that do not vary greatly on the schedule should all be 
made equal since an order calling for only a few different lengths will be put through the mill 
much faster than one calling for many different lengths. 

The following size extras for bars less than ^-in. are standard with all mills and are the 
same for either roimd or square bars: 

Size Extras for Rounds and Squares in Cents per 100 Lb. 

Ji-in. and larger Base 

Jl to 1 He-in 6 cts. extra 

yi to He-in 10 cts. extra 

^e-in 20 cts. extra 

^-in 25 cts. extra 

^e-in 35 cts. extra 

^-in 60 cts. extra 

It should be noticed that a higher size extra must be paid for an odd 16th size below ^~in. 
than for the next larger '^'in. size. This fact alone offsets any advantage in saving steel 
by always calling for the nearest theoretical size whether odd or even. 

Where the character of the work requires small bars a saving in cost is obtained by using 
round bars owing to the difference in size extras between rounds and squares of equivalent area. 

Lengths less than 5 ft. should be avoided, if possible, as they are subject to the following 
cutting extras, whether sheared or hot-sawed: 

Lengths over 24 in. and less than 60 in 5 cts. per 100 lb. 

Lengths 12 in. to 24 in. inclusive 10 cts. per 100 lb. 

Lengths under 12 in 15 cts. per 100 lb. 

All orders calling for less than 2000 lb. of the same size and shape arc subject to the follow- 
ing extras: 

Quantities less than 2000 lb. but not less than 1000 lb 15 cts. per 100 lb. 

Quantities less than 1000 lb 35 cts. per 100 lb. 

61. Deformed Bars. — The following deformed bars are in common use: 

61a. Diamond Bar (Fig. 18). — Furnished by Concrete^teel Engineering Co.. 
New York City. The standard sizes are as follows: 

Diamond Bars 

Sise in inches 

M H 




H li 

1 Ui 


Area in square inches '0.0625 

Weight per foot in pounds. . 213 












Fjq. 18. — ^Diamond btf. 

^t should be noted that the weights and areas of Diamond bars are equal to those of plain 
are bars of like denominations. 

Sec. 1-616] 



616. Corrugated Bars (Fig. 19). — Furnished by Corrugated Bar CJo., Buffalo, 
N. Y. The standard sizes are as follows: 

Corrugated Rounds 

Siie in inches 










Net area in square inches 

Weight per foot in pounds 










Corrugated Squares 

Sise in inches 










Net area in square inches 

Weight per foot in pounds 










Fio. 19. — Corrugated bars. 

61c. Havermeyer Bars (Fig. 20). — Furnished by Concrete Steel Co., New York 
C^'ity. The following table gives the weights and areas of the standard Havermeyer bars: 


Sixe in inches 




Area in 


per foot in 


Area in 


per foot in 


Sise in 

Area in 

per foot in 










. 9770 



















































. 7656 















Fio. 20."Havermeyer bars. 



[Sec. 1-6U 

Special sizes of 1^-in. and l}^-in. square Havermeyer bars can be rolled by special arrange- 
ment, b\it are not carried in stock. A size extra of 10 cts. applies against 1 by H~ii^- ^^^ ^ ^2 
by ^e-^n, flats; all other sizes tabulated take the base price. 

Sid. Rib Bar (Fig. 21). — Furnished by Trussed Concrete Steel Co., of Youngs- 
town, Ohio and Detroit, Mich. The following sizes are standard: 

Rib Bar 

Sise in 

Area in 
square inches 

Weight per 
linear foot 
in pounds 

Siie in 

Area in 
square inchSs 

Weight per 
linear foot 
in pounds 





1.35 1 

1.95 ; 





Fio. 21.— Rib bar. 

60e. Inland Bar (Fig. 22).— Furnished by Inland Steel Co., Chicago. 
Sizes ^ in. to ^ in. inclusive with single row of stars on each side. 
Sizes I j in. to 1)^ in. inclusive with double row of stars on each side. 
Lengths may be obtained up to 85 ft. Supplied in both open-hearth steel and rail carbon 

Standard sizes are as follows: 

Inland Bab 

Sise in inches 









Area in square inches 

Weight per foot in pounds. 









Fio. 22. — Inland' bar. 

Rail carbon steel bars not rolled larger than 1 in. 

61/. American Bars (Fig. 23). — Furnished by American System of Reinforcing, 
Chicago. The following sizes are standard : 

Sec 1-621 


American Baiui 


Hiu in inchea 


Net u« in mnuut 


Wd,ht p.r^oot in 

Nrt km In Hiuire 



3 06 
3 68 
3 38 



1 000 


VBiiite^ pwdnexiCf 

Tin. 33. — .\inn1(«n tun. 
Witm Wabds. — This material ii uned r.o a mnaiderabln ffXtrirtt for ftoors, ^orvfx, wnllo, 
:., aad has been founit M piynem many valiinhte riiiHlities. Wire fnhrk^ 
sWel true* cnwinff generally at riiiht annleR and ^M^tircH nt the \ntf:rae,ct\orm. Th<> 
I ran lengthtriae and am ''ailed Karrymn wires; the lightnr nnen cmm thene and itm 
inting nr tie wires. One diBtinct advantage in the ubr of fahric. m thttt it piwaCTves 
of the sfeel. 

atee! wire \p^ adopted :ts irt«ndsrd for all fitwl wiri- upon rw-ommendiition of the 
iTBS Bureaa of Stnndards m pvm in lln- follnwiim tnhlo; 

*rBBl. WiRB C.A.HK 

Ponnrfs pfl' 



(?. 46875 





3,»31 () 


1 .W3 








■2-,(»<t '1 

1 Tflfl 
1 0^ 

n 4nfl2/; 





J-.l«4 It 

(1 :i.1lft 




1,.^ o 

: .^1 

' Fnmwrly '•aTli-f t h 



[Sec. l-62a 

Steel Wire Gaoe. — {ConJtinved,) 


Steel wire 


Area, square 

Pounds per 

Pounds per 

Feet per 




































































The manner of securing the intersections of wire fabric has given rise to a number of 

different types, several of the principal ones of which are given 

eaa. Welded Wire Fabric— Welded wire fabric, 
Fig. 24, manufactured by the Clinton Wire Cloth Co., is a gal- 
vanized wire mesh made up of a series of parallel longitudinal 
wires, spaced a certain distance apart and held at intervals by 
means of transverse wires, arranged at right angles to the longi- 
tudinal ones, and welded to them at the points of intersection 
by a patented electrical process. Longitudinal wires can be 
Fig. 24.— Welded wire fabric. Spaced on centers of 2 or more in., in steps of }i in. Transverse 

wires can be spaced on centers of 1 to 18 in. inclusive, in steps 
uf 1 in. and on centers of 10 to 18 in. inclusive, in steps of 2 in. The following table shows 






the sixes and areas of the wire used. Rolls kept in stock vary in length between 150 and 200 
ft. and between 56 and 100 in. in width. The wire will develop an average ultimate strength 
of 70,000 to 80,000 lb. per sq. in. 

WixDED Wire Fabric 

Ga^ of 










Diameter of 









Area of one 


wire (square 







Gage of 







Spacing of 








Area per foot of width in lonfiiudinal wires only 

Spacing of longitudinal wires 

2 in. 

3 in. 

4 in. 










































































62&. Triangle-mesh Wure Fabric. — Triangle-mesh steel-wire fabric, manu- 
factured by the American Steel & Wire Co., is made with both single and stranded longitudinal 
or tension members. That with the single wire longitudinal is made with one wire varying in 
size from a No. 12 gage up to and including a H-in. diameter, and that with the stranded longi- 

Fi<3. 23. — Trian^le'tnesh wire fabric. 

tudinal is composed of two or three wires varying from No. 12 gage up to and including No. 4 
wires stranded or twisted together with a long lay. These longitudinals either solid or itniod«ri 
are invariably spaced 4-in. centers, the sizes being varied in order to obtain the dmnd cnmh 
spcxiooal area of steel per foot of width (see Fig, 25). 

The trans^-erse or diagonal cross wires are so woven between the longitudinals that triangU^ 



[Sec. 1-62C 

are formed by their arrangement. These diagonal cross wires are woven either 2 or 4 in. apart, 
as is desired. Triangle-mesh wire reinforcement is made in lengths of 150, 200, and 300 ft. 
and in widths from 18 to 58 in. (4-in. steps). The table following shows the number and gage 
of wires and the areas per foot width when the longitudinals and cross wires are spaced 4 in. 
on centers. 

Triangle-mesh Wire Fabric 


Number of 

wires, each 


Gage of 

wire, each 


Gage of 
crofls wires 


area, long. 

square inches 


area, cross 


square inches 

tional area 
per foot 


per 100 sq. ft. 
























































































32 » 



























12 Vj 
















12 H 













































Elastic limit of regular stock is from 50,000 to 60,000 lb. per sq. in. Ultimate strength is 
85,000 lb. per sq. in. or over. Higher elastic limits and breaking strengths are furnished when 
required. Material may be obtained either plain or galvanized. Unless otherwise specified, 
shipments are made of material not galvanized. 

62c. Unit Wire Fabric. — A rectangular-mesh staple-locked fabric (Fig. 26) is 
furnished by the American System of Reinforcing. The wire used is of high tensile steel and 

* Styles usually carried in stock. 




is flBcured at the interaections b>' No. 14 wire. Standard siseB are ehawn in tbe following table. 
The fabiic ie gahranized and comes in standaid widths of 3. 4, and 5 ft., 200 lin. ft. in a rolL 

Fic. 2G — Unit wire £abnc 

Unit TV ike 


DmtBMkce ceoUfT to eea<«r 

in incites 


Gace of loBsitiid- 
iiuU wires 

QmgB of ero08 wires 

6«etioB*l area in aq. 
in., foot width 




Onma wiree 







: 10 










































1 4 





1 3 





t2(/. I«ock«wov«n Steel Fabric. — ^Lock-woven steel fabric (Fig. 27) is also known 
aa Page Special Process fabric. It is manufactured by the Page Woven Wire Fence Ck)., of 

In.. 27. — Luck- wu ven Bteei fabric. 

Mooeasen, Pa. and is controlled by W. W. Wight & Co. of ^'ew York City. This fabric is 
usually made 64 in. wide with special widths from 18 to 64 in. The longitudinal wires are made 
by a special process whicli gives them an ultimate tensile strength of 180,000 lb. per sq. in. 
with an ebstic hmit of about 70 So of the ultimate. The material is galvanized and is furnished 
in rolls of 160, 300, 460 and tkX) ft. in length. The table on page 60 gives the characteristics 
of tbe different styles. 





IScc. 1-62^ 

Lock-woven Steel Fabric 



Spacing in inches 


area in sq. in. 

per foot width 

strength in 
pounds per 
foot width 

Weight per 
100 sq. ft. 























7 '• 14 
14 1 14 
13 1 14 






1 22.08 
















9I2 H 
9 14 
8 14 
7 14 



1 H 12 

m i 12 
iy2 12 



1 53.43 

1 57.20 



62e. Wisco Reinforcing Mesh. — Wisco mesh is manufactured by the Witherow 
Steel Co.| Pittsburgh, Pa. It is made from the best grade of open-hearth steel and has a high 
tensile strength. All longitudinals are spaced 3 in. c. to c. and cross wires 12 in. c. to c. Stand- 
ard rolls are 150 and 300 ft. in length. Width of rolls are furnished in any multiple of 3 in. 
from 18 to 54 in. Properties of the Wisco mesh are given in the following table: 

Wisco Mesh 




area per 

foot width 

Weight per 
square foot 



area per 

foot width 

Weight per 
square foot 



area per 

foot width 

Weight per 
square foot 












0.395 ' 





63. Expanded Metal. — Expanded metal (Fig. 28) is one of the oldest forms of sheet rein- 
forcement. It is formed by slitting a sheet of soft steel and 
then expanding the metal in a direction normal to the axis of 
the sheet. The principal advantages claimed for this type of 
reinforcement are the following: (1) An increased ultimate 
strength and high elastic limit for low-carbon steel when the 
diamondnshaped meshes are formed by cold drawing the metal ; 
(2) a mechanical bond with the surrounding concrete; (3) great 
efficiency in the carrying of concentrated loads due to the ob- 

Fio. 28. — ^Expanded metal. 

Sec. I'-eSa] 



liquity of the strandB; (4) an increased ductility becauae of the fact that the diamonds or quad- 
rilaterals tend to close under severe loading; (5) a greater slab strength as the effect of closing 
up of the diamonds is to introduce a compression into the concrete at the lower part of the 
slab. Expanded metal and other sheet metal is made according to the U. 6. 8tan(}^rd gage 
which differs but slightly from the Steel Wire gage given on page 45. ' ^ 

tta. Steekrete. — Manufactured by the Consolidated Expanded Metal Cos., 
Rankin, Pa. The designation of the material gives the width of the diamond, the gage of the 
plate and the cross-section per foot of width. Bise 3-9-15 means that it is a 34n. -diamond, 
made out of No. 9 plate, having a sectional area per foot of width of 0.15 sq. in. All standard 
meshes have a diamond 3 by 8 in. The standard sises and gages are given in the following 

^'Stbelchete" Expanded Mktal 

Biae of meali 


No. of 


in a 


No. of 

sq. ft. 

in a 




of mceh 

Width of 
in inches 


Length of 
' in inches 


Section in 

aq. in. per 

ft. of 


wt. per 
wn^. ft. 


Sise of standard 

Wt. per 


to lb 

129 6 

3-13-075 , 






[ 6'0" X 8'0" 


1 6'0" X 12'0" 


194 4 

3-13-10 ' 






f 6'9" X 8'0" 
\ 6'9" X 12'0" 










/ 5'3" X 8'0" 
\ 5'3" X 12'0" 



135 2 






5 ; 

/ 7'0" X 8'0" 
1 7'0" X 12'0" 









/ 5'3" X 8'0" 
I 5'3" X 12'0" 










/ 4'0" X 8'0" 
\ 4'0" X 12'0" 





3-9-30 ' 






/ 7'0" X 8'0" 
\ 7'0" X 12'0" 


123 2 







2 1 

^ 6'0" X 8'0" 
1 6'0" X 12'0" 

96 ' 

122 9 







7'0" X 8'0" 
' 7'0" X 12'0" 










f 6'3" X 8'0" 
1 6'3" X 12'0" 

150 1 









f 5'9" X 8'0" 
I 5'9" X 12'0" 


138 ', 

252 5 








/ 5'3" X 8'0" ■ 
\ 5'3" X 12'0" 








2 19 


/ 4'9" X 8'0" 
^ 4'9" X 12'0" 

76 , 

166 4 

The Consolidated Expanded Metal Cos. also make to order a 64n. mesh, the size of the 
diamond being 6 by 16 in. The gage of plate used is No. 4, or nearly ^ in. thick. Any cross- 
sectional area desired up to and including 0.4 sq. in. can be obtained. The width of the sheets 
depend on the sectitmal area. These companies also make a 4-in. mesh from No. 16 plate 
iiHueh is unexpanded. Any length can be obtained up to 16 ft. The cross-sectional area per 



[Sec. 1-636 

foot of width is 0.093 sq. in. Special meshes can be obtained having diamonds of ^i in., 1)^ 
in., and 2 in. 

636. Kahn Mesh. — Manufactured by the Trussed Concrete Steel Co., of 
Youngstown, Ohio, and Detroit, Mich. The standard sizes and gages are the same as for 
"SteelcrU^." The Kahn Mesh may also be obtained with larger diamonds for reinforcing 
concrete pavements. The sizes of the Kahn Road Mesh follow: 

Kahn Road Mesh 



Siie of mesh 



in square inches 

Width of diamond 
in inches 

Length of diamond 
in inches 












No. of sheets in bundle, 10. Standard width of sheets, 5 ft. Standard lengths of sheets, 8 ft., 10 ft., 12 ft., or 
any equal diviaioDs of these lengths. 

. 68c. Corr-X-Metal. — Furnished by the Corrugated Bar Co., Buffalo, N. Y. 
The weights, sectional areas and standard sizes of sheets are given in the following table: 



Sise of roesh, 

short way 


Nominal thickness 

of metal 


.\pproximate weight 

per square foot, 


Net sec. area per foot 

of width 

(square inches) 














































Sec 1-eZd] 



Standakd Size Sheets 


Lone way of diamond 

Short way of diamond 


6', 8', 9' and 10' 8" 

3', 4', 5' and 6' 


6', 8', 9' and 10' 8" 

3', 4', 5' and 6' 


6', 8', 9' and 10' 8" 

4' and 5' 4" 


6', 8', 9' and lO' 8" 

3', 4' and 6' 


6' and 8' and 10' 8" 

3', 4', 5' and 6' 


6' and 8' and 10' 6" 

3', 4', 5' and 6' 


6' and 8' and 10' 6" 

4' and 5' 4" 



3', 4', 5' and 6' 



3', 4', 5' and 6' 

68d. Econo. — Furnished by the North Western Expanded Metal Co., Chicago, 
III. Standard sizes and weights are as follows: 

Econo Expanded Metal 


Weight per 

square foot, 


Mesh and s&Ke 

Widths, feet 

liCngths, feet 



3"— 16 ga. 


8 and 12 



3"— 12 ga. 


8 and 12 



3"— 10 ga. 


8', 10' 6" and 12' 



3"— 10 ga. 


8', 10' 6" and 12' 



3"— 10 ga. 


8'. 10' 6" and 12' 



3"— 10 ga. 


8', 10' 6" and 12' 



3"— 10 ga. 


8', 10' 6" and 12' 



3"— 10 ga. 


8', 10' 6" and 12' 



3"— 7 ga. 

3' 6", 7' 0" 

8 and 12 



2>i"— 16 ga. 


8 and 12 



2M"— 12 ga. 


8 and 12 



2M"— 10 ga. 


8 and 12 



2K"- 7 ga. 

3' 6", 7' 0" 

8 and 12 

10-1 Ji 


IH"— 18 ga. 
IJi"— 16 ga. 


8 only 
8 and 12 



lli"— 12 ga. 


8 and 12 

15- Ji 


H"-16 ga. 


8 and 12 



H"-12 ga. 


8 and 12 



H"— 18 ga. 


8 only 



H"— 16 ga. 


8 only 

The first two figures in the first column give the area of steel and the last figure gives the short 
dimensions of mesh. Thus No. 30-3 has an area of 0.30 sq. in. per 12 in. of width and has a 
mesh 3 in. wide. 

636. GF Expanded Metal. — Manufactured by the General Fireproofing Co., 
Youngstown, Ohio. Standard sizes are given in the table on page 54. 




[Sec. 1-e^e 

GF Expanded Metal 


Approz. weight per 
sq. ft. in pounds 



Bixe sheets 



Long way of diamond 

Short way of diamond 




6', 8', 9', 10'-8" 

3', 4', 5', 6' 



6', 8', 9', 10'-8" 

4', 5'-4" 




Carried in stock 
in standard sheets 

6', 8', 9', 10'-8" 

3', 4', 6' 



6', 8', 9', 10'-8" 

3', 4', 6' 



6', 8' 

3', 4', 6' 



6', 8' 

3', 4', 6' 




6', 8', 9', 10'-8" 




6', 8', 9', 10'-8" 




6', 8', 9', 10'-8" 




6', 8', 9', 10'-8" 

3', 4'-6" 



6', 8', 9', 10'-8" 

3', 6' 



6', 8', 9', 10'-8" 

4'-4", 6'-6" 



6', 8', 9', 10'-8" 

4', 6'-4" 



6', 8', 10'-8" 

3', 4', 5', 6' 



Five days 

6', 8', 10'-8" 

3', 4', 6' 



to two 

6', 8', 10'-6" 

4', 5' 




6', 8', 10'-6" 

3', 4', 6' 



on size 

6', 8', 10'-6" 

4', 5' 




6', 8' 

4', 5' 




6', 8' 

4', 5' 




6', 8' 




businesB on 

6', 8' 

4', 6' 



6', 8' 

3', 6' 



6', 8' 




6', 8' 

3', 4', 6' 



6', 8' 




6', 8' 




6', 8' 




6', 8' 




6', 8', 9', 10'-8' 





6', 8', 9', 10'-8" 

3', 4', 6' 




6', 8', 9', 10'-8" 





6', 8', 9', 10'-8" 




6', 8', 9', 10'-8" 


"SoTE. — ^Interpret styles as follows: For example 3-10-176. 3 equals short dimension of 
diamond in inches; 10 equals approximate gage; 176 equals 0.176 sq. in. sectional area per foot 



hj xk0 Tpumd Coneiei^ Sieel Co^, and coa- 

tqa n e cied hy fi|^i ctoas mflnbeiflL It is made from « 

cQR«K»tod oa tlie oihw« StHps of the metal adjacent to 

is dnvB ottt into squmre medies (Fig. 29). The sUndani 

of from 2 to Sia. and in all lengths up to IS ft^ The pro(>- 

table vhich foDows: 

N splici 


Fro. SS. — ^Rib metal. 


Rm Metal 


Width of 
' stjuMianl sheet. 

Sq. ft. per 

fioeftr foot of 

standard sheet 

Area per ft. 
width, sq. in. 

Ult, tensile 
strencth per foot 
of width 

Safe tensile 

strenc th ner 

foot of width, 












































Area of one rib = 0.09 sq. in. 

Ultimate tensUe strength = 6480 lb. 

Safe tensile strength » 1620 lb. 

66. Self-centering Fabrics. — ^Permanent centering fabrics (used mostly for reinforcement 
in concrete floor slabs resting on steel beams) are stiffened by rigid, deep ribs which do away 
with the use of dab forms. The mesh is made small enough to prevent ordinary concrct4^ from 
passing through. The centering fabric is laid over the supports, the concrete is poured on top 
and the under side plastered. A simple brace along the middle of the slab span is sometimes 
required to give sufficient strength to the ribs until the concrete has set. The permanent cen- 
tering fabrics may be obtained either in flat or segmental form. 

A serious disadvantage in this type of construction is the difficulty of providing efficient 
fire-protection on the under side of the fabric. Bond with the concrete is also likely to be 


65a. Hy-Rib. — Hy-Rib (Fig. 30) is a steel sheathing, atiSensd by deep ribs 
foimed from a single sheet of steel. It is controlled by the Trussed Concrete Steel Co. of 
YoungBtown, Ohio, and Detroit, Mich. 

FiQ. 30.— Hy-ri 

True of Hr-RIb 

Formerly ulled 


Width o( 

Ok*. N». U. S. 

lJi4n. Hy-Rib 

•Ke-in. Hy-Rib 

>H8-i">- Hy-Rib 

H-in. Hy-Rib 







22, 24, 26 

22, 24, 26, 28 
24, 26, 28 
24, 26, 28 

Standard lengths, 6, 8, and 12 ft. 

Other lengths are cut from standard lengths without chaise except for waste. 

lH->n- and >^B-in. Hy-RJb are shipped in bundles of eight sheets; >^{«-in. and ^-In. 
Hy-Rib in bundles of sixteen sheets. 

Mb. CtoT-Hesh. — Corr-Mesh (Fig. 31) is furnished by the Corrugated Bar Co., 
Buffalo, M. Y. It is a stiff-ribbed expanded metal, the riba being spaced 3^ j in. c. toe. The 

Flo. 31. — Corr-maah. 

height of the ribs is H in. and the width of the sheets is 12 ^ in. c. to c. of outside ribs. The 
standard ^0^68 are No. 24, No. 26, and No. 28, although other gages can be obtained if required. 
The standard lengths are 6, 8, 10 and 12 ft. The sheets are furnished either flat or in various 
types of curves. All metal is shipped painted uoleas specifically ordered otherwise. 

65c. Setf-Sentering.— Self-Sentering (Fig. 32) is manufactured by the Geoerol 
Flreproofing Co., Youngstown, Ohio. It is made up of a series of heavy, cold-drawn ribs, 
•?ifl in. high, always ^mced 35i in. c, to c, connected by a form of expanded metal — all cut 
and drawn from one sheet of steel. Size of sheets — 29 in. wide by lengths of 4, 5, 6, 7, 8, 9, 
10, 11 and 12 ft. Longer lengths up to 14 fL furnished on special order. Self -Sen tering is 
made of No. 24, 26 and 2S-gage meUI. 

65d. Chandath.— Chanelath (Fig. 33), furnished by the North Western Ex- 
W Metal Co., Chicago, Dl., is a type of expanded metal composed of a seriea of heavy 

Sec 1-Ue] MATERIALS 67 

cold-formed steel T-riba connected together by a mesh known as "Kno-Bura" metal Uth. 
The T-ribs are 14 ^- ^K^ ^"^ spaced 4 in. c. to c. The flange o[ the T is >i in. wide. Chane- 
lath is manufactured and carried in stock ready for immediate shipment in the fallowing sises 
of sheete: Lengths— 3, 4, 5, fl, 7, 8, 9, 10, 11 and 12 ft,; widthe— 4, 8, 12, 16, 20, 24, 28, 32, 
36, 40, 44 and 48 in. 

S6e. Rlbplez.— Ribplex manufactured by the Berger Mfg. Co., Canton, Ohio, 
is an expanded metal with ribs 4.8 in. on centers and !i in. high. Standard sheete are 24 in. 

Fio. 3Z.— 3<ilf-«nt«rini. 

wide and are carried in stock in 4, 5, 6, 7, 8, 9, 10, 11 and I2-ft. lengths. Sheets are made 
io 28, 26 and 24 gages. 

S6/. Dovetailed Comigsted Sheets. — Sheete of thin steel corrugated so as to 
form doveUuIed grooves are manufactured by the Brown Hoisting Machinery Co., Cleveland, 
Ohio, and by the Berger Mfg. Co., Canton, Ohio. The first-mepttooed company manufacture 
a plate known as Feiroinclave and the latter company furnish two types of plates known as 

Fio, 33.— Ch«n*l»lh, 

Ferro-Lithic and Multiple Steel. The doveteiling in these plates serve to unite the plat«s to 
the concrete. 

66. Reinforcing Systems for Beams, Girders and Columns. 

Ma. Kahn System. — The Kahn trussed bar (Fig. 34), named for ite inventor, 
is rolled with flanges, which are bent up to resiHt the shear in the beam- For continuous beams, 
inverted bars are placed over the supports in the upper part of the beam, extending over the 
region of tension. Properties of Kahn trussed bars are shown in the following table: 



[Sec 1-066 

Kahk Tbussed Bars 

a X 6 





in inches 



Square Section Bars 




12, 24, 36 

6, 8, 18 
8, 18, 30 

New Section Bars 

IH X 2H 
2 XSH 





12, 24, 36 

8, 18, 30 
24, 30, 48 
24, 30, 48 

Note. — ^8, 12, 18, 24, 30, 36, and 48-in. diagonals are sheared alternately. ' Six4n. diagonals 
only are sheared opposite. 

ft- — b**"-*i 

"••• Q ... ^ 

Fia. 34. — Kahn trussed bars. 

What might be called the Kahn system is illustrated in Fig. 35. The collapsible column 
hooping is shown more in detail in Fig. 36. The hooping is shipped in the form of flat, circular 
coils of exact diameter and accurately spaced by means of special spacing bars. These coils 
spring automatically into a complete hooped column on cutting the small fastening wires. Rib 
bars (see Art. Old) are ordinarily used as vertical reinforcement in conjunction with the hooping. 

The collapsible column hooping is shipped complete with two spacing bars. Sices of wire 
for hooping: ^, ^6i Ht Titt ^^^ K-^^- diameter. Diameter of coils: 9 to 30 in. Pitch: 
1}^ to 12 in. Hooping, where desired, can also be obtained in bundles, coiled to the correct 
diameter, and with separate spacing bars, ready for assembling in the field. 

666. Cummings System. — ^The Cunmiings system is shown in Fig. 37. U- 
shaped stirrups are used on the girder frame shown. They are shipped flat with the longitudinal 
reinforcement, but are bent up to an inclined position on the work. The rods are held together 
by means of a patented chair. In the Cummings hooped column, each hoop is securely attached 
to the upright rods. The hoops are made of flat steel, bent to a circle, with the ends riveted 
or welded together in such a manner that the ends of the hoops protrude at right angles to 
keep them the proper distance from the mold. Reinforcement of the Cummings system is 
manufactured and sold by the Electric Welding Co., Pittsburg, Pa. 

66<;. Unit Syttem. — Figs. 38 and 39 show the unit system of reinforcing con* 
^ by the American System of Reinforcing, Chicago, 111. The girder frames are not stock 





fnuDM but &re built to meet the engineer's or architect's plans. Unit girder frames are pro- 
vided with overlapping rods for continuous beams to reinforce against negative moment. 

Wd. CoiT System. — Cort-bar girder frames (Fig. 40) and shop fabricated ^irala 
(Fig. 41) are fumiabed by the Corrugated Bar Co., Buffalo, N. Y. Aa with the unit system, 
the girder frames are built to meet the ei^neer'a or architect's plans. In the spiral reinforce- 
ment the spacing bars consist of two or — in Ibi^ columns — four apacere made of T-flection 
bars notched to receive the spiral. The spirals are made of cold-drawn wire and are furnished 
in any length, in diameters of 10 to 36 in., pitch 1 to 4 in., and of the following aJEes of wire: 


Di>. »[ 


P«r foot) 










0.4900 ,0.6404 

>4 round 



JIb round 


0.4305 jo. 4943 

^e round 




^ j round 


0.3625 >.3505 

H round 

Fio. 38.— ■■ Unit 

».— "Unifapirili. 

Me. Bnimebiqne SyBtem.^-One of the pioneers in concrete construction in 
Europe is Mr. Hennebique, in France, and the system which still bears bis name is shown in 
Fig. 42. 

(^. Pin-connected System. — Reinforcement in the pin-connected system con- 
sists of ban made into a truss and ready for placing in the forms {see Fig. 43). 

600. Lnten Tnus. — The Luten truss is shown in Fig. 44. The bara are rigidly 
locked together to form the truss by a clamp, with a wedge that is self-locking when driven 
home. The truss is especially adapted to highway culverts and bridges and is put out by the 
National Concrete Co. , 

B6A. Xpantmn System. — The truss by this name is shown in Fig. 45, and is 
applicable chiefly to beams, girden, and heavy slabs. This system is patented by The Con- 
solidated Expanded Metal Cos. 

68i. Shop Fabricated Reinforcement System. — In this system (Fig. 46) manu- 
'''itured by the lAckawanna Steel Co., Lackawanna, N. Y., the standard bar is a troughed 


Fio. W.^Oorr-bu (intaT (nine. 

Fio. 41. — Comicktcd Bu Co.'! (pinb. 

Fio. 42.— Hennebunw lysWrn. 

Fio. 43. — Piii-«oiui«ct«d ■yatem. 

Fla. 44.— Luton 


BectiOD and the auxihair reinfoKuig memben, such u diagoiud Unson m^mbeis, tia rods for 
cohimns, wall% etc., sre flat bsn (^ by Hs ''>') ^'^^ knobs on e&ch edp. Fabrication is 

CraMsactton of !»■ 


effected by placing a portion of the auxiliary flat, properiy bent, within the trough and with a 
bulldoier or other pressure machine aqueeiing the wings of the main bar and also gripping the 
knobs of the flat. The upper or trougbed part of the main bar is a constant. Increased area is 



Fia. 4S- — Shop fU>nc«tcd ninfDrQ«m«at ■yvtem. 

developed by making the section deeper aa required. Tests have shown that the rivet grip, a 
it is called, is greater than the strength of the auxiliary member. 

pROPOxnonsc coscrete 

1. PropertiM o( Caocf«te Defwndeat upon PropMtias *ad ReUtive PioportioiiG of Con- 
ititnent Ibteftala. — Ik»pite pi«valeuc« of csreleea meMureiueat of materialii and arbitrary 
specifying of praportious for coitcrete, it doee &ut follow, as many infer, that such piocedureu 
»re right, either in pi«ctioe or in tbaury, or tltat they should be euutiauod. On the contrary, 
!>uch proceduree are wasteful and inefficient and to meet their eflectM, the aUuuable atreaticii on cuu- 
Crete have been fixed at a Btwtdard so low that actual failure or diulntegratiun cui lesult only 
from flagrant abuMe of tb^e lai and undesirable methods. All consideration of possible per- 
fection amle, it m known beyond queation that proper proportioning of selected materials ia a 
prerequiaile to bucc«h, for concrete ie wholly dependent for its properties upon the properties 
and proportions of its constituent materials, both severally and in combination. 

S. TWinrj of Proportioiiiac. — Considering the composite nature of concrete as revealed 
by a fimctured section (Fig. 1), it will be seen at a glance that this substance is a pudding of 
large stooe particlee set io a mass, or "matrix," of other substancex, which matrix is a mortar 
iif cement and sand. If this matrix be magnified, it will be seen to 
be Mmil^r to the groM section, with sand grains as large par- 
tiflea beld in a matrix of more-or-leas hydrated cement. Theo- 
n^cally, the lai^ e stone particles and also the aand grains should 
lie as closely together as possible, so that if all fragments (suid 
incloded) had come from crushing a scdid cube of stone, tlieir 

reassembly would approach the cube's original volume, density, ' 

and Btnm resistanoe. The aim in proportioning, therefore, is 
f'Scient r«combuiation. 

It is impossible in practice to obtain the doae rearrange- 
Mient and interlocking of the particke that is, on all counts, 

deoiiKbk. There must and do remain between them, as is y^ \ —The puddiDc^Mau 
\-igually evident, inter-particle spaces or voids. To fill spaces or iuueiurt o! oou. reie 

vends between large particles, finer materials of like nature and 

origin are chosen; and on theoretical grounds, it should be possible by measuring Ihme iatet- 
particle spaoes — as, for instance, by pouring measured quantities of water into a given con- 
tainer filled with Btone particles until the container can hold no more^-to add a quantity of 
sand to tbe stone of a volume equivalent to the added volume of water, so us tu render the 
measure almost solidly full of sand and stone puticles, or rather, of stone puticles since sand 
is it«elf disintegrated rock. 

It is to be eipscted, however, as is evidenced by visual magnification, tbat between Ux- 
Rand graiuB must he other and smaller inter-particle spaces in great number; and on the Uieo- 
rptical bans of proportioning, theae should be filled by cement, »o that tbe entire uieasuri' 
might be solidly full of a composite which would clooely approximate natural stone in t«xtur<', 
denatty, and streagUi. But this ideal is not attained, partly because in this hypothe>u« tbe water 
necfwry ior reactuo witii cement is not allotted space, and ptutly b«eauae no allowaoce is 
made for tl»e PBcesaary soiface coating of sand and stone by cement and water, with consequent 
diqiersitti at partidea. Tbe water is assuined either to lie in inter-particle (qmces not filled 


by cement or in inter-particle spaces in the cement itself, or else to be negligible in quantity and 

The foregoing are the assumptions on which the void theory of proportioning is based. 
There is hardly one of these assumptions that does not rely on false premises, so that the whole 
void theory must be and is found at variance with practice, particularly when subjected to 
comparison with results obtained by rough field procedure. Yet the idea behind the theory 
is right, for it suggests by inference that density of natural stone is the ideal to be striven for 
and that it may be obtained: (1) by using a maximum quantity of natural stone in fragments 
large enough to possess unimpaired its inherent properties; and (2) by filling in the inter- 
fragment spaces with maximum quantities of like mineral materials. The error lies not so 
much in the standard thus chosen as in neglect to give proper consideration to the individual 
and combinative properties of the several substances included in concrete, not the least impor- 
tant of which is water, both as a space occupier as well as in its chemical and phjnsical actions 
with cement and with inert aggregate.^ It is always to be remembered that concrete is not 
concrete without water; that its proportioning is equally important with the proportioning 
of cement, stone, or sand; and that it affects by its quantity the proportions of the other 
ingredients that may be placed in any given volume. 

8. The Strengtli Elements of Concrete. — Any concrete will have as an upper limit of 
stress resistance the properties of the most resistant of the materials entering into it. These 
materials are usually stone and sand, with some strength preference in favor of sand, as most 
sand particles are individual crystal units without cementing substance between them, while 
natural stone is an aggregation of similar particles, built up in one way or another. Since 
natural stone is formed under the most advantageous conditions as to arrangement of particles, 
pressure, etc., and since in its composite nature, it is closely analagous to concrete, it may be 
taken as the ultimate ideal in artificial stone made by the admixture of sand, stone, and water, 
with Portland cement. Furthermore, in natural stone as in concrete, the weakest element 
is the cementing substance which lies between the inert grains, for in each material this is 
alterable by various agents and is at the same time of less inherent strength than the mineral 
particles which it unites. 

4. Proportioning for High-strength Concretes. — From the above, it follows that the 
greater the proportion of mineral grains in any stone, brought about through compacting and 
the exclusion of all but a film of cementing material (as in a very compact sandstone), the 
higher will be its strength. In the same way, in artificial concretes, the greater the proportion 
of natural stone (or of analogous mineral matter) and the less the quantity of cement, consistent 
with proper coating of the inert materials, the greater will be the resulting strength, for Portland 
cement in combination with water is the weakest element of this cement-sand-stone combina- 
tion. In this connection it should be further remembered, that while ground Portland-cement 
clinker (commercial powdered cement) is extremely hard and of great inherent strength, the 
same substance in combination with water results in an entirely different product of different 
chemical nature and composition, having relatively low strength. It is, therefore, actually 
true that down to a certain limit, the less cement there is in any concrete the more enduring 
and, at the same time, the cheaper will be that concrete. It becomes, then, of vital importance 
not only to so select inert materials that they shall by their properties be able to endow 
the concrete with high strength, but also to so choose their proportions that they shall be. a 
quantitative maximum in the mixture, the cement functioning in minimum quantities, as 
an adhesive surface coverer and as a void filler. 

6u Weakness Due to Poor Proportioning. — By reason of the remarkable properties of 
Portland cement, carelessness in proportioning concrete has become tacitly accepted, if not 
permitted and sanctioned practice. Fairly good results have been obtained in spite of gross 
carelessness; and this has brought about a belief in the minds of perhaps a majority of con- 
stnictorB, that practically any materials in any proportions in combination with any cement 

-« chapter on ** Wftter*' in Sect. 1. 


will give the desired results. It is also unfortunately true, that in first results and appearance, 
Portland-cement concrete, even when of inferior materials and in improper proportions, appears 
equal to that properly made; but the rapidly increasing number of defective constructions 
which are being brought to light with the passage of time is bringing about an awakening 
in regard to the causes of such disintegrations. Such apparently easy successes have given 
rise to the so-called "arbitrary" proportions in widespread present use, but in probably a 
majority of cases, not the least important of the causes which result in ultimate failure is the 
use of these same arbitrary, "practical" proportions, which actually are not practical in one 
case out of ten, so far as results achieved in making a good product are concerned. 

6. Unit of Proportioning. — In specifying concrete or mortar mixtures a unit quantity of 
cement is taken as a base. The cubic foot is the usual unit of quantity. On large work where 
overhead bins and measuring-hoppers are provided, the cubic yard may be chosen. In general 
the assumption is made that 1 sack of cement weighs 94 lb.; that it is )^ bbl.; and that it 
occupies 1 cu- ft.^ 

7. Arbitrary Proportions. — Despite their inadvisability and incorrectness, arbitrary 
proportions cannot be ignored. With the cubic foot of cement as a unit, these proportions 
are commonly described as 1:2:4 or 1:3:6 or 1:4:8 or some other easily remembered ratio, 
expressed in terms of volumes of sand and stone respectively, to the unit volume of cement. 

Evil as is such arbitrary choosing of proportions, when such proportions are rendered 
still further indefinite by inaccurate measurement of materials, the composition of the resulting 
concrete is indefinite and uncertain to an extent that gives a new respect for the abilities of 
Portland cement. It has been shown repeatedly on test, and confirmed by examination of 
the resulting concrete, that measurement of materials in the usual manner by wheelbarrows or 
shovelfuls, may bring about variations of from 100 to 200% in actual proportion of material 
delivered to the concrete batch. This is particularly true with regard to sand. Since sand 
occupies at least one-third of the volume of any concrete, and as the density and stress resistance 
of the mixture is so largely dependent on the quantity of sand present, the variations introduced 
through change of grading attendant on supplies from diflPerent localities, or through change 
in moisture content, will be seen to be very serious. It is all too true that in commercial work, 
a supposedly 1:2:4 concrete is quite as likely to be 1 : 4 : 8 or 1 : 5 : 9 or even worse, or on the 
other hand it may be 1 : 1 : 2, or other combinations in indefinite number, with a corresponding 
increase of unreliability and cement cost. 

It should always be remembered that each of the inert materials of concrete has surfaces 
and voids peculiar to itself, and the combination of any two is peculiar to them alone. A change 
in either material must, therefore, result in new relations, with change in the burden imposed 
on the cement, which (in combination with water) must function both as an adhesive, as a 
surface cover, and as a void filler. Regardless of the apparent sanction generally given to the 
use of arbitrary proportions in making concrete and to the apparent expediting of work by 
slap-dash measurement, this practice should be prohibited on all work above that of the most 
ordinary grade. 

H. C. Johnson in "What is a 1 : 2 :4 Ck)ncrete7"' states that with maintenance of strict 
proportions with different materials, the cement demanded will range from 100 to 130 bags — 
a variation of 4.5% in a definite quantity of concrete. The table on page 66 summarizes the 
results of his experiments. 

8. Proportioning by Void Determinations. — For reasons given in Art. 2 proportioning 
materials by void determinations is obsolescent practice, not to be more countenanced than 
that of arbitrary proportions, which it resembles. The determination of voids may give a rough 
indication as to the inter-particle spaces existing in any fine or coarse aggregate, but it does not 

^ A number of authorities have approved the adoption of 3.8 cu. ft. of cement to the barrel. This value is 
more nearly exact and gives 100 lb. of cement to the cubic foot or 0.95 cu. ft. per sack. One standard sack, however, 
may be and usually is considered as 1 cu. ft. 

* Concrete and ConttructioncU Engineering (London), Feb., 1915. 



[Sec. 2-8 















2! 8 8| g 
02 8 S S. 

i Tj 2 5 '. 


S2 S S S. 














*4 v^ 














« o 











00 00 













5 5 









*4 ^4 


















^4 ^i4 

















e lo 






















• e 






e 6 g a 

8 . 2 ^ .2 . 2 . 


• ,»*,^,*-p«' 


1 ^ 

•d ^^ 

If 6^ O 

a lo *• 

ft • 8 





•* w 





8 o g S S 

||?sSs2slsi|^s Jill 

^n GO COQD 00 QQ.'vUQ 00 GC 00 00 













CO r^ 

^* ^ 







I ^ 

t Si 
.' 8 5 

S S 3 












S 1 


o — 
O -0 . 

8 -S ^ 



-• « 


»* 00 ^ 

e« ^ 


Sec. 2*8) 



afford a basia on which to proportion the materials. If a cubic foot of broken stone contains 
40% of voids, the void basis of proportioning is based on the assumption that Ho cu. ft. of sand 
should be added in order to fill these spaces, i.e., to bring the mass up to approximate solidity; 
and that if this quantity of sand in turn contains 35% of voids, that ^J^oo X Ho cu. ft. of 
cement should be added (plus a certain arbitrary percentage for coating sand and stone sur- 
faces) to give to the mass actual solidity. 

The inaccuracy of this method is partially due to the variation in voids in sand under differ- 
ent conditions, this variation being sometimes sufficient to make a difference of 30% in the 
amount of cement required. The following table shows variation in voids due to difference 
in moisture alone. The figures in this table are for sand in a loose condition and the differences 
would be still greater if the dry sand had been shaken and tamped. It is evident that the 
method of proportioning by voids is valueless unless the sand is in the same state of compact- 
ness when mixed in concrete as it is when the void test is made. 

Physical Characteribtics of Concrete Aoobeoates^ 
(For aggregates in loose condition) 

Voids (aggre- 
gate contain- 
ing natural 




Weight (lb. per 

cu. ft., aggregate 

containing natural 



1 Weight (lb. per 
' cu. ft., aggre- 
gate dry) 


^ moisture 
in moist 

ity of 

Average of 4 good concrete 


5 2.65 


Long Island washed gravel 


• • 


, , 


graded ^i in. to 1^ in. 


Commercial ^-in. lime- 


• a 




Commercial 1 *'2-in. lime- 


• • 




Trap rock, graded ^^ in. to" 


• • 


. . 


IH in- 1 



In practice, therefore, the proportions obtained by void determinations do not hold. As 
soon as water is added to sand, stone, and cement, very different physical relations are effective 
from those that previously existed. The particles of cement and the finer particles of sand are 
necessarily dispersed by the water which coats and lies between them; the larger particles of 
sand are dispersed by this thin mortar of cement, fine sand, and water; and the finer and larger 
stones in turn are dispersed in a similar manner by a like combination of the finer materials. 
\3 pointed out in Art. 2, this fact is made very e\'ident by the examination of the fractured sur- 
face of any concrete. No matter whether the surface examined is in the gross, showing large 
particles of stone, or whether it is magnified to make visible the ver>' fine particles of sand 
and cement, the same dispersion will be found to obtain, offering visual evidence, confirmed 
by test, as to the necessary inaccuracy of void determinations as a basis of proportioning. 

This may be made evident by a simple illustration. Assume a vessel containing a given 
number of small spheres of varying sizes which may be considered as so many sand and cement 
grains. It is evident that if these spheres fill the measure, a certain quantity of water may be 
added without disturbing their inter-relations. That is to say, it is assumed that these spheres 
will remain in surface contact one with another e\'en after the addition of water. If, however, 
the same spheres be put into a larger vessel and an additional quantity of water added, filling 

* R. E. GooDWTX in ConerHr, Nov., 1915. 


this vessel, then, if the mixture is uniform (which is the condition assumed to exist in mortar 
and in concrete), these spheres must be dispersed and be out of surface contact one with another. 
This represents in exaggerated illustration, but not in exaggerated degree the conditions as 
they exist in commercial concretes, dispersion therein being progressive from the finest particles 
to the largest, each successive grade assisting in the dispersion of the size next laiger, with 
resultant increased demand for cement and corresponding weakening of the mass. 

Rather than proportion strictly on the basis of voids, therefore, a better way is first to 
grade the aggregates, both coarse and fine, by sieve analyses. In this way, the voids are taken 
cognizance of, though in a different way. A combination of materials may then be made such 
as to give a mixture containing these materials in greatest quantities. The proportions of 
aggregates in this mixture having been determined, the amount of cement required will then 
depend very largely upon the strength needed or the degree of imperviousness required of the 
concrete. It can be approximately estimated by determining the percentage of voids in the 
mass, but on account of the errors introduced through the establishment of new conditions by 
the introduction of water, this latter assumption is not to be recommended unless checked on 
actual mixtures. 

9. Proportioning by Mechanical Analysis. — ^Although the foregoing recognizes the exist- 
ence of inter-particle spaces or voids, it properly should be termed '^ proportioning by mechanical 
analysis." Such proportioning is recombination after analyses are made by passing representa- 
tive samples of the inert materials through successive sizes of standard-mesh screens; noting 
the quantity passing and the quantity retained on each screen; and plotting these as a curve, 
with sizes of screen openings as abscissae and percentages of material passing as ordinates 
(see Art. 32, Sect. 1). By this procedure a more or less regular curve will be obtained for 
the sand and for the stone; and its variation from a predetermined curve, such as that of Wil- 
liam F. Fuller,^ or that advocated by the Bureau of Standards at Washington,' may be deter- 
mined. The deficiencies of one material, therefore, can be balanced against the advantages of 
another; and by proper combination of the two, as determined from this curve, a determina- 
tion may be had as to the proper proportions of the several materials. A few trials will give a 
very close approximation; and if the quaUties of the several materials are maintained to sample 
throughout the work, these proportions may be safely followed. It is probable, however, that 
the character of the materials will change more or less throughout the job, so that it is usually 
necessary to continuously check the several shipments of materials as they go into the work ; 
and if they vary seriously from the established standard, to alter the proportions of the concrete 
in accordance with variations noted by repetitions of the processes before noted. 

10. Proportioning by Maximum Density Tests. — A direct test which reproduces actual 
conditions is always preferable to an indirect test based on assumptions subject to variation. 
The result desired in proportioning concrete is a mixture of maximum density, and the most 
direct means to this end is the testing of trial mixtures. These are best made with concrete, 
or the coarse aggregate may be omitted and the mortar alone used. The latter method, how- 
ever, is not always representative, as in this case the voids in the coarse aggregate must bo 
determined and the concrete so proportioned that the mortar will fill the voids in the gravel 
or the stone, with a certain arbitrary excess, thus introducing an element of error. As a factor 
of safety, the amount of mortar should exceed the voids in the gravel or stone about 10 5c* 
Nevertheless, this method has some justification as the percentage of voids in coarse aggregate 
is less variable than in sand, and also because an error in determining them has less effect on the 
quantity of cement used. 

In proportioning by trial mixtures, definite quantities of the materials in proportions first 
determined by mechanical analysis are mixed with a requisite quantity of water and are put in 
a metal cylinder about 1 ft. long by about \}^in.ui diameter, and tamped. The volume they 
occupy is then determined by measuring from the top of the cylinder. 'Wlien this has been 

> See "Concrete, PUin and R«nforcc<i/' by Tatlor and Thompson. 
< See BuU. 58, Bureau of Standards, Washinston, D. C. 


determioed the mixture is removed from the cylinder, the Utter cleaned and e, new mixture made 
and tried out in the same way, with a slight variation of the proportions of sand, stone and ce- 
ment, but with the quantity of water constant. Very soon a mixture which will give the least 
volume for any given quantity of materials will be found and this mixture will give the densest, 
moat impervious and strongest concrete with those materials. 

This is known as a mixture of maximiim dentily, but it is apt to be inconvenient for use 
in ordinary concrete work, inasmuch as it contains so large a proportion of stone that it ia ex- 
tremely harsh and difficult to work. To make it freer-working, more aand is generally added; 
and, although something of strength and density is BacriHced by so doing, the advantages of 
easy working and increased compactness in forms probably compensates for disadvantages 
arising directly from any impropriety of proportions. 

Proportioning by maximum density is very readily applied in the field, all that is necessary 
being an iron pail and a pair of scales. Proportions can be determined for the concrete, without 
the use of a laboratory apparatus or any unusual equipment, by weighing out the materials, 
having due care that the sand is reasonably dry so that too great volumetric errors may not be 
introduced, and then mixing them in the pail until a mi^iturc of maximum density is obtained. 
All tests are useless, however, unless the determined proportions apply to every batch mixed 
and placed in forms. 

11. Checkdiig Materials on the Job.' — When the materials used on the job are from the 
same sources as those tested and from which tests the proportions to be uaed were determined, 
it is a ample matter to check up their qualities. Sand and stone 
from the same source do not vary much in quality, except in so far 
as quality is influenced by size of particles. Having once established 
by t«st the suitability of sand and stone for any grade of concrete 
and having determined the proper proportions in which to use 
them to attain a certain desired result, it is only necessary there- 
after to see that the size, grading, and proportions of these ma- 
terials are reasonably constant to insure uniform quality of con- 
crete. Such a check on size and grading should be bad on each 
and every shipment of materia! and ia easily obtained with a small 
Bet of sieves, or in the case of sand, which is by far the more impor- 
tant material, by means of a self-contained sand t«ster (see Fig. 2). 

The regular and systematic testing of the size of the aggregates 
gives data which will permit the engineer to tell without further 
testa, whether the aggregates will produce a better or poorer con- 
crete than that produced by the original or standard sample. This 
fact is baaed on the well-established principle that, other things 

b^ng equal, the abrogate whose granulometric-analyais curve TMter-^poriableii 

most nearly approaches the line of maximum density will pro- i^Sd*"* "™''"'""' »"'3™" <>• 
duce the beat concrete. This makes it possible to determine with 

reasonable certainty which of two sands of the same kind and from the same aource, but 
differing only in fineness, will make the better concrete. 

To illustrate: Concrete ia to be placed in a certain locality. Thereare to be heavy machin- 
ery foundations and thick building foundation walls and footings below grade, with rein- 
forced superstructure. The engineer in charge secures samples of the available concrete aggre- 
gates, both fine and coarae, and sends them to the laboratory for test. The tests show that 
although the beat available sand haa a strength in 1 : 3 mortar only 70% of that of standard 
Ottawa sand, yet mixed in the proportions of 1:1?^: 3?^ with the cement and coarse aggregates 
to be used, the resulting concrete haa a coiiiprensivc strength of 2600 lb. per aq. in. at 28 days. 
Other proportions give higher and luwnr strengths, ile|M'nding on their riehno&i, but aa the 

>CB*PHaHaDd Johhuh: Bmi. Rtc.. June 12, 19,20, 1915. 

•KolsKh & Co., Fulton Bt., New Yoik. 

veml Sand 


design of the structure requires concrete having an ultimate strength of 2500 lb. per sq. in. 
the 1:1^: 3^ proportion is used. For the foundations and footings, the designs being based 
on an ultimate strength of 1500 lb. per sq. in. in the concrete, the proportion of 1 : 2^ : 4f 2« 
which gave in the test a compressive strength of 1550 lb. per sq. in., is chosen. Under the 
present standard method of specifying sand, this particular sand could not have been used in 

Among the tests advantageously made in the field on a sand are granulometric-analysis 
charts made with the sand tester (Fig. 2). This sand tester has five screens having 6, 10, 
20, 35, and 65 meshes per in. Each screen in succession has openings one-half the width of 
the openings in the preceding screen. The charts are averaged and a special guide chart is 
prepared for the use of the inspector on the job. In making up this guide chart a permissible 
variation of about 2.5% each way from the mean of the tests made on the sample, is allowed. 
A copy of this chart is sent to the job and a copy kept in the office files. 

As the sand arrives on the job the inspector, or some one designated by the superintendent, 
makes tests with the sand tester and compares the resulting chart with the guide chart. If the 
results show greater variation than is permissible — ^particularly if they show the sand to be 
finer than shown on the guide chart — ^then the matter is taken up with the one who supplies 
the sand. 

By this method the quality of the aggregates is recognized and provided for in the selection 
of proportions for the concrete, and enough cement is used to produce the desired qualit3^ 
In this way the uncertainty which is attendant upon separately testing each of the three 
materials, and predicting therefrom the quality of the concrete resulting from their combina- 
tion, is eliminated. The time required for testing the combination is no greater than that 
required for testing any one of the materials.^ 

12. Proportions and the Measurement of Materials. — Proportioning always involves 
measurement of materials. Even with the most exact determinations of proportions, if measure- 
ment of materials in the field is inexact and variable, concrete so made will necessarily be a 
substance of extremely uncertain value. Furthermore, so long as there is prevalent a tendency 
to use excess water, even the most exact measurement of stone, sand, and cement may be 
nullified. Those who seek the best results must use the utmost care not only in the initial 
determination of the proper proportions, but also in the measurement of the quantities of 
each material employed, and in checking the qualities of the materials that come on the 
work. There will be a proportionate improvement in the general quality of concrete as at- 
tention is more generally paid to these matters. 

13. Proportioning Bank-run Gravel. — It is often questioned whether or not a natural 
mixture of sand and gravel as taken from the bank is suitable for concrete work. Inherently 
there should be no objection to this material, provided it is not contaminated by impurities, 
but the proportions of the several grades of sand and gravel in any bank are extremely uncertain 
and variable. Taking bank-run gravel and mixing it with cement in the proportions of 1 part 
of cement to 6 of gravel is not in any sense equivalent to 1 part of cement, 2 of sand and 4 of 
screened gravel, or to 1 of cement, 1)^ of sand or 4}^ of screened gravel, or any other equivalent 

Gravel of itself, if of proper quality, makes a most excellent concrete, equal to that pro- 
duced by the use of crushed stone. Sand in bank-run gravel is often of excellent qualities equal 
in every way to that taken from large deposits of exclusively fine material. However, if bank- 
run gravel is to be used, the relative proportions of sand and gravel must first be determined 
by a series of tests on representative samples in sufficient number so that the average of the bank 
may be determined with fair accuracy. These materials may then be combined with cement, 
preferably by proportioning for maximum density. After proportions are determined in thi.s 
way, it may be found possible to use a bank-nm gravel as it comes. On the other hand, it may 

> Sec r. M. rHiPM^N: "Speri6rtttioiM f..r Conrri'te .\i?i?nKnies/' Proc*. Am. Soc. Te«t. Mat., 1916. 



be found necessary in some cases to screen out the finer materials from the coarse and recombine 
them in proper proportions, the defining limits between gravel, sand, and other grades of 
materials being as stated in Art. 16, Sect. 1. 

This method of screening and recombination is always cumbersome and except on ex- 
perimental or very large scales,' impossible of putting into effective practice. In other cases, 
after existing proportions of the several grades have been determined as above, a measured 
quantity of sand, or of gravel, or of broken stone of a size or sizes lacking in the bank may be 
added to the pit-run gravel, the quantities to be added having been determined by test. In 
this way, the deficiencies of pit-run can be overcome by addition of other substances readily at 
hand, or of certain of the screened-out portions of the bank itself. Only in these ways, how- 
ever, can certainty as to proportions be secured; and it must further be borne in mind that 
frequent tests should be made during progress of the work to insure uniformity. 

Furthermore, great care should be exercised to make certain that silt in detrimental 
quantities, or loam, are not present in bank-run gravel. Sand pits are less likely to contain 
injurious quantities of silty materials than are gravel pits, by reason of the latter being the 
bottom of an old stream bed or a like deposit, with all materials held therein, just as they chanced 
to be when the waters receded. Furthermore, natural disintegration at the surface, with organic 
additions, affects the quality of the material. Stripping away top layers is too often omitted. 

14. Proportioning Crusher-run Stone. — The statements made with respect to pit-run 
gravel apply in lesser degree to crusher-run stone. Different stones crush in different ways 
with consequent variation in the character and quantity of the fine material incident to the 
process. The softer stones give a larger yield of fine materials than the harder ones; and often 
much of the very fine material is of a character unsuited to use in concrete. This applies es- 
pei*ially to limestone, inasmuch as limestone has a flaky fracture in its fine particles, making the 
particles very friable and rendering the adhesion of cement difficult, so that concrete made with 
such material is pervious and of low strength. 

Furthermore, where there are excessive fines in crushed rock materials, some of these fines 
are merely impalpable dust. It is almost impossible for cement to properly coat particles^ of 
this size, as the dust particles then approach in fineness the cement particles. With this very 
fine material in large proportions, a considerable source of weakness is thus introduced into 
concrete, inasmuch as these materials cannot be covered by cement. 

The use of crusher-run materials of undetermined size and grading, therefore, introduces 
an element of uncertainty in the making of concrete, which should not be permitted. The 
course to pursue is similar to that indicated for the use of bank-run gravel — i.e., adequate 
samples of the crushed materials should be taken; the proportions and size of fine and coarse 
materials determined by mechanical analyses; the mixture of maximum density obtained; and 
the proportions noted of each of the several materials required to produce this mixture. Then, 
either by screening or by diluting the crusher run with screened materials or with extraneous 
materials, concrete of proper quality can be more nearly assured. 

16. Proportioning Blast-furnace Slag and Cinders. — It is difficult to proportion for 
maximum density when blast-furnace slag or cinders are used as aggregate. This difficulty 
arises largely because of the porosity of these two materials, cinders being especially absorptive 
of water. A rough approximation as to proportions can be had with blast-furnace slag but with 
cinders it is probably better not to attempt to secure accurate proportioning, inasmuch as the 
use of cinder concretes is so restricted and their strength and impermeability are so low, as to 
render any increase obtainable by refinement of methods of secondary importance. Further 
reference as to the quality of these two materials will also be found in the chapter on '' Aggre- 
gates '' in Sect. 1. 

16. Proportioning Water. — ^Last but not least is the question of proportioning water in 
concrete. This is often given so little thought as to make it considered of either minor or no 
importance but it can be authoritatively stated that the strength of any concrete mixture is as 


dependent upon the proportion of water contained as it is upon the proportions of any or all of 
the other materials. 

Unfortunately, little is definitely known at the present time as to the proper proportions 
of water. It is known, however, that the quantity depends both upon the demands of the 
cement and also upon the character of aggregate employed, upon the surfaces to be covered, and 
the voids to be filled. Research has been recently directed to these lines with highly important 

17. Success in Proportioning. — For success in proportioning, not only must the original 
test determinations be right, and the specifications provide proper authority for their enforce- 
ment, but these powers must be exercised and a rigid compliance compelled. Otherwise there 
is no use in tests, and specifications are empty words. 


18. Mixing Concrete. — Although with careful superintendence hand-mixing will give good 
results, machine-mixed concrete is usually of more uniform quality than that mixed by hand, 
and is less expensive — except, of course, where the quantity of concrete is so small as to pro- 
hibit the expense of purchasing or renting a mixer. The engineer should preferably reserve the 
right to permit hand-mixing if practically unavoidable, but this method of mixing should be 
resorted to only when machineiy is unobtainable or where it is necessary to start work on a 
large job before the machinery has arrived. 

Some contractors mix the materials dry until a imiform color is secured and then add the 
water. Others put the material and the water into the mixer at once. Either way can pro- 
duce good results, except in hand-mixing, where the mixing of the cement and the sand in the 
dry state is the general and better practice. 

The strength of concrete is very largely dependent upon the thoroughness of mixing, and 
much care is needed in this part of the work. No matter how suitable for the purpose the 
materials and proportions of the same may be, insufficient mixing will result in inferior con- 
crete. Time of mixing is treated in Art. 236, Sect. 3, and in Art. 12, Sect. 5. 

The greatest care should also be exercised to make sure that the specified amounts of the 
materials go into each batch of concrete. For measuring concrete aggregates, it is not good 
practice to use the common form of contractor's wheelbarrow because the loads vary consider- 
ably with the variation in the heaping of the barrow. Special barrows constructed with sides 
nearly vertical can be obtained which will give the required amount when level full. The 
proper measuring of materials is discussed in Art. 23e, Sect. 3. 

19. Amount of Water to be Used in Mixing Concrete. — Sufficient water should be used 
in mixing to obtain a concrete of sufficiently mushy consistency to be readily puddled. In re- 
inforced work the amount of water should be such as to make the mixed concrete into a flowing 
paste that will flow readily around the reinforcing steel and require only light tamping or pud- 
dling to bring the mass to a homogeneous condition. A slight excess of water is preferable to 
not enough, but there should not be any appreciable quantity of free water present. Concrete 
is mixed with an excess of water if pools are immediately formed on top of the concrete when 
deposited in the forms. Although the quantity of water needed in different batches will vary 
occasionally because of the condition of the materials, the amount to use can be regulated best 
by measurement. A tank with a float fastened to an indicator on the outside is easily con- 
structed in connection with a concrete mixer. The effect of consistency on strength of concrete 
is discussed in Art. 9, Sect. 5. For harmful effects from the use of excess water, see chapter on 
"Water" in Sect. 1. 

The general t3rpe8 of mixers are described in Art. 22, Sect. 3. 

20. Transporting Concrete. — The transportation of concrete is not only an engineering 
problem, often of first magnitude, hut as a physical operation it is of prime importance in its 
effect on the qualities of material in the manufacture of which transportation and transports- 


lion equipment arc an incident. Briefly, the transportation system must be such: (1) that the 
time interval elapsed between reception of concrete and its delivery to forms will not cause it 
to dry, or to take initial set; (2) that the system shall be tight, so that more fluid portions may 
not be lost in transit; (3) that the mode of transit shall not promote separation of ingredients; 
(4) that the delivery shall be approximately continuous, so that mixtures of varying composition 
may not be caused by stoppage and settling; (5) that it shall be efficient, rapid and economical. 
In this summary of principles, the order of importance is such as to emphasize quality of pro- 
duct delivered, as well as cheapness. 

The varied and various appliances for the delivery of mixed concrete to forms are dis- 
cussed and illustrated in Sect. 3 on ''Construction Plant." To each individual need must be 
applied such means as careful analysis and study indicate, so correlated and systematized 
that the ends desired will best be served. In the proper selection of transportation plant, per- 
sonal experience and judgment enter as factors of such importance that, on large operations in 
particular, profit or loss may depend wholly on them. In default of these, a safe rule is to study 
methods and equipment used on operations of like character, either by first-hand inspection 
or in printed reports; to supplement information so obtained by advice from those who have 
had direct experience; and to so adjust and modify the w&ys and means indicated by the fore- 
going as to suit them to the needs of a particular situation. 

21. Depositing Concrete in Forms. — Responsibility for the character and quality of con- 
crete does not end with ita arrival at the forms. Depositing, or placement in forms is also an 
operation of prime importance and its conduct is governed by elementary principles which are 
similar to those that govern the transporting of concrete. These principles, directed toward 
securing quality with economy, are: (1) that the concrete shall be continuously and evenly 
placed in forms; (2) that it shall not be deposited continuously in one spot, with lateral flow 
and (in wet concretes) gravity separation of lighter, more fluid portions from those that are 
heavier; (3) that it shall not be deposited in forms in a manner tending to promote dissociation 
or segregation of the component materials;^ (4) that so far as possible, forms shall be continu- 
ously filled without stoppage, to prevent laitance, or stoppage planes; (5) that before new concrete 
is deposited on concrete which has set, special precautions shall be taken to secure union between 
the two; (6) that concrete shall be so deposited as to minimize the entraining of air; (7) that con- 
crete shall be joggled in the forms or that forms shall be tapped on the outside after filling, suffi- 
ciently to expel a considerable portion of entrained air; (8) that puddling and tamping shall 
be done sufficient to bring about close filling of forms, and close contact with reinforcement; 
(9) that larger aggregate shaU be spaded away from forms at the concrete rises, permitting a 
dense mortar coat and smooth finish at the exterior surface of the casting; (10) that uncombined 
concrete shall not be deposited through water; (11) that concrete remixed or retempered after 
initial set shall not be deposited in forms; and (12) that no concrete shall be deposited in cold 
or very hot weather unless special and adequate precautions are taken. 

22. ContinuouB and Even Depositing in Forms. — The temptation is very great to localize 
delivery of concrete at one point, with reliance on gravity or hoeing for distribution to other 
parts of a form. By such indulgence, one set-up of spout or barrow runways only is needed; 
and by adding excess water, the form gets filled with less labor than where movement of the 
spout, or movable or multiplication of runways are required. There are, of course, forms of 
such section and dimensions that localized delivery is both permissible and advisable, but where 
the form is long and high, localized delivery brings about a stratification or banding that not 
only mars the appearance of the wall, but also provides fault planes along which seepage and 
disintegration may proceed. 

Furthermore, bearing in mind the harsh nature and heavy weight of concrete, the difficulty 
of manual spreading in forms of concrete dumped on one spot creates a tendency to the use of 
freer-flowing mixtures, and the ease with which a certain degree of flow may be brought about 

1 See N. C. Johnson: Sng, Ree.t Dec. 4, 11 and 18, 1915. 


by the addttion of water gives rise to the use of water in excess quantities in an attempt t<> 
further accelerate the pladng operations. (The coiuequences of such additions are treated a* 
length in chapter on "Water" in Sect. I.) With such sloppy, oroverwet concretes, but littlr 
iniaginstion is needed to conceive what actually happens in the famiB. The more fluid por- 
tions flowing off from the delivery mound can; with them much of the cement, together with 
the lighter portions of «and, and 611 the loner unoccupied parts of the form, there to 8olidir> 

Fin, 3. — "Soup" ia (ornu. Fio. 4. — Looh itons lactioD — ths top of the 

depovtins btsp. 

In a chalky mass of "laitancc" in which is embedded much cement needed by the stripped afc- 
ftregate left higher up. If materials are subsequently dumped into this "soup" before it sets, 
xegregations rmult by reason of the several materials settling through this fluid in the order of 
iheir gravity. 

' Cause and effect are shown in Pigs. 3 and 4. The cause in Fig. 3 is a knee-deep puddle of 
these light materials. The effect, m Fig. 4, is a reservoir wall section almost devoid of cement 
and sand — the top of the heap — while sdjacent to it is a 
lower section that can be chopped like chalk.' 

Knowing the procedures to be avoided, substitute pro- 
cedures suited to individual needs may be evolved. Local- 
ised delivery brings a chain of evil consequences. Distribute<l 
delivery avoids these, at an expense only slightly greater. 
The gain in quality, endurance and value is worth the differ- 
ence in first cost. 

S3. Coatinaons D«po>)ting to Avoid Stoppage Planes. 

— Even in concretes mixed only to a plastic consistency, 

there is tendency for a scum of light, chalky material ("lait- 

ance") to rise. The greater the quantity of water, the 

thicker this deposit, which also is aggravated by ailty sand 

or dusty stone. Such a deposit at the top of a foundation 

block is shown in ¥\%. 5. Hie scrolls were traced by a lath, 

the depth of deposit being about ^ in. and the thickness of 

block about 4 ft. Concrete waasubsequently deposited directly on tUslayer, ssitisinthouaandi 

of other instances daily, but in all of them, this will remain as a plane of weakness, ready to yield 

when stress of proper character is imposed. Visual evidence of such yielding is furnished hy 

seepage of water and disintegrations starting at like planes in concretes on every hand. 

> Bm N. C. JomKih: S<v. B*e., Due. 30, 1918. 

D. A. Aaauii: CimaiU. April, IBIT. Ptk. Am. W. Wlu. Aaor . 1S16. 
Ciu Oatu»: Pne. Am. Sm. C. G., April. IBIT. p. MO. 


24. Bonding Set and New Concrete. — The foregoing is closely related to the problem of 
bonding new and old concrete or, more properly, set concrete and concrete subsequently cast 
upon it. Recognizing that at least on the majority of concretes, a top film or deposit of lait- 
ance exists; that this deposit is loose in texture and non-coherent; and that a portion of it is 
hydrolized cement, it is not to be expected that concrete subsequently placed in contact with it 
shall adhere. It is known and recognized that a dust film on stone or gravel will prevent adhe- 
sion of cement and it must no less be expected that, a like film on solid concrete, often multi- 
plied many times in thickness, will have like effect. Other and more complicated conditions 
also affect the procurement of bond, but those above given are of themselves sufi&cient to ac- 
count for the failure of many attempts (see Art. 50, Sect. 1). 

A first essential, therefore, in procuring bond is to remove this separating laitance film, 
whether the set concrete is hours old, or years old. It is best to remove at least Hm, and pos- 
sibly it may be necessary to remove several inches before clean, sound concrete and aggregates 
are exposed. This surface should then be well washed and preferably soaked with clean water, 
all loose material being removed. A wash of rich neat grout well scrubbed in with clean brushes 
will provide a good bedment; and before this has set or dried, the new concrete should be depos- 
ited, a first thin layer being tamped into place, followed by the full deposition. The foregoing 
gives better guaranty of success than methods usually followed, but it should be borne in mind 
that drying out of the fresh concrete surface, or drying or setting of the cement wash previous 
to applying and ramming the first layer of concrete, or failure to deposit the remainder of the 
concrete before this latter has taken set, will each be suflFicient to cause failure to bond, as each 
can and will duplicate in greater or less degree the separating film which first caused trouble. 

Bonding fluids and compounds are marketed under various trade names, but these cannot 
be successful unless conditions suitable for bond, as outlined above, are first established. 
Hydrochloric acid is advocated by some as a wash preparatory to bonding, but the amount that 
would be required if unassisted by picks or chisels in removing the usual laitance coat to a 
sufficient depth makes its use prohibitive, both in cost and in time and labor required. Its use, 
even when considerable effort is made to wash it away after use, is not to be recomihended, as 
concrete by its porosity, is capable of absorbing harmful quantities. 

25. Removal of Entrained Air. — The customary mixing and depositing processes entrain 
quantities of air. Even when the volumetric air content of a concrete appears low, a con- 
siderable portion of the aggregate may actually be isolated by air, with little or no attachment 
to cement.^ It is doubtful if the weaknesses produced in concrete by the occlusion of air are 
appreciated. The evils of existing practices in this particular are to be deplored. In particular, 
spouting unconfined from a height, or dumping from barrows in like manner probably do maxi- 
mum damage in this particular. The present tjrpe of mixers work further evil in this regard. 
But since many present fixed practices and equipment entail the occlusion of air, with no like- 
lihood of an immediate change, the removal of as much as possible is logical progress. 

To this end, vibrating rammers applied to the plastic concrete, or air hammers rapping the 
outside of forms, or even sledge or maul blows* have been used with good effect. In concrete- 
products plants, vibrated molds have been used to obtain superior density; and in road work, 
vibration by motor, applied to mats on the fresh-laid concrete are said to produce superior 
wearing qualities.' Certainly if the introduction of an objectionable impurity in a structural 
material cannot be prevented, but its removal can be later effected, it is the part of constructive 
engineering to overcome the undesirable effects while seeking to remove the cause. 

26. Spading, Puddling and Tamping. — Forms should be closely filled, and, so far as pos- 
sible, close contacting of form surfaces with smooth, plastic material should be brought about. 
Since large aggregates tend to bridge over, or jam, leaving unsightly surface pocketa, they should, 

t See N. C. Johhbox: Bng. Rec, Jan. 23, 1915. 
C. B. McCuixouob: ConcreUt April, 1917. 
> See H. S. Caxpsittsb; Bng, Bee., March 31, 1917. 
* The Vibrolithie Pavement of R. 8. Stubbt Co., Anatin, Tex. 


aj3 the fomi is filled, be spaded back from form surfaces so that a dense, smooth 
mortar may lie at exposed surfaces. Furthermore, since it is essential for structural 
strength and for preservation of steel that the embedment of reinforcement be adequate with 
close contacting of mortar, puddling of concrete should be progressively carried on as forms arc 

The temptation to use excessively wet concretes to lessen labor in the two foregoing opera- 
tions is prevalent. For intricate remforcement, a free-flowing concrete must be used, but it 
is better to obtain the requisite flow by sufficient mixing, by the use of finer ballast and by pud- 
dling than by indulging in excess water, which so generally defeats the intent of its use. 

27. Depositing Concrete Through Water. — Care should be taken in depositing concrete 
under water that it is not deposited through water, unless confined. 

Underwater concretes are usually deposited by means of a tremie — a tube of about 1 ft. 
diameter at the top, slightly flaring at the bottom and at the start of a length sufficient to reach 
to the bottom. As deposition proceeds, the deUvery end may be raised, but not out of the soft 
deposited concrete, else water will enter, causing washing of concrete subsequently deposited. 
The tremie must be kept full of concrete at all times; and deposition is assisted by moving the 
bottom of the pipe slowly about, permitting gradual discharge. If the charge is lost, and the 
tremie becomes filled with water, it is wise to add extra cement to the next charge, in order to 
compensate for that which will be lost through washing away. Necessarily, a tremie is heavy, 
so that scow, derrick or other handling arrangements must be provided. Care also must be 
exercised in order that waves from passing boats may not lift the tremie as well as the scow, 
causing loss of charge. 

Underwater buckets, which are substantially boxes with bottom-dumping doors, have been 
used in some underwater concreting, but their use is more costly than that of tremies and 
possibly less satisfactory. Tilting buckets are not suited, to underwater work, inasmuch as 
their dumping subjects the concrete to washing. 

Depositing in cloth bags of greater or less size to hold together the mixed concrete in pass- 
ing through the water, has been successfully accomplished. > Paper bags are less successful 
than are those of jute or burlap. The adhesion of successive bags is dependent upon trans- 
fusion between and saturation of the bags with dissolved cementitious products, but in view of 
the great mass in which the concrete is used in such operations, and its gravity functioning, lack 
of strength at joining planes is of little moment. 

28. Remixed and Retempered Concrete. — ^It is erring on the side of safety to reject all 
concrete or mortar which has taken pronounced set, whether initial or final, or which requires 
the addition of water and reworking to have requisite plasticity. The exact actions which take 
place during initial set are not precisely known, but it is probable that in this process is begun an 
interlacing crystallization which is later augmented by other crystallizations and depositions 
of colloidal (amorphous or non-crystalline) nuiterial in the processes of final set and the sub- 
sequent hardening. But whatever the exact process, it is known that retempering and re- 
working of Portland-cement mixtures after initial set is decidedly disadvantageous at best, 
resulting in a loose, unresistant product of inferior strength and coherence. This practice, 
therefore, is to be avoided; and the operations of transporting and placing should never be of 
such duration as to permit initial set, even in hot weather. 

29. Concreting in Hot Weather and in Cold We«ther. — The basis of all concrete is the 
union of inert materials by substances produced through chemical reaction between Portland 
cement and water. Any acceleration or retardation of this chemical process affects the quan- 
tity and quality of binder resultant from this reaction; and any such alteration affects critically 
the quality, strength, and endurance of concrete formed by admixture of this binding product 
with sand and stone. . 

Temperature is known to control the rate of all chemical reactions. In general, heat 

* Proc. Am. Soc C. E., vol. 39, p. 126; and vol. 47, p. 101. 


accelerates and cold retards chemical union. ^ Furthermore, solution of some products, such as 
gypsum (CaS04) contained in Portland cement, is active at relatively low temperatures and 
inactive at higher temperatures, while solution of other products takes place in reverse order. 
Relative evaporation speeds at different temperatures are also to be considered, with correlative 
effect on the strength of concrete produced at any given time. It is reasonable to expect, 
therefore, as is borne out in fact, that hot (weather) concretes are quicknsetting and of early 
strength and that cold (weather) concretes are slow-setting and of low strength; and on forget- 
fulness of these obvious but inescapable facts rests responsibility for many a failure. 

Particularly is this true of cold-weather concreting. At 40T. concrete requires four 
times as long a period to attain a given strength as the same concrete at 50T. ; and at 40^F. 
about nine times as long as at 70'*F. Below 40°F. the ratio still further increases. Many so- 
called "mysterious" failures, in which the concrete is obviously not frozen, are to be explained 
by delayed set and hardening, due to low temperatures alone. Below 40*'F. the set is so 
delayed down to and including 32®F. where rupture by ice formation occurs (requiring a 
later extra period at elevated temperatures to induce reconsolidation in addition to that 
normally required for setting at the average temperature prevailing) that computation must 
be made for each instance to insure safety. 

Using Portland cement of normal hardening rate, the following periods before removal of 
forms in summer weather are suggested as representative of correct practice: 

For concrete in mass work 24 to 48 hr. 

For concrete in thin sections 48 to 60 hr. 

For concrete columns 48 to 60 hr. 

For concrete in beams and girders 12 to 21 days 

For concrete in long span slabs. 14 to 21 days 

The period required in cold weather will be more or less protracted according to the 
average temperature prevailing' both prior to and during the setting period, inasmuch as tem- 
peratures prior to mixing and placing will hold for the aggregates, even though in many cases 
attempts at preheating have been made. 

S9a. Preheatmg Aggregates and Water. — ^Preheating sand, stone, and water 
previous to admixture is an operation difficult adequately to perform. Each cubic yard of 
materials will require approximately 1000 B.t.u. per degree rise in temperature. With an 
indeterminate factor of heat transference, the fuel required on a day's operations may be com- 
puted or, better still, such computation may be neglected and fuel added until the temperature 
of the materials has been sufficiently raised. It is erring on the side of safety to have this 
temperature judged by an unsensitive, calloused hand, rather than by a thermometer. Water 
may more easily be made too hot, inducing flash set when mixed with cement. 

296. Means for Heating Aggregates. — ^An old smokestack section, buried in 
sand or stone, and fired with wood, is perhaps the best construction-job means of heating ag- 
gregates. Steam jets are the least efficient. Water may be heated by either immersed steam 
coils, or by steam jets, or by externally applied heat. A gasoline torch playing directly into the 
mixer drum is sold as a concrete heater. 

i9c. Sodositre and Heating of Forms. — In cold-weather buHding operations 
in particular, enclosure by canvas is desirable. Salamanders, or other heating units are kepi 
burning within to keep the temperatures somewhat elevated. It must be borne in miod, 
however, that at best the temperature of the enclosure is low; and that beat transf^riiviee 
through wooden forms to the concrete is slow. Such precautions, therefore, do not admit tA 
dispensing with preheating of aggregates and water, or of leaving forms in pbee for a 

* Tbe speed of «*«— ««*«J mctioiie b »w>ro¥imat»ty mm the waXh power of the tbmAvUt VumpiwmtL^Htt 
s See A. B. McDAJnKi.: Prme. Am. Con. Inst., Itl5; mbo Art. IC Sect. o. 



29d. Protection Against Frost — The employment of manure in contact with 
concrete is seriously objectionable. The heat of manure is derived from the decomposition of 
its organic portions and in this process, compounds destructive of concrete are formed. Clean 
straw, clean sawdust, or canvas will assist in protection against frost, but in addition, artificial 
heat must be employed for temperatures below 35°F. if assurance of safety is desired. 

296. Freezing of Concrete. — If frozen before initial set, concrete will reconsoli- 
date on later elevation of temperature with seemingly no impairment of strength. This holds 
particularly for sections where there is sufficient hydrostatic head to recompact the mass as 
the expansively disrupting ice is thawed, chemical reactions, in the interval, having been sus- 
pended. It is better, however, to prevent freezing than to take chances. 

29/. Use of Anti-freezing Mixtures. — Common salt (NaCl), or calcium chloride, 
(CaCls), is the basis of most anti-freezing mixtures. Glycerine and alcohol also have been tried, 
but both tend to lower the strength and there is also question as to the propriety of using gly- 
cerine, because of possible organic decomposition and injury to the concrete. Calcium chlo- 
ride or salt added to water will lower its freezing point, and in proportions of CaCU not to exceed 
2% of the weight of cement or proportions of salt from 2 to 10% of the weight of water, have 
been recommended and used, but the ill effects of salt so far outweighs its benefits — as for in- 
stance, by promoting corrosion of steel — ^that it is better omitted. No anti-freezing compound 
is better than salt; and none is equal to adequate heating of materials with proper maintenance 
of temperature during the setting period. 

29£r. Protection Against Heat — Aside from slab and thin wall construction, 
protection of concrete against heat is rarely needed. For such purposes, protecting coverings 
of straw, sawdust, sand, or canvas^ are usually sufficient. Evaporation must be guarded 
against, as must also working after initial set, as in floating floor or sidewalk surfaces. Hot- 
weather evils, however, are less troublesome than are those incident to cold-weather concreting 
and are provided against with corresponding ease. 


30. Object of Field Tests. — The primary object of making field tests of concrete is either 
to obtain information as to the strength of field concretes or assurance as to the strength and 
integrity of a commercial structure. 

81. Limitations Inherent in Field Tests. — Necessarily, field tests of concrete must be made 
on specimens of such section — usually 6-in. cubes, or better, 6 by 12-in. or 8 by 16-in. cylinders 
— ^that the maximum strength to be anticipated shall not exceed the capacity of testing appa- 
ratus available. This limits the size of specimen to a considerable degree, which affects the 
relationship between strength of test specimen and strength of a like section in the structure 
according as aggregates of greater or less size are used. 

Necessarily, also, the strength of test specimens has dependence upon the degree of compact- 
ing and care of molding employed in their manufacture. In actual structures, quite dissimilar 
internal conditions exist, with static head playing a more or less important part in consoli- 
dation, this static head varying continually in each portion of the structure. It is therefore 
difficult, if not impossible, to duplicate in test specimens pressure conditions obtaining in a 

It is assimied, furthermore, that the materials incorporated in a small test specimen are 
representative of, and in like quantities to, those making up concrete in the structure. It is a 
regretable fact that a concrete mix is rarely of uniform composition in its several parts; and that 
so great is this variation found to be that relatively small portions of any mix may or may 
not represent in their constitution and properties when hardened a fair average of the con- 
rrotes in the structure. 

Temperature conditions further increase discrepancies between test spocinion** and strut*- 

> S«c Seel. 4 ou '• Concrete Floors and Floor Surfaces, Sidewalks, and Pavements.*' 


tural concretes. When concrete b in considerable mass, temperature rise due to chemical re- 
actions between cement and water are largely retained,^ and atmospheric variations exercise 
less effect on the proper increase of strength. In small specimens, on the contrary, moicture 
and temperature conditions are subject to abrupt change with consequent variation of proper- 
ties in the hardened concrete.' The mode of applying stress is another factor tending to dis- 
similarity and to misleading conclusions. 

82. Comparative Tests on Field-molded and Structural Specimens. — Two notable series 
of investigations are on record with respect to the value of field tests of concrete. Those of the 
Public Service Commission of New York* give comparative values between field-molded speci- 
mens and specimens cut from the actual structure. Those of Kansas City, Mo.,* were tests on 
field-molded specimens alone, without comparative tests on specimens cut from the structure. 

83. Value of Tests on Field-molded Test Specimens. — The indications of the tests above 
mentioned are not favorable so far as consbtency between laboratory and field is concerned, 
and this is to be expected, as the practice of sampling concrete from a mixer; molding such 
samples more or less inexpertly into small specimens; curing them under conditions dissimilar 
to those structurally existing; and applying a breaking stress, must, obviously, give results of 
doubtful value. Then again, by the time these test specimens are matured and broken, tons 
upon tons of concrete have been piled on or around that portion of the structure of which they 
might have been part, so that the removal of this concrete would be next to impossible, even 
though test results should be adverse and indicate a low strength. The best that might be done 
would be to so vary mixtures or procedures in subsequent parts of the work as to produce more 
nearly the values desired. 

84. Transverse Tests on Beam Specimens. — A variation in form of specimen and method of 
testing introduced in the Welland Canal tests is of interest, though subject to all limitations 
above set forth. In these tests^ the test specimen is a beam of rectangular section 4^^ by 3^ 
in. and 3 ft. long, tested transversely. Such tests give little if any indication as to the ability of 
a concrete to withstand applications of stress other than exactly similar to those applied in the 

85. Core Drill Test Specimens from Actual Structures. — In certain instances core borings 
to secure test specimens have been made in completed structures. Such cores are more rep- 
resentative of mass conditions than other specimens. In taking borings at the Mtmicipal 
Filter Plant, Cleveland, Ohio, both 6-in. and 4-in. cores were taken with a diamond drill bit, 
the cores being subjected to examination and tests of various kinds. Somewhat similar work 
but with a shot drill was done at the Ashokan Reservoir of the New York City water supply 
system in 1916. 

Either a shot or a diamond bit may be used in core borings. The shot bit is slower, cuts 
reinforcing steel more readily, but gives a rough core. The diamond bit cuts rapidly, gives a 
smooth, even core, but the diamond loss may be a serious item of expense, particularly where 
tie-wires or reinforcing steel is encountered. Either bit is almost helpless where segregated 
pockets of loose material are encountered. 

88. Suggested Methods for Makhig and Testing Field Specimens of Concrete. — The fol- 
lowing methods for making and testing field specimens of concrete are taken by permission 
from the report of Committee C-0 of the Am. Soc. Test. Mat., June, 1917. 

The following methods are presented not as final recommendations but as an outline of what in the opinion 
of the committee represents the best practice at the present time. The necessity for greater attention to testing 
concrete in construction, and for the adoption of a proper method for sampling the concrete to repr esent the 
product of the various field operations, is recognised by engineers and contractors, eiipeciaUy in view of the tend- 

1 Paul and Mathbw: TranM, Am. Soc. C. E., 1915, pp. 1225-1267. 
« A. B. McDanikl: BuU. 47. Univ. of 111., 1915. 

WtTBBT: Wise. Engr., Feb.. 1915. 
' Bng, Ree„ Sept. 4, 1915. 
* Bnc. New, Sept. 10, 1914. 
» Bng. Ree„ July 24, 1915, p. 112. 


flDcy in nuny quKrtcn to iu« a wet, sloppy coucreta which mfty give b final etrencth much lower thuk that upon 
which the design la bated. 

ThB tfBia ue dnigned to ptovide an indiaation o{ the quality of the eonerete which ia pl«c«d in the itruclure 
and cliaTaGter of workmanahip in miung. By providing damp aand atorage for the tat apecimena, the variable 
weather oonditiona are purpoaely disrecuded although these lODietiDm greatly affect the final atnngth of the cod- 
erata. In oomparicig the reaulli, the temperature and weather conditions miut be taLen into account. 

Silt and Sliapei^3ptimc7t.—Tbt teat specimen should be of cylindrical form, with leogth twice the diameter. 
When the eoarae aggregate doea not eieeed IH In. in diameter, a S by IZ-in. cylinder may be uaed. although an S 
by l&-in. eylinder ^tb more sonoordant reaulta. For Urger-aiie Mgregste a mohl whose diameter is not leaa Iban 
4 timea the diameter of the largest aiie aggregate should be uaed. 

SjiiCr ^^ SUynJ'J^mB* 

JtfoUi and Appnratu.— Figs, fl to 8 sh< 
show type* whioh are deeiioed (or repeated ui 
or elae has to be reaoldared. While this latter 
in actual eoat, it ia quite ooDreaiest to use 

where apeeimena are to be ahipped at early stasea aa the apedm 
niodificationa of each of these forms will euggeet themsclvea, 

Fro. 7. 

while the mold show 
•Id ia not adapted lo 

Band 7 

Individual plates of 



;h should be uaed in the Geld. I 
n in ^g. 7 ia dntroyed in removii 
coniinuom use and therefore ia more eipenaiTe 
■e to be made and particularly advantageous 

perpend iculai to the vde. 
> plane surface on which to mold the ipecimeB*. 
1 surfaeea about 2 in. larger than the diameter 
e mold, may be used, placing ope under each tqt piece- 
ccfl of wai paper should be provided to place under each 
specimen to prevent the concrete from adhering to the 
9, or the plate may be oiled, 
A central place abauld be aeieeted for molding th« 

. in damp sand at described below, to prevent undue 

Samptine Oit Cantnlt. — Concrete for the teat specimens 
lid be taken immediately after it has been plsced in the 

luld be taken from one place. A 

imber of ai 

^11 ^*i> jSJiW 

ould be Uken 
m which 


a fail 

.mple ii taken 


of the ec 

e miMed, Cw« should be tak 
It that place. 
JtfoUiair Ik* Sptriwun, — Tbe.paila containing the aamplea ol coi 
(or making the teat pieces aa quickly aa possible. To offset segrtcat 
sample should then be dumped out of the pail into a non-abaorbant 
mixing immediately plaoed in the mold. DiSenot aamplea should noi 

o( the work. The location 
should be clearly noted for f 

In seeuring samplss. the concrete is Uken up irom the 
maaa by a ihovel or aimilar Implement and placed in a large 
pail or in some other receptacle lot trmnsportitig to the plar« 

U the to 
Crete should be taken to the place selected 

be mixed together, but etch sample should 


te UDUod Ihfl edga of tba udaa ot 

nold, ■ 3-iD. n 


1, BbouJd 


Ihe mold b( 
of pl*(e b1s» or mac> 

be takea to removg sir I>ocketi. The freahly made apMinieii 
p( tbe tarra. The epecimeu should prBloribly be capped in the 
t«lin( machine. After the concrete haa stiffened »ppt*ciably 

1 then be worked around qd the top of the mortar until it rfsts 
be placed between it and the concrete. If the forma 

metal plate ah o old then b 
on the form. Thia plate should be oiled or a inece of wax pa| 
ore oarefully made, this will give top and bottom aurf aeea pi 

buried in sand whilo the specimen ie beiag molded. 

At the end of 48 hr. the apecimeDB should be removed from the mold and buried in damp aand. In case tbe 
molds shovn ia Fig. 8 are used, Bpeclmcns may be buried in damp eand without the removal of the forma, thus 
permittioa ahipment of the apecimene in the molds. Teet epecimens made in the mold shown in Fif. S may be 
removed by opening the soldered joint with a aharp tool. 

r«(>n«.— Ten days prior to the date of test, apedmena should be well packed in damp aand or wet ebavinga 
and shipped to the testinc laboratory, where they should bo stored either in a moist room or in damp sand until the 
date ot the test. It is assumed that ordinarily s 23-dHy teet will be msde. although tests st 7 and 14 dsyg will give 
soma iqdications of the results to be eipected st 28 days. In ease 7-dBy tests are made, the teet pieces ahoukl re- 
mun at the job as lone as poedble to hsiden, and should be shipped so a* to arrive st the laboratory in time to 
make the test on the required dale. 

The foregoing recommendations of the Am. Soc. Test. Mat. are subject to revision, it 
being recognized that they arc, by the nature and coat of equipment speciGed and require- 

Fio. f 

— Paraffinod paper cs 

SB. — Slitting 

ments for curing conditions, more nearly allied to laboratory procedures than to testing in 
the field. To overcome these limitations, the author has, with uniform success, used stock 
cartons of paraffined paper, such as those shown in Fig. 8A for field molds. As will be 
seen, they are simply stout cartons with caps and they may 
be had in quantities at prices ranging from IJ^ cts, for 3>^ 
by 7-ui., to 6 eta. for 6 by 12-in. sizes. 

Molding and puddling arc accomplished in tbe uaual 
manner, the mold retaining its shape, and when full, capped, 
with identifying data written directly on the cap. When the 
specimen has matured, the mold ia sUt down the side with a 
sharp knife, aa in Fig. 8B, and the shell removed. This leaves 
a perfect specimen, as in Fig. 8C, whereon, it will be noted, 
the top and bottom cardboards remain as cushions for the testing machine heads. 

To insure even bottoms, it is well to set the empty cartons on loose aand during mold- 
ing and until set. For such bottoms as are sprung, or out of true, a little melted paraffin, 
or of cement grout (if time to set is permitted) poured in with the mold vertical, will ensure 
a bedmcntso even as to make plaster preparation unnecessary (see fig. 8D). 

The advantages of this mold for field work are : 


1. It is readily and cheaply procurable anywhere. 

2. Temperature changes excepted, curing conditions for all specimeiu are always uniform 
and alike, without bedding in sand or immersing in water, aa the waned carton retains all- 

3. Shipment of specimens from job (where molded) to laboratory (for crushing) may be 
made in any manner convenient, curing proceeding uniformly throughout this period. 

4. There arc no molds to clean or to rc-ehip. 

ST. Pre-use Tests of Haterials.^lt is to be observed that recognition is accorded in the 
above recommendations to the questionable commercial values of field tests of concrete. The 
art is at present in a transition state- 
Some Geld tests, however, are of value. One of these is as follows: It seems to hold true 
that the strength of concrete is directly dependent upon the size-grading of its aggr^ate; that 
this is particularly true as respects the fine ag- 
gregate; and that of a selection of sands, con- 
crete will be strongest when made with that 
sand whose sum of percentage passing a given 
series of screens is lowest. 
.£ In Fig. 9 is shown a series of curves prc- 

A pared by C. M. Chapman illustrative of this 

g point. Percentages passing in these curves 

f9 are taken from record cards of the Universal 

'•£ Sand Tester and illustrate the field practice 

b of Westinghouse Church Kerr & Co. in the 

i selection of sands. It is obvious from this 

e chart, that if the percentages passing is known 

^ for any sand, the strength of a mortar made 

% from it in given proportions at any given age 

£ (in thb case, 28 days) may bo read directly. 

^ Although it has been held that the strength of 

a mortar is not a measure of the strength of a 
like mortar in concrete, the relationship is not 
entirely misleading. On the contrary, it now 
seems probable that pre-use tests of materials ; 
ra> Its oo "» "» tM w to 3w H5 HO mm jjjg establishment and maintenance of correct 

Co»rficie"* or tmrforrtiity 

Sim ef'Birar^piBxrj-scrmefifif^fv'Si^lhrrr proportions; and refinement of processes of 
Fto. e. mixing and placing will afford the greatest de- 

velopments in the concrete art.' 


SB. Heaning of "Waterproof." — "Waterproof" as applied to concrete may, in its literal 
sense, give rise to confusion and misunderstanding. "Water-resistant" to a speciSed degree, 
or "impermeable" might more nearly define and delimit the abilities of concrete to withstand 
attack from or permeation by water.' 

n. Resistance of Concretes to Water Action. — Few concretes are free from one mani- 
festation or another of water action. Except for minor surface attack, such action follows 
water penetration, which latter may result from an actual hydraulic head, as in a dam, sewer, 
aqueduct, or reservoir; or from a negative head induced by evaporation from an exposed 
surface, as in a retaining wall, subaqueous tunnel, or sidewalk; or it may be caused by surface 
wetting and mass absorption, as in a concrete building, or in stucco. Chemical attack by snl- 

' tec R. R. QoOBWls: Ci,KtrHr. .Vov„ 1915. 

'3m M. O. Wpthbt; "Permubilily Te.1 ol Gravel CoiKretc." #•'«■, Wotani 8oc. Encr'i.. IBM. 


vents excepted (but inclusive of secoadary frost action) the effects of penetrant water on con- 
cretes are generally alike, though differing in degr^. Water penetration is directly or indirectly 
the eause of the majority of disintegrations in concrete and the degree to which water pene- 
tration is permitted by the texture of any concrete is a direct measure of its strength and 

40. Resistance of Concretes te Water Penetration-^Penetration of water into concrete is 
readiest by an actual physical passageway or paasagewayB. Obviously, a given quantity in a given 
time may enter by one large passageway, or by a multiplication of minute passageways. 
Securing resijrtance to penetration is, therefore, to be ac- 
complished by reduction of sucfi passageways to a minimum, a 
both as to size and number, or by seating them off. It 

follows, therefore, that concretes of given materials are 

water-tight and water-resistant, ils well as strong and en- — a 

during, in proportion to their absolute densities. Con- 
versely, concretes are weak, permeable, and of low endu- 
rance in proportion to their porosities.' 

41. Degree of Impermeability Attainable. — Absolute 
freedom from water penetration is probably impossible of 

attainment in the commercial manufacture of concrete. ^^^ lO — Medin* saiiditooe sh 

Certainly, the average results of present practice warrant ini pot» which render the itane nit- 
that belief. An improvement in present-day work is not "°' " ' ^'°' uhob.) 

only an imperative necessity, but, fortunately, practicable as well. 

lUufltrative of the difficulty of obtaining absolute impermeability in artificial concretes, 
the structure of sandstone [Fig. 10) is worthy of study. This has before been cited' as an ideal 
concrete in structure, in that it has in combination silica (sand) particles closely compacted, 
withaminimumofcementitious material between them. Yet sandstone of this grade is known 
to be absorptive of water and to weather (disintegrate) rapidly. Arrows (a) and (j>) indicate 
the minute passageways in the cemerUing material between the silica particles throaghwhichw&tei 
entrance is secured and at which disintegrations center. And in further likeness to artificial 

Fio. 11. — Pioholea poaMKewayi in comiiiefcial rio. 12. 

concrete. (Magniiied 10 dUms.) 

concretes, the more cementing material in any sandstone, the higher its porosity and the lower 
its strength and endurance. 

42. Porosity of Commercial Concretes.— Necessarily, because of limitations imposed in 
artificial concretes by inadequate compacting and consolidating processes, a density equal even 
to sandstone cannot be obtained. Dispersion of aggregates in concrete, both coarse and fine, 
with corresponding increase of cementing material between and around them, has been before 

'8eeoh»ptmoi. ■■Piopoptioning Concrete," .Sect. -2. and an -Properti€= uf Plain CtincrHc," S*M. 5. 


noted.' I»>Iatioii of aggregates by water* and occlueioD o( air by muting and placing methods 
in current vogue' have been pointed out at various times. The value of proper proportioning 
as an aid to water-tightness has been the subject of frequent papers, discussions, and writings- 
Field methods, however, provocative of undesirable conditions, have remained unchanged. 
This argues either an apathy not creditable to the engineering profession and inimical to con- 
crete, or else a confession of inability to remedy recognised evils. A present lack of adequate 
presentation of the problem of impermeable concrete may be one reason for this state of 
the art. 

Search in commercial concretes for passageways capable of conveying water need not be 
protracted to meet with reward. Such passageways vary in size from "pinholes," indicated 
by the surface shown in Fig. 11, to those of finger size in Fig. 13. "A wall you could throw a 
cat through" is verbatim repetition of a field characterization which is not infrequently ap- 
plicable. It is often objected that "pinhole" passageways are not continuous, but no proof of 
such assertion is offered; and while the converse is equally difficult of proof, the porosity of 
pinholed concretes under test and the presence of like pinholes throughout any and every sec- 
tion of such concretes gives warrant for belief that they are, by their multitude, of great im- 
portance when their combined water-conveying abilities are considered. 

43. Excess Water as a Cause of Porosity. — Aside from segregated pockets of stone (which 
also are caused by excess water), water voids are, however, quantitatively more important than 
are airholes as passageways for penetrant or percolating water. 
Water voids are relatively massive, approaching segregations 
even when not so classified ; and inasmuch as the water once 
lying in them has either flowed away, or been evaporated, con- 
tinuity of passageways for subsequent water flow is strongly 
indicated, if not proven by the fact of this loss. An example 
of such water voids and flow passages in a commercial concrete, 
intended to be of superior grade, is shown in Fig. 13. This is 
typical of innumerable passageways of like character found 
'^n '^^Si" (M^ Sfi3*^ generally in overwet concretes. 
dUm.) ' 44. Shrinkage Cracks. — Shrinkage cracks in concrete are of 

a general type and so univeisal as to be viewed by the majority 
either with eyes unseeing, or regarded with the contempt that conies from familiarity. Their 
importance both as a condition and as an indicator of internal processes is, however, of the 
greatest importance. 

Shrinkage cracks are of a general type, irregular in line and radiating from a common center, 
usually a pore of greater or leas size. This is to be expected, inasmuch as such a pore is at least 
a possible point of egress for water from the mass immediately sur- 
rounding. Further, flow is freest from such an open center, so that 
under evaporation or other processes it soon becomes a point of 
dryness; and inasmuch as it is already a point of weakness, relief 
planes radiate from it gradually as drying proceeds inward, until 
shrinkage stresses ore balanced by internal resistance. 

44a. Types of Shriiikage Cracks. — Perhaps the com- 
monest type of shrinkage crack is three-branched. This is to be 
seen on every hand in concretes and stuccos, both in the gross Pio- 14. 

(Fig. 14) and in microscopic sixes (Fig. 15). Where numerous 

pure centers exist, comphcated systems build up by the junction of a multitude of like radiat- 
ing cracks from "crazing", with oftentimes, deeper and more acriuus disruptions, as in Fig. 
16. In both Fig. 14 and Fig, 16 the centers have been outlined in circles. Like cracks in 

> Ssa shaplH oo " PnqnrtioiuDi Conenta " ia Btct. 2. ud dd "Witu" in Sect, 1. 

• N. C. JOBNaOH: S<w- R'c Jbd. 23, IBIA; Dw. 30, IBIB. 

• N. C. Josmoa: B<v. Bte., Dm. 4, I6IA. 


drying eartli are everywhere to be observed, the three-branched irack prrmitting Bpherical 
contraction and relief with minimum disturbance. 

M6. Shrinkage CrackB and Porosity.— Necessarily, suc^h shrinkage cracks in 
concrete are open passageways for water. This is attested on every hand by the cryatalMne 
filling of dissolved salts left behind in such shrinkage cracks. Fig. 17 is typical of such condi- 
tions, which exist where the fluid supply is, or becomes, somewhat limited, so that super- 
saturation and crystallitatioa may be brought about. Moreover, as is evidenced by such 
crystalline fillings, these cracks once conveyed fluid through the concrete. Cracks of like 
formation are equally potent to convey other fluid; and, if the supply is ample, to remove 
soluble portions of the concrete, with mechanical dislodgment and removal of inert particles 
released by such solution. The original passageway is thus speedily enlarged, possibly to harm- 
ful proportions and certainly to increased wat«r-carrying capacity, evidence as to such quanti- 
tative removal of material is given on sheltered surfaces of concrete, such as inspection galleries 

of concrete dams, where deposits removed from the concrete and aggregating many tons are not 
infrequently piled on floors and cling to waUs. 

Uc. Prevention of Shrinkage Cracks. — Shrinkage cracks like the foregoing are 
difficult of prevention. They may be minimized in number and severity by: (1) use o( 
graded materials; (2) avoiding the use of excess water; (3) adequate mixing; (4) careful placing 
to avoid segregation; and (5) curing (annealing) under proper conditions of moisture, so that 
shrinkage stresses will be developed only at a rate commensurate with the slow increase of 
strength in concrete. This, in connection with slow drying and hardening of colloids, largely 
explains the high strength of concretes cured under water, and conversely, explains the easier 
disintegration of concretes in which too rapid drying occurs. 

But though the foregoing five principles, if made effective in practice under skilled direc- 
tion, would result in better concretes as to water-tightness, with correlative strength and en- 
durance, their field observance is so limited as to be negligible. Adequate mixing will not be 
had, so long as engineers countenance and tacitly, if not openly, approve inadequate mixing in 
the interest of quantity output. Graded sand and stone will not be used unless cheaper than 
other available materials, so long as the same price per yard in forms is paid for one as for the 
other. Excess water will be used until insistence iz hod for the use of lesser quantities. Careful, 
uniform placing needs standardization and enforeement by engineers of such standards. 
Moist curing, or annealing of large sections is often a physical impossibility, but often it can 
be done if required. The practice of curing commercial concrete is now extending; a fact 
which offeia much encouragement. Though many problems yet remain unsolved, our present 
knowledge is suflicient for great improvement, once engineering sentiment for right practice is 
brought to the point of putting them into effect. The cause of pervious concretes lies not so 
much in lack of knowledge as to how to make concretes that are impervious, as in neglect to 
put into effect the knowledge at hand. 


4Jk POTTkras Concretes and Luitance, — Laitanoe — ^the porous, chalky material which rises 
during deposition to greater or less extent at the surface of concretes — Is a chief foe of water- 
tightness. The deposit is particularly deep with excess water, or too fine or dusty aggregates^, 
or both. Concrete subsequently placed on this laitance fails utterly to bond; and seepage 
readily takes place along this construction or "day's work" joint, often followed by later dis- 
integration. Laitance an inch or more thick, scrolled with a lath, on top of a newly poured 
foundation block 4 ft. in depth, is shown in Fig. 5, page 74. In sections of greater height, as 
in reservoir walls, especially where overwet concretes are indulged, the deposit will be of greater 
depth, usually extending through the body of the wall to form a horizontal joint, open to per- 
colating water. 

Occasionally such laitance deposits are localised, forming pockets; and inasmuch as such 
pockets are formed from the finer material assumedly lying distributed in inter-particle spaces 
throughout the concrete mass, their isolation implies and often proves the existence of segre- 
gated and open pockets of ballast at other points. This is essily to be understood when continu- 
ous deposition in one part of overwet concretes is observed in field woric with runoff of lighter 
materials as the mound grows. 

In all concrete work subject to water action in any degree, laitance planes may be sub- 
stantially avoided by filling forms without permitting set of one portion before the portion next 
above is deposited. Running off ''soupy" portions from the top of forms while the mass is 
fluid is a palliative measure that may be used, or removal of laitance after setting may be at- 
tempted. No measures yet devised are wholly adequate, in that none basically remove the 
cause of complaint. 

46. Effect of Temperature and Atmospheric Effects on Water-tightness. — ^Temperature 
effects in finished structures, and those due to atmospheric changes, may result in opening 
passageways capable of conveying water. The coefficient of expansion of concrete is approxi- 
mately the same as that of steel (see Art. 32, Sect. 5). Adopting a linear unit, the movement 
at any point may be found by multipl3ring the distance expressed in terms of this unit of this 
point from a fixed point by the degrees change of temperature experienced or anticipated. In 
massive masonry, the interior experiences little thermal change. Actual dimensional changes 
in such structures are probably compensated for by internal flow.^ 

Variations in moisture content, even after prolonged set, affect the volume of concrete 
from 0.05 to 0.08% in the usual atmospheric range. This may be outwardly evidenced by 
cracks, but is more generally taken up in internal stresses in the concrete and reinforcement. 
Such changes are to be anticipated from our knowledge as to the colloid content of concretes 
and the absorptive and dessicative properties of such materials. 

If consideration of anticipated temperature and moisture changes indicates that a given 
structure will be liable to cracking by stresses thus induced, expansion joints must be provided. 
Their use is preferable to their omission in most cases, but they must be most carefully formed. 
Copper or lead flashing, or asphalt or elaterite mastic and fabrics have been found efficacious 
when properly applied, but all precautions must be observed as to obtaining density in the sur- 
rounding concrete. With mastic compounds, dryness of the adjacent concrete must be secured 
in order to obtain proper bond, else leakage at the joint will occur. 

47. Integral Waterproofing Compounds. — ^The foregoing paragraphs have given an in- 
sight into the causes of porous or leaking concretes. Where penetration and flow of water occur, 
it has been pointed out that an actual, physical passage or passages exist, that these passage- 
wajTB may be: (1) pinholes, or pores, resulting from occluded air; (2) water voids, or spaces left 
by excess water; (3) shrinkage cracks radiating from a pore, or hole of greater or less sise, with 
joining of a myriad of like cracks into complicated systems of cracks in concretes that dry too 
rapidly; (4) segregation due largely to excess water, with open pockets of stone; (5) laitance, in 
pockets or strata, due largely to excess water; and (6) temperature or other cracks, due to at- 
mospheric changes. 

t F. R. McMillan: BuU. Univenity of MinsetotA. 


It is to be expected that the customary violations of the natural laws governing concrete 
which, bring such passageways into existence should cause wide demand for something purchas- 
able which would afford relief from consequences. If any agent exists or is to be found, which, 
when added to concrete made with lack of care, is capable either of preventing or of closing the 
passageways through which water penetration or transmission is brought about, without detri- 
ment to the strength or other properties of the concrete, it may be ranked as a noteworthy dis^ 
covery. There is some question, however, as to whether any substance, particularly in econom- 
ical percentages, will accomplish this end to any save a minor degree. Consideration of the 
open void volimies and segregated areas in many concrete structures reflect the magnitude of 
the task. It is probable that proper practices and materials and they alone are adequate as 
well as unsurpassed in securing water-tight and enduring concretes. 

47a. Integral Waterproofing Classification. — Integral waterproofings now on 
the market may be grouped under four heads : 

(a) Special materials added to the mixing water. 

(6) Special materials added dry to the cement at the job. 

(c) Cement to which has been added the special materials during manufacture. 

(d) Special materials and cement applied as a plaster, this being intended to so bond with 
the concrete surface as to become integral with it. 

The special materials employed in the foregoing are substantially as follows : 
(a) Various forms of metallic salts, such as chloride of lime; oil emulsions; lime soaps, sus- 
pended in water; and like compounds. The actions of oil emulsions is to form soaps in combi- 
nation with the lime of cement; that of soap solutions as lubricants and formers of insoluble fillers 
by reaction with cement. Lime chloride has a catalytic action difficult properly to define, but 
tending to hasten set rather than either to lubricate, or to form pore-filling compounds. 

(6) Dry powders of floury consistency, formed of metallic stearates, such as lime soap, 
often with alum and hydrated lime. Their properties are claimed to be void-filling and 

(c) like substances, or glycerides of limes, mixed with cement during manufacture. 

(d) The same as (c), used as a surface plaster. 

47&. Value of Integral Waterproofing Compounds. — There is no general authori- 
tative conclusion yet determined as to the value of integral compounds. Field testimony differs , 
probably according as the methods and materials of one use have, through inherent excellence 
or weakness, proven either adequate or insidequate to produce impervious concrete. The 
most extensive work that has thus far been done is published in Tech. Paper 3, by R. J. Wig and 
R. H. Bates of the U. S. Bureau of Standards. A majority of present commercial waterproof ers 
were tested in the coiirse of the work therein detailed, but a subsequent series requested by 
several manufacturers of tested compounds, with concretes to be made under commercial 
conditions, has not yet matured. 

The conclusion of the foregoing tests is that no additive, proprietary, or open, will of itself 
overcome initial, serious deficiencies of material, or admit of defective practices; and no addi- 
tive so far known is superior in results to an excess of cement and the use of graded sand of 
proper quality with a little water as circumstances permit. 

47c. Rendering Defective Structures Impervious. — It is often necessary to ren- 
der an existing structure as nearly waterproof as possible. The end to be attained is, of course, 
the closing of all water passageways. The proper method to use is dependent upon the size, 
character, and origin of the pores or passageways in the concrete. If the pores are very small, 
some inert filler such as clay or silt may be sufficient; or a soap and alum mixture, such as that 
employed in the Sylvester process, may be applied. If the pores are of slightly larger size, paraf- 
fine or a paraffine-carrying oil, or bitumen, or an asphaltic oil may be successfully used. Paraf- 
fine may be applied either hot or cold. If applied cold, it is dissolved in a volatile carrier in 
saturated solution. Applied to the surface of the concrete, it penetrates to a greater or less 
depth according to the dryness and porosity of the concrete. Within a short time the volatile 


carrier is evaporated, leaving the paraflinc in the holes. Paraflinc may also be applied in a 
molten condition and, to render successful its use, the concrete must first be rendered sufBciently 
warm by artificial heat so that the melted parafline may be thoroughly rubbed in. Hot parafiine 
treatment is one of the most durable of waterproofing methods for work exposed to weather, but 
it requires considerable experience to secure a successful result. 

Bitumens of one grade or another are applied either in solution or hot, as in the parafiine 
surface treatment. They also may be incorporated in paints which are applied to the surface- 
Any bitumens employed must possess a high degree of elasticity and durability and must 
have considerable bonding ability with the concrete. To this end all concrete surfaces to which 
bitumens are applied should be thoroughly dried and preferably should be warm at the time of 
applications. Material should be well rubbed into comers and recesses; and the waferproofing 
film should be continuous throughout. 

Id applying bituminous paints and solutions it is a prerequisite to success that the coating 
shall be applied on that side of the concrete gainst which the water pressure is exerted. If this 
is done, the materials will be carried into the water passageways, but if this iB n^lected the 
materials will be forced out so that their application is waste. 

18. Waterproofing by Cement Grouting. — Neat cement grout has often been tried as a 
waterproofing coating, applied either as a surface plaster or as a surface wash. It has also been 
used as a crack filler, but inasmuch as it is virtually impossible to make a coating or filling of this 
kind adhere to set concrete, its use is rarely, if ever, attended with success. In difficult situa- 
tions attempts have been made to use cement grout under pressure as a waterproofer, but the 
instances on record where this has been successfully done do not indicate generally satisfactory 
results. It seems to be requisite that any waterproofing mixture shall be more or less plastic 
and viscous and that it shall be so applied as to deform and closely fill passageways in the con- 
crete under pressure of the water. 

49. HembranouB Waterproofinga. — Membranous waterproofing is an elastic, continuous 
sheet or membrane completely covering or surrounding a structure to be waterproofed (see Pig- 

T" .... J 

■ ., i iji i ?i\50M/pi:j::.... . 

18). Tlus membrane is laid in several overlapping layers (Fig. 19), impregnated (uod fastened 
down with some bituminous compound. The membranous system of waterproofing is adapted 
principally to the waterproofing of structures in course of erection, such as subways, tunnels, 
building foundations, retaining walls, arches, reservoirs, etc. 

Tlie bituminous materials employed as sealing compounds in the membrane method of 
waterproofing are: (a) coal-tar pitch applied hot; i_b) asphalts applied hot; (c) asphalt mastic 
applied hot; (d) especially prepared asphaltic compounds sold under various trade names. 

He membranes to be used with the above sealing compounds are: (a) tarred felt; (b) 
asphalted felt; (c) burlap; (if) burlap saturated with asphalt or tar; («) combinations of canvas 
and felt, or canvas and burlap, or felt and buriap. 

4Sa. J^pUcation of Hembranous Waterproofing. — Success of membranous 
waterproofing depends largely upon the care with which the materials are applied, tt is neces- 
sary first to prepare the concrete surface. It must neither be too rough, nor too wet, nor cov- 
ered with dirt or foreign substance; and it must not possess a glaie due to richness of cement 
surface. It is, therefore, necessary: (1) that all dirt and foreign matter shall be removed before 

Sec. 2-49&] 



waterproofing is applied; (2) that when it is anilied the oanrrete gIulD be rendered dry, either by 
drainage and evaporation or by the apphcation of artificial heat; (3) that the concrete shall be 
thoroughly set (as is indicated in the requirement for diyness) ; (4) that any glased surfaces «h%11 
be picked or rubbed down in order that the mateiials may adhere; (5) that fonn ties or other pro- 
jections that might puncture the waterproofing shall be remoTed; and (6) that any metal sur- 
faces encountered shall be dry, clean, and free from rust or dirt. 

496. Continuity of Membrane. — Lack of continuity may be fatal to the success 
of any waterproofing membrane. The waterproofing sheet must, therefore, be applied con- 
tinuously over the whole surface to be treated, footings and foundations included. All joints 
in the membrane must be broken at least 4 in. on cross joints and 12 in. on longitudinal; and 
at least 12 in. of lap must be left at comers to form good junctions with adjoining secticms. 
Where it is necessary to stop work, a lap kA at least 12 in. shall be provided for joining on new 
work. Each layer of bituminous or other material must com- 
pletely cover the surface on which it is spread, without cracks or 
blow-holes; and the fabric must be rolled out smoothly and pressed 
over the cementing material so as to insure its sticking 
thoroughly and evenly over the entire surface. 

49c. Protection of Wateiproofing. — ^After the 
waterproofing has been put in place it must be properiy pro- 
tected from injury. Such injury may occur when backfilling 
with earth; when depositing concrete against the waterproofing 
(see Fig. 20) ; when laying brick or rubble, or from careless piling 
of materiab on the completed waterproofing work. Injury from 
workmen's shoes is not infrequent. It should be remembered 
that a completed membranous waterproofing is usuaUy soft and 
liable to injury and the chances of so doing should be reduced to a 
minimum. A single point of entry for water, particularly if inaccessible when the work is 
completed, may render ineffective all precautions against leakage. 

The following table gives the numbers of ply of waterproofing required with various heads 
of water: 

60. Rules for Making Concrete Imper- 
vious. — (a) To make concrete that shall be 
impervious, the rules basically governing the 
making of dense concrete apply with special 
force. These are: 

1. Use proper tfuUerials, i.e., clean and 
preferably graded sand; hard, durable stone; 
cement that conforms to standard; and clean 

2. Use proper proportions of proper man 
terials, t,e., avoid arbitrary proportions; use 
careful measurement for each batch; test 
each shipment of sand for uniformity of 
grading; if variation is found, properly com- 
pensate by variation of proportions. 

3. Properly and adequately mix the ma- 
terials ^ i.e.f not only stir together the several 
ingredients, but prolong the operation suffi- 

iProm "Modem Method of Waterproofing/' c^ently to secure the needful consistency and 
M. H. l^wis. distnbution, particularly of the cement. 

Fio. 20. 

Number or Plt of Watebpboofing Required 
FOR Varying Heads of Water' 

Head of 




Coal tar 

and felt 

felta and 





in inches 









































4. Vm a minim%tm of vxiter thftt wiU permit adequate fiUing of fonns and contact witb 

G. Plact earefuiiy informs to avoid segregation or unequal distribution . 

6. Expd at much oa ■pottU:^ of occluded air by puddling, or vibrating, or jarring of forms a.^ 
filling proceeds. 

7. Fill forms amlinuoutly to lop, TTeferoUyover/Iauitnjr, to avoid stoppage, or laitance planes. 

8. Properly protect amcrele agaiiiat rapid evaporation and against unusual heal or cold during 
the setting, hardening, and curing periods to avoid shrinkage, frost, or other crackings and 

B. Remove visible segregations as soon as discovered, replacing with good concretA, well 
rammed into place. Do not rely on surface plastering of defects. Such attempts are unwork- 
man-like and ineffective. 

10. Construct expa7\sion or contraction joints with extreme care. Do not rely for water- 
tightness on any supposed bond between abutting concrete sections. 

(6) Integral tealerproofings cannot be relied upon to avert the consequences of improper 

(c) Membranous ipaierproofings are of service in closing water passageways in existing de- 
fective structures when, and only when, carefully applied. 

U. Character of Surface Finish Desired. — The character of finish desirable to produce on 
concrete surfaces is determined by the ends to be served. The majority of requirements for 
special finishes are architectural, varying according to the character of the structure and with 
the location of the surface. 

S3. Removing Form Harks. — For all concrete work exposed to view, forms should be 
exceedingly well constructed, producing plane surfaces and straight, sharp lines and true angles 
in the finished concrete. In work of this 
character, extra care is usually taken to ob- 
tain even-textured, dense surfaces by using 
a mixture of proper consistency and by 
careful spading and puddling. Such sur- 
faces necessarily reproduce all defects of the 
mold, 90 that after-treatment is necessary to 
remove the form marks, as well as to relieve 
the "dead" color due to excess of cement at 
the surface, with oftentimes efBorescenee, 
or other whitish deposits, indirectly occa- 
sioned thereby. 

It is elementary optics that blemishes 
arc least visible on a non-uniform light- 
diffracting "matt" or stippled surface. 
Form marks, therefore, are concealed by 
producmg such a surface. Hiis may be 
done by tooling, by rubbing, by brushing, 
f™. 3i.-Boi«y eoncm. .urf«« b«o« u«d. Kou, ■>■■ ■>> sand-blasting, and such treatmenU 
coBtrut between Gniabed and unfiniabed iuKkc. have a further advantage of modifying the 

dead color above referred to by exposing a 
multitude of sand grains to light, so that by reflection from their facets, the gray-green of 
cement is relieved and brightened. 

Sia. Tooling. — Tooling concrete surfaces is more or lees costly, depending upon 
the length of time the concrete has set and upon its hardness. Bush4iammering, erandalling, 
->r axing may be done by hand, or a pneumatic ur electric tool may be em|doyed at considerable 


advanta^. One typeof surfacing tool permitting of tooling or of KTindiog and the method of its 
use is shown in Fig. 21.' Its capacity ia rated st 60 to 70sq. ft. of surface per hour on concrete 
from 1 to 21 days old. Pig. 22 shows a hand bush-hammered surface of colored aggregates, and 
Fig. 23 a picked surface. 

Only small-ained aggregate should be used in facing material to be tooled, as it is hard to 
dress and to obtain uniform results on surfaces where large angular stones are eacountered. 
Tbe concrete should be thoroughly hardened before work is commenced, especially if sharp 
clean surfaces are desired. The concrete should preferably be about 2 months old, although if 
it is allowed to stand too long the labor involved will be unnecessarily great. 

A variety of surface effects may be obtained by tooling, as the effect produced in any given 
case depends upon the kind of tool used. Some variation ia the appearance of the finished sur- 
face may also be obtained by the manner in which the tool is handled. By striking a perpen- 
dicular blow no lines or marks are left in the surface, whereas with a glancing blow, tooth marks 
are left which can be made parallel to each other or at various angles. Tooling cannot ordinarily 

Fio. 22. Pia. 23. 

■h obtained bj' hand buah-bammerini. The contrast of ahadea <■ piodimd by using dif' 
The ooncreta in the dsrk portion was made with red ■andatone. while the light por- 

Fio. 23 — Hcked eurface. 

be pwformed satisfactorily on gravel concrete, as the pebbles will be dislodged before being 

fiS6. Rubbing. — If a nibbed surface finish is desired, the coarse aggregate should 
I>e well spaded back from the face of the work and the forms should be removed before the 
concrete has set bard, preferably in a day or two after the concrete is poured. It is necessary 
with green concrete to use care in removing forms to avoid spalling, as it is very difficult, if not 
iropCssible, to adequately repair such spalls by patching with mortar. It is necessary, also, 
to remove form wires or other projections before rubbing and to point with mortar any pockets 
or open places in the surface. The process of rubbing consists in grinding down the surface 
of the concrete sufficiently to remove all impressions of the timber or other irregularities, using 
a brick of carborundum, emery, concrete, or soft natural stone. In connection with the rubbing 
(which is accomplished with a circular motion), a thin grout composed of cement and sand 
should be applied to the surface, well rubbed in, and the work afterward washed down with 
clean water. The grout is used simply to fill surface imperfections and care must be taken not 
to allow it to remain as a film on the surface. 

This method of treatment produces a comparatively smooth surface of uniform color much 
superior to that obtained by the all too prevalent method of painting with a grout, which almost 
invariably erases, cracks, and peels off. Rubbing is a very acceptable, cheap way of finishing 
concrete surfaces. 

< Tbe Beig Electm llotary Surlaccr, Etevatot 8uppiiea Co.. \, Y. 


4. U«e a minimum of water that will permit adequate filling of forms and contact with 

5. Place earefuUy in forma to avoid segregation or unequal diatribution. 

6. Expel at much as poesibie of occluded air by puddling, or vibrating, or jarring of forms a^ 
filling proceeds. 

7. Fill forme continuougly to top, pre/'ernUi/ouerflowinff, to avoid stoppage, or lajtance planes. 

8. Properly protect ctmcrele against rapid evaporation and against unusual heal or cold durirtg 
the setting, hardening, and curitig periode to avoid shrinkage, frost, or other crackingB and 

9. Remove visible segregationn as soon as discovered, replacing with good concrete, well 
rammed into place. Do not rely on surface plastering of defects. Such attempts are unwork- 
man-like and ineffective. 

10. Constrvct expansion or contraction joints ivith extreme care. Do not rely for water- 
tightness on any supposed bond between abutting concrete sections. 

(6) Integral teaterproofinga cannot be relied upon to avert the consequences of improper 

(c) Membratume waierproofingB are of service in closing water passageways in existing de- 
fective structures when, and only when, carefully applied. 

01. Cboracter of Surface Finish Desired. — The character of finish desirable to produce on 
concrete surfaces is determined by the ends to be served. The majority of requirements for 
special finishes are architectural, varying according to the character of the structure and with 
the location of the surface. 

03. Removing Form Uarks. — For all concrete work exposed to view, forms should be 
exceedingly well constructed, producing plane surfaces and straight, sharp lines and true angle:' 
in the finished concrete. In work of ibis 
character, extra care is usually taken to ob- 
tain even-textured, dense surfaces by using 
a mixture of proper consistency and by 
careful spading and puddling. Such sur- 
faces necessarily reproduce all defects of the 
mold, so that after-treatment is necessary to 
remove the form marks, as well as to relieve 
the "dead" color due toexcess of cement at 
the surface, with oftentimes efflorescence, 
or other whitish deposits, indirectly occa- 
sioned thereby. 

It is elementary optics that blemishes 
are least visible on a non-uniform light- 
diffracting " matt " or stippled surface . 
Form marks, therefore, are concealed by 
producing such a surface. This may be 
done by tooling, by rubbing, by brushing. 
^^^ or by sand-blasting, and such treatmenU 
face, have a further advantage of modifying the 

dead color above referred to by exposing a 
multitude of sand grains to light, so that by reflection from their facets, the gray-green of 
cement is relieved and brightened. 

BSo. Tooling. — Tooting concrete surfaces is more or less costly, depending upon 
the length of time the concrete has set and upon its hardness. Bueh-hanunering, crandalling, 
or axing may be done by hand, or a pneumatic or electric tool may be employed at considerable 


advtmtage. One type of surfacing tool permitting of tooling or of gnnding and the method of its 
use is shown in Hg. 21.' Its capacity is rated at 60 to TOaq. ft. of surface per hour on concrete 
from 1 to 21 days old. Fig. 22 shows a hand bush-hamroered surface of colored aggr^ales, and 
Fig. 23 a picked surface. 

Only small-eized a^regate should be used in facing material to be tooled, as it is hard to 
dress and to obtain uniform results on aurfaces where large angular stones are encountered. 
The concrete should be thoroughly hardened before work is commeni^ed, especially if sharp 
clean surfaces are desired. The concrete should preferably be about 2 months old, although if 
it is allowed to stand too long the labor involved will be unnecessarily great. 

A variety of surface effects may be obtained by tooling, as the effect produced in any given 
case depends upon the kind of tool used. Some variation in the appearance of the finished sur- 
face may also be obtained by the manner in which the tool is handled. By striking a perpen- 
dicular blow no lines or marks are left in the surface, whereas with a glancing blow, tooth marks 
are left which can be made parallel to each other or at various angles. Tooling cannot ordinarily 

Fio. 22. Fia. 23. 

Fia. 22. — Surfua finiah obUined by bBOd buah-bumineriRiE, The contrut of ihadH ia produced by uaiii) dif- 
tennt colored ■sfregsln. The concrete in the du-k portioD wu mada witb red nDdstoDe. while Ihe light por- 

Fio. 23.— Picked eivfio.' 

be performed satisfactorily on gravel concrete, as the pebbles will be dislodged before being 

B26. Rubbing. — If a rubbed surface finish is desired, the coarse aggregate should 
be well spaded back from the face of the work and the forms should be removed before the 
concrete has set bard, preferably in a day or two after the concrete b poured. It is necessary 
with green concrete to use care in removing forms to avoid spalling, as it is very difficult, if not 
impossible, to adequately repair such spalls by patching with mortar. It is necessary, also, 
to remove form wires or other projections before rubbing and to point with mortar any pockets 
or open places in the surface. The process of rubbing consists in grinding down the surface 
of the concrete sufficiently to remove all Impreasiona of the timber or other irregularities, using 
a bnck of carborundum, emery, concrete, or soft natural stone. In connection with the rubbing 
(which is accomplished with a circular motion), a thin grout composed of cement and sand 
should be applied to the surface, well rubbed in, and the work afterward washed down with 
clean water. The grout is used simply to fill surface imperfections and care must be taken not 
to allow it to remain as a film on the surface. 

This method of treatment produces a comparatively smooth surface of uniform color much 
superior to that obtained by the all too prevalent method of painting with a grout, which almost 
invariably crates, cracks, and peels off. Rubbmg is a very acceptable, cheap way of finishing 
concrete surfaces. 

I The Beff F.lectrit- liolary iSurlacer, Elevilor Supplits Co., N. Y. 


joints, also, should be made tight enough to prevent any material leakage of the liquid mass, as 
such leakage will mar the appearance of the finished work. 

Forms should have sufficient strength to properly support the loads which they are called 
upon to carry. Horizontal members, such as floor sheathing and supporting joists, should be 
able to support the weight of the concrete and the construction load. Vertical members, such 
as wall sheathing and supporting studs, should be designed to resist the hydrostatic pressure 
of wet concrete which is about 145 lb. per sq. ft., for each vertical foot of height. 

60. Bconomical Considerations. — The cost of forms constitutes a large item of expense in 
the building of reinforced-concrete structures and this cost vanes to a considerable extent with 
the kind of form construction adopted. Formwork, of course, should in every case leave the 
finished concrete true to line and surface, but even with this accomplished a great deal can be 
done in so designing forms and in so planning the detailed methods of their construction that 
erection and removal will be greatly facilitated without undue waste of lumber. As a general 
rule, the most important consideration is that of ease in form removal as great economy may be 
effected by using form units over and over again with a minimum of repairs. 

Simplicity and symmetry in formwork should always be given consideration. In buildings, 
uniform story heights shoidd be selected whenever possible in order to prevent continual re- 
making of column forms, and frequent changes in column sizes should be avoided (at least in 
the case of light-floor construction), not only on account of the column forms themselves, but 
on account of the beam and girder forms or slab forms which frame into them. Also, where it 
is feasible, beam sizes should be so chosen that local standard widths of lumber may be employed 
without splitting; and, at the same time, the sizes and spacing of the beams should be made so 
uniform that the contractor may use his forms repeatedly, thus greatly reducing the expense for 
lumber and eliminating the cost of making new forms. In many cases it will be found that a 
slight excess of concrete will save many times its cost in carpenter work and lumber. 

61. Lumber for Forms. — ^The kind of lumber to use for forms depends upon the character 
of the work and the available supply in the local yards. Spruce seems to be the best all-round 
material. It can readily be obtained in almost any locality and is undoubtedly an excellent 
lumber to use for joists, studs, and posts. For sheathing, however, white pine is better than 
spruce by reason of its smoothness and its resistance to warping, but this wood is generally too 
expensive (except for cornice and ornamental work), and spruce makes a good substitute. 
If white pine is to be used for sheathing, the fact should not be overlooked that this kind of 
lumber has little durability on account of its extreme softness and would not give good results 
if the forms were to be used many times. 

Aside from spruce or white pine, Norway pine and southern pine are generally the most 
available and give satisfaction. Hemlock is not usually desirable, especially for that part of 
form work which comes into contact with the concrete, but it is sometimes used for ledgers, 
studs, and posts. This wood is too coarse-grained to be suitable for sheathing and is liable to 
curl when exposed to the weather or to wet concrete. 

It is safe to say that lumber which is only partially seasoned should be the kind employed 
in form construction. Kiln-dried has a tendency to swell when soaked by the concrete, and 
this swelling causes bulging and distortion of the forms. Green lumber, on the other hand, 
dries out and shrinks if allowed to stand too long before the concrete is placed; fortunately, 
though, this tendency of the green lumber to check and warp may be prevented to some extent 
by keeping the boards thoroughly saturated with water. When using natural, well-seasoned 
lumber care should be taken not to drive the work up too close, since forms should always be 
left in a position to experience some slight swelling without any undesirable results. 

Sheathing liunber should be dressed at least on one side and both edges, even for non-ex- 
posed surfaces^ as the removal and cleaning of the forms are greatly facilitated thereby. In face 
work, where a smooth and true surface is quite important, the lumber employed should be 
dressed on all four sides. Lumber which is dressed in this manner b easy to work up and place, 
and this fact alone usually more than offsets the cost of dressing. 


advantage. One type of aurfacing tool permitting of tooling or of grinding and the method of its 
use ia shown in Hg. 21.' Its capacity ib rated at 60 to 70aq.ft. of surface per hour on concrete 
from 1 to 21 days dd. Fig. 22 ahowB a band buah-hammered surface of colored aggregates, and 
Fig. 23 a picked surface. 

Only amall-fliied aggregate should be used in facing material to be tooled, as it is hard to 
dress and to obtain unifonn results on surfaces where large angular stones are encountered. 
The concrete should be thoroughly hardened hefore work ia commenced, especially if sharp 
clean surfaces are desired. The concrete should preferably be about 2 months old, although if 
it is allowed to stand too long the labor involved will be unnecessarily great. 

A variety of surface effects may be obtained by tooling, as the effect produced in any given 
case depends upon the kind of tool used. Some variation in the appearance of the finished sur- 
face may also be obtained by the manner in which the tool b handled. By striking a perpen- 
dicular blow no lines or marks are left in the surface, whereas with a glancing blow, tooth marks 
aie left which can be made parallel to eachother or at various angles. Tooling cannot ordinarily 

Fio. 22. Fio. 23. 

Fto. 22. — Surfue finish obtained by hand biuh-Iiiimnieiiiig. Thicontrut otghsdielg produced by UBlngdif' 
(cKot colored ivfreEmteg. The oancrete in the dark portion wu niuls iiitb red undatone. while the light por- 

Fio. 23.— Picked ■urlace. 

lie performed satisfactorily on gravel concrete, as the pebbles will be dislodged before being 

S36. Rubbing. — If a rubbed surface finish is desired, the coarse aggregate should 
lie well spaded back from the face of the work and the forms should be removed before the 
concrete has set hard, preferably in a day or two after the concrete is poured. It is necessary 
with green concrete to use care in removing forms to avoid spalling, as it is very difficult, if not 
impossible, to adequately repair such spalls by patching with mortar. It is necessary, also, 
to remove form wires or other projections before rubbing and to point with mortar any pockets 
or open places in the surface. The process of rubbing consists in grinding down the surface 
of the concrete sufEciently to remove all impressions of the timber or other irregularities, using 
a brick of carborundum, emery, concrete, or soft natural stone. In connection with the rubbing 
(which is accomplished with a circular motion), a thin grout composed of cement and sand 
should be applied to the surface, well rubbed in, and the work afterward washed down with 
clean water. The grout is used simply to fill surface imperfections and care must be taken not 
to allow it to remain as a film on the surface. 

This method of treatment produces a comparatively smooth surface of uniform color much 
superior to that obtained by the ail too prevalent method of painting with a grout, which almost 
invariably erases, cracks, and peels off. Rubbing is a very acceptable, cheap way of finishing 
concrete surfaces. 

> The Berg Eleclric Itutary ^urlacet. Elevaior Supplier Co.. N. V. 



[Sec 2-63 

using them. The following rules are recommended by the Illinois Department of Factory 
Investigation :^ 

TiMB Required Before Removinq Forms 

Above W¥. 

60° to 


AOP to 50<»F. 


1.688 than 4ff*F, 


Within 3 days 
Within 4 days 

Within 4 days* 

Within 14 days' 

5 days 

6 days 

8 days 
18 days 

Not less than 10 days 
Not less than 10 days 

Not less than 14 days 

Not less than 14 days 

Not until tests 
have been 
made indicat- 
ing that the 
concrete is set. 

Side forms for girders 

and beams. 
Bottom forms of slabs (6 

ft. or less span). 
Bottom forms of beams 
and girders (less than 

14-ft. span). 

Form work in buildings should be so designed that the column forms may be removed 
without in any way disturbing the supports of the beams and girders. This practice bares the 
concrete of the column to the hardening action of the air and permits a defect to be detected 
and remedied before any load is brought to bear upon the column. The beam and girder sides 
are next taken down, but this is not done until the slabs are strong enough to stand up without 
support. Sometimes, however, as a matter of safety, the slab supports (namely, the sheathing 
and joists) are replaced with temporary uprights bearing against a plank on the underside of the 
slab. The beam and girder bottoms and the posts supporting them are left in place longer 
than any of the other formwork, and should not be taken down until there is absolutely no 
doubt as to the strength of the members supported. Walls are usually built separately from 
the other parts of the building, and a wall form may be removed whenever the concrete in the 
wall is hard enough to bear its own weight. 

83. Number of Sets of Forms in Building Work. — The number of sets of forms which are 
necessary in building work depends upon the weather, the variation in the shape of the members 
from floor to floor, and the floor area. Under reasonably good summer conditions 1}^ sets of 
forms (forms for l\i stories) will serve the purpose and allow the work to progress at the usual 
rate of a story in a week or 10 days. If the floor framing, however, is particularly complicated, 
one set of forms may be the more economical if the speed of construction is not the leading con- 
sideration. In such a case, of course, extra lumber must be used for beam and girder bottoms 
and for supports which must be left in place. Where the floor area is large and the construc- 
tion is fairly uniform throughout, even less than one set of forms with additional beam bottoms 
and posts may be suflicient. 

64. Examples of Form Design. 

Ma. Column Forms. — ^Fig. 25 shows a typical form for a rectangular column. 
Two sides of the column are held together, as shown, by bolts, and the two opposite sides by 
hardwood wedges between the bolt and the form. The sheathing or lagging boards run the 
entire length of the colunm and are made up into panel units by means of the cleats, which serve 
also as clamps. A somewhat similar colunm form is shown in Fig. 38, page 102. 

A common type of rectangular column form where wooden wedges are used to do all the 
clamping is illustrated in Fig. 26. The tightening is made possible by the use of stop blocks on 
each clamping piece. When a reduction is made in the size of column, these blocks must either 
be ripped off or additional blocks nailed on. The boards comprising each side of the column 
form are usually battened or cleated together so as to form a panel unit. This practice allows 
the clamps to be put on separately and thus permits stop blocks on the clamps to be easily 

^ Sm mho rolea giren in "Conerete, Plftin and Reinforoed/* by Tatlor and Tbomfson (1016 edition). 
* Add 1 day for each additional foot of span. 


changed when the column form is remade. It is a common prautice Ut muke up the sides of 
column fonns with narrow Btripe of sheathing in order to facilitate the reduction in sise of columns 
from floor to floor. A method of reduction sometimes employed by the Aberthaw Construction 


Co. is ^own in Rg. 27. The column form represented here is also illustrated in Plate III, 
page 114. The use of narrow sheathing strips applies especially to interior column forms since 
exterior or wall columns are usually of the same width from basement to roof and change but 
little in thickness. 

Fig. 28 and I^te II, page 113, show a method of bracing column forms by means of 4 
by 4-m. "whalers" or poelA. The use of Henderson clips should be noted. Sometimes the 
boiting ia done directly to the whalers. 




.^'iayginff plania fyar *lM 

■Firvt Raduction 5«coni] RMucttan 

ekaf cut dsmi ana liS' *ik> and lil'miM, 

art m at ttnin t^vnt tthind cintm 



Other types of rectangular column formi in common use &re shown in Figs. 29, 30, and 31. 
The cleats which fonn part of the clamps are nailed to the sides when making. The clamp 
shown in Fig. 31 is patented and known as the "New England Column Clamp," controlled 
by the N. E. Column Clamp Co., Boston, Mass. It consists of angle irons punched at frequent 
intcrvala with bolU for holding the form together wherever needed. Angles and bolts may be 
liought in the open market. The New England Column Clamp Co. sella only the rights to use. 

Fig. 32 Qlustrates still another tjrpe of rectangular column form. The use of horizontal 
sheathing should be noted. The sheathing is nailed to the vertical pieces before erection. 

A patented steel clamp, known as the "Gemco Column Clamp" and manufactured by the 
Gcmco Mfg. Co., Milwaukee, Wis., is shown in Fig. 33. These clamps are described by the 
manufacturers as follows: 

A Gcmco Squm-eoloiBD CUmp ii oompoKd of (our, stnicht, int«Khui«*kbIe, ateil bui, eub 2 in. wide 
^r ^* in. thick. One eod of OAch lur u proTided witb m hibrdcned toothed iockia^ dog pivot«d between two pLaUe 
wbicb arc firmly riveted to the oppoeite lidee of the bmr. These plmle* fully protect the loclcinE dof from dftmece. 
t>ut sufficient spAce ii allowed to permit the ban to elide frcdr whcu the locltinf dog ift not eimAced. 

The cUmp ii Kt by piwure on the tichunioc levet whish alido over the tree cod of the ben. u ihown in 
Flf. 33. When the bui u* dnwo to position, the loekiui dog i* Kt by preasun at the Gaccn or licht tap of a 
Hammer and it poaitively loelu the clamp. 'Hcbtcmnc lever ia detachable and can be uattlcin any Gemco Stjuarc- 

A Gemco Clanp can be aiiembled, tifbtened and set in 1 min. It automatically aquares a column sod 
^an be ti^hteasd from all ^dea so ■■ to paaitively dose all cracks, thua nrtsinint all th« Tsluablc part of the miiturc 
vbich vnuld othsrwiae Sow out with the water. Gemco Clamps will square a rectangular pi»r or column in th^ 

These clamps can be removed in the same time which it takta lo adjust them. The ti(btenin( levrr ■■ 
is«d to partially rvlievc the Btfaici when the loekioc doc can be diaencaced by tbe use at a claw hammer and a 
lail or punch inserted in tbe hole at the corner of the dof . After two diaconaliy opposite contcra are loosened liif 
Futire clamp ia Free. By working two men oa opposite sides ot the column and at diaconai corners of tbe clam|M. 
be work of seltini or wreckinc can be completed in a very short time. 

When knocked down these st[>i(hl. interehauieable ban ol ileel are easily transferred IrMn floor lu Out 

To avoid iiiiui»<Mii J expense by usinf large damps on amaU ecdumna the Gemco Bquaie-calumu riarnfM 
Lre manufactured in three stock sises- Special aiaee made only on specific order. 


Stock No 

Column »K, 






C-13. 'HcbteDiDi lever 

Another patented type of all-eteel column cUmp which requires ao wood framing is shovrn 
in Fig. 34 — manufactured by the K. & W. Clamp Co., Minneapolis, Minn. 

All square or rectangular-column forms should have bevel strips in the comers. Sharp 
comers in concrete work should always be avoided, not only because it is difficult to tamp 
concrete bo as to obtain perfect comers, but because the edges are likely to be chipped oS when 
the forms are removed or while the concrete is still green. The bevel strips are usually nailed 
to two opposite panel umts when the forms are being made up. 

Octagonal and round-column forms with wooden clamps are not susceptible of ready 
arrangement into units, and are clumsy and quite expensive to make. The problem seems to bo 
solved in present-day practice by the use of either iron bands or chains. Fig. 35 shows a 
form clamp manufactured by the Sterling mteelbarrow Co., Milwaukee, Wis., which can be 
used on any type of column form. The band-iron placed around the form (view a) is passed 
through the clamping head, and drawn tight. The manufacturers describe tlie operation as 

Tba opat*tor srupa ifae band with one ha 
And r< l » * w the bmnd, drtviag it pflrfe«(ly tifht d* 
form. Whan he tmm thai the bund i> tighl encui 
erimpins tb* band in BUeb ■ way that aUppace ii in 

e lever (ever*] 

d the bud wu 

■nee lb« lever do*™ c]a« 

id loelu the clunp bj inacr 

Thii ratehMins crip* 

beiikc put arouod the 

the form, tfaui 


ODC o[ tha ofiwiiDai in ths lockini davics (visw r|. Wbsn kaockiai down ths (arm. il is D«««ury ulnply to ntswa 
th« levw FroQ] tb» lookinc deviae. return the clAmpiDj bead ia pwtioiit and pull tha baad out. 

Flo. 31.— K. k n 

A Gemeo Clamp tor round i-olumns (Fig. 36) is deacribcd by the manufacturers in the 
following m&nner: 


X Oamo'o Rouad-ealumD CUmp a m»de of h gtrip of IM-in-. ie-g*«e buxl iron, ona Bad of which curies i 
□ulleAble-troo cnatiria which holdfl the hardeaed tooth lockiDC dog. When in use the free e^d of the kwnd iron i 
•lipped under the lockinB doa uid th? duiip »a be quickly and eoaily drawn to and eet at any dtaired deare« o 
(ennon with a tiEhteniog lever. The lever i« detachable and can be used on any aise ol Gemeo Round-eolumi 

A clamp may be struck or wrecked ia lees lime than required [or Htting, and in this operation thedetachabl. 

Gemco Round Clamps are made in two slock eiiea— for 24-iii. and 30-10. columns. Special giiM Dadi 

Dck No. 




Weight, pounds 





C-Z2 Tiibteni 

,uver.;:: ::::::; 

Another patent«d clamp for circular columns is shown in I^lg. 37, manufactured by the 
Uoiveresl Form Clamp Co., Chicago, 111. These clamps are made in one siie only, i.e., size 

C-1 for H-in. rods, 6 in. long overall. A tightening wrench shown in Fig. 60, page 117, i; 
used with this clamp. 


Win And rod cUmps deacribed under the heading "Wall uid Pier Forms" (pagft 111) tt 
ttaed to some extent on forms for colunuu. 

Mb. Bemm tnd Girder Forma. — Fig. 
Joists support the slab sheathing and are carried in 

(8 is an isometric view of a typical floor. 
turn by ledgers or joist bearers (see S\ao 


Fia. 4a 

Plates I and IV, pages 112 and 115 respectively). The method of clamping the beam and 
prder forms is shown more in detail in Figs. 39 and 40. Separate blocks are sometimes used 
instead of a joist beater. Plate II, page 113, shows a method of 
supporting slab forms by notching out the beam cleats to receive 
the joists. 

A number of different methods are employed to clamp beam 
and girder forms and prevent them from spreading due U> the 
pressure of wet concrete. The method referred to above is per- 
haps the most common, but bolts either above or below the beam 
bottom are sometimes employed. Clamping by means of ribbands 
is well illustrated in Plates I to III inclusive, pages 112 to 114. 

Typical girder-^orm construction is Bhown in Fig. 40. Some- Kiu. ii. 

tiroes the girder sides are not erected eomph te, the portion between 

the beams bang erected in separate sections, as shown in Plate I, page 112. The lower part erf 
the girder sides are made of thick plank in order to provide a suitable support for the beams. 

A common method of bracing the out«r aide of a wall beam ia shown in Plate I. *"-•»—■ 

r 1 • 








method is by means of a cantilever brace. Methods shown in Fig. 28 and Plate II, page 113, 
should also be noted. Fig. 41 shows a form of ordinary post or shore in common use. Haunch 
design may be seen in Plates I and IV, pages 112 and 115 respectively. 

6ic. Slab Forms. — Two types of slab forms need to be considered for ordinary 
construction where beams and girders are employed: (1) the panel type (the only type of form 
construction so far treated) ; and (2) slab forms made up into box shape, comprising the joists 
and the sides of the beams and girders. 

In the usual t3rpe of construction, designated above as type 1, the beam and girder forms 
are either erected complete and the poets set in place after the erection, or the beam and girder 
bottoms are spiked in place to the post caps, and the sides are erected as separate units. After 
this much is accomplished by either method, the joists are then put in position and the panel 



-lar^ Sf/x/fr 

^'Oafvaniztef Iron 

Fta. 42. 

units are placed on top of them without nailing. Notching out of the slab panels is usually 
necessary at the columns. When the forms are to be used but once, the slab sheathing is not 
made up in advance into panels, but is lightly nailed to the joists after they are in place. 

Fig. 42 shows one method of making up slab forms into box shape — ^the form construction 
being that used at the University of Wisconsin. Planks extend in two directions supporting 
the cells. The steel and continuous stirrups are set for demonstration only. 

A collapsible core box is used by at least one prominent builder (Fig. 43). A plank resting 
on cleats on the sides of the cores forms the bottom of the beam mold. The main girders are 
molded in similar spaces between the ends of the cores in one panel and those in the next panel. 
The molds are made in two equal parts with a hinged joint through the longitudinal center line 
of the upper surface, and are held open by transverse struts between the lower edges. When 
the forms are to be removed, the transverse struts are knocked out and the boxes are collapsed. 
The use of these molds presupposes a standardized layout. 

Fio. 43. 

Forms for flat-slab floors are much simpler than for the ordinary beam-and-girder construc- 
tion — with the exception, of course, of the forms for the flaring column heads. Typical de- 
signs of forms for flatnslab floors are shown in Fig. 44 and Plate V, page 116. Corrugated 
iron has been used to a considerable extent for sheathing where the ceiling surface does not 
need to be absolutely smooth. 

A new and novel system of form work (Fig. 45) has been devised by Jesse E. HodgeSf of 
Cincinnati, which is applicable to both beam-and-girder and to flat-slab constructions. 
Edward O. Keator k. Co. of Cincinnati, Ohio, are sole agents for this system. 

The slab forms consist of very light metal sheathing held on a matting (usually wooden 
striiM 2 in. by 2H in* by 4 ft. 6 in.) which is supported by stringers placed about 4 ft. on centers. 


Thit matting is fomiMl bj raanertiitg the sii>*D wooden atripn hy • fight netaJ duua nnv «*fit 
end •(>■• to aOownn opening o(>bixit3'i in. between tbeitrifw. TV mata an m«(fe of *linn 
leogtha so that tbey can be rolled into bondlea and not be taoheavyfnrnneortwnimen Uif^rry. 
VsoaDjr one ttringer b osed halfwar between beana, and lerlKen on the beam nrfes IttM tfk« 
end* of the mat. Tk atrag en do not bare to be efaanged in length for d i flerent jobs M lb*7 
eaa be aaed in long lengths and lapped. 

Tfea^Bfei&bi&iB»(I%. .M) anmadB&vn^^ Uuahee, wvA » 4 liy 440. pme anC eoo 
3 bf 4-fn. sde pieces, ami oie iidiuBCBhlK fmm 9 to 14 ft TIuB rangB is lotflinieiu; (bv aittioM^ 
work but [nngBT liinfws isw. nf nomme. hn hiiilc fiir ^^fMrial wnrfc. ia Knv'axuf nne of rJu>)H> ^okm, 
tbe (tlamfe ar« baMnad and the -t b? ( rfnvwn tiU M Ien({(^ itenisnaSat b;p it measumft pnls 
in tbc EBooI manna. Cbmpa an t^en tigfaCened and Che abore* raised. Ths tbsCunt of thi* 

rvpo [£ Aunt i> a. pwlios alamp tvhinti mob itaslf flrhen liie nam i* in dritfjnff pfMrion imt 'Mf 
-ttaum i» under \amL T!u» .tfaon alaa ba* a. riMMtfaoble bmnl whiHi m^ b« Inft' nn, in -wndMr 
-^iiBiimiers.lMun.iirremavBdin'iMnirttae.'AoiBsfiirSBC-^Ah tfnnm. With th" hmri rt^iifMil 
4ie fnnii Mnn{|Bs mC in a.fliiii madehy rbe l;woiipp«rpten«»<if rfop '4inr'^ tWii*'l1 iflAlM* ■' »>'- 
naaoh.«teafrtiBiiii«*:ift»i»'Hi»j|i»M*i»>« (P«" 


Wedges m uiuieceasaiy with the Hodges shore as leveling the centering for the floor msy be 
nccomplished with a kind of rackel>-]ack arrangement shown in Fig. 47. To adjust the lenglli 
of shore a steel bar which is notched to form a rack is attached to one f&e<- of the 4 by 4. A fork- 

ended lifting lever is then fitted over the timber, with its fulcrum bolt resting in one of the slots 
of the bar and the ends of the fork engaging the lower ends of the 2 by 4-in. sticks. With the 

cams kxMened, the upper portion of the shore is raised by bearing down on the lever twndlr, 
and when in proper position is locked by the cams. 

Another type of adjustable shore is shown in Ilg. 48, manufactured by the H. W. Roa 

See. a-64cl 



Co., Ckicinnati, Ohio, and known as the "Rooe Self-lock Adjustable Shore." It is made up 
for average use witha6-ft. piece of atandard pipe and two piecea of 2 by 4's dressed aide and edge, 
8 ft. long. Only the necessary castings are sold by the company above mentioned. The opera- 
tion of the shore consists in taking the extension member by the two legs and lifting it to the 

!k adjustable ihore. 

Fio. 46.— Gemco sdjustabli 

desired height. The slightest pressure on the two legs of the extension member toward the pipe 
causes the yokes to grip and the shore is ready for the load. The greater the load the greater 
tlie grip. For finer adjustment the shore can be raised or lowered with the jacking device 
which can be attached to any point under the two legs of the shore by turning the handled screw 

and the upper member can be aet at its final height with the adjusting lever. A turn of the 
thumb-screw lock sets the yokes and prevents any possible release through outside jarring. 
The jacking device is then removed for adjusting the next shore, 

"Gemco Adjustable Steel Shoring" is shown in Pig. 49. The shores are adjustable in 



[Sec. i~Md 

height from 9 ft. to 12 ft. 6 in. For adjuHting the height or raising the load, a detachable lever 
is used. These shores are manufactured by the Geroco Mfg. Co., Milwaukee, Wis. 

Here are many other types of wooden floor forma but those described above illustrate the 
main features in floor-form design. For steel floor forms see page 135. 

/^' ^ood ^orm 

Section A-B 

Sid. Column Heads. ^Thc tops of column forms are sometimes made separate 
from the column forms proper. Generally this is done either to avoid remaking the column 
tope to fit varying sizes of beams and girders, or to facilitate the construction of special form 

ximed fv true 


hettds. Fig. 50 shows an assembly of column heads fur regular interior eolumas of the 
Larkin Go's, building, Buffalo, N. Y., designed by the Abertbaw Construction Co., the contrac- 
tors. These heads are also shown, but not in detail, in Plate IV, page 1 15. 

The problem ctf the column head in flat-«Ub construction has been met practically from 


Fio. 63.— Bt«I column form med in Reid-Murdock C 

Fio. S4.— Steel column head to fit octAgon 



[Sec. ^-Md 

the first by the use of metal in some fonn. The metal mold used by the Aberthaw Construc- 
tion Co. in the Massachusetts Cotton Mills at Lowell, Mass., is shown in Fig. 51. Note that 
the head is very simple, a square where it leaves the column, with flaring beveled comers, which 
finally form an octagon at the ceiling level. 

Fig. 52 shows a type of column-head form used in constructing the Larkin warehouse at 
Chicago, HI. It is one of the standard types of flaring column caps manufactured by the Des- 

Jop Dhmefer 

)^' Apron Strip 


^Opomgy extension 



Plan of Fbrm/ng of 
24-Inch Column 

'^-. _ 

Lower umt'^^ 

Upper uniP^" 

Plan of Joint Where 
Units Telescope 

Section of Complete 
Column Form 

5tandord iS^Head 

Fxa. 55. — ^AdjuBtable steel column heads, HydrauUo Pressed Steel Co. 

lauriers Column Mold Co., St. Paul, Minn. The columns themselves were molded by steel 
forms and these forms were placed after the erection of the upper floor falsework. The column- 
cap section was suspended from the floor forms and the remaining column sections bolted to it. 
Figs. 53 and 54 show adjustable column heads as manufactured by the Blaw Steel Construc- 
tion Co.; conical and octagonal heads only are illustrated here, but this company also makes 
rectangular head molds and molds which spread from a rectangular section at the column to an 

Sec. 2-Me] 



octagonal shape at the floor slab. The conical head is in two parts; the top portion remains the 
same for all columns while the bottom is fitted to columns of various diameters. The octagonal 
mold is also adjustable at the lower end and may be used in connection with either octagonal 
and circular steel column forms or with wooden octagonal forms. The Hydraulic Pressed Steel 
Co., Cleveland, Ohio, manufacture round steel column forms with both standard and adjust- 
able column heads (see Figs. 55 and 56). 

r- Sheet niefal decking for depressed pane/ 

Cover opening with-' 
guaHer strip 

^-Fafse tverk b/ock nailed 
fo framing for- support 
(f head form. 

Frame togire from Ijl'foi" 
clearance fw* actual dlamel'er 
of head 

'Bolt fogether 

Fig. 50. — Method of forming depressed panels. 

64€. Wall and Pier Forms. — Forms for bridge piers and for walls of appreciable 
height are usually constructed of either 1-in. or 2-in. plank nailed to studs and held by horizontal 
waling pieces, with tie bolts extending across the pier or wall between opposite wales. The 
wales, which often consist of two planks fastened together but separated by spacing blocks, 
are set edgewise against the form studs and the tie bolts are carried through the openings which 
occur in the waling pieces. Wire is sometimes used for bracing and is tightened either by wedges 
or by twisting. The wire pulb against spreaders which are inserted between forms and which 
are removed as the concrete level rises. 

Where bolts are employed in pier and wall construction, a number of different methods are 
used for withdrawing the bolts. One method is to cover each bolt with old pipe cut somewhat 
shorter than the inside dimensions of the forms, and to place a wood washer at each end of the 
pipe. When the forms are taken down, the bolts are easily drawn out of the pipes, the wood 
washers are then cut out of the face of the concrete, and the holes pointed up. Another method 
is to make the bolts in three pieces, with the middle piece occupying the same position between 
the forms as the pipe above described. This middle section is connected with the end pieces by 
means of ordinary unions. When the concrete has set sufficiently, one turn releases the end 
sections and the holes left in the work are plugged with mortar. Still another method is to use 
a patented casting with set screw, also a tightening wrench and rod puller. The tightening 
wrench is a device for the purpose of exerting a pressure against the form to draw it in line or to 
desired dimensions. The rod puller is for removing or pulling the tie rods from the concrete 
after the concrete has set sufficiently to stand alone. 

The second method described is shown in Figs. 57, 58 and 59. (In Fig. 59 wire is used in- 
stead of the middle piece of bolt.) The bar couplings shown in Fig. 57 are manufactured by the 
Marion Malleable Iron Works, Marion, Ind., and are stocked in seven sizes threaded to take 
bolts J^, J^, J^, %, 1, 1>^, and \y^m.. "Universal Cone Nuts," shown in Fig. 58, are manu- 




•-•=: = 


iS " at, s. 








.< ° 

I I 



t, utd rod pullw, HihI 



{*) _ ,, (0) 

Fra. M.— Typical piv lonu wad in bridcn ol 


f actured by the Universal Form Clamp Co., Chicago, HI. Cut Bhowe waahera used with cones 
and countersunk into forms, but thia is not necessary. Universal rod clamps (Fig. 60) can bi; 
used on the outeide of forms in place of nuts and washers. The device shown in Fig. 59, known 
as "Tyacni," is being marketed by the Unit- Wall Construction Co., New York City. 

A patented casting, rod tightener, and rod puller such as used in the third method above 
described are shown in Pig. 60. These devices are manufactured by the Universal Form Clamp 
Co., Chicago, III. The rod clamps are made in six sizes, namely; No. 1 for K-in. and Hs'''*' 
rods; No. 2 for %-ia. rods; No. 3 for ,'i-in. rods; No, 4 and No. 4 extra heavy for 5^-in. rods; and 
No. 5 for ^-in. rods. The No. 4 is recommended for column clamping and No. 4 extra heavy 
for large retaining walla. 

Patented wire form clamps are illustrated in Mge. 61, 62 and 63. The wire form tightener, 
afaown in Fig. 61, is manufactured by the Marion Malleable Iron Works, Marion, Ind., and is 
made in three sizes, threaded to take \i, 3i, and 1-in. bolts. The tightener is simply attached 
to s wire slipped through from one side of the forms and a bolt is screwed into it from the face 

irfrt iO prrtrnt 
^rtatllf^ vf wall 

side. By turning this bolt any degree of tension may be obtained in the wire. The " Universal 
Wire Clamp" is illustrated in Fig. 62. The clamp ia placed against wales or studding and wires 
bent as shown in (a). The clamp should be placed in a position so as to allow the wires to enter 
the slots at nearly right angles. For locking all that is necessary is to pull handle down so that 
it takes a position as shown in (6) and (c). The "American Wire Clamp" is shown in Fig. 63 
and ia similar in operation to the clamp just mentioned. 

Where especially good work is desired, forms are lined with galvanized iron. For high 
piers or walls, the forms are constructed in large panels. After the concrete has been 
constructed to a proper height and the laat course has set several days, the panels are discon- 
nected and hoisted to a higher position and then reassembled for concreting, and ho on. 

When the ends of bridge piers are rounding, special forms are necessary. In the construc- 
tion of the Atherton Ave. bridge over the Pennsylvania R. R. tracks in Pittsburg, the forms 
for the curved ends of the piers were built of 1-in. by 2-in. strips nailed to horizontal segmental 
wales. These wales were nailed to the wales of the side forms. The rounding forma were kept 
in place by wiring to dowels set in the foundation concrete. Before starting the erection of the 
form work, a flexible panel was made by nailing galvanized-iron sheets to the 1-in. by 2-in. strips. 
This panel was bent against the wales, which acted like hoops. Curved pier forms used in 
bridges of Luten design is shown in Fig. 64. 


Fig. 66 illustralee n simple method employed by Itie Aberthuw Cunstructiuii Co. for build' 
ing ft wall of considerable height by means of movable forms. A simply method is alao afaown 
in Fig. 66. 

Walt forms of small height may be braced by battered posta outaide. Pig. 67 ahowa forma 
for a coacret« foundation wall for smalt buildings where no cellar is neceaaaty. Such forma may 

Fio. 86. Fis. 07. 

either be conatructed in sections or built in place. It should be noted that the forms are sus- 
pended over the trench and are not allowed to reat upon the new concrete. Figs. 68 and 69 
show other methods of form construction for low walla. No outside form is needed in the 
lower part of such walls as shown in Fig. 68 when the earth is extremely hard and firm. 

Fig. 70 shows a common type of form design for curtain walls below windows. Curtain 
walls for complet« wall panels are of similar construction. 

U. Deaign of Forma. — Although form design in most caaea has been left entirely to the 
discretion of the superintendent or foreman on the job, it has been found by a number of en- 
gineering contractoia that it pays on large or important work to have the forma designed in the 
drafting room, provided such designing work ia done under the direction of a man of practical 
ability. This plan is not only more economical, but also eliminates the possibility of an error 
bains made in the proper siie and spacing of form members. 


MaMjr ieUitttd Ivm iJinicK ikpaM 

wtinty ttpuM i<i i%m >t «Md w tp »riw K »! but 

of pasta for fctiii btMBk. aai (wdm aie c«|mM« «i{ b#a^ drt^nBiBttl bv !iMe«ltfe v<kll^ubtt^« 

tioo allowed. In no caae should the fipacing of the eupporta be greater than a safe ipan for 
the sheathing. Deflection should always be considered in order to give sufflciont stifTneas and 
thus prevent partial rupture of the concrete. 

6Sa. Values to Use In Design. — In calculating floor furms, we must add to the 
weight of the concrete itself, a live load which 
may be assumed as liable to come upon the 
concrete while it is setting. This live load is 
usually taken at 7S lb. per sq. ft. except in 
cases where the floor is made a storage for 
<«meut or sand. 

The pressure of concrete against forms 
depends principally upon the rate of filling 
and the temperature. The accompanying 
table' is from testa made by Major Francis 
R. Shunk, Corps of Engineers, U. H. A. 

Testa on the pressure of wet concrete 

ifnoi Ens. Kk.. Ju. is. 1610. 


t™„.„™ I 






































1,130 1,240. 1,440 

1.74(l. 1 



[Sec. 2-65a 

poured rapidly have been made under average conditions by the Aberthaw Construction Co., 
Boston, Mass. The following table^ gives the results observed in the pouring of two columns 
of small cross-sectional area. It should be noticed that the wet concrete exerted a hydrostatic 
pressure equivalent to that of a liquid weighing from 140 to 150 lb. per cu. ft. or very nearly 
that cf a liquid having the same weight as the concrete. 

Description of concrete 



Gb. per 
sq. It.) 

(lb. per cu. 


Time of pouring, 9 min . . 

Mixture, 1:1^:3 

Stone, 1-in. run-of-crusher 

Wt. of mixture, 152 lb. per cu. ft. 
Time of pouring, 14 min 

Mixture, 1:1:1 

Stone, 1-in. run-of-crusher. 

Wt. of mixture, 149 lb. per cu. ft. 



No. 1 








[ 15.08 






No. 2 













From the results of the tests just mentioned it would seem that 145 lb. per sq. ft. per foot 
of height is a rational value to use for the lateral pressure of concrete in the design of forms. 
This same value was also arrived at from laboratory and field tests made under the supervision 
of Prof. A. B. McDaniel and N. B. Carver at the University of Illinois in 1913, 1914, and 1915. 

Opinions differ as to the proper coefficient to use in the moment formula when making com- 
putations for the strength of floor sheathing and joists. It is probable that the formula 

Af B Tjr may safely be used in all cases except for single-span jcnsts where the concrete is conveyed 

to place by small dump cars on a portable track. With dump cars the concentrations on a sin- 
gle-span joist may be so heavy that the negative bending moments of the dead load will be 

relatively very small as compared to the positive moments and the formula M » -^ should 
undoubtedly be employed. When the bending moment is figured as M — r^, the deflection 

should be determined by the deflection formula D = «^. — 1^ (see "Notation," page 124). 

This formula is the ordinary one for calculating deflection except that the coefficient is taken as 
a mean between ^84 ^or a beam with fixed ends and ^84 for a beam with ends simply supported. 
For lumber commonly used in form construction E may be assumed at 1,200,000 lb. per sq. in. 
The fiber stresses allowed in form design may well be higher than those it would be desir- 
able to use in more permanent construction. The following values are usually employed : 

Maximum fiber stress in spruce or equal : 

for timbers 1,200 lb. per sq. in. 

for column yokes 1,800 lb. per sq. in. 

Horizontal shear for spruce or equal 200 lb. per sq. in. 

Crushing perpendicular to grain in spruce or equal 400 lb. per sq. in. 

« From ConcretfCtment Age, Oct.. 1913. 


The crushing of form lumber perpendicular to the grain should not be overlooked. Although 
a 3 by 4-in. or 4 by 4-in. post may be braced at least every 6 ft. in height and have apparently an 
allowable compressive strength of about 800 lb. per sq. in., such a post should not be permitted 
to carry more than one-half such a load due to crushing of the lumber perpendicular to the grain 
in the crosspiece or girt over the post. Especially is this true where a granolithic finish is to be 
cast with the slab, as settlement would ruin the slab surface. Large hardwood wedges should 
be used to prevent any settlement due to crushing imder the post. 

The maximum deflection which may be allowed a form timber is not definitely known. 
Some designers allow very small deflections for joists and use full live and dead load in design- 
ing. Others reason that any serious deflection will be caused only by the dead weight of the 
wet concrete and for that reason allow comparatively large deflections. Deflections allowed 
in practice vary from an arbitrary maximum of }{ in. for all timbers to maximum deflections 
of f^60 of the span. 

666. Drafting-room Methods. — The following article is taken by permission 
from a paper presented at the Twelfth Annual Convention of the American Ck)ncrete 
Institute by R. A. Sherwin, Resident Engineer, Aberthaw Construction Co.: 

The first study and drawings to be made form a general assembly. Usually several different combinations of 
timbers and methods of assembling the panels are sketched up and compared as to cost. Other things being 
equal, of course, the cheapest design is adopted. A record of costs of the various units which go to make up the 
complete centering scheme is therefore necessary. 
Points to be remembered in this study are: 

I. Joists and girts should be in as few lengths as possible to save time in sorting on the job. 
2t Use stock sises and lengths of lumber. 

3. Keep number of panels ^nd pieces to a minimum. 

4. Provide easy stripping. 

6. Allow clearance enough for slight inaccuracies in making up and erecting, swelling of panels, etc. 

6. Panels should be a whole number of boards in width, if possible, for ease in making up. 

7. Units to be as big as can be handled and joists used as panel cleats where possible. 

8. Provide for re-use of panels. 

9. Beams to be handled as trough units when the job is regular and imits can be re-used. 

10. Consider use of floor domes or inverted boxes when beams are close together. In this way beam sides 
and slab are erected, stripped and moved as a unit. When either of the last two systems is used the beam sides 
should be given a slope to prevent hard stripping. 

II. Provide for reshoring if necessary. 

12. Have bracing above the men's heads. 

13. When four beam haunches occur at a column consider making haunches as a unit similar to a column 
head in flat-«lab construction. 

14. Consideration of steel forms. 

A general assembly shows how the various parts are to be put together, supported, tied, and braced to resist 
the concrete pressure and the weights coming upon the members. This drawing in its final shape can be made on 
tracing cloth with soft pencil. The dimension arrows, lettering and figures should be inked so that clear prints 
can be made to avoid errors in the field. These assemblies should be filed permanently for reference on future work. 

When it is necessary to strip the floor centering for re-use before the concrete has been in place long enough 
to gain its full strength, provisions for proper support of the green slab must be made in the design of the floor forms. 
The safest and cheapest method of doing this, in the writer's opinion, is to place boards between the floor paneb 
and wedge poets up to a bearing under these boards before the centering poets are knocked out. In this way the 
slab is never left unsupported, as b the case when posts are replaced after all the centering b down. These boards 
should be placed according to a plan and shall be located so as to shorten the spans of the main reinforcing bands. 

After the general scheme for the forms has been decided upon the detail panel drawings are prepared. By 
panel, in thb paper, b meant several boards deated together into a unit to be used as a form for some part of a 
concrete member. Every different-sued panel b first sketched roughly on standard 6 by 9-in. sketching pads. 
These pads have holes punched at the top, so that as sketches of the different kinds of panels are made they can be 
bunched together and brass rivets put through the holes. These sketches are transferred in more detail onto thin 
tracing paper. The standard sheet b 25 by 33 in., divided by a 1-in. space into two halves of five spaces each. 
These detaib are used at the job mill for making up the paneb. The blue-prints sent to the mill man are cut up 
into 5 by &-in. units, with one detail on each, and are given to the carpenters at the making-up benches. On these 
detail sheets for each panel there b given the panel mark (which b stenciled onto the finbhed panel), the number 
wanted, and the floor on which they are to be used. The sbes of the stock, dimensions and alterations, when panels 
are to be re-used, should be clearly marked. 


A system of symbols as follows is easily learned by the workmen and should be kept standard: 

Bt beam side; BB^ beam bottoms; F, floor slab paneb; P, plinth forms; H, haunch forms; C, column sides; 
W, wall panels. 

Thus B18 means beam side number 18, and its location is shown on the key plan. 

These detail sheets can be drawn up entirely with soft pencil on paper, as they are not valuable after the forms 
are once made. Rubber stamps for a great deal of the lettering will save time. 

Where time allows and plans are complete it is advantageous to get out one kind of panel at a time. In any 
case the first panels to be sent to the making-up mill should be the typical floor panels. These can be used economic- 
ally in building foundation walls, etc., and in this way the job can get an extra use out of them. 

It is impossible to get a satisfactory cost unit for doing this form drawing. A price "per square foot or per 
detail varies widely according to the number of details required and the number of square feet that can be made 
from one detail in different buildings. A big building where a large duplication of the different kinds of units is 
possible will cost much less pa* square foot for this work than a small building which is badly cut up. The average 
cost per sheet has been found to be about $4.50 complete. This includes time spent on studies, assemblies, key 
plans, soheduling, and checking. The number of sheets required on buildings of different sixes of the same type 
is fairly uniform according to the sise. The number varies from 30 sheets on a small building up to 100 or more .on 
a large building with irregularities. By reducing the cost of the sheets to drafting-room hours, by dividing the 
probable total cost by the average rate per hour, a fairly close estimate of the number of men required to turn out a 
job in a specified time can be made. This is useful on rush work. 

The key plan mentioned above is really a diagram of the fioor plan upon which are shown the locations of 
the various fcn-m panels. This plan is the only one which the workmen consult in erecting the formwork. It 
must, therefore, be complete and clear. 

This plan can best be made by tracing the floor plan on cloth from a blue-print, indicating the beams and holes 
in the floor and the columns and walls in the story below. The lines should be heavily inked and the figures large 
enough and spaced so that the tracing can be reduced and the photo reduced to convenient sise for the foremen to 
handle on the job. A big blue-print is inconvenient and fades in the strong sunlight. As a rule, to avoid mis- 
understandings, a key plan should be prepared for each floor in the building. 

The key plan may be supplemented by a letter to the job in which should be noted the assumptions made in 
preparing the details. These notes might include such information as clearances allowed, grades at which forms are 
to start, scheme for re-use, etc. 

66. Tables and Diagrams for Designing Forms. 

66a. Notation. — The following notation is used in the tables and diagrams : 

h — breadth of member in inches. 

d » depth of member in inches. 

{ = span of member in feet. 
I" * span of member in inches. 

w » uniform load per linear foot. 
v/' = uniform load per linear inch. 

v/ B total load on floor in pounds per square foot — dead weight of slab per square foot 
plus 75 lb. per sq. ft. live load. 

h =B head in feet. 

D s deflection in inches. 

E » modulus of elasticity in pounds per square inch. 

/ = moment of inertia in inches^. 

/ a maximum fiber stress in pounds per square inch 
M B bending moment in foot-pounds. 
Mr ^ resisting moment in inch-pounds. 

s s spacing in inches. 

n B number of spaces. 

n' 8 number of spaces between column yokes. 

h B largest dimension of column in inches. 

P s concentrated load in pounds. 

V B total maximum vertical shear in pounds. 

V 8 maximum unit horizontal shear (pounds per square inch) ^ o \b^) 

666. Fiber Stresses Allowed. — The fiber stresses allowed arc those given in 
Art. 65o. 

Sec S-66r] 



€6c. Ponnnlas Used. — The spacing of joists was determined in liable I by the 

For flexure 

For deflection 

For horisontal shear 

s = 2000— r,, 

IT 7 * 

D = 0.0225 


When D = Vs in 

When D = I4 in. 

/ - 

/ = 




^Tien D = ~. 



H = 3200 ^'/, 

IT / 

(Depths for joists are taken ^4 in. less than nominal sixes.) 

In Table II, Part A : 

For flexure 

s = 2000 -7}i 

For D = H in. 

a = 11,100 


For horizontal shear 

s = 3200—. 

IT 7 

Table III: 

Deflection of 3^ in. governs. 

In Table II, Part B: 
For flexure 

8 = 1600-7r, 

* = 1780^, 

For horizontal shear 


8 = 3200 Tt 

tr 7 





In Table IV the concentrated loads from the joists were considered on a simple span in 
calculating the bending moment. The worst case was assumed — ^that is, when one joist comes 
at mid-span. Since a girt is usually continuous for at least two spans and the full live load never 
reaches it, the moment of resistance of the timbers was multiplied by 1.2. Nominal aises of the 
girts were used in the computations. 

For flexure 


I = 

240W« -h ^P{n* 4- 2n) 
3P (^-fl') 


Horizontal shear is to be considered separately. From the above considerations it would seem 
that an allowable horizontal shear of (200) (1.2) — 240 lb. per sq. in. may safely be used. 


Diagram I: 
For flexure 

For deflection 

Dutgram II: 
For flexure 






8 = 238(h 

For deflection (D • }i in.) 







When d - 2 in. When d = 4 in. When d = 6 in. 

V - 28.3 in. l" = 40.0 in. V = 49.0 in. 

These results are based on a value of A; - V and the values given may at least 
be increased to 30, 42, and 50 respectively. 

For shear (o * 200 lb. per sq. in.) 

r' - W - (r - k) 

Table V: 

, W")' 
^ " 19,060 

A - p Vl9,060n' 

Formulas to use with table: 

Diameter of bolt — \ oi'f ^ t 
Net area of washer 

2/" "k 

iLifUtTBATiTS PBOBLBin. — 1. Required the proper spaoins of 2 by 8-iii. joists bavins a span of 7.5 ft. to 
support forma for a 54n. slab in beam-and-girder oonstruotion, assuminc 1-in. sheatliing. 

Table I shows that 2 by 8-in. jobts spaoed 31 in. on centers will give sufficient strength if the bending moment 

is assumed SQual to t^* For this spacing the tablci indicates that the deflection is somewhat over Hia. but lees 

than H in. and muoh less than Hso span. Accurately, D ■- 0.0225 ,'-^ « 0.16 in. From Table III we find that 

the spaeing eannot be greater than 80 in. without the deflection of the sheathing exceeding )i in* • which is not 

«p<i (7 5)* 

For M m -g-, the spacing would be (0.8) (31) - 25 in. from Table I and the deflection D - 0.030 ^75 * 

0.22 in. FVom Table II, the spacing would be 22 + H(8) « 24.7 in. 

For M ^ j^ and the deflection limited to H in., Table II shows the proper spacing to be 21 + H(8) « 23.7 in. 

2. Assume Joists in the preceding problem to be supported midway between beams. Det^mine their 

economical sise and proper spacing, assuming ^ '^ Tq <^d deflection limited to H in. 

It will be sulBoiently accurate to assume the span as 4 ft. From either Table I or II, we find that 2 by 4-in. 
joists may be emplosred spaced 25 in. on centers. 

8. Determine sise and span length of girt in the preceding problem to support the joists midway between 

Load ooming from each joist is (3.75) (*Mt) (137.5) - 1075, or say 1000 lb., accurately enough. From Table 
IV w« And that a 8 by 4-in. girt with posts spaced 3.8 ft. c. to o. could be used, or a 3 by &-in. girt with poets 5.6 
ft. o. to c.i or a ^ by 5>ia. gbt with posts 6.6 ft. e. to e. 

Hovisontal shear must be considered separately considering a joist to occur dose to one support. Assuming 

a 4 by 6-in. girt with posto 6.6 ft. on centers, F- 2260 lb. and t - H-jtct^ * 142 lb. pcrsq. in., which is less than 

the allowable value. 


3 I 

I i 11- 
.1 I i 

~ 1 




s I 



Sec. 8-66c| 



A 3 by 4-iii. post oould Bustain (3) (4) (400) « 4800 lb. without injuring the dbera of the girt. It would only 
be required to support 3450 lb., consequently this siie of poet is suitable. 

4. Assuming the cross-eection of beam below slab as 14 by 18 in. (23 in. total depth) determine the safe span 
for the beam bottom to be made of 2-in. plank. 

Live plus dead load on beam bottom is 75 + 'Ma (150) - 3625 lb. per sq. ft. Diagram I shows the maximum 
span to be 43 in. 

5. Find the proper si>acing of 3 by 4-in. posts to support the forms for 8 by 16-in. beams (orosa-eeotion given 
bdow slab) spaced 6 ft. on centers with a 4-in. floor slab. Assume that no girt is placed at midspan of joists. 

Total load on beam per Unear foot is (125) (6) + (^) . (^^)^^^) . gg3 ly^ ^^^ considering the weight of 

the forms, which may be neglected. Safe bearing of post on fibers of cap ■■ (3) (4) (400) ■" 4800 lb. Safe spacing 

of posts — go«- — 5.4 ft. 

6. Determine the sise and spacing of joists, girts, and posts to support an 11-in. flat slab floor. 
Assuming 1-in. sheathing we find from Table III that the joists cannot be placed more than 27 in. on centers. 

Table I shows that for 2 by 8-in. joists the spacing of girts may be made 6.5 ft. The load coming from each joist 
is (6.5) mU) (212.5) - 3110 lb., or accurately enough 3000 lb. From Table IV we find that for 4 by 6-in. girts the 

posts may be placed 3.8 ft. on centers. Horisontal shear on girts must be considered separately, v ■■ Htttts^ ~ 273 

lb. which is somewhat greater than the allowable value and a 6-ft. spacing of the girts is necessary. 

Table III.^-Safb Span for Floor Sheathing 



Based on M 

(4) (6) 

■rrr- With dcflection limited to ^i in. 




























4 in. 
























6 in 






II in. 


















7. What spacing of vertical studs is required for a wall form with l^^in. sheathing and a height of 12 ft. 
Diagram I shows the spacing to be 18 in. 

8. Assuming a 24 by 24-in. column with V -■ 37 in., determine the spacing of the column yokes. 
For 2 by 4-in. yokes placed on edge, the value of T' to be used in Table V should be 

(37) (1.23. ^(2) (37) (24) - (24) ] _ ^^^ ^ ^^ ^^ .^ 

For 4 by 4-in. yokes, the value of l" to be used should be (37) (0.87) (0.035) - 30 in. 

For shear the actual value of /" must not be taken less than Od - (2" - ik) i- (9) (4) - (37 - 24) - 23 in. 
for either the 2 by 4-in. or the 4 by 4-in. yokes. Evidently the shear in the yokes will be less than the allowable. 

Other things being equal. Diagram II shows that 1-in. sheathing would be more economical than IJi-in. 
when 2 by 4-in. yokes are used. The table shows that the same number of yokes would be used in the two cases. 

The spacing center to center for the 2 by 4-in. yokes for a 10-f t. column with 1-in. sheathing would be as follows 
in inches starting at the top: 30-20-14-1 1-10-9- 8-7. 


Where the strength of yokes governs their spacing, the bolts must have a diameter 

.^(0.096) (2) (16) 

using actual 

values of k and V, Thus for the 2 by 4-in. yokes, diameter of the bolts must be at least 

3M« (3) (2) (16) 


0.25 in. 

The net area of washer should be 

1.92 sq. in. 

2/" - k 50 

9. Assuming a 12 by 12-in. column with V — 22 in., determine the spacing of 2 by 4-in. yokes placed on 

The limiting value of I'* for shear on yoke b (0) (4) — (10) - 26 in. Thus actual values of t' and k must 
be considered as 26 in. and (26 - 10) - 16 in. respectively, which gives a value of 2" to be used in Table V of 

(26) iiaS) .y pr(26) (16) - (16). ^ _ ,3 , ,„ ..^ ,, .„ 



Sec. 2-67] 



67. Systematiziiig Fonnwork on Bufldings.^ 

•7a. Sawmill i^d Yard. — The sawmill should be located after a study of the job site in a positiou 
where there is plenty of space for piling the stock and finished panels. The sawmill shed and equipment should be 
standard so tha^ it can be erected and put in operation as soon as possible in order to reduce hand-eawing to the 
minimum. A sawmill at its best is dangerous. For this reason every precaution should be taken to protect the 
workmen by efficient saw, machine, and belt guards. The mill should be near the building so as to reduce the cost 
of moving panels. This moving cost is an important item in obtaining low erection costs and will amount to a 

Diagram I 

Spacing op Vertical and Horizontal Studs (or Column Yokes) at Any Given Depth 

Below Surface of Concrete 

(Based on strength and deflection of sheathing) 

M M -rrr- DoflecUon limited to yi in. 



20 25 30 35 

> - sSpac/ngr h inches 




considerable item when the mill is poorly located, either from lack of planning or from lack of space for the plant 
around the building. 

The mill yard should be so arranged that the stock will go in one direction from the lumber piles to the saws, 
to the making-up benches and to the finished panel piles, which should be nearest the building. 

The lumbor when received is cheeked as to quality and quantity. The stock is then sorted and piled accord- 
ing to sise and length, each pile having its sise plainly marked. 

It is then an easy matter for the millman to make a lumber ledger of all the stock in his yard. Thus at all 
times it ia possible to know the stock available, for as the lumber is used deductions can be made and the running 
total kept up to date. 

•76. Shop Procedure. — The millman divides the panel details into boards to make the required 
width of the panel. The number of boards of each kind needed to make the number of paneb wanted are ordered 
moved to the saw to be cut to proper length and thence to the make-up bench, where cleats of the proper sise and 

1 By R. A. Sherwin, Resident Engineer, Aberthaw Construction Co. From paper presented at the Twelfth 
Annual Convention. American Concrete Institute. 



[Sec. 2-676 

Diagram II 

Spacing of Column Yokes oh Horizontal Studs at Any Given Depth Bbw)w Surface 

OP Concrete 

( Based on strength and deflection of the yokes or studs) 
Based on M ^ —^ Deflection less than H ^n. 

10 /5 aO 25 30 35 40 45 50 

s - Spacing of cokjmn yokes or horizonfai studs in inches 

Directions for Uunc Diagram II and Table V 

For 2X4-in. yokes or studs (flat) multiply actual l" by 1 .73 before usitig diagram or table. 
For 2X4-in. yokes or studs (on edge) multiply actual V* by 1 .23 before using diagram or table. 
For 3X4-in. yokes or studs (on edge) multiply actual V by 1 .00 before using diagram or table. 
For 4X4-in. yokes or studs (on edge) multiply actual /" by 0.87 before using diagram or table. 
For 3X&-in. yokes or studs (on edge) multiply actual V* by 0.67 before using diagram or table. 
For 4 X 6-in. yokes or studs (on edge) multiply actual /" by . 5S before using diagram or table. 
For 6 X 6-in. yokes or studs (on edge) multiply actual V* by 0.17 before using diagram or table. 

For 6Xd*in yokes or studs (on edge) multiply actual V* by \ -~ before using diagram or table. 

For co lumn s the value of i" to be used in diagram or table should be the value of I" as found above multiplied 

by '\\,, • in which expression the actual values of i" and k are to be substituted. 

In determining spacings from the diagram or Uble for {| ^ tiS:.*4**X 6^i^*"ind } ^"^"^ ***'**' 
values of l*' greater than { 50 q } in. will give a deflection of yokes greater than H io> and actual values of I" 

greater than \ 72 1 *'** ^' ^^^ '^ deflection greater than >« in. 

In determining spacings from the diagram or table, aeiital values of I" must not be considered as less than deter- 
mined by the formula 

/" « 9rf.(/" - k) 

otherwise horisontal shear will be greater than 200 lb. per so. in. The corresponding actual value of k (which will 
be called i')8hoald be determined by subtracting the value of a (see sketch^ from the value of V* found by the above 
'^""-ttla. The value of i" to use ia diagram or table should then be found as explained above and finally multiplied 



lencth 19 ftlnady waiting, hmvinc been fmrioaely erdved tliroa^ the mOL AnotlMr erdir to the beneh car- 
pentcn, which k dipped to the blue-print sketeh of the panel, enablei them to eleat tocether the boarda into the 
finiahed paneL The panel ia then taken away by a laborer, atencikd with iti location mark. oHed and piled until 
ready for use in the building. 

AH thia ti done by orders written on atandard order forma of three kinds, one for movinc the atock, another 
for sawing, and the last for *»**^"g up. Duplicates of the orders are placed on the miUman's progress board so 
that he knows at all times how the work is progressing, for when an order ia completed and returned to him its 
duplicate is taken down. Each kind of order ia of different color ao as to be easily identified. 

There are several points in connection with making up panela which may be considered here. 

As much assembling as possible should be done at the bench. Rangers tor wall bsami can be attached and 
ledgers to support joists can be nailed to the interior beam side cleats at the proper depth. When inacrts need to 
be placed in the sides of beams, the holes for the bolts which hold them can be ^ocated on the details and the holes 
bored at the mill. The groove strip for steel sash and also the corner fillet can be put on the column sides and beam 
bottoma. Berd key strips for walls should also be nailed lightly to the paneb when necessary. 

Heavy cleats for big wall columns should be cut and bored, but not attached to the paneb. The panel can be 
deated with IH-in. boards, and is thus made much lighter and easier to handle in erecting. Cleanouts at the 
bottom, on two opposite sides of all c<Jumna, should be made at the null. Reduction strips should be nailed at the 
edge of all pands which reduce in size when reused. 

An small pieces liabl > to get lost or used for other purposes can be dipped in red paint ao that their small sise 
m-ill be respected by the carpenters when looking for looee boards. 

Much waste in making up ]^i-in. panels can be prevented if the floor is laid out to make the majority of the 
panels a whole number of boards wide. Roofers come 5H to 5H in. wide, and it is an easy matter to ]4an the 
paneb to come, say. seven or eight boards wide, which means no waste in ripping one board to make the width. 
The lengths should be planned as near stock lengths as posdble Paneb made oi spliced boards are expensive to 
make and are easily broken. 

New or Uttd Materialt — It b not eeonomieal in labor to make up paneb from lumber which has already been 
in contact with concrete several times. If. however, paneb are in good condition, they can be cleaned and repaired 
at a considerable saving, both in labor and material. 

•Te. Cleat Spacing. — The cleat spacing for the various kinds of paneb should be kept uniform so that 
tne strips on the benches, for spadng the cleats, w 11 not need to be moved for every set of paneb. 

For beam sides 2 by 3->in. cleats, flat, can be used on paneb up to 30 in. deep: 2 by 4-in., flat, up to 42 in.; and 
3 by 4-in.. on edge, above 42 in. deep. Naab should be specified as follows: 

, Wire naib should be used for f ormwork because 

No. SUe Width of board.'inches °' «*^ »° driving and drawing. The holding power is 

Ai^ ^ 73^ ''^ sufficient. Double-headed naib should be used 

I ™ «!: *f r^ in securing all boards which have to be looeened bc- 

2 ]^ 2?i to W forestripping. 

* ]^ I^ i*r ,'^ ♦ 'W- PHumim of Field Work.-An im- 

It J J ,. u ? : ,. . ^^ J- portant feature of the fidd work b a planning depart- 
So coated and clinch ends in ^-m. boards. «.^_* «.U«— U. -i 2* • 4,^ I. aL I • J 

^ ment. wboee busmess it is to plan the work in advance 

so that at all times the several gangs will have definite 
taaka to do and be supplied with suffident material with which to do the work. 

The moving boss hss an important position in thb fidd work. He b responsible for getting the proper paneb 
to their correct location in the building and for having all otho* stock called for on the assembly plan on the job 
ahesd of the erection carpenters. When the forms are stripped he must move the paneb, which are to be remade 
before their next use, to the remaking benches, and thence to their next location. It b expensive to have high- 
priced carpenters waiting for stock cm* using stuff not suited for the job they have to do. 

The erection of the centering and the assembling of the paneb, after the latter have been moved to their 
proper location in the building, are done by carpenters. It b economical to employ good carpenters and to keep 
them employed constantly. A gang of men that understand form work on concrete buildings, the methods used, 
and the grade of work desired, b a valuable asset for any firm specialising in retnforced-concrete work. 

A competent foreman should be in charge of a|l carpenter labor. He should be consulted by the planning 
department so that "team work" will prevail. 

Quality in concrete building work b usually more desirable than low costs. Good lines and surfaces will be 
remembered after the cost b forgotten. These results can be obtained only by the careful supervision of the erec- 
tion of forms. 

•Ts. Stripping of Formsw — When forms are to be reused, sab usually the case, stripping should be done 
as carefully as posdble. An intelligent foreman in charge of a stripping gang b a good investment. Any man cmn 
wreck forma cheaply, but a man who can strip the forms and leave them in good condition for reuse b in the 
end the cheapest stripper. He saves the time of the high-priced carpenters in remaking and fitting broken paneb 

The question of when to strip b one for the building dedgner to decide, and hb instructions should be car^ 
fully followed. It b cheap insurance to make enough forma ao that no ehancca need be taken of weakenii^ the 
structure by stripping while the concrete b still green. 

The paneb which need to be altered before reuse should be remade from the blue^prints showing the neces- 
sary changes. A bench for thb purpose can be set up in the story where the forms are stripped. It b sometimes 

oESESAj, ueTHom or cox/afsvofiaif 136 

'tint V«liut ijd Uj^ lU^ lfiV,JttX. 'i UJ 

d^ iU MCl IIP tif tilt viow vovaMiW kiwk W oooat upoi. tUtdiL, frw ^^ ^im^t^ b^uM be itudu^'iuUutt'edui i^ Lo^- 
1 sind ■OM' DiH^ tin- *i»aiUiuii oiore lAiOHiut nuii ic 6 auutl iuuimiw tuiMiiUuuut 

W. BiMl Ffldus. — Btoul {uTHit Wve aLtt-uyE U«aii used tuu<« or it^e for £eweis, «UJ:t», &ad 

1-1.; ;i ~IUi>n iuLi >ilJ ^ruir uu luusiklwu wall W pUntoi Uubinu^A^ Co.. l'>tuljuis>i. ^'i- 

b: on cxeeedisely rnpid rau. bw*.-l )ufiusHr<;itluiuei utiivereally employ e<l (ur circuljii' culuiuus 
&□<. lor flftiiaf^ (.'oluuiii IttituLi. Wall furuis lu^ ako iiiucli u««d luid uuvc given UttUfiJucUuii. 
eapeciaUy m rseideaovs and oUter atruuiurt^ of tiiu iuundatiuii-wall iyp<'. 

IspvctbI iype» oi iloor forms are un tL«: mArkfi nad hti: uaed to euiuo extent. Tbe aaoie js 
•ru'- of fnma for recbWRUbv und ucUi«o>ii>i toluDiii^ 


Hie ftdjustment in height of steel forms for circular columns ie obtained by tekscoping 
the ends. Such forms are usually made up of a series of panels of thin galvanised steel held 
rigidly in pUce, like staves in a barrel, by means of stiff steel bands. The panels are somewhat 
Sezible and are sprung in or out depending upon the siie of the colunm. 

Fia. 72.— StM aontyln. 

Fig. 53, page 100, and Plate V, page 116, show steel forms for circular columns, in place, 
and ready for the pouring of the concrete. The conBtruction around the column heads should 
be noted. The form illustrated in Kg. 53 is manufactured by the Blaw Steel Constnictioa Co. 
and ia so designed that all variations are taken care of with a single set of forms. Vertical ad- 

Flo. 73. — Ulsd flontyica with end d^m. 

jiutmont is obtained by telescoping one form section inside of the section below it, an adjust- 
ment of 18 in. being permitted between each two sections. Diameter adjustment is provided 
for by the use of form panels of various standard widths. The edges of the panel sheets are 
bent back to form flanges, and these flanges are slotted. The panels an kept to any desired 

curvature by segmental steel bands which slip through the slots in the panel flanges, and arc 
drawn tight by means of keys or wedges. The bands are not adjustable, and a complete set 
must be provided for each diameter of column to be built. The form panels are made in three 
Standard lengths, 6 ft., 4!^ ft., and 3 ft. The steel column forma are not used to support any 
part of the floor, as is usually the case in wood-form construction, so that the floor forms may 
be erected complete before the column molds are sot up. 


The " Hydraulir " roluinn (oniia aluiwii in Fig. S9, pace 1 lU (inaii<jfiu'tur»^l liy Lhc llyJrau- 
Prencd Steel Co., ClnveUncl, OhioJ ron«Jit of Kalvanii^U-iruii giiils li^lij tfigethsr at tlu: 

Fk. ?&.—<:. F. HMl-Ule. 

jumts witb qiiiuk-«(Tttii{ i^ltuups uud (ibti>HKl u'iUi st«e1 rui£. Any beighit «tf ixduaui Oif fliee g 

bt»d mar br uln4iiiHKl. 

Fif. Tl ^DWE BUw li^il wftll fofUB lor fouiidati(in-«'ii.U 

work wbere wiiv ties ktf uaed. The standard walUforui patiul 

■e 2 fl. aquare sad pnn-ided witb boliK in the fuur fitutgiog 

BDgls for the paaaay irf faaUmtne, Hlute are ubu provided in 

the plBte to permil rf the inBertioo of the wire tit*. The borl- 

■ontaJ snd -vartic-Hl iinere are used Ui keep the [utki straitfht Mid 

to ^omMd panels tugctber so tbat tbey iiiftj' be«hift«d in larger 

unitE tfaas Biugle paiiele. This type of furm is oot limited to 

two-cDUTK wtrtkins. but mtiy be umkI in pouiing anj' desired 

height irf wall at imt opemtiun. 

Vol] and foundation fonuE manufactured by the Hydrauli<' Pressed 8t«el Co. ootisist of 

ujiTigfate vbitih aif aligned and uci'iiniiely apac«d 8 ft. 3 in. c. t« c, <if iKeflsed st«el liawe. 
Between (beie uprights ateel-faoed plat<ie are 

A type of permanent (ooa-removabkj 
8l«e! form for flow (.-onatruction wluch ha^ a 
similar futtctiun to boUuw' tile in t«ffa-ciitUi 
, boliow tile fluoB, iiianufautured by (be TruiiiKxl 
' Concrele ljt«e] Co., ie shown in Figs. 12 and 
73. Tliie type of floor form is also shown in 
Fig. 35, page 3fl. "tjtsel Kjoredomes'' shown 
in Fig. 74 are manufaetured by the same oom- 
pany for two-way coostrwrtion m whicli the 

loads aae tecrned in two directions to the supports. The metal dumes art deeply corrugated 

t(> secure stiffness and are only <^»en on the uudvcsid)' su that the joislA extend on -" ■"■<-= "* 

the docDG. Tbe siaiuiard h^ghte of "m«el Floreiyke:'* fi, H, 10, I'i, and 14 ir 


mate width at base: 20}^ in., exclusive of 1-in. flanges along bottom edges, which add 2 in. 
to this dimension. Standazd lengths (nominal) of all sixes: 4 ft. and 3 ft. — actual lengths are 
4 ft. 1 in. and 3 ft. 1 in. to provide for end lap. "Steel Floredomes** may be obtained in 
depths of 6, 8^ 10, and 12 in. and 21 by 21 in. base. Permanent steel floor forms similar to 
''Steel Floredomes" are furnished by the General Fireproofing Co., Youngstown, Ohio (known 
as G. F. Steel-Tile, Fig. 75). Steel floor forms which are removable and may be re-used in succes- 
sive floors of a building are furnished by the Concrete Engineering Co., Omaha, Neb. (known 
as Meyer Steelforms, Fig. 76) ; and by the Witherow Steel Co., Pittsburg, Pa. (known as Wisco- 
forms, Fig. 77). 

69. Construction Notes. — Nails should be used sparingly in the construction of forms ex- 
cept in those sections which are to be used over and over again without change. Unnecessary 
naUing not only adds to the labor of wrecking but is liable to render the lumber unfit for con- 
tinued use. Where nails niust be used in the connection of form sections, the heads should he 
left protruding so that they may be drawn without injury to the lumber. A special form of 
double-headed nail is now on the market and gives satisfaction. 

The location of column forms from floor to floor of a building should be determined by 
means of the transit, and special care should be given to the erection of these forms in order to 
make sure that they are set true to line and level. Column forms should be held in position by 
diagonal braces in two directions nailed to adjustable or slotted blocks which are bolted to the 
concrete slab — the bolts being placed when the floor is poured. Sometimes small pieces of 
plank are employed instead of blocks, and these are nailed directly to the floor slab within 2 or 
3 days after pouring and while the concrete is still green. 

Trouble in the erection of floor forms may usually be traced to inaccuracies in form measure- 
ments. If the column forms are of the proper widths and if the beam and girder forms are 
cut to exact lengths, no trouble of great consequence can arise. A variation of more than ^^ 
in. from the sizes shown on the drawings should not be permitted. 

One method of erecting forms for rectangular columns is to nail three of the sides together 
lightly before raising them to place, and then to set the remaining side afterward. This 
enables the column reinforcement to be put in place before the form is set. Another method in 
common use is to assemble the column ftfrm complete before raising, in which case the form 
must be raised above the projecting reinforcement (belonging to the footing or column below) 
and then lowered. 

All forms for concrete require a coating of some lubricant to prevent the concrete from 
adhering to the wood and making a rough, unpleasing appearance. Crude oil or petroline is 
used to a considerable extent and preserves the forms against damage by alternate wetting and 
drying. The forms should preferably be oiled before they are set in place. 

Oil should not be used on forms against surfaces which are to be plastered, as oil prevents 
the adhesion of the plaster. Wetting with water in such cases will be sufficient. 

Beam and girder forms should be raised slightly higher at the center than at the ends in 
order to prevent sagging. If this is done, deflection and compression of the supports will 
finally leave the beams and girders in a level position. A deflection equal to ^ in. ip every 10 
ft. of length is usually provided for. 

The sides of beam and girder forms should project over the edges of the bottom plank. 
By so doing it becomes possible to leave the beam and girder bottoms in place after the side.*; 
have been dropped. 

Slab forms of the ordinary panel type should be made in sections (usually four to a floor 
panel) in order to prevent binding and permit easy removal. A splice of ^ in. is usually 
allowed between adjacent sections, and this space is covered with a strip of sheet metal, thus 
giving some leeway in fitting the sheathing panels into place without unnecessary cutting. 
Sometimes it is a good plan to provide for a loose board between two panels near the center of 
the span so that temporary uprights may be used to support the floor slab when the forms are 

Sec. 2-70] 



All posts or shores should rest on large hardwood wedges, driven in pairs to an even bear- 
ing. Hard driving of wedges should not be permitted as it is sure to injure the concrete which 
is setting under them. In some instances wedges have been placed at the top of the posts 
instead of at the bottom, as is the usual custom.* The disadvantage of this method lies in the 
difficulty of driving wedges while standing on a temporary support but, on the other hand, 
by top-wedging, the shores or posts may be permanently braced before the form work is leveled. 

Concrete, when poured under horizontal or inclined forms (such as in footings), will exert 
an upward pressure, and such forms should be securely anchored. 

When posts are placed on plank sills, great care should be exercised to avoid settlement be- 
cause of the Hkelihood of hollows coming under the sills due to unevenness of concrete floors or 
to thawing out of frozen ground. 

Deflection should be carefully guarded against at a window head as a slight deflection at 
this point will cause considerable trouble and expense in setting the window sash. 

Forms which are to be used again should be cleaned as soon as they are taken down. 

In removing forms the green concrete must not be disturbed by prying against it. 


70. Checking, Assorting, and Storing Steel. — Steel should be checked, assorted, and stored 
as soon as it is delivered at the site. It should be blocked up several inches from the ground 
and should be stored in such a manner that those rods needed first may be easily reached. 

71. Bending of Reinforcement. — Such a simple matter as bending of rods for concrete 
reinforcement might seem to be almost too unimportant a subject to be worthy of very much 
attention. However, when it comes to the proposition of making thousands of bends per day, 
factors of time and expense in this work are most important. Of course, the structural design 
should be such as to require the steel bends to be as few in number and kind as possible, but much 
can be done to lessen expense by the manner of making these bends. In addition to economical 
considerations, care should be taken to see that the bends are made true to Une and plane, and 
that the steel is not injured during the operation. 

71a. Types of Bends. — In general, there are five different types of bends to be 
made with reinforcing rods which may be indicated as follows: (1) bending of heavy beam and 
girder rods; (2) bending of the vertical reinforcing rods of columns at or near floor level where 
columns change size; (3) bending of stirrups and column hoops; (4) bending of slab reinforce- 
ment, and (5) the coiling of rods or wire to form spiral column reinforcement. In making each 
type of bend, the work should always be so arranged that all rods of the same size and shape 
are bent at the same time. This avoids remeasuring and resetting of templets. 

716. Hand Devices. — ^A simple method of bending heavy rods is shown in Fig. 
78. Either steel bars or steel plugs 5 or 6 in. long are placed in holes in the bending table, these 
holes being bored at the points where the 

C/eats'on uncferstde*.^ 


1 1^ 





1 1 





Fig. 78. 

bends are to be made. A piece of plank 
is nailed to the table, as shown, in order 
to hold the rod in place while being bent. 
The rod is then placed in the position GU 
and bent around the plugs C and Z>, and 
then around the plugs B and E until the 
ends AB and EF are parallel to GH. 

The manner in which rods are to be bent is generally left to the discretion of the steel 
foreman. The general practice is to bend beam and girder reinforcement by using a heavy pipe 
slipped on the rod and then making the bends as described, using either steel plugs or angles 
bolted to the table. 

In building two large reinforced-concrete buildings for the General Electric Co. at Schenec- 
tady, N. Y., the Stone & Webster Engineering Corporation, of Boston, accomplished the bending 




uf the heavy rods by means of (wu -^^-in. Htecl ptalcs itiouDted one on top of t ho other on it bend- 
ing table, and, in these, holes were drilled 3-in. on centers in both direrlions. Stwl pegs wen- 
dropped in the holes the proper distance and angle npart, and the rods were then bent by hnitd 
using a 2'in. steam pipe. 

R. C. Hardmao, in Engineering Reeorii, describes as follows a small homemade bar bender 
for bending cold bars up to I 'i-in. diameter, where not greater than 90-degrce bendsare require^l : 

Tbe ippantua coiuuta nwntutly ol 
Bt*fti Itvtt fkatened to the pUte by muns ' 

The cwt-itoD pUtc can be cut ol coi 
iptaeODtHtbeiacmschined. Tbsbalt h- 

« i 

o[rx ii'?'* 



ri iw""'"T{" "*'" 

o X— 





KD luis beta 

reen whieh the bar i. pla(«d. a 


cU. and 


cnindrv. the 


e pUte «ith which (he lever com« 

, certlin 


nnl. to allow free aetian <rf the 


lu» Ufa 


be Blightly 

he largnt aiic bar lo be bent. 

The fiUer in . 

dciH« mad 



by any blK^kamilh. which bta 


(he lei 

o insure a ciiht fit for baia 



the matin. 



'a can be made at amall coat. 

To in. 

good fit the 



ugarouod which the ■■Slier" 

it plM«l oh 

uld b« mac 



ever ia made of 1 by flat 

■ tMl forged 

o .hape. m 

nEagins the bar alightly upeel 

on the 


side, ttale 



4 ft- 

The oper-tion of t 


an be readily Been in FU- T». 

in Rhic 


o be beat 

nd a RUer are ihown dolled. 



la ia futened 

to a be 

r itaelf oi 

Several hand and power devices are on the market for the bending of steel reinforcement and 
for the making of spiral coils. These all have their merits and have given satisfaction. 

Fig. 80 shows the l.'niversal bar bender which may be fastened to any bench or plank. 
It is a light, portable device weighing about 60 lb. and capable of bending all ordinary sites of 
reinforcing bare to any angle desired, without any adjustment being necessary. The top half 

.Lniveraal bar bi 

of the bender can be removed and used to bend bars after they are in place. The bar rolls 
around the pin in bending, thus distributing the strain along the bar and reducing the chances 
of fracture at the bend. 

The bender is equipped with a 5-ft. crowbar for a handle, which may be removed and used 
for other purposes. To bend large bars easily, the handle should be lengthened by using aik 
iron pipe over the crowbar. 

A bar bender designed for heavy work and manufactured by the Wallace Supplier Mfg. Co., 
Chicago, ni., is shown in Fig. 81. This machine haa an auxiliary ratchet lever which operstrs 
ft Dinion against a series of teeth in the frame at a large ratio, thus developing great power. The 


CESEHAL \fETtlOhS i»f CiiSM'tHri'ttdS 


r>tf bet p*nri nt^j be Xhnma tmx i^-pntfgi^aimnt. anrl loivhrAff i/pKHiiMl wifb (Ih< 'i^flM kv" 
fur tight w-orfc. 

A bolder maimfartansl bjr ib« W«i»™t« (:*jft««n»'fi'*i *'. , WXWiyW,',. (/«**. *■ rtrfW**, ,i, 
Fmc S2. This bender bewli rnnfr^nrimt t<i*r!) np tfr atn't m^lij^bft* ('i »*, '(■>>* «hw(«ia* hi (npi 
niattrd «nb a detaehahk batHtle 7 (t. ["fna; ('■( '■.'.nv*iM>-Tu-<- ir. td>wlljh«, 

■»r header dM»g»wli#>i»«d»(if ^^•►'rfnfiftfi'OTirasfr'wtf.&sW'. iKlJwtfl'r/.P.* iwbwJIa PrtMMiH*'.)ln»i*.- 

! niH diAin ujiC 111' ii-nitiniffiiiT-n itiiinln w iijiHii-fl jimiiiHilv iiii( i.if t*.r,i , i^^j it.^rir.if'-- 

ucinoG' Jna' enoBilfiii. 

: ramumtaaia n£. .mnHir. .nmuu' juMUir- f iu* to^r' i 

< MUfdtae ««Bi>t: w 



(Sec t~lU 

attempted. Square cold-twisted ban frequently give trouble in bending, due to the fact that 
they are apt to twirt out of their proper plane. 

Tie. Bending of SUb Reinforcements-Slab reinforcement is usually bent aft«r 

il is in place on the floor. A tool for this purpose is shon-n in Fig. M. Sometinies, however, 

slab rods are bent before being placed. This latter 

_ _ method aeems preferable ainrc the bends can be morr 

ULUi ' accurately made. To keep the coot of such bending aa 

low as possible, the bending machine should be con- 

■-■ ■ — ■ structed so that the bends may be made with great 

rapidity. Such a machine is shown in Fig. S5 and was 

•■"'"■ ** employed on the Geneial EHectric Go. 's buildings retetrpd 

This machine consists of two uuita which can be placed any distance dpart and securely 
bolted down to a table. Each part consists of a steel plate on which is mounted two amallor 
platefl having angle stops attached. Between the stops is lorated a casting, which is pivoted »> 


lio. 8i. 

that it can readily take any desired position between the angle stops. \Mth the steel rod in 
position, this casting can be turned by means of a long lever attached so as to move one of the 
mounted plates. On the upper side of the casting are four lugs, and two of these lugs are cod- 
otructed so that adjustable collars can be slipped over them, thus making a tight fit for the rods 
which are to be bent. The amount and the angle of the ofFset 
can be regulated by changing the distance through which the 
lever is turned. This machine offBets a rod parallel to itself and 
with one pull of the lever two bends can be made. 

72. Placing of Reinforcement. — Steel should be thoroughly 
cleaned before being placed in the forms in order to obtain a posi- 
tive adhesion of the concrete to the steel. A slight film of red 
rust is not objectionable, but no rod should be set in place on 
which nut scales have formed (see Art. 54, Sect. 1). 

All reinforcing metal should be securely fastened in correct 
positions by wiring or otherwise before the placing of the con- 
crete is begun. Particular attention should be given to loose-bar 
reinforeement — that it is accurately and property supported in 
position and that it is not disturbed until the concrete is poured. 

The advantages to be derived from placing beam-and-girder 
reinforeement in frames has been considered in Art. II, Sect. 11. 
With steel in frames the erector has simply to line and level 

themintlieforms.placehraccswherenecesaarj-, and make end connections with abutting frames. 
Toluran reinforcement should be made up into frames, the same as for beams and girders. 

Slab and wall rods should be tied with wire at their intersections to prevent them from slip- 
ping or getting out of place. The usual method has been to use a pair of pliere and to cut ihf 

Sec. 2-73] 



Tvire into convenient lengths. A device known as the Curry Tyer has recently been placed on 
the market for wire tying which has made a great record for itself in a very short time. Fig. 
86 shows this to be a simple and practical device. The ties used for the binding are uniform in 
length and the mechanical action of the tying tool gives the wire a uniform number of twists. 
The tie is looped around the rods and the ends are placed on the hooks of the tying tool, then a 
quick upward jerk of the wooden handle whirls the teeth and draws the wire up tightly. 

Another method of tying slab and wall rods together at their intersections is shown in Fig. 
87. These Bar-iya are manufactured by the Concrete Steel Co., in the three tjrpes illustrated. 




FiQ. 87. — Bar-tys. 

The tys are quickly put in place and when once snapped on the bar will resist a tremendous 


73. Devices for Supporting Reinforcing Bars. — There are many excellent devices on the 

market for supporting reinforcing bars at the proper distance from the forms. Some of these 

devices not only support the rods but give them the proper spacing and lock them in position. 
Beam spacers sold by the Universal Form Clamp Co., of Chicago, 111., are shown in Fig. 88. 

They are made in three sizes — ^namely: 4 in., 5 in., and 6 in. — and space the bars accurately 

both from the sides of form and center to center of bars. 
Easy chairs sold by the same company are adaptable to 
either a one-way or a two-way system of slab reinforce- 
ment. Fig. 89 shows the chair before being placed in 
position. The upright tying finger enables the chair to 
be readily seized and put in position. Fig. 90 shows the 

Fio. 88. — Beam spacers. 

FiQ. 89. — Easy chair. 

FiQ. 90. 

chair applied to a one-way system of reinforcement. After the chair has been placed in posi- 
tion the flexible tyipg fingers are bent over the bar from opposite sides by hand, firmly 
securing the chair to the bar. This chair can readily be applied to a two-way system of re- 
inforcement. Easy chairs are made in three sizes so as to provide for reinforcing bars varying 
in diameter from ^ to 1 in. 

A device for supporting slab rods, known as the Securo locking spacer^ is a light bar passing 
beneath the rods and having depending lugs which rest on the bottom of the form and so keep 
the rods at the desired height. At the location of each rod the bar has two flexible clips which 


Pia. SI, — Recum sUh bar sputar. 

Fig. H.— Ty-cfaur. 

Fia. 03.— Kual-chwr. 

Fla. U7.— Ujr-ebau. 

Sec- 2-73] 



are bent around it, and these serve to insure the use of the proper number of rods. Three strips 
of Securo slab bar spacers are used per panel for ordinary spans. Two additional upper spacers 
are used where slab bars run in two directions. Upper spacers are also used in flatnslab con- 
Btruction. The device is shown in Fig. 91 and is made by the Metal Building Materials Co. 
of Chicago, 111. 

Securo supporting and locking spacer for beam bars is shown in Fig. 92. Three spacers 
are used per beam for ordinary spans. Upper beam spacers are employed when beam rods 

Fia, 98.— Chair spacer. 

Fia. 99.— -Continuoua slab spacer. 

are to be placed in two or more layers (Fig. 93). Securo beam spacers are furnished for any 
widths of beams and for any number of bars in lower and upper layers. 

Supporting and spacing devices manufactured by the Concrete Steel Co. are shown in 
Figs. 94 to 100 inclusive. Ty-chairs, shown in Fig. 94, are made of spring steel ¥dre for tying 
any combination of reinforcing bars. These chairs are made in the following standard sizes: 

No. 2 

No. 3 

No. 4 

Combinations of cross bars, 

Combinations of cross bars, 

Combinations of cross bars, 




H and % 

H and Vo 

^i and H 


li and H 

H and % 

H and H 

H and H 

^i and 1 

H and H 

H and H 

H and IH 


H and H 

li and 14 
li and 1 

The EaseMuiirs shown in Fig. 95 are designed particularly for terra-cotta or steel tile and 
joist construction. Single chairs supporting one bar are made 2 in. wide, and double chairs 
supporting two bars are 4 in. wide and will fit any size bar. The standard Easel-chairs space 
the underside of the bar 1 in. from the form. 

Bar-chairs (Fig. 96) are used for supporting single bars. They are easily sprung into posi- 
tion, locking on the bar with a strong tension grip. Bar-chairs are made for each size and shape 
of reinforcing bars. Standard distance from underside of bar to forms is 1 in. 

Hy-chairs are illustrated in Fig. 97. They are made from 2 by )^-in. flat steel with any 
required height, and are used for supporting single reinforcing bars of any type or size. They 
are particularly useful on flat-slab buildings for rigidly holding the column bars. 

Chair spacers and cantintums slab spacers are shown in Figs. 98 and 99 respectively. 
Lne "chair spacers'' are made of spring steel wire. The "slab spacers" are made from J^-in. 
t'old-rolled angle and the prongs are so flexible that they can be readily bent around the bars by 
nand. In both types of spacers standard distance from underside of bar to form is 1 in. 

Adjustdbie beam saddles (Fig. 100) are made from sheet steel and used in beams and girders 





for accuratdy spacing the b&ra and holding them the lequired distance from the forma. Stand- 
ard dtstAnpe from underaide of bar to forms ia I }-i in. and 2 in. 

"Hie cltair lock shown in Fig. 101 is manufactured by the Electric Welding Co., Pittsburg, 
Pa. Stock sizes are as follows: ?^-in. round cross rod by ^i, M, %, ^, and J^-in. round main 
rods. Any aize, however, can be furnished to 6tanyBhapeof rod. CAoir p>ncA«rs are f umiahcd 
in two sUes for fast«Ding the locks to cross rods. 

Slapie cAoirt shown in Fig. 102 are made from extrastiS sheet steel, cut and bent to develop 
two pairs of pointed proi^s projecting in opposite chairs. The chain are driven into the 
formwork as far as is desired and in the exact position the steel is to occupy. The bars arc 
placed and the upper prongs bent down over the bar with a quick blow of the hammer. The 
driven-in points do not aeem to make form removal difficult. 

Fid. 100.— AdjiBtable bar 

Fm. 102.— Suple ehuc 

Bt Habvbt Whipple' 

71. Derelopment of the Industry. — Although the development of the concrete products 
industry, embracing, firstof all, the manufacture ofbuildingunitS,b^an earlier than reinforced 
concret« building construction, it has lagged behmd it. This has been due, in general, to a lack 
of trained management. And the trained management has been lacking, perhaps, because of 
the apparent sLmplicity of operations involved in the manufactui« of concrete building units. 
In the early days, the use of cheap machines of questionable value was common and many of 
them were sold on such a basis as to tempt the more slovenly artisans. Men who had made 
failures of other work sought easy money by making block beside a gravel bank and following 
the inadequate directions which came with the machine. 

Attracted by the seeming lack of any complexity in the essential operations of block 
manufacture, by the cheapness of the machines with which the work was to be done, and by the 
seeming lack of any necessity for any auxiliary equipment, it is nut to be wondered at that there 
were many enterprises in concrete products manufacture of woodshed and backyard magnitude. 
The result was that many cheap buildings put up with this new product soon developed the 
weaknesses which brought down upon an entire industry the condemnation which was earned 
by its incompetent putt^rers. 

Many of these early products were not even sound structurally. Many of them were 
extremely porous. The appearance of a concrete block wall after a rain and the slowness with 
which the moisture disappeared in subsequent sunshine, gave rise to the early belief that 
i-uncretcos a building material makes for dampness and is unfit for dwellings. The fact that 
the early block had hollow spaces giving a cored wall equal to 25 to 50% of the cToa»«ectionnt 

' MmiJiiint E<lilDr Cnri-^i.. AuihcT -r-nfrrtr SloM Minufutute." 


area, led to the extravagant claims not only that no moisture could get through the wall from the 
outside, but that such a wall was sufficiently insulated so that there could be no possible danger 
of condensation on the interior wall surfaces. The blockmakers had the architect against 
them. The commonest type of block was an imitation of pitch face stone, which the architect 
objected to, not so much because it was an imitation, as because it was a very poor imitation 
of the real thing. Architects also objected to the early \mits because of their proportions in 
height and length, commonly 8 in. high by 16 in. long. 

The conditions which are gradually driving the incompetent out of the industry, and which 
are closing down those plants which are inadequately equipped, poorly managed, and under- 
capitalized, are bringing into the industry men who were first unattracted by > what seemed to 
be a "small fry" business. The possibilities of concrete stone manufacture have been de- 
veloped very remarkably in the last 10 years, and much more rapidly in the last 4 or 5 years, so 
that many of the leading architects are now specifying manufactured stone on an equal basis 
with natural stone, and in some cases, in preference to natural stone in important building 

There are very few who doubt the value of well-made concrete building units and it remains 
only for intelligent manufacturers to develop their business along lines entirely different from 
those which characterized the industry's early efforts. 

A most important thing for them to appreciate at the outset is the difference between 
concrete units which are suitable for exposed walls and for the trim of first-class buildings, and 
those other units which demand no architectural consideration and which are used to replace 
common brick in foundation walls and such walls as are to be faced with some other material. 

76. Two Mam Lines of Work. — There are in general two types of concrete building units. 
Their manufacture involves two distinct lines of work. One is in the production of standard 
units in quantity; units which are structurally sound but have no special claim for use where 
any architectural purpose is to be served. This is a simple bulk proposition, where the constant 
flow of materials and the quantity of manufactured output are largely the determining factors 
in the success of the enterprise. The other line of work is in the manufacture of trim stone or 
standard units which are specially faced or otherwise surface treated to make them suitable for 
exposed walls or trim in competition with natural stone, face brick, terra-cotta, and other 
well-known building materials similarly used. 

The success of one enterprise or the other depends quite as much upon the natural supply 
of other building materials in the community as upon the enterprise and ability of the particular 
manufacturer. A successful enterprise in the manufacture of trim stone and ornamental work 
necessitates the employment of modellers, pattern makers, mold makers, workers in glue, 
plaster, and wood; it involves the use of selected aggregates, more skilled workmen in the 
molding department, and frequently the employment of stone cutters in the finishing 

76. Methods of Manufacture. — In the production of concrete building \mits there are the 
wet process and the dry process. There is no well-understood definition which sharply distin- 
guishes between a dry mixture and a wet mixture and the consistency of the concrete used varies 
all along the line from that mixture which will just stick together when squeezed in the hand, to 
the other extreme, a mixture which is of a soupy consistency. 

There are three general classifications in manufacturing methods, each calling for the use of 
diflFerent equipment. These three are the so-called dry-tamp method of manufacture; the so- 
called pressure method; and the wet-cast method. 

76a. Dry-tamp Method. — The dry-tamp method is the one most commonly 
employed. It is the method whose products have in the main given concrete block its early 
))ad reputation; it is a method whose uses arc quite satisfactory when in intelligent hands and 
where there are methods of curing the products which contain so low a percentage of gaging 
water; and it is the method whose abuses have resulted in many of the bad products that have 
brought unmerited criticism in some cases of the entire output of the concrete products industry. 


The dry-tamp product may be made ia an ordinary wooden box, but as commonly known it is 
made in the mold boxes of simple machines, ao familiar everywhere (see Fig. 103). The mix- 
ture should have just as much water in it as will permit the quick removal of the product from 
the mold in which it is made. The abuse of the method is in using too little water. 

766. Pressure Hethod.^Preaaure machines are not so common in the field as 
they were at one time although their use is undoubtedly increasing at present. There are in use, 
however, machines applying pressure hydraulically, 
and others in which the pressure is exerted mechani- 
cally by means of toggles operated by hand (see Fig. 
104). It has been urged by some that the applica- 
tion of pressure which is exert«d evenly over one en- 
tire face of a product does not result in so dense a unit 
as is possible through tamping, the contention being 
that this even pressure, allowing less free displace- 
ment of individual particles than by tamping with a 
small-headed hammer, induces an arching action, 
particulorly when crushed stone is used, this arching 
action among the particles of stone resulting in voids. 
There appears, however, to be little practical evi- 
dence of the correctness of this belief. 

76c. Wet-cast Method.— Wet -cast 
concrete products are not easily defined nor are the 

Fia. 103.— Huid Ump block mscbioc. , . i.- u . ■ . .i i a1- j -m. 

factors which enter mto them eaauy outhned. There 

is more variation in the equipment used and in the methods pursued in wet-cast work than in 
the other two methods. The quantity of water varies a great deal in wet-cast work, as be- 
tween the consistency used in sand molds and that used in metal molds, for instance. With 
sand molds, it is possible to use a very high percentage of water in the mix, providing the 
is (Kiastantly agitated before being deposited in the molds, because the excess moisture 

Fia. 104.— Prcnure block nwchinc 

is, in large measure, taken up by the sand of the mold. The use of a very wet mixture in 
metal gang molds (Fig. 105), or in a fairly tightmoldof any kind, would result in a poor prod- 
uct, due to the fact that much of the moisture could not readily escape and would be almost 
sure to result in a )>oraiis product. Tlic tcniicncy lo l>c overcniiif is Ihc use of two much walcr. 
77. Consistency. — The ronsidcralion of the various processes of manufacture has involved 
some thought of consistency which, in a measure, defines those processes. It is now fairly well 


established in tlie concrete fteld that the ideal consistency is in that mixture which will barely 
rotain He shape when the forms are removed immediately after the concrete haa been deposit-ed 
and pressed into shape. This is just a little wetter than can readily be used in dry-tamp ma- 
chines. It is just a little wetter than is ordinarily used in pressure machines and it is a great 
deal drier than the mixture which is ordinarily used in the wet-cast concrete. Since, however, 
various other conditions which contribute to the production of good concrete are more suscep- 
tible of control in the case of factory work than in field work, the concrete products maaufao- 
titrer has a somewhat wider latitude in the matter of consistency. 

Concrete products manufacturers using dry-tamp equipment have coifstant difficulty with 
employees in trying to get them to use a mixture of the wettest consistency which it is possible 
to use in the mold boxes of their machines. 
The drier the mix, granting that it is just wet 
enough to stick together under tamping, the 
easier it is to remove the product from the mold 
without dam^e. The quantity of water which 
gives the mixture an ideal consistency resulting 
in water marks on the outside of the product 
when it is removed from the mold will, it i a gener- 
ally contended in the field, cause the product to 
stick to the face plates, resulting in damaged 
products and in retarding the work. This stick- 
ing is undoubtedly due to a combination of the 
water and fineness of the facing material in caus- 
ing a suction on the face plate which mars the 
fresh product. A wetter mixture with a slightly 
coarser facing material is not bo liable tx) cause 
damage in removal from the mold. 

The results of dry mixtures are not nearly 80 ^o- ^05. — Metttlgug molds mounud on ™r. 
bad as might be expected, 'When the products 

are removed from the molding room promptly and put into a steam curing room where a warm 
atmosphere saturated with moisture does not permit the evaporation of any of the moisture 
which has entered into the block in the first place, very good products can be obtained. 

Mixtures so wet that they would give very unsatisfactory results under ordinary conditions 
give high-class products when poured into sand molds. A recent tendency among sand-cast 
stone manufacturers is toward less water and toward longer mixing, the additional mixing 
serving in the place of so great an excesa of water in obtaining a smoothly flowing mixture. 
High crushing strengths and a high degree of density are obtained. Numerous architects show 
a preference for the manufactured product over the natural product because of less tendency 
to discolor through the absorption of moisture and dirt. Where a mixture is poured, however, 
into a rigid and non-absorptive mold, as in the use of steel gang molds, In block and brick manu- 
facture, it is important that the moisture content be kept down just as low as ia possible consist- 
ent with a ready flow of the material into the forms and around the cores. 

78. Commercial Holds. — Commercial molding equipment is, for the most part, very sim- 
ple, the essential requirement being a mold box from which the product can be readily removed 
when shaped. An idea which has been almost inseparable from concrete block from the incep- 
tion of the molding machinery for its production, is that it shall provide a partially hollow 
wall. This is very clearly shown in Fig. 106. The various sketches show how the designers 
of different machines have varied the proviaiona for air space either in the unit itself or in the 
wall OS the block are laid up. The block shown at A, B, C, D and / arc in one class, being 
complete unite in each case providing the entire wall thickness. At E are two separate thin- 
wall slabs which, when laid in the wall, are held by metal ties. The unit shown at H is for 
light residence or other light wall construction, providing the plain outer wall surface on what 



[Sec a-79 



I J K 

106.— Horizontal cros8-«ections of representative types of 

concrete block. 

is here shown as the upper side of the sketch. The projecting lugs provide a base for attaching 
furring stripe for lath and plaster. Blocks F and G each consists of two separate parts held by 
metal ties, cast in the blocks. The broad U-shape block J is designed for an interlocking 
arrangement as laid in the wall, the straight faces of this block forming both interior and 
exterior wall surfaces, giving a complete and continuous air space in the wall. The block 
shown at / for a similar construction, the lug on the block giving the desired bond between 
the two sides of the wall. Still another type of block has three rows of vertical ducts, and two 
horizontal. The center row is filled with dush concrete as laid up and provides for reinforcing 

The hollow space serves to economize in material, to make a lighter, more easily-handled 
building unit, to provide as laid up in the wall, either a series of vertical air ducts, as in the block 
shown at A, B, C, D and /, or a more nearly continuous air space throughout the wall, as in 

the block shown at E, F, (?, H, J and 
K. In the varied manner of provid* 
Ing air space in the wall lies the chief 
difference between many of the ma- 
chines on the market. 

79. Operation of Machines. — In 
the operation of the machines, con- 
123 side ring more particularly the mold 
boxes, the general type shown in 
Fig. 103 is made either to make 
block face down, face up, or with the 
face at the side. If block are to be 
made with a special face design, as 
for instance, the lamentable example of the rock-face block which is poorly conceived as an 
imitation of pitch-face stone, then the face plate, which is to give this design, is usually at the 
bottom. If the face of the block, however, is to be faced with a separate mixture of material for 
a special texture or color, then the machine will be of either the face-down type or face-up 
type. If block are to be produced solely for structural purposes, without face design, or with- 
out a facing material used on the face side, the so-called stripper machines have given very 
satisfactory results. In these machines, the block is produced upright in the mold just as it is 
used in the wall and the cores of the machine are introduced and removed by upward and down- 
ward movements. The sides of the block are thus always perpendicular to the bed of the 
machine, hence the block is stripped out of the mold with a troweling action. Other types 
of machines tip the block over before it is removed on the pallet. 

In some machines the cores are removed with a downward motion after the block has been 
tipped over. The common way, however, is to withdraw the cores while the block is still fa<'e 
down, leaving the hallow spaces lying in horizontal position. After the cores are removed a 
block is turned over on the pallet. All this is done by the movement of two levers, one to 
remove cores and one to tip over the mold box, or by automatic mechanism which is set in 
motion by one lever movement. 

One of the pressure machines has a sort of track of equal length on each side of the preasure 
head (see Fig. 104). Mold boxes travel on this track, one box at each end. The box is filled 
clear to the top, if it is to be a plain block, or if it is to be a faced block the backing material 
is struck off at a depth of K u^- below the top and the facing material is put on and struck off. 
The box is then rolled on the track to the center of the machine under the pressure head. 
When the levers are released, the box is rolled back to its first position. A pallet is placed on 
top and clamped in position. The box is then turned over and lowered so that the pallet rests 
upon a stand placed to receive it. The block is thus released face downward on the pallet. 

79a. Tamping. — In the use of machines in which block are compacted by means 
of tamping, this tamping is done in three ways: (I) by hand solely; (2) by hand-operated pneu- 


Tiiatac taj&pcn; oi ^) b^- uieiine of macdiine tsdOfiera auepexuied abov€ tJ;ie xuold box. iiajud- 
t :i mpiac; is most ccmmiaii, the oonunoD type <af <««Dj)cr (Fiig. |A)7; beuog a douUe-cuded JAi^ti u- 
rxjent villi a oeuttsr bar lor a handle, a<cid with one jusurrow £^pa4eiULke he&d luxd uoe broad ilat 
hf!»d, the ikfljTotw head bewg used beweoa oo&ee and the broad head over the full block aiea 
before aikd after the ooree are mtroduoed. In hand work, eveo-ythiug depends upon the oper- 
ator and most manulactiHrerB maiortain that the ^i^^wa^tor jcilaxee his efforts to a cooaiderabie 
exiicait toward the end of the day. lu apite of this, a man who is accustomed to this work gives 
e&oeDent remihs with hand^tamping- 

Machinee for tamping have to be built to withstand severe jarring. They are supported 
f'lther by a fmmewufk from the floor of the factory or suspended from a framework above. 
Sut^h tampers have several feet at the ends of plungors so 

ammged as to fit the type of macliine in use. These f r | -j — ^ \J 

plungers strike the concrete with equal force over the "^ 

entire open area of the mold. The plungers are so ad- ^*" 107.— Hand tamp. 

justed as to be responsive to the depth of the utaterial 
in the mold box, so that the length of stroke varies as the mold box is filled. 

BO. Gmg MbMs for Wet-<aMt Products. — A typ<' of equipment for making wet-cast 
products that is coming into more general use consists of gang molds. One kind is mounted 
upon cars (Fig. 105^, which are run on tracks to receive the concrete at the inLaker, and from there 
to curing tunnels. Another kind is located in such a way throughout a stretch of floor as to 
lequire the mixed concrete to be brought to the molds. This process thus involves either the 
use of a veT>' large number of individual molds in which the products must remain arranged 
over a ver>' laqse casting area for from 12 to 48 hr., or else it involves the use of molds set up 
in such a way that they nuiy be conveyed to a curing place when they are filled. When the 
products have become hard, the molds are simply taken down by a removal of the core pieces, 
stdss and division plates, and are oiled and set up again, either on platform or car. 

81. Ms^rriiris. 

81a. C«mflBt>-^orage aad Conveying. — ^As a rule concrete products manufac- 
tUFRS arc^ satisfied to use any standard brand of Portland cement, which can usually be de- 
pended upon to conform to the specifications of the American Society fur Testing Materials. 
There is, however, a disposition among some manufacturers, particularly those making a high 
class of trim stone and more particularly als(» where a rather wet mixture is used, to select their 
brand of cement with some care. Some of these manufactiu'ers beUeve that some brands have 
a tendency to cause crazing, which is one of the bugbears of the concrete stone manufacturer. 
No manufacturer has been found, however, who can explain just exactly the reason for his pref- 
erence for one brand over anotlier, except so far as his experience has seemed to show that the 
use of one brand resulted in less crazing than another. Ther(3 is a beUef among some manu- 
facturers that a cement which has lioen aged much longer than is ordinarily demanded is desii- 
able in concrete stone manufacture and it is said tliat this older cement is less hkely to givt* 
hair-checking or crazing. In other respects, the quality of cement required in concrete produet s 
manufacture scarcely differs from tliat in general concrete work. 

In a small plant the ceuM^it is ordinarily stored on a platform a^ near a^ possible to th<' 
level of the hopper which feeds the mixer. This may be, in some plants, at the bccond floor 
level, chutes being used for cement as well as aggregates, or it may be at a level between the 
fiiBt and second floor, <letermiued by the level of the mixer itself. If stored in bags on the second 
floor level, an elevator of some kind is provided unless the plant is so small as not to warrant 
the use of equipment of this kind. It is possible, sometimes, to have a railway siding on a 
trestle and to use gravity conveyors with ball-bearing rollers to carry pallets bringing bags of 
cement. With an arrangement of this kind, the cement is brought mto the storage space 
direct from the car with very httle handling. In another plant the track may be slightly 
above the level of the floor on which the mixer stands; or, even with it oii the same level, it is 
possible to build a platform at the level of the car floor and to pile bags of cement in such a 


way that they may be emptied into a cart filled with gravel from a bin along the side and dumped 
into a mixer which stands just under the platform. 

Bulk cement has not been used extensively in producU manufacture but has been very 
succeasfully used by a few. The cement may be scraped from the door of the car into a chute 
feeding into the bottom of an elevator boot, the elevator lifting the cement into a bin in the 
top of the plant from which it falls by gravity into a measuring box above the mixer hopper. 
In another plant the cement has been loaded from the car to wheelbarrows, handled over a 
runway to the hopper of a mixer. Where the output of the plant is large enough and it haa 
been possible to hold a car to use up its entire contents, this handling of the cement has been 
very economical. When this has not been possible, the cement has been dumped from the wheel- 
barrows into a bin close to the mixer from which it is shoveled into the mixer hopper. 

816. Aggregates — Kind and Quality. — In the main, the aggregates used in the 
general field of concreting, are suitable, except as to size, for concrete products manufacture. 
In general, fine materials are used throughout the concrete products field. Better, cheaper 
products can of course be made when larger aggregate can be used, the maximum size equal 
to one-half the smallest dimension of the product. 

Aside from quality as to cleanness, hardness, and so on, the concrete products manufacturer 
has been most concerned with the consistency of the mixture with which the size and quality 
of the aggregate have a great deal to do. A very dry mixture on one hand and a very wet mix- 
ture on the other — ^both of them more common in the field of products manufacture than any 
intermediate mixture — ^have both been an influence in favor of rather fine materials. Most 
manufacturers using dry-tamp equipment appear to be convinced that coarse materials cannot 
be successfully used in a mixture containing little water because of a tendency of coarse mate- 
rials to fall out of the product on removal from the molds and to cause a high percentage of 

Crushed limestone, especially when there is a rather high percentage of fine material, un- 
doubtedly permits the use of more water than does sand. It is still a question whether or not 
the excess of moisture used in a mixture of such material, becomes available for the hydration 
of the cement in the curing period which follows the molding. 

The prevalent belief is that a high percentage of fine materials should not be used, the 
usual specifications being that a percentage no higfier than 5 or 10% passing a 100-mesh screen 
shall be used in the fine aggregate. 

The most desirable qualities in concrete building imits are, of course, strength and the 
quality of resisting the attacks of the elements, to the end that they will not disintegrate the 
concrete nor spoil its beauty through absorption of discolorative agents. 

Bank-run and crusher-run materials should not be used. It is important that a concrete 
products manufacturer's output be of even quality. To this end he should maintain aconstant 
supply of an aggregate of uniformly high quality, grading in size from fine to coarse. Any 
amount of tamping or pressing, or care in puddling and pouring, or in placing the concrete, 
is entirely unavailing if these operations have not been preceded by scrupulous care in the choice, 
grading, and mixture of the materials, as essential to securing density in the product. It is 
strongly recommended in connection with this chapter that reference be had to the chapter 
on "Aggregates" in Sect. 1 and the chapter on "Proportioning" in Sect. 2, as these chapters 
treat in detail of the selection and grading of the materials in order to obtain the best results. 
No manufacturer should determine his proportions arbitrarily but should first examine or have 
examined samples of the material which he proposes to use and which he has reason to believe 
will come to him in unvarying quality. To make sure that this quality is unvarying, frequent 
tests should be made to determine the grading which should precede any decision as to the pro* 
portions in the mixture. 

88. Mixing. — No one in the concrete field has a wider latitude in selecting mixing equip- 
ment than the concrete products manufacturer. His plant is stationary, the requirements of 


-the plant are more or less regular; and he has no problems of getting about, here and there, 
under varying conditions. 

82a. Mixers — General Type. — ^The commonest types of mixers in concrete 
products plants are those with the simple cylindrical drums in which a comparatively dry mix- 
-ture of concrete ordinarily is turned out, and the continuous mixer, for a long time much despised 
in the general field. In many concrete products factories where a wet mixture is being turned 
out in large quantities, types of mixers are used similar to those found in large construction 
iTTork in the field. In addition to these types, the small cylindrical mixer for handling facing 
materials is common in almost every concrete products factory. There are frequently two or 
three of these so that they may be used for facing mixtures of various kinds without change. 
In connection with the mixers employed there is one thing which is perhaps unique in the prod- 
ucts field and that is the adaptation of some of the best-known makes of continuous mixers to 
the specific needs of the factory in which they are used. Local mechanical ability in each case 
has been able to set up these machines so as to give very satisfactory results. Not only has 
the flow of dry materials been fixed under close control but the water is added in definite 
quantity to give a continuous flow of a like mixture. It is almost invariable that the use of a 
continuous mixer in a concrete products factory requires the erection of bins and hoppers over 
those with which the machine is equipped. 

826. Mixing Dry and Mixing Wet. — It is coming to be more general practice to 
mix the concrete materials dry in one mixer and to add the water in a second mixer. In a plant 
where stone of a very high quality is manufactured, materials are stored on the second floor. 
They are shoveled into a car in definite proportions with the cement on top. This car is shoved 
along the track in front of the bins from which the materials are obtained and is dumped into 
a cylindrical drum mixer where the inaterials are mixed dry. They are then dropped through 
a chute to a continuous mixer on the first floor where the water is added, the quantity of water 
being very carefully gaged to give a like consistency all the time. 

Long and thorough mixing is particularly important in concrete products manufacture 
where a homogeneous mixture of like color throughout is particularly desirable. Mixing the 
materials dry and then adding water in another mixer is almost sure to make for greater 
thoroughness in a combination of the materials. 

82c. Agitation Subsequent to Biixing in Wet-cast Work. — Where materials are 
mixed wet for casting in sand molds, it is highly important that the agitation of the mixture be 
continued so as to prevent segregation of the materials. Even in mixtures where the size of 
stone is little more than }^ in., it is impossible in the wet mixture which is used, to prevent rhis 
stone settling to the bottom of the receptacle as it is poured out from the mixer. For this reason 
it is common practice to provide some means of keeping this mix agitated up to the time it 
is placed in the molds. In the largest plants this is done in an auxiliary mixer travelling on an 
overhead crane, driven by an electric motor — ^really a mixer in itself. It takes the materials 
from the mixer proper, keeps them constantly agitated until the mix is deposited through a 
pipe 3 or 4 in. in diameter to molds in a sand bed. In smaller plants where such equipment 
seems unwarranted, it is common to use a large wooden cask either swung from a travelling 
hoist or mounted on a truck moving about on tracks through the casting area. A workman 
simply turns a crank operating paddles to keep the mixture agitated while it is being run off 
through a spigot into the molds. 

82(i. Mixing Facing Materials. — There are small cylindrical drum mixers specially 
provided for mixing facing materials in small batches. Thorough mixing of the facing mixture is 
highly desirable so that there may be, in the case of special aggregates or the use of color in any 
form, a thorough distribution of the material in order that it will not be spotted or in any way 
uneven either in color or in texture. In the coarse-textured concrete stone, which is becoming 
more popular, the mixture used is comparatively lean in cement and thorough mixing is, there- 
fore, necessary to be sure of cementing the particles in place. These products are afterward 


bnuhed and it in important that the mixing be thorough so as to be mire of embedding the stone 
particles . 

8S. PUciilC. — In a small concrete products plant where perhaps but two hand-operat«d 
block machines are used, sjid where only two men may be employed at this work, it is Dot 
uncommon to operate the mixer for a short period; pile up a batch of concrete in front of the 
two machines and shovel it from the fioor direct in the machines, each workman serving him- 
self at this labor. This, however, is not the way of the modem concrete products plant which 
does away as much as possible with wasteful hand methods, and by increasing thecapacity 
and output of the plant and increasing the quantity of machinery used, lowers the coet of the 
product and in most cases improves its quality. 

83a. Buckets and Hoppers. — In the majority of concrete products factories, 
the mixed concrete is conveyed from the mixer by a bucket travelling by a trolley system to serve 
the trim stone department and a row of block machines. From this travelling bucket the cod- 
crete is deposited in 4 to 5-cu. ft. batches in hoppers feeding to sloping tables just behind the 
block machines (Piffi. 108 and 109), the operators of the machines scraping the mixed concrete 
into the mold boxes with very little lost effort. 

. Fia. 108.— Faclory layout. [Block dipuUnonC in bwikcrouilil.) 

S - BecoBd Boor, or nthcr an inlcrmcduU floor abore euriot roonu where it battery o[ raiien (5) ue locaMil. 

S - Elevator far bom o( mixed (acinc lOBteriaL 

B - BuekM (s part of an elMtric inonarail ayatam) for delirarini miied coDeretii to nuDbiiH* and at banker*. 

H • Hopper* to reeeiva minid eoncrete. 

C - Blodrcar, 

In a large factory, where there are several block machines, and a large-dimension stone 
department, there is a battery of mixers at an intermediate level between the first floor and the 
second floor, so arranged that the raw materials come in by conveyor belting at the level of tho 
hoppers over the various continuous mixers. The mixed concrete is fed into buckets which an' 
hooked to n monorail system operated electrically so that any workman anywhere in the large 
molding room can have a box of mixed concrete delivered to him suited to his special work, 
by means of tracks and switches controlled from the starting point. The empty buckela are 
then returned to be refilled. This is an elaborate system and on expensive installation. Fen- 
existing factories probably have an output to warrant it. In a factory making but one typr 
of units a conveyor belt brings the mixed material from mixer to machines. 

The mariiet offers machinery which couples the block machine with elevating equipment 
and delivers concrete direct from the mixer to a hopper in the top of the machine and drops it aa 
required into the moid box under the tampers. Such equipment is usually coupled with machine 


tlietr ^ian cxi d m ataka i Bt^. to find thai E n ncTe w if iumdted cfaia&y in whee)lxTTDW&. One 
uf tbe lusMt wet-cftEt aumv faetatiee- in the ^Bt. anij diif of the laifieBt fantiineE nmlfhir dir- 
EUHK' in t^ MiddtF Wesi. nHto uw of wlieelbaTTDWE in tranaportinfi the 
o the idoIcIe. Id tbccase of tlirwm emicrate, fmirpaife KTe caziied in a Iibtidii . 
Tlip ntiutioai is, quite diSerani. havrerer, in v&riniE^ InciklmeK in nqwcl tn labot. and thr aitUB- 
ticm 'm aieo afieeted hy the fsci that, in & itaxtry mnirmr dimoision stone whicb k to aell in 
i-ompetituD with natural stoDV. the actual IuIkit in uotdini:. tanqiinf;. and firwh inp ie much 
wan pel cubic foot of concrete thaii it. tlie ardmiir^- standard imiduct made in mBChins. The 
l>ulk of tbeconcTM*' handled i&. thei«fan;. of I^f^ ironaequenee than wbsn' standatd uiitnan' tiir 
ehief productE. 

Be. PaUttt. — The paliett used in standard block machines and brick mactuuet^ 
urt commaalT of tnri idade — wood and iron. The paltetf stand Tcpeatrd cbanfT^ from wel tii 
drr and are suhjeci to esrere wear. If iron pallet «if- negieeied. tbny liecome coated wHh nisi 
tind concrete bd as to be uaeieBB. It i$ recommended that to keep inm pelletE in prDpcr eon- 
(liiion tlie^- diould be kcqit coated nith paraSne oil. or that tlipy be dipped in a mixtun' uf 

R«l.. Tbr.cimpti--(laiike''t»iutfaeiiualUnMlanplHadaitlK(«ratt^nsliI,>Bilupilad 

ktTuaeue and axk gteaac. when the nmciete con oaailj' be wiped o& at the end of eacli day't 
wurii. liu: e^BOt treatment ie rccomiaended for the met^l parte of the block machine. An- 
(itber matudactUTCT euggeste thai iron pallets be dqiped in a Boiuiion of 1 part lard oil and 1 
part keroKiw, to keep the pallets clean and to prevent rusting. With wood pallets, tlie 
difficultiffi are from awelline. Eplintering. warping and eo on. It is neceaaari' to use wood which 
IF an little Eulijecl to warping aa paesible, and have the palletE well treated in order to overcom*- 
warp Bcj far ai4 it can be done, it ■£ pointed out that pallete should nol be made of one solid 
piece but of stripe not taoK than 4 in. wide, with slightly open joints to allow for aome expan- 
aion. Botue manuiaeturerg recommend difqnng wood pallet into hot linseed oil. Difficulty 
fnini sli^t swelling of the pallets ie not inqiortant except in maehineE where paltebi must hi 
vcri- accuiBteh- an ie- not the case with ntOBl tamp machines. T^ere are mocliiiiee, howswr, 
where the palletE must fit witli groat nicety and in such caaee eonaiderabte difficulty has been 
expenenced in icettai^; a pallet which will raaiat the severe treatment. One manufacturer 
flufiering such conditionE finally adopted a combuiation wood and metal pallet. 

Moet bloeki: are delivered on n pallet ri(riil aide tip, face perpendirtilar as they wil) ' 


in a wall. The condition of the pallet in this eaae is not so important. Where, however, the 
block is turned over face down and delivered on a pallet in that position, it is very important, 
particularly for face block, that the pallet be very true, and metal pallets are recommended 
for such work where a true, smooth face is required. For rough-textured block, this is, of 
course, not necessary. 

88d. Bankers. — PracticaUy all dimension stone is made in special molds, these 
molds resting on heavy pUnk pallets, supported in turn by bankers which may be of reinforced 
concrete to minimize vibration. Such bankers are used in factories where dimension stone is 
made by the dry-tamp process. For work of this kind, the placing of the concrete is much 
slower than with wet-cast work, as the facing materials have to be built up vertically 2 or 3 in. at 
a time on such faces of the stone as are to be exposed, besides placing on the bottom, and the 
backing must be tamped in as the facing is brought up. Thus for the convenience of the 
workmen the mold is placed on bankers at a height which is convenient for his work. 

84. Caring. — In from 20 min. to 1 hr. after water has been added and the mixing of con- 
crete completed, this mixture must be placed and it must not only be so handled subsequently 
as not to disturb the hardening process but it must be kept in a condition which will aid that 
process. Conditions for curing must be such that the product will not be rapidly dried, yet 
as temperature influences the hardening — ^heat quickening it, cold retarding it, and freezing 
interrupting the hardening process for the period in which the low temperature continues — it 
is important that these things be considered in caring for products when they have been molded 
or cast. 

The problems of curing do not present themselves as so serious a matter to the manufacturer 
of wet-cast stone as to the manufacturer of dry-tamp products. In the wet-cast work where 
there is already an excess of moisture, sufficient heat is practically the only essential, with con- 
ditions which will prevent too-rapid drying. In tamp products, and for the most part in pressed 
products where the moisture entering the mixture is only just about (or even a little less than) 
that actually required for thorough hydration, it is very important that none of this moisture 
be permitted to escape before complete hydration. 

In curing wet-cast stone, block, and brick made in gang molds on cars, one method is to 
move these cars on tracks to curing tunnels which are heated by steam. The tunnels are made 
of concrete and just high enough to admit loaded cars. They are heated to give a rapid hard- 
ening of the concrete. The cars are usually removed 24 hr. after they arc placed in the tunnels, 
the molds are taken down, and the products are carefully piled under sheds. The molds are 
oiled, set up again on the cars and returned to the mixer to be refilled. In other wet-cast work, 
where gang molds on cars are not employed, or where other molds are used with a wet mixture 
on a large casting floor, it is common to leave these molds in place until the product is suffi- 
ciently hard to be handled. The molds prevent a rapid escape of moisture in the early stages 
of hardening, and particularly is this true in sand cast work, where the sand is always damp 
from having absorbed the excess moisture of the mixture. 

84a. Natural Curing. — The recommendations in the old '^ Standard Practice'* 
of the American Concrete Institute with respect to natural curing are usually regarded as sound. 
They are as follows: 

Natural Curing. — The concrete producUi shall be protected from the sun and strong eurreota of air for a period 
of at least 7 days. Throughout this period they shall be sprinkled at such intervals as ia necessary to prevent dry- 
ing, and maintained at a temperature of not less than 50° F. Such other precautions shall be taken as to enable thr 
hardening to take place under the moet favorable conditions. Products must not be removed from the yard until 
they are 21 days old. 

Where products are cured in this way, it is necessary that racks or cars be used so th.nt 
block on the pallets may be piled up in tiers. As standard practice requires that producU U> 
Hprinklcd for 7 days, it is obvious that there must be curing shed space for 7 days* output ami 
a very large yard storage space in order to keep products until 21 days old. Building rcgula- 


lions in numerous cities require that products be at least 30 days old when cured in this way, 
without regard to the crushing strength. 

Inasmuch as this method of curing would require the use of a very large number of cars, 
on which to tier up the products, it is common, when natural curing is used, to apply the rack 
system of storage. It will be obvious that this can be used only in rather small plants. The 
racks are usually built of 2 by 4's, each rack 16 to 20 ft. long. One rack can thus readily 
accommodate 26 blocks 8 by 8 by 16 in. These racks can be piled, about four high. Four 
rows of racks will thus accommodate about 400 blocks. 

With natural curing, the products are either moved on cars and placed upon racks, or are 
carried on pallets direct from the machine to the racks. This latter method can only be used 
in the smallest plants because the labor of carrying the block the distance required by racking 
in a large plant would make the cost excessive. 

Sometimes products are made which are too large or too heavy or of too awkward shape 
for removal to curing sheds. Until hardening has progressed to a considerable extent precau- 
tion should be taken to see that these products are kept under suitable conditions to attain 
strength. When left in the molding room, the products should be covered with wet cloths 
and the cloths kept wet. This applies particularly to tamp products where the molds are 
removed a day or so after manufacture. Sprinkling may be done systematically and thoroughly 
with a nozzle, which gives a fine, well-dififused spray. The nearer the spray approximates a 
floating mist the more thoroughly it will do the work, reaching all the surface of products 
stored in tiers on racks or cars. 

Where the products are removed to sheds and cured in the natural way, it is obvious that 
considerable labor will be required to use the hose on the products and it would be difficult to 
use the hose in any effectual way. It is common, therefore, to install permanent sprinklers 
in the curing sheds. 

In an Eastern factory turning out high-class products from wood and plaster molds, the 
time for the products to remain in the molds is from 5 hr. for small units up to 24 to 48 hr. for 
the larger and more complicated pieces. Until the molds are removed it is not necessary to 
apply additional water to prevent the escape of the moisture contained in the product because 
there is considerable water used in the mix. As soon as the molds are removed, sprinkling 
begins, using a fine spray. This factory, which is of old construction, has wooden columns 
about 10 ft. apart in each direction. Pipes have been placed so that there is a water outlet at 
every column. A man is kept busy all day sprinkling the products and another man continues 
the work at night. Great care is taken to keep the products moist until they dry out evenly. 
In some ornamental work, manufacturers frequently make use of total immersion of prod- 
ucts to cure them. It is common in such cases to cover the product with wet cloths just as 
soon as it has been removed from the mold and allow it to remain in this way until it has at- 
tained sufficient strength to be handled and placed in the tank. Other manufacturers do not 
use the immersion system but simply use the wet cloths. 

846. Steam Curing. — General practice in steam curing makes use of a wet steam 
and a low pressure to create a dense warm fog with all the moisture which can be introduced at a 
temperature in the curing rooms between 100° and 130°F. While common practice in the field 
does not warrant the use of high-pressure steam and while investigations with high-pressure 
steam in curing have not gone far enough to suggest its adoption on a commercial basis, there 
has been some investigation tending to show that steam under pressure up to 80 lb. can be 
used with success. This steam has to be employed in steam-tight compartments. The ordi- 
nary curing rooms of the concrete products plant are not steam-tight. They are wide enough 
to accommodate the cars which usually run on 24-in. track and high enough to permit four 
decks of standard block to be piled on the cars, leaving some room for the accommodation of 
special products to which it may be necessary to give steam-curing treatment. The curing 
tunnels of the plant are usually about 60 to 90 ft. long, built of concrete block, with arched or 
" A-ehaped " ceiling — preferably a ceiling made by applying Portlant-cement plaster to a ribbed 


reinforcing mesh. The ceiling is built so as to carry the condensed moisture of the room 
to the sides and away from the fresh products which dripping would damage. Curing 
rooms are sometimes built wide enough to accommodate two or even three tracks, so laid out 
that in the case of very wide products to be admitted to a curing room, only the middle track 
can be used, allowing plenty of room for the projection of the products at the sides. At one 
end the curing rooms usually open into the molding department as convenient as possible to 
the machines supplying the greatest number of products to be cured, and the other end fre- 
quently opens into a passageway connecting with the yard, or to the yard direct. 

The construction of the doors has given concrete products manufacturers considerable 
difficulty from time to time, due to the fact that metal doors rust, unless kept in perfect condi- 
tion, wooden doors swell with the steam on one side while they remain dry and of their original 
stxe on the other side, and canvas curtains are very short-lived. These curtains roll up and 
are fastened at the sides, usually, with carriage buttons. In using them, allowance has to be 
made for shrinkage. Galvanized sheet metal for doors lightly framed with wood, or small 
angle irons, have given satisfaction. 

The following recommendations of the American Concrete Institute with respect to steam 
curing were made some time ago but are still considered as representing good practice. 

The products sluJl be removed from the molds ss soon as conditions will permit and shall be placed in a 
steam-corins chamber containing an atmosphere of 8te«m saturated with moisture for a period of at least 48 br. 
The curinc chamber shall be maintained at a temperature between 100* and 130"F. The products shall then be 
removed and stored for at least 8 days. (This does not apply to hich-pressure steam curing.) 

From an excellent discussion of the proper use of steam in curing concrete products, the 
following by W. M. Kinney in Concrete is quoted: 

The principal object in curins concrete products with steam is to accelerate the hardening by means ol heat 
without endangering the concrete through loss of moisture by evaporation. Saturated steam will provide not only 
heat but sufficient moisture to insure against injury from drying. 

In the early history of concrete products manufacture it was customary to use exhaust steam for heating the 
curing chamber, but as a sufficient quantity was not always to be procured, the natural resort was to use steam direct 
from the boiler. The records show that in many cases large quantities of good concrete came to grief due to its 
drying action, which was not at that time explained. 

Especially difficult is the maintenance of a sufficient quantity of steam in a boiler at low pressure to best 
sufficiently any number of curing rooms. Coupled with this difficulty is the danger of the pressure rising con- 
siderably above that necessary for proper curing. With this in view we have been recommending steam under 
pressure, that is, around 30 to 45 lb., provided it be admitted through water. 

The most satisfactory way of admitting steam in this manner is through a perforated pipe embedded in a 
trough of water running through the center or along the sides of the curing chamber. The floor should be so sloped 
that any water or condensation on the products or on the walls of the curing chamber will be returned to the trough 
for re-evaporation. In this manner the trough is automatically kept full of water and we have yet to record s case 
of trouble when this method of curing wss employed. AU of the heat which the steam contains on being emitted 
is taken up immediately by the water and the result shown by evaporation. The temperature of the water is at 
the boiling point due to the fact that steam is continually being forced through it and what heat is taken up by the 
water is used in evaporating the water. This water evaporated at 212^F. is just ss useful in warming the room ss 
is the steam at the same temperature. 

To explain the drsring action of steam under pressure when admitted to the curing chamber, let us assume that 
we are taking steam from a bailer operating at 30-lb. gage pressure. The temperature of the steam in the boiler 
is 252* atmospheric pressure. It is, of courae, understood that steam under atmospheric pressure and having a 
temperature of 252* is in an abnormal condition which we technically call superheated. This steam is in a rimilar 
condition to that which would be obtained if water vapor at 212* were heated up to 252* away from contact with 
water. Nattirally the first thing that rteam in this condition does is to avail itself of the first opportunity to reach 
normal conditions and the most ready way i^ to absorb water from anything in its vicinity which is so possessed 
The result is a drying out of the concrete products which happen to be stored in the vicinity. 

Another method of introducing steam into curing rooms is described by A. E. Cline sj* 
follows : 

The simplest way is to have a main pipe over the top of the curing-room doors, then from this lead a icparatc 
l>i-in. pipe to each room. Run this along one of the walls close to the floor but with slant enough to drain to the 
farther end. Join each length with a T. having the third hole of the T reduced to H in. and turn this at right angles 


to the wftll. The W-in, pipe can be reduced to 1 in. at half the length of the room, and thia on a supposition that 
the curing room is 80 ft. long. If using steam for power, have this main pipe connected with the exhaust during 
the day and with the boiler at night. If the boiler pressure at night is greater than 10 lb. it is best to have a 
reducing valve in the main line as high pressure is not wanted. 

85. special Molds. — Although the commercial market affords a wide range of machines for 
molding various concrete units and a large number of special molds — ^not only for siUs, linteb, 
and so on, used in building, but for various ornamental objects and special architectural pieces 
— the progressive concrete products manufacturer no longer considers that his plant is fully 
equipped until he has a department for making molds to meet the demands of architects in 
turning out dimension stone according to speciid designs as may be required. 

85a. Wood Molds. — The material most commonly used in making these molds 
for dimension stone is wood and a large plant will have an extensive wood-working department 
with power saws, planers, etc., for cutting down the labor cost. By far the greater part of 
dimensiouHstone work in most factories will be made in wood molds — ^preferably white pine. 
This will include sills, lintels, belt courses, cornices and so on. In general, for a plain piece of 
work the mold consists of side planks resting on a pallet with end pieces fitting inside (see 
Fig. 110). 

All pieces have to be carefully finished, allsmall holes or cracks filled, ordinarily with 
plaster, the entire work shellacked and then oiled before use. By clamping the side pieces 
firmly, the end pieces are held in place on the outside against cleats. 
When the facing mixture has been placed on the bottom of the mold 
and up the front side and part way on the ends, as the case requires, 
the backing follows. When it is tamped all the way up, the work is 
struck off at the top and a plank very carefully bedded on top by means 
of a layer of bedding sand. This plank is then clamped in place, the 
clamps passing over the bottom pallet and the entire work is turned 
over so that when the mold pieces are released the stone is right side up, 
having been tamped in place in an up-side-down position. It is not fxo. no.— Cross-aection 
necessary to use the bedding sand on very small products which will bed °^ simple wood mold, 
readily on a smooth plank surface. 

In wet-cast work in which the molds cannot be removed immediately, it is common to 
use a large casting floor smoothly finished with concrete. This is shellacked and oiled as an 
ordinary mold surface would be treated and on it are set up side rails for plain work, with the 
necessary insert pieces and dividing partitions to produce the plain units in necessary lengths. 
Sometimes for greater convenience the bottom of these molds is provided in a bench or table 
with a concrete slab top. 

Where long side rails are necessary channel irons of proper widths can be used to advan- 
tage. Properly cleaned and oiled they will give long service and the principal thing to recom- 
mend them is the fact that they do not warp as wood does frequently unless the grain is very 
heavily filled and the surface shellacked and oiled, nor do they spring out of line from the weight 
of the concrete. This is something that has to be carefully considered in long work, where 
even a slight wind due to the springing of the form frequently prevents the use of the stone on 
nice fitting work. 

The market affords a great many standard molds made of metal for various ornamental 
pieces and standard architectural units. Manufacturers who are catering to discriminating 
architects will not depend upon standard units, however, but will be prepared to meet the de- 
signs of architects. 

856. Plaster Molds. — Plaster molds are very extensively used in concrete stone 
manufacture. In fact, plaster is used not only in making molds but in making models, and 
not only in plaster molds themselves but in making molded inserts for wood and metal molds, 
l^ig. Ill shows how, by means of a template of thin metal, stiffened by a wood frame operated 
on a smooth oiled surface, it is possible to make moldings for various purposes in mold manu- 


facture. The tempUM is guided by a straightredge and moves on the table, pushing aside thr 
Hott plaster, except in the desired section described by the template. Similar work Tequirint; 
curved outline ia handled by mounting a template at the end of a pivoted arm at such a length 
OS to describe the required arc. The template then has a circular movement. 

Plaster is readily used in making models from architects' drawings. When this is doaf 

the plaster is cast in a lar^e block from which the model can be carved with knives and suitable 

chiM-la after the details have been outlined with a pencil on the various faces of the block. The 

plaster model is then shellacked and oiled and the mold made 

over this model, the mold also being made of plaster. 

Flat panels in moderate reliH without undercut may be 
cast in draw molds. The model should be framed with wood 
strips or clay "fences" to control the plaster. It should then 
be given two coats of shellac (orange) and when dry, should be 
greased, using a mixture of 1 part Bt«arine and 2 parts kerosene 
(combined hot). If a clay model is being reproduced, grease 
with lard oil. Sift plaster into a pan half full of water, until 
plaster lies about an inch below the surface of the water. Stir 
thoroughly. When it has thickened to a creamy consistency, apply to model, going over 
the entire surface thinly at first and jarring the model which is best supported by a rigid 
frame. The jarring eliminates bubbles and pinholes. Then pour in the plaster, reinforc- 
ing as may be necessary with strips of burlap, excelsior or wood frame for heavy work. Jars, 
urns, Capitols and similar objects must have piece molds, the model surface being divided 
by clay fences against which the plaster is applied. When a section of plaster mold is hard, 
the fence is removed and notches arc cut in the plB8t«r edge to key the adjoining mold aec- 

Fia. tl3.— PlHter moldi. Ttwduk modded p«rt i« the plulcr mold. The put* A and fi am tba Snt two 
pteeca of m idA«t«r mold. 

tion (see Fig. 112). The edges are then shellacked and greased for ready separation of 
the mold parts. When the parts are set up for casting the concrete, the mold previously shel- 
lacked and greased is held by rope or by chain and tumbuckle, or by clamps, as the siie, 
weight, or shape of the work may necessitate (see Fig. 113). 

Ue. Glne Holds. — The use of gelatin or glue molds is necessary in all work 
where there is intricate undercut in the model to be reproduced, it not being possible to remove 
a rigid mold over these undercuts unless a plaster mold, for instance, is made in many pieces 
to join at the undercut and thus pull away. When a glue mold is to be made, the model it 




Kreaaed and rovcml, fir^it with pnpcr a,rnl tlien with modeling rlay to the thickness necessary 
fur thi' thickncKs of thi- glue mold. This clny covering in then greased and plaster is ttpplied 
over it, to [urm a kIioII with a hole or several holes at the top with air vents at various 
hcixhts. This in illustrat^Kl in Fig. 1 14, the lion head being the model to be reproduced. The 
space immediately around it is first filled with clay and over this is the phtster shell with 
a funnel at the top. When the plaster mold is hard it is removed and the clay and paper 

m.— Pluler mold with plank pallets clamped o; 

ult. befor. 


cleaned from the model. The plaster shell and mode! i 
shell is then fitted in place over the model and the b[ 
is filled with the glue, the vent holes being stopped t 
ready handling, the plaster shell is usually divided : 

re shellacked and oiled again and the 
ace between the model and the shell 
ith clay as the shell is filled. For 
a number of pieces, together with the 
glue mold, the division being made with a knife. Glue should be of the best quality and should 
he melted in a double receptacle with very little water used in the vessel with __ 

the glue. Over-heating takes the 'life" or elasticity out of the glue. 

86d. Combinatioii Molds. — The manufacturer who has a w 
developed model and mold department will have workers in wood, in clay, f 
in plaster, and men also familiar with the making and use of glue or gelatin 
molds. As the work progresses, a resourcefulness in meeting special require- ^°- ^^~^f^ 
ments will lead the manufacturer to make combinations of various mold- 
making materials as, for instance, plaster inserts in wood molds, and small glue molds in con- 
nection with plaster molds to take care of small areas of undercut in the model to be reproduced. 
86e. Waste Holds.- — Ornamental pieces, especially when there is to be consid- 
erable duplication and rapid work is necessary, are sometimes made in so-called "waste" 
molds of plaster. If the model is intricate, a glue mold is made first and in the glue mold a 
glue model is cast. From the glue model as many duplicate molds are made of plaster ss there 
arc pieces to be cast. Wben the concrete has become hard within the plaster mold, the plaster 
is cut away and the concrete surface cleaned. Panels, without undercut, are reproduced in 
a similar way without the necessity for first making a glue model. 

86. Sand Holds and Casting in Sand.— Sand-molding of concrete has not been in eiten- 
sive use except by a few manufacturers particularly in the East, until recently. Patents 
covering important features of sand-molding are just about to expire. 

For ordinary' work the sand, or the mixture of sand and stone dust with a little loam or 
other ingredients to make the sand particles adhere, as in iron foundries, is used in large beds 
on a big casting floor. The sand is packed around models; the models are bo made that they 
can be withdrawn at the top and the molds are filled with a very thoroughly mixed concrete 



[Sec 2-^: 

at a flowing consiatency in which there are usually no particles larger than H ^ &&<! most of 
tho aggregates much smaller than this. Less water is used than the consistency might suggest. 
The ideal mix has no free water. The nuxture is constantly agitated after leaving the mixer 
proper so as to prevent segregation of materials. Mixing frequently is continued for from 10 
to 15 min. which greatly facilitates the smoothness of the flow. Ordinarily, the concrete is 
(Urposited through a spout and some means adopted (as for instance holding a small board near 
the bottom of the mold) to prevent injury to the sand mold by the heavy stream of concrete. 
To preserve the edges at the back of this stone, which is usually the surface upward, and to 
pf'rmit subsequent troweling of this upward surface, it is common to use wood strips placed in 
the sand to give a more stable edge against which to work. It is customary to fill the sand 
molds throughout one floor area; allow an hour for the very wet mixture to stiffen and settle 
find then trowel the upper surfaces, filling in a little additional concrete on the backs of the stones 
where it has settled. 

When balusters, capitols, and similar pieces are to be made, having no large flat surfaco 
to which the model parts will ''draw'' on the 
upper side of the sand bed, it is necessary to use 
the flask method. A box for each of two or more 
(Kirtions of a pattern supports the sand for a sec- 
tion of the flask, llic boxes are assembled to 

Arch sfone- Twt? or three pftct pafttrn 
depending on dfrection orl/nstw' 

No*-*.- Ineasepaffwm 
n arawn from iand as 
Sftcwn ty arro¥¥ 'A' t*ooi 
bo* IS >Xft rtQwtrtd ar'di 
p(x**fn sphf mthnt 
party g$ jh^^ft 
by sohd I rf0$ 






^0*» - m ant am 


■ spilt OS % 
ttfd hn*s and wocdbot >$ used 

bv arrow a if IS spilt OS s^'on<^ ty 
aotttd lines and woq6 ' 
fo take care of cheek 


Firj. 1 lA. — flplltting pattern for making land mold. 

Fio. 116. — Two wajrs of splitting a pattern. 

coinplpto a mold, llie sand is sometimes mixed with a very small percentage of plaster to 
Kivo stiffness to the mold. 

A great deal of the skill in successful stone manufacture with sand molds is in making thr 
iiKKiols, around which the sand must be packed. Take for instance a stone like that illustrator 1 
iti I* ig. 115. The molded surface at the right b to be down and the mo<lel drawn from the sand 
in the direction indicated by the arrow. It is necessary, therefore, to make a four-piece pattern 
MM indicated by the numbers 1, 2, 3, and 4. The parts are held lightly when first made by nails 
lis shown in l^g. 1 16, which shows two ways of splitting a pattern somewhat similar to that in 
Fig. 1 15. The nails are loosened when the pattern is packed into the sand. Note the small 
|mtt<'m part at 4 in Fig. 115. It is the custom among many manufacturers to eliminate tho 
undercut in such a small part in making a pattern, it being cheaper in such a case to let the stone 
('Utter put in the undercut after the concrete is hard. If a great number of these pieces were to 
l>e made, it would, in most cases, be cheaper to make the pattern complete to avoid so much 

There is a tendency in practically all sand-cast concrete, for some of the sand in the sur- 
roimding bed to be held by the cement and so leave lumps and uneven places on the casting. 
S\irface treatment of sand-cast concrete stone b, therefore, necessary, in most cases. 

87. Surfaces. 

8Ta. Face Design in Standard Units. — Design of units is a matter best left to 
the judgment of architects who are specialists in such matters. Some eariy manufacturers of 
block machiner)* and of concrete block were led astray by the ease with which the face of a 
block could be cast to imitate anything. Fare plates were supplied to imitate pitch-face stone, 
in>bhle stone, bush-hammered stone, tooled stone and with all sorts of pands and border?. 
Not satisfied with this, manufacturers of block and special stone, still drunk with the plastic 
noMibilities of iio tractable a material, impressed it with the designs of rubber matting and 

rd-steel ceilings. These errors the industry b outgrowing. The most persbtent crime 


against good taste is a rock-face block which is bad chiefly because it does not imitate success- 
fully one of the least desirable types of natural stone. This much should be of record in this 
connection. If block nuLkera will now devote, as many of them are doing, as much thought to 
making concrete look like itself as their predecessors have to making it look like the least desir- 
able natural stone, a great future seems probable for the manufacturers of the material. 

In making special units, the manufacturer will do well to follow the designs which archi- 
tects work out for him. In making standard units, he will more frequently have the architects 
on his side if he eliminates face designs and makes a plain unit whose title to beauty is in the 
honesty of its appearance, in the tones and colors of its exposed aggregates, or in the light and 
shade of a rugged texture. 

87b. Facing Materials. — Most concrete block and special-dimension stone is 
not of the same mix throughout. When such stone is made of a dry-tamp mixture, it is a simple 
matter to face it, on the surfaces which are to be exposed, using a special aggregate which will 
give the desired qualities in color and texture in the finished product. The facing mixture can 
he backed up with plain concrete. When stone is cast in sand molds — in fact^ in the manu- 
facture of most wetrcast stone — the conditions in the work necessitate that the concrete shall 
be the same all the way through. In such work, therefore, facing mixtures are not used. There 
are some exceptions to this rule in the use of commercial equipment for the manufacture of con- 
crete block in gang molds. One exception is in the coating of face plates with a thin film of 
glue on which, after the glue has become sufficiently thick and sticky, the facing aggregate, 
with no cement, is sprinkled so as to form a complete layer over the face plate. This plate, 
placed in the bottom of a mold is filled with a wet mixture of concrete. The water loosens the 
glue and the facing aggregate bonds with the wet backing. Another method of facing a wet- 
cast material is in gang molds iq which the product is made face up. The molds are filled with 
a slushy mixture, not quite to the surface, then the facing mixture is spread on after the backing 
coQcrete has partially hardened. This is troweled into place on the surface In general, 
with regard to facing mixtures, it is common to make them of a 1:2 proportion of cement and 
a special aggregate, or 1:2^ or 1:3, depending hot only upon the grading of the mixture but 
upon the effect desired in smoothness or roughness in the product. Materials commonly used 
include special sand, in white, buff, and so on, crushed marbles in the more choice work, crushed 
granite, trap rock, even crushed cobble stones (which are within the reach of almost anybody 
in a glacial country), and crushed limestone. Micaspar crystals, so-called, are used by many 
manufacturers in producing a gray granite effect in either dark or light shades. Other facing 
materials are on the market containing mica which gives a sparkle and life to the finished product. 
It has been very common for most manufacturers to use fine sand for facing material, particu- 
larly in tamp work, and to make a special effort to get a very smooth, fine face on their products. 
Some architects are encouraging work in another direction by the interest which they have 
shown in products having a rough texture. This is secured by using coarser aggr^ates not so 
well graded — that is, not so much fine material to fill up all the spaces — and with just as little 
cement as can be depended upon to bind the aggregates thoroughly in the face. It is common 
in such work to use 1:3 mixtures and to use facing aggregates as large as ^ in. The results 
which can be produced in this way are limitless and their effectiveness depends largely upon 
the taste and judgment of the manufacturer. The use of yellow marble or black marble 
(crushed trap rock is frequently used in place of black marble), red granites, and so forth, lead 
to possibilities in colors in concrete stone which only need experiment to prove. 

If a smooth block or polished finish is desired, the product should be made on as smooth 
a surface as possible so that there will be few projections to work down. Where a well-graded 
aggregate of a polishable material is used — as, for instance, granite or marble — and this material 
is well distributed over the surface area with few spaces between, it is possible to polish a con- 
crete product just as the granite or marble itself is polished. A recent development in the manu- 
facture of concrete stone lies in the direction of obtaining a smooth surface by means of a sheet 
of very heavy paraffined paper of glossy finish, which is placed in the bottom of the mold of 
whatever kind is used; the facing material being placed in on top of this and the backing tamped 



[Sec. 2-87r 

behind it. This gives a very smooth surface which requires only a minimum of nibbing; to 
give an excellent finish. 

Aside from the special methods and devices in obtaining various products which have 
already been described, facing mixtures are used chiefly with tamped products. In such work, 
the facing mixture is generally a relatively dry mixture, placed on the bottom or face-side of 
the mold box, and the backing tamped in behind it. When it is necessary to fill up the facing 
material on the side of a mold of any kind, this can be done by piling it up 2 or 3 in. at a time 
and filling in the backing behind it, or it may be done by the use of a thin piece of metal of a 
sise equal to the side of the mold which is to be faced. With the use of this dividing plate the 
facing mixture is placed on one side and the backing on the other, the plate being gradually 
raised as the two are tamped together to form a bond. This facing mixture, while relatively 
dry, must neither be too dry nor too wet. It must be just sticky enough to hold its position when 
pressed into shape. 

Whenever a facing mixture is used, it is desirable to finish the work in such a way that the 
special aggregates of whatever nature are used, are exposed to lend color and to give better 
texture to the work by the removal of the cement which covers the surfaces, leaving the cement 
in the matrix to bind the aggregates in their position. Various methods for finishing the stone, 
not only that which is faced with a special mixture, but also that which is of like character 
throughout, will be considered. 

87c. Colors. — Great care should be exercised in the selectioa of colors. They 
cannot be used in rich mixtures without destroying a great deal of the binding value of the 
cement with which they are mixed, and the strongest colors cannot be obtained unless rich 
mixtures are used. For both these reasons there has been some tendency to discourage the 
use of mineral colors in cement mixtures. 

J. H. Jackson, authority on colors, writes in Concrete as follows: 

Mineral colon of the highest degree of purity are the only ones to use in coloring cement. The pernianenry 
of shade or color obtained depends upon the elimination by the color manufacturer of anything in the color that 
the oement itself will destroy. Few contractors realise that the more intense and brilliant are the colors, tiie more 
quickly they fade, and, in more instances than one, help to disintegrate the concrete, for none of the mineral colors 
useful in oement work, is found naturally brilliant and the addition of chemically prepared colors or the treatment 
of native colors chemically to give them intensity is a positive detriment under all conditions. All true cement 
colors will withstand the acid treatment (I part acid, 5 parts to 6 parts water), scrubbing and troweling, but in 
polishing after setting and trowel polishing before setting, special care should be taken when yellow, green and 
similar colors are used for the metallic polishing is likely to darken the color. Never give a smooth finish to outside 
concrete when color is used. 

The standard proportions for colors generally used are 6 to 6H lb. of color to every 100 lb. of cement. The 
amount of color can be increased slightly if a deeper shade is desired, but you should not use more than 10 lb. of 
color to every 100 lb. of cement, for an excess of color reduces the binding power of the cement. 

A table of color quantities and the results by L. C. Sabin from his ''Cement and Concrete" 
is as follows: 

Colored Mortars 
Colors given to Portland-cement Mortars Containing 2 Parts River Sand to 1 Cement 

Dry material 

Weight of dry coloring matter to 100 lb. cement 



2 1b. 

4 1b. 

Lamp black 

Prussian blue 

Ydlow ochre 

Ultramarine blue.. 

Burnt amber 

Venetian red 

Chatt. iron ore.... 
Red iron ore 

Light slate 

Light green slate 

Light pinkish slate. 
&1ate, pink tinge. . . 
Light pinkish slate. 
Pinkish slate 

Light gray 

Light blue slate 

Light blue slate . . . . 

Pinkish slate 

Bright pinkish slate. 


Dull pink 

Blue gray. 
Blue slate 

Dark blue slate 
Bright blue slate 
Light buff 
Bright blue slate 

Blue slate 

Dull lavender-pink . . 

Light dull pink i Dull pink 

Light terra-cotta . . . . | Light brick red 
Terra-cotta Light brick red 


Coloring by abeorptioii is described in the Concrete by Adolph Schilling: 

To color in reds and in browns use sulphate of iron in a solution of 1 lb. of sulphate to 1 gal. of water; for 
greens 1 lb. sulphate of copper to 3 gal. water. The older the concrete the longer the bath must continue. It 
must be borne in mind that the coloring stops to a very great extent the hardening of the concrete. Tho'efore, 
it ia neoeasary to permit the concrete to attain such strength as is necessary in the ornamental work before im- 
mersing it. Immersion may continue for a few minutes or for several days, depending upon the age of the con- 
crete and upon the depth of color desired. The effects may be varied by using an aggregate which is not highly 
absorptive so that this stands out while the matrix surrounding it is colored. Effects can be heightened in elaborate 
deaigna by "picking out" certain parts, coloring them with a brush with cement stains. 

Rough-textured work is sometimes beautified by having stain stippled on or brushed over the high points. 
The manufacturer, of course, will always have to use his judgment in the application of these colorings. The great 
danger is in overdoing. 

Aside from the use of white cement to obtain light colors in concrete, it is possible to 
make an ordinary gray cement surface lighter by rubbing with a concrete brick made of fine 
materials. The work should be kept wet while being rubbed. 

A method of obtaining white surfaces is described by John Oursler, in Concrete^ as follows : 

A wash made of 1 lb. of concentrated lye, 4 lb. of alum and 5 gal. of water, with enough cement added to make 
the wash of a good consistency for spreading with a brush, has been used to give a white surface. 

87d. Spraying. — One method for removing the surface film of concrete block 
and special stone where there is a special facing mixture, is the spraying method, described in 
ConcreU by the early user of the method, E. J. Thompson: 

Immediately upon removing the block from the machine, place it where there will be a good light on the face 
and Bfpmy it, using a fine vapor spray, such as is used in spras^ng fruit trees. The outlet holes in such sprays are 
about the nse of an ordinary pin, and, for the best results, should be used in connection with a water pressure of 
40 lb. or more. This spray nossle attached to a length of ^i-in. hose is all the equipment needed. The spraying is 
a simple operation. It can be done, after a little practice, by any intelligent laborer. The effect of the spray, 
which lasts only momentarily, is to wash off the surface film of cement and expose the aggregate (see Fig. 1 17) . 
The spraying must not be continued until the surface begins to run or furrow, but just a little practice teaches the 
operator when to stop. 

Some manufacturers prefer spraying their products when they are lying face up; others 
prefer to have them with the face perpendicular. The perpendicular method is the conmionest, 
but the face-up method has a tendency to leave all the cement on the surface, to wash it into 
the pores. One manufacturer who advocates this method urges that it not only exposes the 
Aggi^&te but makes tke face of the product more dense. When this work is done on standard 
concrete block, it is done very rapidly and adds very little to the cost of the block. 

876. Brushing. — ^Brushing the surfaces of concrete stone in standard and special 
units is particularly desirable where a rough-textured effect is desired (see Figs. 117 and 118). 
It is conunonly used where a graded aggregate, with larger particles than is ordinarily used 
in facing mixtures, is employed. The brushing is done, ordinarily, while the product is com- 
paratively green, using a fiber brush with stiff bristles and using considerable water while 
the brushing is in progress. Care must be taken that brushing is not started too soon so that 
the face of the products will be damaged, and on the other hand, the work must be done before 
the product is too hard, or the work becomes expensive. The manufacturer will find that the 
time for brushing depends a great deal upon the conditions of curing. For surfaces which have 
been allowed to become partially hardened, a brush about 4 in. wide made by clamping together 
a number of sheets of wire cloth has been found more effective than the wire brushes which are 
ordinarily sold for this purpose. Care must be taken in brushing not to injure the edges of 
the products. Sometimes it is desirable for the operator to use a small frame or at least a 
straight-edge to protect the edges of the product while brushing. 

Fig. 119 shows three views of turning stand for handling green concrete block to be brushed. 
This stand is used for handling products which are delivered on pallets face down. A block 
on pallet is placed on stand as in the first position. The shelf is turned over, so that the block 



[Sec. a-87/ 

rests on its aide as in the second positioa. The shelf which fiist supported the block is then 
turned donn leaving the face exposed for brushing or other treatment. 

Where steam-curing is employed, it is common to run the block into the steam rooms for 
a tew hours and out again, when the brushing is done, and the products are returned to the cur- 
ing rooms, 

B7/. Rubbing. — Rubbing, as a finish in concrete stone, is commonly used where 
a limestone finish is desired. In wet-cast work when air bubbles and slight imperfections occur 
on the surfaces of concrete stone, it is the usual practice to pour on the surface a creamy grout 

and GDarM-cnuned 

of cement and water. This is nibbed in first with a brush or swabbed on with a cloth, and a. 
few minutes later rubbed in with a small wood block. When hard,Hhe stone is rubbed down 
with abrasives. The rubbing removes the effect of the painted surface. It is common to ust' 
cement brick made with very fine material in rubbing concrete work. Ordinary commercial 
abrasives are also employed. 

For finishing marble concrete which may be cast in large blocks and cut up by gang saws 
into slabs, or cast in thin sections in pressure machines as for floor tiles, the methods arc like 
those in factories where natural marble is handled. The pieces are first smoothed down on a 
revolving rubbing bed, sand and water beinfc 
fed to the grinding surface. Then power- 
operated carborundum discs are employed, 
or a second rubbing bed with a finer grit, 
for a dull gloss surface, or when a polish 
is desired, the work going under hand or 
power-operated mops, employing oxalic arid 
and putty powder in the process. 

VIg. TooUng. — In considering 
the tooling of concrete stone in various sur- 
face treatments now put on it, there might 
be considered all of the methods which are common to the finishing of natural stone. The 
best quality of manufactured stone admits of the same treatments as are given to natural 
Htone, with similar results. The means at the disposal of the manufacturer of high-class con- 
i^rete stone include the ordinary hand work with lasps and chisels, bush hammers, crandalls, 
planers, polishers, and so on. To make pomible the use of tools of this kind in obtaining 
natisfactory finish, it is necessary that the concrete have a uniform texture, with no aggregates 
of unusual hardness; that is, the materials shall be of a like character throughout. The aggre- 
gates used in such work, ordinarily pass a H'Ui. screen. They include crushed limestonp, 


m&rbley granite, and trap rock. While faced products are frequently tooled, it is more common 
to find such finishing methods applied to products which are of like character throughout. 

The use of pneumatic tools in bush-hammering and crandalling concrete surfaces is becom- 
ing more common and resulting in great economy over the use of hand tools. A finish in parallel 
f^Tooves, very common on natural stone, b put on concrete stone using power equipment which 
revolves a gang of thin carborundum wheels mounted together, so that considerable surface 
is covered at one time. Methods of cutting which are common in the natural stone field and 
which are used, though are not so common in the concrete stone industry, include the cutting 
up of large slabs of stone by means of saws which make deep grooves, so that a slab is easily 
split up on the job just before the stone is laid, thus preserving the edges and keeping them 
clean and sharp. 

For fine ornamental details and for finishing undercuts which are not easily included in 
patterns and molds, expert stone carvers are employed in some of the, best plants making a 
high quality of concrete stone. 

87A. Mosaics. — The subject of the surface treatment of concrete would not be 
complete without the mention of the possibilities which lie in the use of mosaics in laiige and 
Hmall ornamental surfaces. These may be glued into a mold on strips of paper where the 
concrete surface is to be flush with the mosaic itself; or the inlay, consisting of tile or bits of 
marble or other stone, may be set in places which have been provided in the products when cast. 
.\fter the product is complete and before the curing period is entirely ended, the grooves or 
spaces left where inlays are to be inserted are thoroughly wet and grouted in to hold the inlays 
in place. 

87i. Efllorescence. — ^Efflorescence is ordinarily considered in connection with 
surfaces because it is a surface disfiguration. The cause, however, goes far back of the surface 
treatment and lies in the fact that the concrete mixture is not so dense as should be obtained. 
Efflorescence is, in reality, a disfiguration which is common to brick, to concrete and to 
natural stone. It occurs, as a direct result of porosity in the material on which it appears. 
An impervious substance is not subject to efflorescence. When a substance is porous, water 
which is absorbed dissolves certain salts found in the material; as for instance in concrete 
block and brick, the salt which is dissolved is principally lime carbonate. A concrete product 
which soaks water like a sponge after a rain, subsequently dries out generally from exposure 
to the sun. In dr3ring out, any salt solution is brought to the surface and left there when the 
water evaporates. The best remedy is in prevention by maintaining a dense mixture. Efflo- 
rescence can usually be removed with a solution of muriatic acid and water although this is 
not always successful. 

87;. Air Bubbles. — ^Air bubbles, like efflorescence, are a part of the subject of 
density. They are formed when there is insufficient care in placing the concrete. A wet 
mixture should be spaded against the forms so that these little air pockets will not form. They 
are also prevented by tapping the molds, or by vibrating platforms, actuated sometimes by 
machinery. The object of such equipment, however, is not so much to avoid the pinholes, 
which become surface blemishes, as to produce a dense concrete by consolidating the mass. 

The removal of air bubbles is frequently accomplished by tooling the surface when the 
concrete is removed, and sometimes it is done by filling the surface a cement paste being rubbed 
in, as is described in connection with rubbed surfaces. 

87^. Crazing. — Crazing, or the formation of hur-cracks in the surface of con-> 
crete stone, is more frequently encountered in wet mixtures than in dry mixtures; is more 
common in rich mixtures than in lean mixtures; and it is said by some manufacturers that the 
tendency to surface crazing diminishes greatly when a well-aged cement is used. The writer 
has never known but one manufacturer using a wet mixture of concrete who claimed entire 
freedom from crazing in his products. This manufacturer uses a graded mixture of trap rock. 
The reason for the hair-checking is in the greater tendency of the surface of concrete stone 
to undergo temperature changes, leaving the interior of the stone comparatively imafifected. 


These surface cracks do not cause any structural harm, as they extend scarcely more than 
^^2 u^* u^to the stone. They are entirely removed in ordinary tooling treatments to which 
much of the wet-cast stone (in which crazing is most common) is subjected. While these 
cracks are believed by some manufacturers to be characteristic of the very rich surface skin of 
cement which forms on spaded mixtures of concrete or on very rich mixtures of fine material, 
other manufacturers find that even in tooling this thin mortar skin on the surface, the erasing 
is not entirely done away with as it may craze again later on. Its only disadvantage is in the 
fact that it provides minute lodging spaces for dirt. It does not seem possible to say just how 
this crazing can be avoided in concrete stone, because it is maintained by many manufacturers 
that two pieces of stone made in the same way, from the same batch and cured in the same way, 
will not show the same results, one piece being covered in a short time with surface cracks and 
the other piece entirely free from them. A manufacturer of concrete monuments says that ho 
obviates surface crazipg by keeping his products buried in damp sand for 30 days after manu- 
facture. In manipulating the drier mixtures of concrete, manufacturers frequently insist that 
workmen shall not trowel the surfaces of products after they are molded, this troweling having 
a tendency to bring the fine cement particles to the surface and result in crazing. 

88. Specifications of the American Concrete Institute. — Newly adopted (1917) specifi- 
cations and building regulations of the American Concrete Institute for manufacture and 
use of concrete architectural stone, building block and brick provide as follows: 

1. Concrete architectural stone and building blocks for solid or hollow walls and concrete brick made in 
accordance with the following specifications and meeting the requirements thereof may be used in building con- 

2. Te^. — Concrete architectural stone, building blocks for hollow and solid walls and concrete brick must 
be subjected to (a) compression and (6) absorption tests. All tests must be made in a testing laboratory of recog- 
nized standing. 

3. Ultimate CompreMtive Slrenoth.—(a) In the case of solid stone, blocks, and brick, the ultimate compressive 
strength at 28 dajrs must average fifteen hundred (1500) lb. per sq. in. of gross cross-sectional area of the stone as 
used in the wall and must not fall below one thousand (1000) lb. per sq. in. in any case. 

(b) The ultimate compressive strength of hollow and two-piece building blocks at 28 days must average one 
thousand (1000) lb. per sq. in. of gross cross-sectional area of the block as used in the wall, and must not fall below 
seven hundred (700) lb. per sq. in. in any test. 

4. Oron Crou-»ectional Areat. — (a) Solid concrete stone, blocks and brick. The cross-sectional area shall 
be eonsidwed as the minimum area in compression. 

(6) Hollow building blocks. In the case of hollow building blocks, the gross cross-sectional area shall be 
considered as the product of the length by the width of the block. No allowance shall be made for the air space 
of the block. 

(c) Two-piece building blocks. In the case of two-piece building blocks, if only one block is tested at a 
time, the gross cross s ec tional area shall be regarded as the product of the length of the block by one-half of the 
width of the wall for which the block is intended. If two blocks are tested together, then the gross cross-sectional 
area shall be regarded as the product of the length of the block by the full width of the wall for which the block is 

6. Abaorption. — The absorption at 28 days (being the weight of the water absorbed divided by the weight 
of the dry sample) must not exceed ten (10) % when tested as hereinafter specified. 

6. Sample*. — ^At least six samples must be provided for the purpose of testing. Such samples must represent 
the ordinary commercial product. In cases where the material is made and used in special shapes and forms too 
large for testing in the ordinary machine, smaller specimens shall be used ss may be directed. Whenever possible, 
the tests shall be made on full-sised samples. 

7 Comprt—ion Te»U. — Compression tests shall be made as follows: The sample to be tested must be carefully 
measured and then bedded in plotter 0/ Pari* or other ceroentitious material in order to secure uniform bearing 
in the testing machine. It shall then be loaded to failure. The compressive strength in pounds per square inch 
of gross cross-sectional area shall be regarded as the quotient obtained by dividing the total applied load in pounds 
by the gross cross-sectional area, which area shall be expressed in square inches computed according to Art. 4. 

When such tests must be made on cut sections of blocks, the pieces of the block must first be carefully meas- 
ured. The samples shall then be bedded to secure uniform bearing, and loaded to failure. In this case, however, 
the compressive strength in pounds per square inch of net area must be obtained and the net area shall be re- 
garded as the minimum bearing area in compression. The average of the compressive strength of the two portions 
of blocks shall be regarded as the compressive strength of the samples submitted. This net compressive strength 
shall then be reduced to compressive strength in pounds per square inch of gross cross-sectional area as follows: 

The net area of a full-sised block shall be carefully calculated and the total compressive strength of the block 

Sec 2-SSl 



«ili hr obtained by xnu1tiplyiu£ thi» area by the net compressive strength obtained above. This total cross coin- 
piraniTT strenctb afaall be divided by the utobb cross-sectional area as figured by Art. 4 to obtain the compressive 
Btrracih in pounds per square inch of gross cross-sectional area. 

When testing other than rectangular blocks, great care roust be taken to apply the load at the center of 
eravity of the specimen. 

8. Ahmarplion TetU. — The samples shall be first thoroughly dried to a constant weight at a temperature not 
to c»e eed two hundred and twelve (212) degrees Fahrenheit, and the weight recorded. After drying, the sample 
ahall be immersed in clean water for a period of 48 hr. The sample shall then be removed; the surface water wiped 
off, and the sample reweighed. The percentage of absorption shall be regarded as the weight of the water absorbed 
divided by the weight of the dry sample multiplied by one hundred (100). 

9. lamnH of Loadino- — (a) Hollow walls of concrete building blocks. The load on any hollow walls of concrete 
bloeka, including the superimposed weight of the wall, shall not exceed one hundred and sixty-seven (1G7) lb. per 
■q. in. of gross area. If the floor loads are carried on girders or joists resting on cement pilasters filled in place with 
slush poncrete mixed in proportion of one (1) part cement, not to exceed two (2) parts of sand and four (4) parts 
of gravel or crushed stone, said pilasters may be loaded not to exceed three hundred (300) lb. per sq. in. of gross 
c f o aa s e ctional area. 

(fr> Solid walls of concrete blocks. Solid walls built of architectural stone, blocks or brick and laid in Port- 
land-oement mortar or hollow block walls filled with concrete shall not be loaded to exceed three hundred (300) 
lb. per sq. in. of gross cross-sectional area. 

10. Girdrrs and Joixts. — Wherever girders or joists rest upon walls in such a manner as to cause concentrated 
loads of over four thousand (4000) lb. the blocks supporting the girders or joists must be made solid for at least 
eight (8) in. from the inside face of the wall, except where a suitable bearing plate is provided to distribute the load 
over a aufl&cient area to reduce the stress so it will conform to the requirements of Art. Q. 

When the combined live and dead floor loads exceed sixty (60) lb. per sq. ft., the floor joists shall rest on a 
steel plate not less than three-eighths {H) in. thick and of a width H to 1 in. less than the wall thickness. In 
lieu of said steel plate the joists may rest on a solid block which may be three (3) or four (4) in. less in wall thick- 
ness than the building wall, except in instances where the wall is ei^t (8) in. thick, in which eases the solid blocks 
shall be the same thickness as the building wall. 

11. Thieknets of WaiU. — (a) Thickness of bearing walls shall be such as will conform to the limit of loading 
given in Art. 9. In no instance shall bearing walls be less than eight (8) in. thick. Hollow walls eight (8) in 
thick shall not be over sixteen (16) ft. high for one 

■tory. or more than a total of twenty-four (24) ft. for Table I 

two storifls. 

(6) Walls of residences and buildings com- 
monly known as apartment buildings not excee<ling 
four stories in height, in which the dead floor load 
does not exceed sixty (60) lb. or the live load over 
sixty (60) lb. per sq. ft., shall have a minimum thick- 
ness in inches as shown in Table I. 

12. Variation in Thickness of Walls. — (o) 
Wherever walls are decreased in thickness the top 
rouFse of the thicker wall shall afford a solid bear- 
ing for the webs or walls of the course of the con- 
crete block above. 

13. Bonding and Bearing Walls. — Where the face wall is constructed of both hollow concrete blocks and 
briok« the facing shall be bonded into the backing, either with headers projecting four (4) in. into the brickwork, 
every fourth course being a header course, or with approved ties, no brick backing to be less than eight (8) in. thick. 
Where the walls are made entirely of concrete blocks, but where said blocks have not the same width as the wall, 
every fifth course shall overlap the course below by not less than four (4) in. unless the wall system alternates the 
cross bond through the wall in each course. 

14. Curtain Walls. — For curtain walls the limit of loading shall be the same as given in Art. 9. In no in- 
stance »hi^ll curtain walls be less than eight (8) in. in thickness. 

15. Party Wall*. — ^Walls of hollow concrete blocks used in the construction of party walls shall be filled in 
place with concrete in the proportion and manner described in Art. 9. 

16. PaHUum Walls. — Hollow partition walls of concrete blocks may be of the same thickness as required 
in hollow tik, tetra-cotta or plaster blocks for like purposes. 






















, , 

• • 






, ^ 







8ECTI0K 3 
CUJffiiatUtTAOST riiAST 

1. Srepantion of CnssbedHrtone AggR^ste. — ^The prepamtioii of erudhed-stoac abrogate 
buF grown to \n' on industry' of sucb size thttt marked refineoiente in raethods h&ve been in- 
tnidueed in recent yearb in the better plants. The general flcbeaie, however, Ls (1) the breaking 
down of the ledge, by one means or another, into pieces which can be readily' bandied and fed 
into cniahing machinery' ; C2) the breaking of these lai^^ pieces in erusb^s of one typeor another ; 
and r3^ the Beparation of the crushed material into various sizes. 

la. Prepatt thm of Site for Q tiai ' iyiiig . — When a lec^ is located for quanying, 
V ^ neeeeaar^- to strip off the overburden in order to expose clean rock and to prevent the work- 
tngs from being filled with dirt and d^ns. Quarryii^ is carried on in moreK>r4eas distinct 
iovelF one above another, the overbunkn being stripped back a distance from the top and the 
i^dge quarried down a convenient depth, ^n^idually working backward into the face. A lower 
t'^rrace will then be started, working backward for a distance, until the vertical face of the first 
portion i<^ encountered. Successive levels may be quarried in this way, the top ledge being 
successively stripped back to greater distances and the lower ledges being worked again in the 
same sequence, it is usually sought to quarr>' inward from the side face of a le4ge whose top 
i^ i\ considerable distance above the skie of the cniahing plant, in order that the rock may be 
brought within reach h>- gravitj'. 

16. Qn aiij f i^g . — ^The general method of quarrying is to bring rock down by 
chaiy^es of explosive set' off in holes dniled in the rock ledge. These holes are put down to the 
deptii to whicii the rock Ls to be spht. The requisite amount of powder is charged into the 
hole, coveretl b^' sand, and fired b>' means of a fuse or by dectrieitj'. In laiger operations 
chancer in a line of drilled holes are fired simultaneously by electricity. Gunpowder is the 
"Tpiosive mostly used, although nitroglycerine and dynamite are often preferred botJi because 
o! the laiger quantity- of rock which can be brought down per batch and also because of th(> 
siiatteiing effect of these quick-acting explosives. Various refinements in minor details arc of 
great importance and have a distinct bearing on the effect of the explosive charge. Even sucli 
an apparonth* small matter as the form of the bottom of the drill hole has a very marked effect. 
When bored witli a hand drill, the hole is triangular at the bottom anil the blast in such a hole 
wH: break rock in three directions. Explosives in a squared bottom hole have a more distinctly 
lateral effect. An expert rock man will shoot approximately that portion of the rock which In- 
desires to bring down. 

It. BdMs. — The jumper is> a driU similar to that used for drilling holes for plu^ 
ajid feather wodc in dimenaionnitone quarries, except that it is larger and longer. It is usually' 
held hy cme man, who rotates it between the alternate blows from hammers in the hands of 
tw<» other men. Chtam tkiUs are long heav>' drills meaauring from G to 8 ft. in length. The}- 
an- raised b>' a workman, let fall, caught on the rebound, raised and rotated a Uttle and then 
dropped again, thus cutting a hole without being driven by hammer. They are more econom- 
ical than jumpers as they cut faster and make iaiger holes. Machine reck drills bore much 
more rapidh* than hand drills and are more economical in most operations for preparing rock 
for ooncrete aggregate, where the work is of sufficient magnitude to justify the preliminary out- 
lay. The>* drill in any direction and can often be used in boring holes so located that they 
coukl not Ixi bored by hand. They are worked either by 8l«am or by compressed air, and may 



be either percussion or rotary. The action of a percussion drill is the same as that of a chum 
drill already described, a piston moved by steam or compressed air being attached to the drill 
in such manner as to make a stroke at every complete movement of the piston, an automatic 
device rotating the drill slightly at each stroke. Rotary driUs may be either shot drills or dia- 
mond drills and they are more often used for prospecting than for drilling holes for explosives, 
inasmuch as in their use a core is obtained which is of value mainly as indicating the strata 

Id, Stone Crushers. — Crushers are of two general kinds : jaw crushers and g3rra- 
tory crushers. The former type is better adapted to small or portable plants, while the latter 
is used in larger operations. A convenient size of jaw crusher for a portable plant is about 
10 by 16 in. This will crush from 50 to 100 cu. yd. per day, depending upon the size of the stone 
to be crushed. 

Both types of crushers have means for regulating openings so that by using a proper open- 
ing together with a proper crushing plate, almost any size of crusher product can be obtained, 
the size being limited by a small opening at the crusher-plate end of the machine. The output 
of any crusher will depend to a large extent upon the plant arrangement. Necessarily also, 
the more finely a stone is crushed the more work must be done upon it and the less the output. 
In deciding upon the t3rpe of crusher to be installed at any plant, it is best to get comparative 
estimates, costs, and tables of weight and output from manufacturers of various types of ap- 
paratus, balancing the advantages of one against those of another and finding the machine bci>t 
adapted to the purpose in mind. Machinery of this kind is constantly being improved and 
changed in t3rpe, so that accurate data representative of the latest practice is difficult to give. 

le. Screening and Grading of Crushed Stone. — As stone comes from the 
crusher, it is carried by some elevating means, usually a bucket elevator, to revolving screens 
fixed over bins. Elevating and screening plants can be furnished in either portable units 
(in which case they are so arranged that they can be readily dismantled for transportation) 
or in fixed units with the machinery more massive. The usual type of screen is a rotary screen 
inclined on its longitudinal axis, screens of various-size holes disposed successively throughout 
its length forming the screen barrel. The stone as received from the elevating buckets is fed 
into the fine screen end. Through the openings in this screen the very fine materials, dust, 
etc., are taken off; and as the stone progresses down the screen barrel, the several sizes fall 
into bins arranged below them, from which they are drawn off into conveyances as required. 
The storage bms vary in size from those having a capacity of 13 tons to those having a capacity 
of 50 tons. In some of the modem types of bins, provisions are made so that a bin may be 
raised to a height sufficient to permit wagons being driven under gate spouts. 

1/. Washing Crushed Stone. — Crushed stone is often covered with a tenacious 
film of dust of which it is very hard to get rid. Although seldom if ever done, it would be 
advantageous to wash stone after crushing and screening, inasmuch as this dust is of such size 
that it is impossible to coat it with cement, and so tenacious that it prevents the cement from 
being in contact with the aggregate. 

Ig. Crushed Limestone. — ^Limestone crushes with a flaky fracture and a con- 
siderable amount of dust. If the finer screenings are to be used, it is well to roll them between 
rollers, inasmuch as this flaky fracture renders them extremely friable and unsuited to the 
production of concretes impervious to water or of high strength. 

2. Screening of Sand and Gravel. — Screening of sand and gravel may be done by hand or 
by machinery. Hand-screening is adapted to small jobs and light work. Power-screening is 
adapted to handling laiiger quantities of material. Screening of gravel or sand containing large 
amounts of coarse material can be done more cheaply by mechanical than by hand means, 
using either revolving screens or fixed screens placed upon an incline. The type of screening 
equipment is largely determined by location and natural topography, and the availability of 
power. Revolving screens are most effective and economical for large quantities, the material 
being conveyed by bucket elevators to the screens and then falling into bins provided with 
' -^ convenient for unloading. 


Large and elaborate plants for screening of sand and gravel are being installed in increasing 
numbera in situations where dredging of these materials from river beds is practicable. Many 
of these plants are now turning out aggregate of exceedingly high quality, the screening and 
grading operation being incidental to the elevation of this material. Necessarily also these 
oiaterials are washed while being screened and graded. 

3. Washing of Sand and Gravel. — Gravel is not. infrequently coated with a tenacious film 
of material which, if not removed, may greatly reduce the strength of the concrete. Sand also 
is not infrequently contaminated .with clay, loam, or slit coatings of organic matter. Such 
coatings are responsible for a great deal of difficulty and many defects in concrete; and their 
removal, while not easy even by washing, is decidedly essential. 

Various means have been proposed for washing sand and gravel. Attempts have been 
made to wash them in piles with a hose but this is alwa3rs difficult and usually impossible to 
carry out properly, inasmuch as the materials washed from the upper layers are carried down 
by the stream of water to the lower portions, where they are rarely dislodged. Another scheme 
is shoveling sand into one end of a V-trough, washing it down with a hose and endeavoring 
to carry off the finer materials in the runoff water. For large quantities of material a combina- 
tion of a trough of this kind with an ejector has been successfully used. A concrete mixer has 
also proven adaptable to this kind of work, water being turned in and the mixer allowed to 
overflow while the drum is in rotation. Necessarily all of these processes are somewhat ex- 
pensive and add to the cost of the aggregate, but where tests indicate that the quality of concrete 
will be seriously affected by uncleanness of sand or stone, it should be undertaken without 
hesitation even at the increased cost. 


4. General Considerations. — The handling of materials, and their storage and disposal is 
of great economic importance in concrete work. Hundreds of tons are removed, loaded, trans- 
ported, unloaded, piled and elevated oftentimes to considerable heights before the making of 
concrete has begun; and all this material as concrete must be rehandled one or more times 
before it is delivered to forms. The engineering problems prior to placement are, therefore, 
highly important; and their adequacy not only determines rate of progress, but their eflficiency 
may decide whether there is to be a quick turnover or a slow one, with corresponding profit or 
possibly loss. 

It is axiomatic that gravity should be employed in such work whenever possible by utiliz- 
ing natural advantages of site to the fullest extent. It should also be borne in mind that there 
is approximately twice the quantity of stone to be handled as sand; and twice the quantity of 
Hand as cement. In planning a job, careful routing and proportionate disposal of these materials 
should therefore receive early and adequate attention. 

It is also trite and almost unnecessary to say that suflficient storage room should be pro- 
vided for supplies of materials, ample to insure continuous prosecution of work when desired. 
Shipments may be held up through any number of unexpected and unforseen happenings, so 
that unless there is reserve supply, a shutdown must result. Specifications often stipulate 
that a certain reserve quantity of material shall be maintained on the job, but even where this 
salutary provision is omitted, contractor and owner's engineer alike, for their individual and 
mutual protection, should have regard for this very important feature. 

It would seem that remarks as to providing proper care for materials on delivery to the 
work were equally unnecessary, yet disregard of these important matters is seen every day. 
Cement always reacts with water, whether such water comes from the mixer measuring tank, 
or whether it comes from rain, or from condensed steam, or from dew, or from water absorbed 
from the ground. Cement, therefore, should always be stored in weather-tight houses, having 
floors raised at least 6 in.* above the ground; and any cement directly at the work should be 



(Sec. 3-5 

Bef vp em moAtr/af is 


kept off the ground and carefully covered with an impervious covering, whether or not the 
atmosphere seems damp, or rain actually falling, or the weather threatening. 

6. Storage and Care of Stone. — One of the essential qualities of large aggregate for con- 
crete is cleanness. Stone, therefore, should not be dumped indiscriminately, so that when it is 
rehandled, dirt and rubbish are carried into the concrete. A thick layer of sand, or preferably 
a platform of planks should be placed on the ground before stone is deposited ; and not only will 
this precaution be found to keep the material clean, but it furthermore will often pay for its 
own cost, both in the quantity of stone saved by this means in rehandling and also in assuran<*c 
as to freedom of the concrete from deleterious foreign matter. 

6. Shoveling Materials Directly from Cars to Ground. — Wliere a siding extends to tho 
construction site, sand and stone may be shoveled directly from the cars to the ground, which 
should have been previously smoothed. That portion of the ground designed for the stone 
should first be spread with a layer of sand at least 1 in. thick, this layer serving to keep the stone 

clean and also working economy in subse- 
quent shoveling. In order that materials 

Cf^erts short or continuous . m j i. • i_ j ai_ a i l i ^ 

nailed ioenintie^ which ^fiay be piled high and the track be kept 
may be spaced about rctae, clear, a bulkhead may be built of a double 

row of 2 by 12-in. plank with 1 by 3-in. 
f» securs yrtater 9fobi/tfy cross-ties, having stops as indicated in Fig. 
the bass o^ knfftr courses 1. This method requires no fitting or cut - 

may be mode three it ibur ^j^g q£ ^^ plank. 

feet wide Crib shouid be ^ Storage and Care of Sand.— Sand 

should not be dumped directly on the 
ground, but for like reasons, should receive 
care similar to that described for stone. 
Although sand is a common material — 
"common as dirt'' — all dirt is not sand and 
dirt rarely, if ever, makes good concrete. It 
is only by attention to such seemingly small matters that a job can be well organized, made 
profitable, and the best results secured. 

8. Conveyance Economics. — ^The type of vehicle in which sand and stone are conveyed 
exercises a large influence on the economy of a job. As an excellent employment of gravity to 
cut labor costs, bottom-dumping conveyances, whether horse-drawn or motor-driven, or 
railroad cars, are conspicuous for economy. In certain situations, of course, gravity cannot be 
employed. In these the use of a locomotive crane with grab bucket is increasingly prevalent; 
and the type of car or barge used to convey materials within reach of the bucket will have a 
pronounced effect on the amount of material the bucket can handle in a given time. 

As an example of comparative conveyance economies, consider the following with respect 
to two types of railroad cars: 

Two quarries are located on different railroads. Railroad "A" is prepared to supply 
only hopper-bottom cars; railroad ''B" can furnish only flat-bottom cars. Assume gravity 
dump to be impossible, and that the plan of operation involves unloading the stone by shoveling. 
Assume also that one quarry, on railroad No. '* A" quotes $1.30 per cu. yd.; and the other, on 
railroad No. '' B" quotes $1.32 per cu. yd. It will be found cheaper in the end to order material 
at $1.32 per cu. yd. in flat-bottom cars for the reason that more than the difference in fiist 
cost can be saved by lessened labor in unloading. 

The reason for this is self-evident. A good man, under efficient superintendence, can 
unload 2 cu. yd. per hr. from a flat-bottom oar, as against 1)^ cu. yd. per hr. from a hopper- 
bottom car — a saving of at least 3 cts. per cu. yd. — and this proportional saving increases an 
less efficient kibor is employed. These figures represent average results. A good man working 
under a yardage or carload system will perhaps average 50% more than the above, but the 



Fra. 1. 

S«e.»-»1 iX}SBTRVCTIOS FLAK!' 175 

ratio win not change betweem the two typee of can. The ofteu-oeglected mfttter of idKiveliog, 
tbocfore, beoamee & matter of monwnt. 

i. UniiMdiiic E co ma ni eB. — Low-eide care are nioie economical for unloading than higli- 
lode. except where a wapm loader (Fige. 2 and 3) is used. IMiere Buch a loader ie employed, 

(lie car aide to which it is attached ahould either be of sufiicient height to give wagon cteftnuce, 
or Btakee must lie provided to raise the hopper to the proper level. The use of a hopper es- 
pedilee unloading the car, relieving poBsible demurrage charges and cutting down team and 
wagon houiB. One, two, or three hoppers may be attached to a single car and one or more men 
put in the car for each hopper, depending 
upon the number of teama available. 
EioppeiB should be filled, ready to dump 
when the teams tange alongside the car; 
and the more quickly materiale are di»- 
r-barged into wagons, possibly even with- 
out stopping, the more economical will 
lie the pnx'eiifi. 

10. Proper Size and Type of BlioveL 
— Although refiucmeute in efDciency can 
be carried to extremes, a potent factor 
in securing a proper output of work in ihj 
simple an operation ae shoveling ii> a 
proper siie and type of shovel. A good 
worker will alwaye look for a good shovel, 
which of itself Lh eloquent testimony an 
to the importance of this tool. Frederick 

W. Toytor found that a shovel adapted P,„ 3 

lo a load of 21 lb. gave the best results. 

Such a ahovel corresponds to the standard No. 4 sice sttpvet. The No. 3 shovel is somewhat 
smaller. The size of shove) should always be chosen with reference to the weight of the uiu- 
li^rialE to be handled. 

The better the quality of material in u shovel, the longer will it lost. A shovel is u i-on- 
\eying tool and. at the same time, it is a cutting tool no leas than a chisel or a drill. No con- 
tractor would consider using an inferior steel for such cutting instrumenlM, yet ui biiying sho\'el- 


maQy consider first coet only, having tittle regard to performance or endurance. It is ele- 
mentary economics to balance the value of a shovel costing %& a dozen which lasts about 2 
months, against the value of a shovel costing (5 a dozen, 'which hists 1 month or even less, par- 
ticularly when its performance, in addition to its absolute existence, is taken into account. 

11, Clam-shell Buckets.— ^^'hen gravity dumping is preclutled, a most efficient unloading 
device is the clam-ehcll bucket (Pig. 4), hung from derrick or locomotive crane. With this 
I ombination a tskillcd engineman can unload a large quantity of material in a day, but care 
should be taken that the bucket chosen is one suited to the work. Some buckets tend to ride 

Fio. 5. 

the material when it is at all resislanl, while others are so desipncd thai Ihcy till to the capacity 
at each bile, with proportionate efficiency and economy. The cuxt of equipment of this 
type is necessarily large, so that proportionate care should be taken in choosing the type of 


:iucket. No matter how excellent the derrick, or crane, or engine, or operative, a poor bucket 
will negative them all and run up handling coats at an alarming rate. An orange-peel bucket ia 
1 digging rather than an unloading tool and is not well adapted to the uaual rehandling of 

12. Bucket Unloaders and Conveyors. — There are an almost endless variety of elevating 
bucket unloaders each suited to some particular need. When the quantity of materials to be 
handled and stored is considerable, a bucket elevator installation (Fig. 5) may prove very 
economical, the more particularly as it makes possible the use of gravity discharge from con- 
veyance to elevating buckets as well as gravity discharge from storage bins through measuring 
hoppers to the mixer. The installation and power equipment required tor such elevating 
mechanisms is not neceaaarily expensive or extensive. Some lighter types are shown in Figs. 
6 and 7, and from these, the plant may range up to the heavier types, capable of hpudlne very 

'S^ quantities. In each individual layout, the needs should be studied and the economy or 
lack ot it determined. No hard and fast rcpummendation can be made that will apply to all 

W. Belt Conveyors. — In certain situations an endless belt, grooved to V-form by pulleys 
wer which it mns, furnishes a rapid and excellent means of handling raw concrete materials. 
' '*"iot, however, raise the materials so nearly vertically as can the bucket conveyor, its 
fWating slope being limited to about 1 vertical foot for every 2^ horizontal feet. The ap- 
putauong o( conveying belts arc without number, but, as in the case of the bucket conveyor, no 
"nliPt recommendation can be made. The manufacturers of conveying machinery are in 
W^ition to advise with respect to types, applicability, and relative economies of various types 
i^MveyoH fot any given set ot conditions, and should bo consulted for individual needs. 

U. Storage and Handling of Sack Cement. — Cement, because of case in handling, is usu- 
1 otdered in cloth or paper sacks, each holding 84 lb. Cloth sacks are less liable to rupture 


thftn ate paper socta, but their cost is greater. When sack cement is leeeived, a chain of laborers 
is formed between cam and storage houM, each man shouldering one sack. Economics in such 
procedure may be introduced by insuring ready entrance and exit both to cars and to atot^ge 

houaea, with adequately wide gang ptanka to cars, so that confusion and int«rferenoe may be 
prevented. The speed at which men will work is then a question of the personality and driving 
force of the superintendent or foreman. 

PlQ, a, — ProiMr meUiDd of bandtinc s 
Upptr U/l! A buibdl« of fiO c«] 
jpc of ibout S It., nstini od lo^ 

Lffvrr Irfl! After the ifaort rope* hmve bnn tied, the bundle is -_-.— _.„, ^- 
n<I cTO^^ in thft middte oF the bundle, enrwnf fint the shorter rapcA. 
lAHtrr rifhl: Bundle ol SO nnicnt uelu tied and tined ready [oi ibipioeDt. 

the pile, and witli ft lontei 

tyinc ti^Uy 

IB. BundUnf and Stotage of Empty Cement Sacka. — Inasmuch as each cement sack has a 
'4um vahie of 10 cts., it is important that none should be lost, damaged, or destroyed. Pur- 

Set. »-lfi: {!ONl<TRl'CJ-}Qf. fiLAAV i7V 

iberiDtnc. -titr deBmne. bimdling, and riiippine at fhwe smIce beooane an opMHttoo «f imiKK- 
tsuDF on laipv jobe. Much w men i e- iost wlmi iwge ate tDaufinii«iitt.v ehakwi, and tluii i«Ukiii(.-ii 
mroeni furtlin Kdcfe to the wei^i and bulk a! tttekt bundled tor retiun. I'-br proper metbod 
of bundlmf! nokt for return ahipmcai b" sboim ni h'if.. S. io tjliuwing u bundlt of aO saokt-. it 
faaf been wtlh the purpofli' of emphatiiinp Itn- ftnattw ctmvtmieDUC iu bMtdlin^ witli equal ud- 
THuUita! K B countiiif! unit of tbe aO^sok bimdk. iwtber Iban tbe lUO-twci( bundle. 

U. D tn i M gc aad ThMlKne •! 'W«t«.— 'Is moat inatoDvet ws(«r in ptj>«e ■« wiOiin KtMli u( 
& (lancrett iob. Wbert' tiiisisiiateoaiiilwaitfrinust be pumped, nawrvokt uf adequate (^p»cnv 
p fupe riy protected afai^ coatamiQaUaii tiiould be Mipplied. From ducti nMarrvaioE, wawr 
cam be dmrifantcd in ptpe.--' tobf^uMdngaeedvd. Tbe aUitagE' of w^UtrieM) simple wtd no eusilv 
canied am that Uk' nmd^ of fvdi individual MtuAtion are of greuMf delefHUoiug imporUttii:' 
-dian any pRTticular tjp<' of pumptu^ or distributinf; spparstUE. 

n. A "rfpifikl hittrilttii r— Fip . !> illustraCe:!> storage ftrmufcemeate aitd bMtdlJng laeiliUt^ 
m inMoUed in ccmieetkm -intJi ix Isgp.- Uuidinr for Ibt Hm^f^ ManuCactutii^ Co. at Cluabetha- 
port. K.3. Tbf Avjjbiikk' ttonge tp^ntoin thii^kM^itK- wat^noitnoK aUip'AHt. vide^Mlinetui 

tke Tt^Toac! tnck. and Iki'- buildiDg, of wbteh Uit' work iii queelion (oifued an extenskiD. A 
tranch wa.'' made, ncteudiug tJi< leagtli of tbi^ availaUo ipaec. and tbi.'^ was BbwAlied and 
cmverted mio il lunael by u cuvoriit); of !i-iii. plank. In tb(! l>cAtom of tbf drench was isjd u 
Btitable tmck for the operaliiiu of u ttkip («r. Un Ibo ledgt^r pteoeH were uiuunLod two ean- 
toothed nKHUTiDR hoppers, Sutd with wbu^. 

Matenal^ were uok>aded {foni tJit cars aad piled over tbe trencL, eaad and sloae iii aller- 
iiaAe pike. Tbe BawHoolbeii nieasuriai; hoppers were l&ept continualiv' asaineii th<! it>o of tL' 
aand and stoac; piic:) t'l facilitatii eharKiai; b.v breakiuf; duwM of titi- bu-e of the piie!>. I'uur 
man in \hi» way lumdled tiif- luaterial? fur about 30 i-u. yd. of uonurelJ' per br. Material- 
wore Bntomat4i;3lJy dropped from Uih uieaaurtuK boppers iulu lJi& okip car uti ii. patKied b('iuv^ 
tb'- iioppeni oil us way lu ibe oivx<rr. Thv J-irouiu aliowti iu Fig. 10 luis been UJ>eil tu ^<jo<t 
advanUf^e as a sul>3titute for the u«iKb. With lliis ariajigeiueal tuie man only aa» ueeded ii' 
2 hopper. 




16. Plant Economics. — ^Neeeasaiiiy much of the plant for the handling and storage of 
materiak must be considered as a part of the concreting plant proper. The economies of the 
whole plant therefore depend upon the individual and collective economies of its elements. 
Tbe main fiactors affecting these are fint cost, cost of insiaUaium, cost cf operation, cost of main- 
lenance, cosl of re m ova l , sahnige, and interest on the inoestmeni. 

IBa, Wast Cost — ^First cost of plant includes many items in addition to prices 
for machinery. Bonus for quick delivery, where equipment is required in a hurry; express 
chai^ges; tracing charges; and a himdred other items all swell the quoted figures. And it is not 
always best to consider fiist cost too cloeel3\ A plant that is cheapest in first cost may not be 
the most economical ; and labor losses due to delay speedily offset differences in price between 
good and poor equipment. Furthermore, low costs of operation and maintenance with higher 
salvage returns still further reconcile any disparit^^ 

186. Cost of Installation. — Cost of installation varies with the character of the 
plant, cost of labor, location of the woik and a variety of factors which must be separately 
considered for each situation. 

18c. Cost of O perat i on. — Cost- of operation depends both upon plant arrangement 
and upon organization. The concreting plant should be of a type, size, capacity, and arrangement 
to permit continuous operation during working hours, assuming an organization so coordinated 
as to make this possible and desirable ; and the character and arrangement of plant will depend 
to a large extent upon local conditions, such as contour of the ground, class of construction, 
manner in whi^ materials are delivered to the site, total yardage to be placed, time limit, bonus, 
penalty and other financial considerations which permit the use of equipment more or less 
expensive and elaborate. In addition, attention must be given to the time of the year during 
which the work is to be done, the normal temperature at that season for the particular locality, 
and the amount of land available for plant and material, since ston^ is always a factor in 

ISd. Cost of Maintenance. — Cost of maintenance includes upkeep of machines, 
repairs, oil, etc., this being greater or less according to the mechanical excellence of the plant 
and to its disposition and treatment. 

18e. Cost of Removal. — Cost of removal includes clearing the site of the plant 
and its appurtenances after completion of the work. This cost will vary, according as more or 
less of the plant is sold or jimked, with proportionate lessening of care and labor required in 
loading on cars, and of transportation. 

18/. Salvage. — The salvage value of machinery is always problematical. It 
iisually is worth what can be obtained for it. Certainly, depreciation on contractor's machinerj*^ 
is very laige and most estimates of salvage value should be liberally discounted. 

19. Balancing the Plant. — ^The general layout of the work will probably be the determining 
factor in the choice of means adopted for carrying out each portion of the work. The total 
yardage of concrete will also have a pronounced effect, possibly suggesting two or more separate 
installations of medium size, or a single installation of greater size, or a number of smaller 
mixers placed on different parts of the work. Various factors must be balanced one against 
the other and various layouts planned, with a following through from delivery of raw materials 
to delivery of concrete in the forms, with juggling of one scheme with another until the most 
advantageous result, consistent with allowable cost, is secured. Careful planning of plant 
before starting the job is well repaid in results, and a well-balanced plant is far more profitable 
than one poorly balanced. Installation of a mixer of double the capacity of the charging 
facilities, or of a fraction of the capacity of the handling facilities for the mixed materials is 
sheer waste. 

SO. Typical Plants. — Some typical examples of plants which have proven successful in 
service, are given in the following paragraphs: 



iSec. 8-20 

Fig. 11 shows an arrangement on the work of Cramp & Co., Philadelphia. It will be not«d 
that materiala are delivered in bottom-dump wagons upon the incline, and paas by way of 
bucket elevator to the bins above the mixer. Once in the bina, it is a gravity process through 
measuring hopper to mixer. Mesara. Cramp & Co. report 289 cu. yd. with this plant in 8K br.. 



with ft cr^w of eight men, and covering tdl handling from wagons and cement storage to delivery 
bin on hoiat tower. The mixer ia >i-yd. capacity. 

Fig. 12 illustrates an arrangement adopted by the Turner Construction Co. on the Bush 
Stores, South Btooklyn. Materials were deliveicd to the work in standard care, and unloaded 
by ibovel into special hoppers aa indicated. Theae hopper* were readily portable, and each 


hopper had a capacity for approximately 1 cu. yd. Materials were drawn off as required into 
cstB with a capacity of 6 cu. ft., and wheeled to the mixer which was set at a lower level eo that 
the fixed hopper was on a level with the ground. The bumping poet A facilitated discharge 
of carta into the hopper. Carts were wheeled up against the bumper when a alight lift on the 
handles did the trick. 

Fig. 13 illuatratea the plant used in the erection of the buildiiigB for Foster Armstrong Co. 
at Despatch, N. Y. The work on these buildings was carried on through the winter months 
and the bins indicated provided the readiest roe&na for heating the materials to the desired 
temperature of 90 to 100°P. An auxiliary measuring tank took caie of the salt solution used 
in the concrete mixture. Steam coib also served to warm the water used in mixing the coaerete. 
Wben mixed, the concrete was placed immediately; in no case more than 10 min. elapsed. 
When the concrete had been placed, it was protected against the action of frost by a solid wood 
coveriDg, blocked up at least 6 in. above the surface gf the floor in a manner to pennit free 

ciroulation lA air beneath the covering. Heat was introduced beneath the floor (or in the 
case of ground floors, beneath the board covering) by means of steam coils and salamandera, 
provision being made for the escape of sufficient steam beneath the covering to prevent prema- 
ture drying out of the concrete. Salamanders were sprinkled freely with water, thus producing 
the necessary amount of moisture, and small openings were left in the floor slab to permit 
the warm air to circulate over the upper surface of the floor. The sides of the floor were pro- 
tected by canvas curtains which extended downward; to the floor next below. 

Tliere were placed beneath the floor and beneath the panels on top of the floor, at intervals 
of to ft,, aelf-iegislering thermometera, which in no case showed lower than 32°. This tempera- 
ture was maintained until the test cubes which hod been allowed to set on the floor and beneath 
tbe top covering showed the strength used as a basis for the design. 

The extra plant involved in carrying on winter work involves considerable outlay, and work 
in frceiiog weather should not be undertaken without a thorough understanding of oil that is 

Fig. 14 indicateti a more or lees elaborate plant, designed for large work, or for cramped 


quartera. Tbe plant consists of a suitable bucket elevator, designed to handle the oniiri' 
aggregate. This elevator diachargea the materiab into trough screens as Indicated. While 
passing through these screens, the coarse mnterialB arc washed by a flow of water applied at thi- 
head, and the sand is still more thoroughly washed by passing through the water boot and in- 
clined worm. The washed materials are discharged from the first flight of trough screens upon 
inclined troughs leading to a fixed measuring hopper at the mixer. The lower end of the 

troughs it fitted with a gate to control the flow. Such excess material as cannot be cared for 
by the measuring troughs falls to the ground in piles as indicated. Two belt conveyors operat- 
ing in tunnels beneath the piles permit ready draft against these reserve piles, delivering materi- 
als to the original elevator, and thence to the measuring troughs. With such a plant it is a 
■jiRT^ mBtt«T to handle upward of W cu. yd. per hr. with four men. 

Sec. »-20] 



Figs. 15 and 16 illusuale two set-ups of pnctically the bmhc plant, and iUustmte foiciblv 
the UDponamce of proper arnjifement. Wiih Ibe airangeinenl sfaa«-D in Fig. 16, IT mc^ 
haodted 72 cu. yd. in iU hr. With thp Bmagement shown in Fip. 15. loir*- '-—-"-^ '•- 

1 4>^ hr., a BBvine of approximately 13 fts. per cu. yd. In Fig. 15 the runway and plat- 
art too email, with thp result that the men interfere with each other, and cannot work to 

uiivBDUme. Kurlhermore, liie ase of the Ivtsi rbute involves assembly of the batch in the 
iiuMT druDi. \Sal«r waa fed tu the uiuciune ii hueketful at u tiioe, requiring an extra man for 
tills puipoHe. In Fig. lt> a charging hopper te substitutL-d for the feed chute, and both niatfarni 


uid lunway are increased in mm. Water ia fed to the mixer through & pipe, and all operating 
levere are controlled by one man. Provision is also made to take care of any material working 
down beneath the mixer, with the result that wear and tear on joumale, etc., is reduced. 

FM. 10. 

The difference in cost of the two arrsngemento amounted to S59. The illustrations an? 
taken from actual experience, and indicate results secured under different superintendents. 

31. Hadiilie vs. Hand-mizing. — Except in relatively small quantities, hand-mixing of 
concrete is not to be economically considered. Furthermore, hand-mixing is inferior to ma- 

Pia. 17.~Drum miur. 

chine-mixing, with no comparison in quantity output. The province of a mixing machine U 
caaentialty the thorough incorporation of materials — one of the fundamentals in the producti(» 
of sound, enduring concrete. Mixing, therefore, should be accorded the respect due its impor- 
uid the best possible means choaen for its accomplishment. 


22. Types of Hizers. — The general types of miiiera which have endured and are on the 
ni&rket at the present time may be claaeified aa drum mixers, trough mixtrs, ffravUy mixers, 
and pneumatic mixers. 

22a. Dnim Mixers. — Drum mixers (Fig. 17) are esaeutialiy o! a type, differing 
tiiainly in excellence of mechanical construction and arrangement. The action of all of them 

Fia. 13. — Low chw^aj drum miier. 

is about the same so far as mixing ia concerned, the operation being accompliabcd by agitation, 
lifting, and pouring of the several materials by blades and scoops attacbed to the inside of the 
mixer drum. With the exception of tilting mixeis, discharge of the materials from the drum 
is accomplished by inserting a trough into one aide of the drum, in such position as to cabrh 
the concrete as it is poured from the mixit^ buckets. Minor differences in cbargU^ mechan- 
isms and arrangements are to be noted in different makes, but the action of all is essentially 

PlQ. IB.— Small pot miier. 

the action of a chum, in which capacity they would function if filled with cream, instead of 
with atone, sand, cement, and water. 

Of the low-charging mixers, the mixer shown in Fig. 18 is typical. Small pot mixers 
BUch 08 shown in Fig. IB are excellent for email work. 

336. Trough Mixers. — Trough mixers are paddle mixers of one type or another. 


Thejr may be batch mixers of the shoveling type (Fig. 20), or continuous mixers (Fig. 21), in which 
a sectional screw rotates in an open trough. Continuous mixera have not met with general 
favor as have batch mixers since many engineera object to these mixers on the grounds of 
uncertainty of mixing operation. 

33e. Gravity Mixers.— Gravity mixers are easentially a series of large funnels 
or pans suspended one above another with bottom gatea which can bo opened s 
permitting materials to flow from one into the other with incidental mixing to a 

Flo. 20.— Batch miier of the shoveling type. 

less extent. Gravity mixers are often u:^ed in preference to power-driven mixers on grounds 
of cheapness in operation and low first cost, permitting their being scrapped when worn; hut 
inany engineers do not advocate their use because of the inherent uncertainty of their mixing 
operation and oftentimes the requirement of detcrimental quantities of water to prevent the 
mass sticking in the pans. 

22dL Pneumatic HixerB. — Pneumatic mixers have been dcvelofiod by various 
inventors. At the present time there are two main types on the market. In some of thciic 

Fio, 21.— Continuoiamiier. 

machines premixmg is had before delivery, either mechanically or by the agitation of tur pres- 
sure, while in others the charge is introduced into a chamber, dependence for mixing being placnl 
on what may happen in transit througti pipes untlcr the delivering air pres.sure. Pncunmlic 
mixers have their own particular field — that of placing concrete in fomis where access is par- 
ticularly difficult — but because of the large compressor plant which must be installed for 
each mixer, and for other reasons which are valid and of importance in many classes of work, 
'w is relatively restricted. 

Sec. ^23) 



23. Machine Mixing. 

2da. Time of Mixer Operations. — ^Considering the concreting plant propec as 
an installation for mixing together raw materials to form concrete, the plant cycle can be 
considered as complete in three operations, viz., charging, mixing, and discharging. 

In charging and discharging the mixer, a time limit is imposed both by the physical laws 
governing the flow of materials from one container to another, and also (in the case of power- 
loading, or side-loading hoppers in particular) by the physical limitations of operatives and of 
the mechanism itself. As plant refinements are given consideration (particularly with regard 
to the gravity loading of measuring or charging hoppers from overhead bins) this loading 
time is diminished; but when a side-loading hopper, or a measuring hopper is charged by wheel- 
barrows, the time is lengthened more-or-less according to the perfection of the runway arrange- 
ments and the speed at which the men work. 

In the following table is given the result of timing of different types of mixers on different 
classes of work. These studies were made both with a seconds clock, motion picture camera, 
and with a stop-watch, and arc the summary of a large number of observations. From this it 
will be seen that the loading periods vary greatly; that the unloading periods have an equally- 
great variation; that the mixing periods are usually dependent upon the time taken in loading 
and unloading; and that successive operations overlap, the endeavor of the mixer man being 
to get out his material on as near a batch-a-minute schedule as is possible. In a number of 
instances it will be noted from this table, such a procedure gives a negative mixing time. 

Summary of Timing Data on Concrete Mkers 
Time given in minutes and seconds. 















Kind of 










Time of 





Lakewood, 1 yd . . . 
Koehrlng, 1 yd ... . 

Smith, ?iyd 

Foote, M yd ; • 

Foote, }i yd 

Chain Belt, J^ yd. . 
Koch ring, H yd.. . 
Lakewood, 1 yd . . . 
Ransome, 1 yd. . . . 
Ransome, }i yd.. . 
Chain Belt, H yd . 
Ransome, K yd. . . 

Batch hopper . 


Batch hopper . 


Batch hopper . 


Side loader. . . 


Side loader . . . 


Side loader . . . 


Side loader . . . 


Side loader . . . 


Batch hopper. 


Side loader. . . 


Side loader. . . 


Side loader . . . 


Side loader . . . 


Average - . 

Average, omitting 8 and 9 




2 m 




























































Note. — Mixer 8 had very poor blading. Mixer 9 was fed by derrick bucket. Long mix- 
ing due to inability to get raw material and to dispose of mixed concrete. 

23&. Time of Mixing. — Insufficient time is given to the mixing operation itself 
in most commerical work. Too long a period may possibly be indulged, but it usually is not; 
and no fear need be entertained of injuring the concrete by a mixing interval up to and in- 
cluding 30 min. The mixing operation proper comprehends not only admixture of materials, 
but also reaction between cement and water with distribution of the products of this reaction 


over the aurfacea of &and and stone. The time required for such thorough incorporation, 
to a certain extent, for the hastening of the reaction between cement and water, depends 
the adequacy of the blading and cleannesB of the mixer. Oftentimei mixera are put on 
with the drum (Fig. 22) half-cbolced with concrete or full of holes, or the blading so worr 
they cannot handle the materials. Neceasariiy such mixers will not produce the same 
as a clean mixer, properly bladed and having a tight drum. Also, mixers are not all eq 

So many factors ent«r into the making of good concrete, that a hard and fast rule appli 
to all cases cannot be made, but in general it may be said 30 aec. or even 1 min. of mixing 
adequate. It is far better, when it is desired to do a thoroughly firat-claas job, lo cr 
more mixers even at a higher first cost for equipment and work them on a loi^r schedule, 
it is to attempt with one mixer to get out concrete on a rapid-fire schedule. The latter ox 
often brings a chain ot unfortunate consequences, for not only is the concrete inadcqu 
mixed and the cement insufficiently used, but also excess water ie nearly always added in 
to make the mass fiee-working and to diminish the labor of mixing. 

33c. Drum Speeds. — Extended experimentation has established stni 
drum speeds for various sizes of mixers. Engines and motors as supplied with them n 
regulated as to maintain these speeds practically constant. Necessarily, as the art of ror 
making advances, changes will result, but the present rotational speeds of standard n 
seem suited to the requirements of average practice. Obviously, a slower drum speed i 

Fio. 22.— Mini drum. Fi 

result in less thorough incorporation of materials i 
to stick to the drum through centrifugal action 
ratings prescribed by manufacturera unless suet 

as any concrete mixer that will perform its operations better and more quicUy than 
petitors is sure to have correspondingly greater sales, it is safe to assume that mixc 
facturers have adopted for their product the maximum speed consistent with proper o| 
It is not well, therefore, for the user to attempt economLes by changing speed of the inii< 
SSd. Loading the Hizer. — There are many time economies that may bf 
in loading the charge of materials into the mixer. Various types of loading mechnni 
beep designed to meet different conditions of service and the time cycle of each is ■ 
A study of each type will show its adaptability to particular needs. 

Charging Hopprri. — Where a charging hopper mounted on the mixer frame cai. 
as in Fig. 23, the limitation to charging time is dependent upon the design of this ho] 
the slope of its sides and upon the site of opening from hopper to drum. Inasmu 
type of charging device is usually loaded by gravity from superposed measuring ho. 

> For (tudica of iriuT aclioiu He M. C. JoHnKH: Enif. Bre.. Dot. 4, 11)15. 


cooaidemtiona must be tctlceii into account in their deeif^; and always there must be prompti- 
tude ia releasing of gates, et«. In some very large operations euch aa the Elephant Butte 
Dam, pneumatic opening devices have been installed with an interlocking ayaiem, so that a 
sequence of operations is carried out with almost perfect regularity and great efficiency. 

Power Loader!. — Side loaders or power loadera aie often attached to mixers in order to give 
the advantages of low loading, as well as those of relatively high discharge of mixed materials. 
The general tjpo of mechanism employed is shown in Fig. 24. The type of loading hopper 
or skip varies with different manufacturers, some hoppers having a raised back, requiring a 
nlight incline for wheelbarrows that must be dumped into the hopper, while others permit 
running wheelbarrows directly on to the hopper back itself. Through a friction clutch, the 
power loader is elevated by the same motive power which drives the mixer drum. Inasmuch 
as it is required to hoist such loading skips to a considerable height before materials will run 

Flo. 24, 

from them into the mixer drum, it a essential that sufBcient power be provided to hoist this 
skip rapidly, asotherwiseanundueamountof time will be consumed in this elevating operation. 
The mixers of different manufacture vary widely as to speed of hoisting; and it will gcneraUy 
be found that the more expensive mixere have a better and more rapid hoisting mechanism, 
in addition to their other economies, than have the cheaper types of mixing machines. 

LmD-duaying Mixen. — Low-charging mixers (see Pig, 18), particularly in smaller units, 
have of recent years been meeting with favor. In such mixers the opening at the charging 
end is relatively larger than in other types of drum mixers and blading about this opening on 
the interior of the drum is so disposed as to draw the materials within the drum from a relatively 
small hopper of low height into which they are charged by wheelbarrows. With such mixeis 
an inclined runway platform of 2>j to 3ft. in heightisrequired. Their advantages, therefore, 
consist in a simplification of charging and the absence of hoisting mechanisms rather than is 


any particular efficiency ot mixing operation. Furthermore, these machines are Tclaliveiy 
low in price and a number of small units, gasoline or electric motor-driven, arc often very 
ftdvantogeous when distributed about the work. From a standpoint of thorough mixing and 
flexibility of operation, there is much to recommend this practice, inasmuch aa the needs of 
one part of the work can be supplied without reference to other parts or causing an overdraft 
on any one machine with consequent speeding up of operations as is the case when all parts 
of the work are demanding concrete at the same time from a single, centralized plant. Without 
disparaging in any way the importance of the time-cost clement in concrete mixing operations, 
it is yet to be regretted that considerations of quality and ultimate satisfaction of the customer, 
rather than first cost, do not more often govern both the selection of the plant and its opcration. 
23e. Measuring Materials. — It is often taken for granted that measurement 
of materials for a concrete batch is of little or no importance and that it can be accomplished 
in almost any way. It is probable that the average mix varies at least 50% in its proportions 
from those desired, and for this rcHHon alone it is not to be wondered that much concrete found 
on every hand is so variable in quality. 

Materials should be measured either in bottomless boxes placed on wheelbarrows, or like 
devices, or else in a barrow pan pcrmittinf; of struck measurement. A measuring barrow of 
known capacity permitting struck measurement is shown in Fig. 25. At the same time these. 

Fto. 2.-1. 

barrows are adapted by reason of their balance, to the conveyance of considerable quantities 
of material at one time. Measuring hoppers of known capacity, if carefully filled, can be made 
to function quite accurately; but where they are not struck, or where there is pronounced 
variation in the moisture content of the sand, the quantities of materials obtained per batch 
will be found surprisingly variable. 

It is difficult to convince the average contractor that economies can result from careful 
measurement. It tnay seem a useless task to confine field men to struck measure of sand and 
stone, but if the comparative quantities of cement required for accurately proportioned and 
inaccurately proportioned mixes were taken into account, the cost balance would usually be 
found in favor of the careful proportioning and measurement. And in addition, there should 
be considered and there will be considered with increasing force with passage of time, the 
ultimate performance and endurance of the concrete produced. The time is not long distant 
when owners will demand of contractors guarantees as to the quaUty of the product which they 
ate to receive and only by careful proportioning and measurement and placing of the concrete 
can a reasonable basis for such guaranlPCH he established. 

3S/. Discharge of the Mixer. — \ further economy of time can often be had by 
-nving attention to the proper and rapid discharge of materials from the mixer. Many mixets 


have inadequate and insufficient blading due to having bevrome worn with passage of timi-. 
Many also are partially choked by n)ncret« hardened inside the drum. Both insuflkient 
>>lading and choked ini:ter drums mean A relatively stow discharge of materials from the dram 
which cuts into the essential mixing operation. 

X4. Tranaportini; and Placing of Concrete. — Providing means for transporting mixed 
FODcrete and for placing it property in forms is an art in itself. These operations both in fiist 
coat and in ultimate effect rank equal in importance with the operations of conveying, pro- 
portioning, and of mixing raw materials. In mL\ed concrete, not only are the raw materials 
to be handled and oftentimes conves^ to considerable distances, but in addition this muatbc 
done at low unit cost and in such a monneT and so expeditiously as to protect the mixed mass 
from injury. 

The means usually adopted for the conveyance and placing of concrete are some sort of 
bucket or cableway, or else open spouta or chutes through which the concrete flows- by gravity, 
or else in barrows, carts, or cais. The particular means adopted in any case, will depend upon 
the siae of the operation, upon the physical conditions attendant and upon the financial limi- 
tations to plant imposed by commercial considerations. 

S4a. BaiTOWB. — As affecting perhaps the great bulk of concrete used today, 
it will be proper to first consider the use of barrows or carts. This method involves less original 

Fia. 2S. 

plant outlay than the otbera before enumerated. In many instances, the cost of installation 
of an elaborate plant would cover not only the cost of the barrows themselves, but a great 
part of the entire cost of distribution of the concrete by other means. 

The ordinary wheelbarrow (Fig. 26) having a flat pan is not well adapted to the distri- 
bution of concrete. In such a barrow a man can handle about 1^ to2cu. ft. of mixed concrete. 
This load he can wheelAbout 25 ft. every 3 min., the objection to the pan wheelbarrow being 
(bat the man's working rate ia necessarily cut down by the care which is required to keep the 
materials from slopping over the sides. Furthermore by the des^n of the barrow a large 
proportion of the weight of the load is on the man's arms, rather than on the wheel. Deep 
pan barrows have been designed to overcome this difficulty, but have not wholly accomplished 
the desired end. 

246. Concrete Carts. — Two-wheel concrete carts (Fig. 27) are better adapted 
to this work than wheelbarrows, both because they can carry a larger load and also because this 
load is balanced on the wheels themselves with little or no strain on the man. The usual 
two-wheel concrete car is of 6-eu. ft. capacity in which about 4}i cu. ft. of mixed concrete can 
be carried by one man. 

In this comparison there are, however, certain cost offsets to be made. Wheelbarrows 


require leas RCftffolding th&n do the heavier and wider carts, so that the cost of this runway must 
be carefully eetimated. When runways must be elevated, the showing becomes more favorable 
for carts, as bents or supports for wheelbarrows must be practically of the saoie size aod strength 
as thoee for carts. Turnouts and gangways must in both cases be of ample width so that there 

Fio. 27. 

may not be congeetion in the passing of full and empty carts going to and retumii^ from the 

He. Bucket*. — There is a great variety in types of buckets adapted to the 
dirtribution of concrete. Some of these buckets are straight-side skips, as in Pig. 28, adapted 

Fia. SS— TJltinc bucket Pio. 20.— Round Mlf-tilttni biKkel. 

to dump by overturning. Others are bottom-dumping buckets operated by a man at the form ; 
and thew bottom-dumping buckets may be of various pattem.4, adapted to some particular 
UM. Anezampleof this sort of bucket is sho¥m in Fig. 31, in which the bottom is so eonatiucted 


us to form a long narrow opemng, actuated through a powerful lever mechaniam. A great 
variety of these devices is on the market and the needs of i^sch particulnr situation must be 
studied and met by as specialiied a product for that use, as financial conai derations will permit. 
Md. CablewajB and Buckets. — Cablewaya usually require large initial outlay 
but on large operations they may be found very economieal. Usually they consist of a strong 

Fm. 30.— Bottom-dump bucket [or Fia. 31. — Battam-dump bucket for 

Una (oTioi. narroR lotma. 

messenger cable or cables (carried between either fixed or movable towers) with actuating 
cables to hoist and carry the buckets to any desired spot on the work. Cablcway buckets may 
be of various types, but that shown in Fig. 33 has proven its worth in many constructions. 
Tfae bucket shown in this illustration is a 2-yd. bucket. Its deep upper body end steep sloping 
bottom provide capacity and free flow, while a gate t»>ntrolled by levers regulates the discharge 
of concrete. 

Fio. 3i.~Towu-bout bllins bosket. Fin. 33.— Bottotn-dumi) cnbleway bucket. 

He. Spouts or Chutes.— The handling of concrete through spoute or chutes 
is a development of the last S years. This system at the present time is in more extensive 
use than any of the foregoing methodsof distribution, with the possible exception oi^distribution 


in carts. The economic features of spouting are undeniably attractive. To raise concrete 
vertically in a tower by means of a skip bucket and engine located at the central mixer plant, 
then distributing by gravity through channels which can be arranged in convenient sections 
to cover any area with a radius from 10 to 300 ft. from the base of the tower, appeals strongly 
both to engineering and to business sense. Further, the ease of handling by gravity is usually 
greater and the time cost per cubic yard for placing is usually less than in transferring the same 
quantity of material in hand-barrows, in cableway buckets, or in cars. Yet in spite of 
its many good points, the convenience of spouting has brought about many abuses. 

For instance, it is obvious that in order to flow readily through chutes, concrete must be 
smooth and plastic, whereas the materials of which concrete is composed, with the exception of 
water, are all exceedingly sharp and gritty. It is not to be wondered then that lubrication 
and ease of flow secured by increased wetness, has encpuraged the use of excess water, especially 
where for reasons of cost, it is desired to erect only a relatively low tower causing the angle of 
the spout to be comparatively flat. Furthermore, many spouting equipments have been in- 
stalled with ease of distribution alone in view, the first cost of plant and rapid deterioration not 
being taken into account, so that saving has later been sought by cutting comers to make 
up for the initial mistake. 

In all spouting installations, care must be taken to have the chutes at a workable inclina- 
tion. Furthermore, it is important to maintain a uniform pitch throughout the entire line, 
in order that the flow may be thorough and uninterrupted and not subject to slackening at one 
part and accelerated flow in another. The pitch also must be greater when the material is to 
be carried to a considerable distance than when it is to be carried only a short distance, for as 
the distance increases, the friction of the concrete in a chute tends to overcome its initial mo- 
mentum. Whereas, therefore, a wet concrete will flow 50 ft. with the pitch of 1 in 6 it becomes 
necessary to increase this pitch to 1 in 4 for a distributing distance of 100 ft., while a distance 
of 300 or 400 ft. will require a pitch of 1 in 3. The slopes as above described are based upon 
chute rigidly supported having uniform pitch throughout; and it would be even better to in- 
crease this pitch in order that concretes of a drier consistency may be used. 

Various methods have been proposed for increasing the ease of flow of concrete in chutes. 
Hydrated lime in one proportion or another has probably proven the most effective, but there 
is no standard procedure in this regard, nor is the exact quantity of hydrated lime required for 
any given concrete prescribable without experimental knowledge of the aggregates separately 
and in combination. Hydrated lime added to concrete has some undesirable features, but 
even aside from these it is an expensive diluent of inferior streng;th, and inasmuch as practically 
the same effect through the same agency may be realized by a longer mixing of materials, the 
wisdom of its use is not yet beyond question. 

The unfortunate tendency, as before pointed out, is to add more water to spouting con- 
cretes to make them flow freely. This, however, defeats its own end, inasmuch as segregation 
takes place very readily from wet mixtures, so that there is initially a rapid rush of semifluid 
materials down the chute, with afterward a slow dribbling of the heavier and harsher materials, 
oftentimes requiring men in the rigging with hoes to keep the un watered sand and stone from 
stopping. With sloppy mixtures, therefore, not only is the quality of concrete impaired but 
aLso the cost of delivery and placing is very largely increased. On the other hand, thoroughly 
mixed concrete without excessive water may be successfully delivered through spouts disposed 
at proper pitch without segregation or the loss in value attendant upon the use of excessively 
wet mixtures. 

84^. Sections Used in Spouting. — It is desirable that concrete spou ting be arranged 
in a series of units which may be assembled in various combinations. Continuqus-line spouting 
should be changeable to swivel-head, or swivel-head to continuous-line, as the conditions of 
the work require, it being necessary, of course, to have in stock a supply of the necessary units. 
This interchangeability is of great value in service, for spouts wear at the head and foot of each 
unit of length. By reversing a trough section, end for end, when showing heavy wear at one 
~t, a new, unworn surface may be put at point of greatest wear. 


A standard trough section, Hg. 34, is made of No. 14 gage steel, forming a trough 8^ in. 
deep by 10 in. wide on top. The bottom is curved to practically a eemicircle of 4-in. radius, 
the upper part of the sides being straight and tangent to the curve. Each section is punched 
with standard spacing, arranged for connecting all oC the various attachmenta. 

Fio. 34, 

The hopper head, Fig. 35, attached at one end for receiving the concrete from the bin, or 
from an upper trough section, forms one point of support of the next trough section. At 
the other end is the splash hood, Fig. 36. By fastening the hopper head to the trough section 
at one end, and the splash hood at the other, we have the complete trough section, Fig. 37. 

Pio. 3S. 

These 24 by 24-in. hopper heads, as well as the splash hoods, can be bolted to either end of any 
standard trough section. 

Standard trough sections are joined for continuous-line spouting by bolting together their 
angle-iron yokes or flanges and bolting on the compression plate part. Thus, several sections 
are joined together, with a hopper head at one end of the entire group, and a splash hood at the 
other end. 


Fig. 38 shows the swivel-hook used in making the flexible joint between niccessive trough 
sections for swivel-head spouting and shows one of these jointe, in which the upper line of 
spouting is supported by a fall and tackle attached to the btut on the splash hood; while the 
lower line is supported by the swivel-hook, connecting the lower hopper head with the Bplash 
hood of the upper line. The swivel-hook b kept clear of the path of the concrete. 

In some cases it is desirable to have a flexible joint in continuous-line spouting. In this 
case the two sections are put together in a different manner, ¥1%. 80, where both the hopper 

Fin. se. 

head and the splash hood are dispensed with. The hanger plate is here used in conjunction 
with a special yoke, after one of the angle-iron yokes has been removed. This allows a slight 
movement sideways, without requiring the attachments for the swiveljiead operation. 

Various types of spouting have been tried, ranging from round pipe to rectangular troughs. 
Best results have been secured from the use of 5-in. pipes, or UHn. open troughs, the latt«r 

Fio. 37. 

having the preference for flat slopes, and the former where there is necessity for varying pitch, 
with a likelihood of steeper pitch than named above. 

With open spouting the use of remixing hoppers (Fig. 40), in connectioQ with flesdble 
spouting {Fig. 41), accomplishes satisfactorily the neceasaty L'hanges in pitch. 

The greatest items of expense in spouting plants are first cost, installation, and mainte- 
nance. Maintenance charges are particularly heavy. The ordinary slock spouting whieh is 
made of No. 14 gage metal will seldom handle more than 2000 cu. yd. without rtnewaL This is 


doe to Vut ftbraaive vtion of the material, especially as aSectiog the rivets whii'h join the vmrioui 

A recent devdopment is a spout made up of two or more longitudinal BBctions of the ibape 
indicated in ¥ig. 42. The various aectiona ate interchangeable, and there are no bolts or riveta 

extending through the spout, all jobt boHa or tivet« passing throacl> ^^ flanges, and tbe 
vvrioae kmgitodina] joints made secure by fish plates. This type of spouting has tbe further 

advantage dl ""'■-"«c pueeible renewal of i 

Inigtfa of spout as a whole. This trpe fuithermoie e 

difcctiooa, a poinl tt c ouaid er a ble nDportaaee. 


X4ff. Hoists.— Whether the distribution is by spouta, by carts, or by burows, 
it has become general practice on all work extending above ground to hoist the concrete. For 
Ihis purpose a tower is practically indispensable. 

It will ordinarily be found advisable' to install the hoist at the beginning of operations. 
sinc« by so doing the mixer may readily be set so that the operation of rharging may be facili- 
tated, principally by cutting out inclines, with resultant saving in labor. 

Tuwers are coiutruct«d of at«cl or wood. The hoist bucket should be constructed on 
the simplest lines without catches or trips. A substantial bail made of two 3-in. Z-bars 
back to back, is arranged tu operate brtn-ccn two 5^4-in. wooden guides, and is fittml at the 

S«u»-2<sl COKSTRUCTIOX PL.iST 201 

l>>ic«r end with jouhiaIs in which rests the bucket truimioa. In setting up the tower uid 
t'uckrt, it is advisable in all cases to set the bucket bo that it is balance, and to this aid the 
front guide should be so set as t^ be almost in contact with the nose of the bucket when the 
latter is pushed back to a point where the load will tend slightly to press the stops on the sides 
of the bucket backward a|^inat the bail. Friction of the nose against the guides ia, by this 

means, cut down. By removing the front guide at any pamt in the height of the tower, and 
placing a block on the back of the latter, the bucket is canted forward so that it will drop its 
•■ontents out through the opening made by the removal of the front guide. The bucket auto- 
matkalty rights itself, and b pulled back into position by the weight of the bail when the 
iiperator raieases the brake. 

A typical hoist is showu in Fig. 43, operating in connection with a mixer, the power being 
taken from an extension of the nuKer shaft. The power equipment of the latter should be of 
sufficient capacity to opeiate both mixer and hoist at the same time. A variation of this plant 
showing a diiect-connected hoidt is shown in Fig. 44, but for ordinary conditions the first- 
described arrangement is preferable. 

At any desired height a bin or hopper is set, into which the material is discharged by the 



hoist backet. From this point difltribution may be effected by wheelbarrow, cart, car, or 
spout, either separately or in combination. A concrete bin such as shoi^oi in Fig. 45 forms the 
upper end of a spouting system, the gate of the hopper under manual control regulating the 

tf. Spootfaig Plants. — Spouting plants may be classed as hoam ptanls^ guy-line plants j and 
tower pUmiB. 

26a. Boom Plants. — ^In boom plants, the first and second sections of spouting 
are mounted on a bracket attached to the hoisting tower, the free end being moved by tag lines 
to the position desired. This rig offers advantages of flexibility and freedom of movement 
not often obtained in placing concrete. Oftentimes open-throated booms (through which the 
first section of spouting is carried) are used, these having the advantage of lending lateral 
stability to the spout itself as well as of economizing space. 

856. Guy-line Plants. — In guy-line plants, the spout is suspended by blocks and 
fails from guy lines, or special cables suspended between towers, or other supports especially 
set up for the purpose. The advantage of this type of spouting plant lies in its ready adapta- 
bility. It is limited, however, in lateral movement unless its deficiencies are supplemented 
by take-offs at various points with small boom plants or supplementary guy-line plants. 

25c. Tower Plants. — Tower plants are of like general feature, but the spouting 
line is supported at ends of successive sections by movable towers or tripods. A plant of this 
kind is leas flexible than a boom plant, but is more flexible than a guy-line plant, inasmuch as 
the various supports in the line may be moved successively, rendering possible the covering of 
a very wide area from a single hoisting tower. A guy-line plant, on the contrary, requires 
under like circumstances that the whole line be dismantled and set up again in the new location. 

26d. Combinations of Spouting Systems. — Combinations of the above systems 
are used advantageously in one way and another in order to surmount special obstacles. Among 
such combinations may be mentioned a rehoisting tower which permits covering a wider area. 
In such a plant the concrete is distributed from mixer and first tower through chutes to a hopper 
at the base of the second tower, when it is again elevated and distributed throughout the work. 
A careful study is required in order to make spouting plants thoroughly effective; and this study 
should always be made before the job is started to make sure that the proper radius of delivery 
and best arrangement is secured. 

26e. Regulating Flow of Concrete in Spouting Plants. — It is quite essential for 
the proper operation of the spouting plant that concrete should be uniformly and continuously 
carried down the chutes. To this end a receiving hopper is placed at the head of the elevating 
tower, with a man in control of its gate. Upon this man then depends to a large extent the 
Bucoeas of the operation. If he permits a proper amount of material to flow into the chutes, 
they can usually be relied upon to carry it freely providing they are disposed at proper inclina- 
tion. If he sees the line becoming choked, upon his slackening or shutting off the delivery 
depends either a speedy clearing of the line with relatively continuous operation, or shutting 
down for an indefinite period. 

No matter what type of mixing equipment, or what system of distribution is adopted, 
there should be kept in constatnt view the object to be attained, namely, the economical pro- 
duction and placement, not merely of materials which will fill form spaces with possible accept- 
ance, but rather of materials which in forms will solidify and endure under stress, whatever the 
nature of such stress may be. Each yard of ix>or concrete carelessly placed gives concrete a 
black eye. Each yard of good concrete properly placed b testimony as to the abilities of this 
inherently wonderful material. 



1. The Coftcrete-flDor l^blem. — Canpf«t4> floors in modern rommcrciul liuildiiigs irn- oi 
IH-^-uliar iraportAnre. Not only is the floor slab an inU^ral part of the stnirtut«, niakitiR pon- 
Kihlp ite iisefulnesB by supporting applied loads — such as maphinery, or storod icoodi> — bul enrh 
floois must further perform unusual service in wilhstandinf; HC\-erc ronrcnttat)^ atrpBsee — »«, 
for instance, those due to paasing trucks heavily loaded — and particularly must they withstand 
at their top surfaces not only the crushing above referred to, but., in addition, a constant 
and aevere abrasion through the impact of shoes, or the movement of loads, with ofU-ntimcH 
attack from chemicals used in manufacturing processes. 

No part, therefore, either of aggregates or of the cement matrix in which they are embedded, 
may give way without "dusting," or progressive destruction, of the floor to greater or loss 
degree, causing not only annoyance and inconvenience, but possibly more serious consequcncCH 
hy reason of the released particles being carried into machinery and manufactured products. 
TlteBe particles are abrasive, gritty, and may be chemically Injurious. I'he problem of satin- 
factory concrete-floor surfaces becomes, therefore, essentinlly the problem of producing a 
concrete 1,1] of a strength requisite to resist compression and shear due to floor loads; and (2) of 
BufGctent top-surface resistance to withstand the mechanical attack of normal service. Chem- 
ical attack must be provided for by special supplementary treat- 
ment with a resisting paint or varnish. 

He requisite first named is met with relative ease. E>en- 
sity and strength through use of proper aggi«gate8 and good 
cement, thoroughly mixed, without excess water, and carefully 
placed are the procedures to be followed (sec chapters on " Aggre- 
gata" and "Water" in Beet. 1 and on "Mixing, Transporting, 
and naeing Concret«" in Sect. 2). Porous floor slabs, such as 
that shown in Fig. 1 do not make either for strength or for any of 
the qualities desired in good concrete. 

The requisite second named — that of producing a resisting 
top surface is more difficult to meet with full satisfaction. If the Tia. 1 — Itauih Uvtmr 

surface of concrete floors (or of concrete roads or sidewalks) could J^^S^a ii^™) """ '''''■ '""' 
be natural stone of proper quality molded with the same ease as 

is concrete and made Integral both as a monolith and by tying with steel to the rest of the 
structure, there would be little cause for complaint on the score of dusting, wear, or struc- 
tual fimctioning. Yet the aggregate employed in concrete is natural stone in fragments. 
Since reinforeing steel does not affect wearing qualities at the surface, (he difficulty must, 
therefore, lie either in the choice of the natural materials, in the proportion of those matttrials 
eipoaed as resistants to abrasion, or elKe in the quality or quantity of the ccmcMiting material 
holding them vise-like against the abrading forces. 

The ideal in artificial stone floors is the tcrazzo floor in which an even surface (W6% or 
more of which is natural stone with 5% or less of cementing material) is prcaent^il to wear. 
Grinding concrete floora is necessarily expensive, but removing a surface layer by such means 


prodncee & superior result tttat is found to jiutSy the Mvt. The litiiifii iir of ^tam f at-tT. 
bigetber with what is knom of the general top-surface thvmetet of courreteE. hsds to "lii- 
coatlusion, which is fully borne out. that in the eztmne top ]ani> td a tuueiHe floor it ti' '- 
found much of whatever difficulty is experienced. 

1. "Ducting of Concrete Floon. — Research faAaBhowntlut"daMiBg"'floanaMeaBcr<-i^-r 

which are at least locally poor, such local weftkneae (ntost evident in tbe top mu bciag dtcnrr. 

typically in Fig. 2, whemn the sand grains are seen to be uncertainly held in a looae aatd ckeI: ' 

abraded matrix of what appears to be of tbe nature of effltHcacoiee. AD dosting-Aoair iniif»cws. 

however, are not identical with tbe one shown, nor are the causes of dnsluif; BBccaeanlT \ti' 

same. Nevertheless, tbe process of progress of progreiEive destruction or "AosSmi^" k nmc-L 

the same in most ca^es and is largely due to a loose conditiiMi of tbe cenwnl binds' wfaid pRrmnf 

the resistant sand grains to fall out, exposing fresh surfaces of scrft. hydiated o^boiI 1o aianciaii 

with repetition of the process until remedies are applied, or until resistant strata at irregnKr 

depths below the surface are reacfaed. Tbrae adiortc aiv Siiir- 

menled at times by chemical action, atha' from luuitliui. or fron. 

other atmospheric or fiuid agencies. 

S. Malring Good Coocretc FkMcs and Fbor **— '"t^ — A 
good concrete Boor is easentiaUy a good concrete. Tbe primln let. 
of ma king good concrete — the right materials and the rii^i pmfKir- 
tions of same, including water; thorou^ mixing; carnal placing 
careful curing — epitomise tbe rn«kit.g of good floors. 

To these axiomatie and seU-evidest gowiml prineipkc Atnii^ 
be added the following: 

1. Whenever possible, rwit(9 coat nod buEtogHber- Bxkiit 
Fio. z— DiBtiu nrfm □( IB uot possible OT advisable, remove t<^ surface of base to a drptL 
«««. floor. C^inified "> of at least H ««»- Then, before placing t<^ coat, iw^Mi tbe haae 
surface well and wash clean, using hose or bradies. This paT>- 
cedure is necessary to procure a proper bond between top coat and base. 
2. Use coarse, rather than too fine material in the top coal.' 
4. SpedaJ Snrface Finishes. 

4a. Surface Grinding. — "Granolithic" is a term aj^ilied alike to wmu r w 
floore having cement and sand finish, and to those having a surface layer of cradied gntiiTr. 
or other hard, enduring rock bonded with cement. As above not^d, a desirable fmA i-r> 
such floors, a finish that gives a pleasing appearance and removes much of any tcadatcy ihcrr 
may be to dusting or surface disintegration of any kind, may be prodund by surface giiDdiitf> 
when the concrete is from 4 to 7 days old by means ai a machine similar to that empkiytid in 
grinding terszio floors. Such grinding removes any laitance or loose material fram tbe sartaer.. 
produces a smooth though not polished surface and, by selection of agxregates befbtc tayin^ 
with special reference to color, gives an unusually pleasing effect. 

4h. Integral Pigments. — Pigments of one coloratioa or another, dMOKaUr 
inert toward concrete, can be had lA a number of dealers. Loasmuch as surface color ooly n 
desired, it is advantageous to apply them only in a relativriy-thin mortar Uycr M tbe lop 
surface. This layer should be truly integrmi with the layers below, ebe it will scale oC. It 
nhould further be borne in mind that the coloring value in concrete of any inlcgial pigmmt 
will be affected strongly by the color of Portland cement, which n itself a pigment, white vhm 
hydrated, gray-green nhcn unhydrated. According, therefore, to tbe color added, tbe (4ect 
of pigment in concrete will be more or less intense according to its percentage preseat* a.< 
related to the percentage of cement in tbe mixture and to the degree of hydratiou of tbe lattet. 
The color of aggr^ates may also affect tbe result. Care should, therefore, be exercised in ii- 
tempting to secure a given intensity of final color, not to use pigments in such quantity aa to he 
detrimental to the concrete. Small trial batches will aid in securing tbe effect desired. 

1 Be* L C W»Mx: fm"' A SMB., 1«H. p. KM. 


4c. Finish Produced by Removal of Water From Surface. — A process of finishing 
floors said to give excellent results is the abstraction of excess surface water through absorption 
V>y ^Ty cement laid on webbing over the soft floor. ^ This method should be advantageous 
in maxiy instances, particularly where excess water is used. The process is proprietary. 

Ad, Integral Hardeners and Surface Compounds. — A number of compounds 
a. re ozi the market designed to be incorporated with the surface to make it resistant. One of 
these is carbide of silicon (carborundum) under one name or another. This material unques- 
tionably has great abrasive resistance and if properly held in place by cement should produce 
a surface capable of withstanding severe traffic. There is, however, no reason to expect better 
conditions of manufacture attending its use than would obtain where it is omitted; and as 
good quartz sand or crushed durable rock properly bedded in cement is capable of supplying 
xnost needs; and inasmuch as the sparkling, glistening effect incident to the use of carborundum 
is often objectionable, the advantages to be expected from it should be carefully looked into 
before it is employed. 

Common iron, powdered, is extensively marketed as a surface hardener for concrete floors. 

A variety of claims, many of them conflicting, a:e advanced by its advocates. Some assert 

that the iron oxidizes (rusts) with expansive filling of pores and prevention of further moisture 

penetration. Soil-ammoniac may even be added to promote this rusting. Others claim no 

rusting with the virtue residing in the superior hardness of such iron as remains at the surface. 

It is unquestionably a fact that the average iron in contact with moist air will rust. This 

produces a characteristic red stain of rust in the concrete, but it is doubtful if as an incident to 

mixing or placing, this iron rust can be so directed, distributed, and placed as to constitute 

a reliable pore filler; and it is further more likely to be an attractor of moisture than a preventive 

of moisture penetration, since rust (Fe208-X H2O) is deliquescent. Further, iron is so inferior 

in hardness to common quartz sand as to make the ratio of comparison about 230 (for quartz) 

to 18 (for iron),* so that with equally satisfactory embedment in cement, iron should prove 

inferior to ordinary sand. In addition, factory grease is often not removed from the iron, so 

that attachment of cement is hindered, if not inhibited. 

There is no top coat superior in all-around qualities to good quartz sand of proper size and 
grading, nor is there any additive at present known qualified per se to overcome initial deficien- 
cies resulting from faulty manufacture or inferior materials. 

5. Causes of Common Defects in Concrete Floors. — A statement of defects commonly 
found in concrete floors and the causes which give rise to them is conversely an aid to the pro- 
duction of floors of proper endurance. Avoidance of wrong practices is the surest guaranty 
of success. Such a listing of the causes of defects, therefore, follows: 

(a) Poor Cement. — This cause is infrequent. It is true that defective Portland cements 
are occasionally manufactured and that they are marketed, but misuse of cement is more 
frequent than deficiencies in the cement itself. Other causes should be sought and eliminated 
before blame is attached to the cement. 

(6) Poor QtuilUy of Sand. — This cause is relatively frequent. Sands, as before noted, 
are derived from the breakdown of natural rocks; and in most sand deposits the grains have 
existed for millions of years, so that their inherent quality and endurance is vouched for, but 
decomposing cementing materials, such as clay (uniting very small mineral particles to form the 
larger grains), or organic matter, or dirt, or other impurities, may render the best sand unfit 
for use in concrete. Be sure of the quality of sand before laying the floor (see Art. 30, Sect. 1). 

(c) Poor Grading of Sand. — This cause is relatively frequent. It needs no demonstration 
to prove that the greatest quantity of sand in a given volume will be obtained when the 
particles are so graded in size that smaller grains will lie between the spaces of larger grains, 
progressively down the scale of sizes. Likewise, the least quantity of sand in a given volume will 
be had when all the grains are of one uniform size. This truth is sometimes stated by saying 

^8ee P. M. Bruner: Proc. Am. Con. Inst., 1015. 
' Scleroecope scale. 


that '^ minimum voids" of ^' maximum density " is obtained with graded sizes of grains; and that 
'^ maximum voids" or "minimum density" is obtained when the grains are all of one sixe. 
Sand for concrete floors and particularly for the top coat should, therefore, be graded in size 
in such manner as to give "maximum density" with maximum quantity of enduring silicious 
material of proper grain size at the surface. Be sure of the grading of sand before laying the 

(d) Dirty Water. — ^The occurrence of defects due to the use of dirty water is relatively 
infrequent. Dirty water is not merely water that carries fine silty matter in suspension, but 
more particularly contaminated water, carrying organic matter, such as stable or barnyard 
drainage. Organic matter of this kind is very injurious to concrete and may even cause failure 
to set, or total disintegration. Use only water of unquestionable cleanness. 

(e) Too Much Water. — The use of too much water is of general occurrence. With excess 
water in a mix, the fine particles of stone, and the fine particles of sand, and the finest particles 
of floor cement separate and rise, forming a thick scum at the top of the slab. This is where 
the slab should be most enduring, but if too much water is used, a material having about the 
resistance and character of chalk is substituted for the enduring materials desired. Avoid 
excess waJter. Use thorough mixing to obtain the plasticity desired. 

(J) Wrong Proportions of Materials. — Arbitrary proportions in concrete making have little 
except careless convenience to recommend them. Through lack of understanding and be- 
cause of the supposed difficulty of proper proportioning, they have remained in practice. It 
should be borne in mind that each sand and each gravel has properties peculiar to itself; and 
that the proportions in which they should be used in combination with cement and water 
apply to them only, so that such proportions cannot be taken as a criterion for the use of other 
sands or gravels. Proportion the materials so as to obtain maximum density (see chapter on " Ag- 
gregates" in Sect. 1 and on "Proportioning Concrete" in Sect. 2). 

(g) Insufficient Mixing. — One of the reasons that excess water is so commonly used in con- 
crete is that it renders mixing easy. The desire for ease and rapidity of working tends to carry 
the speeding-up process beyond allowable limits so that insuflicient mixing is more often in- 
dulged in than is realized. Whether the mixture is sloppy^ pltistiCf or dry, mix it not less than 1 
min, in a batch machine, or an equiwilent amount if other form of mixing is employed, 

(h) Too Much Tamping. — Concrete may be tamped too much. With a medium wet con- 
crete or concrete of plastic consistency, tamping to a certain degree is desirable to compact the 
mass. But tamping to more than the required degree brings unfortunate results. It should 
be recognized that sand and stone and cement in a fluid or semifluid concrete are non-coherent 
and that by the agitation of tamping, heavier materials sink and lighter materials rise. This 
causes separation or "segregation," of necessity putting fine, chalky materials, not adapted 
to resist abrasion, at the wearing surface. DonH flood the surface by tamping. 

(«) Too LitUe Tamping. — It is frequently the case that a concrete floor is deposited in such 
haste and with so little care that it does not compact; and particularly, it is not sufficiently 
joggled to remove air from the mass. In all mixing operations, air is stirred into the plastic 
mass, much as it might be into a stiff batter; and if such air is not removed in placing, a honey- 
combed structure will result. Further, the tendency of entrapped air is to concentrate at or 
near the upper, or wearing surface, so that when the cement has set, it will be molded as bubbles 
in the mass. Needless to say, such bubbles are holes in the concrete; and holes offer very poor 
resistance both to stress and to abrasion. Tamp enough to compact the concrete, buX noi enough 
to flood ii. 

(J) Too Much Troweling. — After a concrete floor form is fiUed and screeded, the surface 
is rough and irregular. When the slab has taken its initial set, the finisher rubs or floats the 
surface to an even finish with a steel or wooden trowel. This process brings considerable water 
to the surface, acting in a manner analogous to the tamping operation before referred to, so 
that the very fln« particles of cement »nd s^nd will rise to the surface. Dan*t flood the surface 
fm loo much IrovfUng, 


(A;) Use of Cement as a Surface Dryer, — It is not infrequently the case, particularly when 

very ^ret concrete is used and it is desired to hasten finishing operations, that dry cement is 

sprinkled over the surface of the partially set mass, and worked smooth with a trowel or float. 

In. this case, the liquid it is sought to absorb is a strong solution brought up from the body of the 

sla.b ; and to this solution the new cement further contributes like substances. This results 

in a deposit of fine silicious material from the sand and hydrolized cement a little below the finish 

with a skin of nearly pure cement directly at the surface. Necessarily, therefore, the body of 

the concrete slab and this thin veneer of neat cement at the surface are separated by a loose, 

non-adhering laitance film, so that scaling in patches of greater or less size soon results and will 

continue indefinitely until loose portions are entirely removed. Use less waief in mixing and 

avoid ctddxng cement as a dryer either neat or mixed with sand, 

(0 Use of Reiempered Concrete. — ^The setting of Portland cement takes place in two stages: 

(1) a gradual stiffening, known as initial set; and (2) an attainment of rigidity, with later slow 

hardening with passage of time. If the concrete for a floor has been mixed for some time, it 

may have taken its initial set. Adding more water and reworking renders the mass again 

plastic, so that it can be deposited much as might freshly mixed concrete. A certain valuable 

property, however, has become lost by this retempering, or rewetting, inasmuch as among 

other actions interlacing crystallization had already begun, and by rewetting and remixing, 

these crystals have been damaged and their reticulation destroyed. Such a floor, therefore, 

will be of inferior strength and density and possibly crumbly. Avoid losing concrete which has 

taken its initial set, 

(m) Loosening of Top Coat from Base, — Where floors are deposited in two layers and par- 
ticularly where the under floor as deposited is overwet, the upper layer (top coat) will be sepa- 
rated from the base by a scum of laitance deposited at the top surface of the under portion 
previous to setting. If the top coat in such floors is of good quality and sufficiently thick, it 
may stand ordinary shocks and wear, but if it is thin, or is subjected to sufficiently intense 
stress, it will become loosened and at best be unpleasant in use, giving a hollow sound when 
struck, or walked over, or in extreme cases, will become shattered and -crumble into pieces of 
greater or less size. Cast top coat and base at one operation wherever possible; and in other 
cases, remove surface of base to a depth of H in. or more and wash thoroughly, bonding to 
top coat with layer of rich cement grout. Do not permit this latter to dry out or set before 
top coat is applied. 

(n) Placing Concrete in Freezing Weather without Protection, — Contact between Portland 
cement and water results in a chemical reaction. As is well known, the speed of a chemical 
reaction is a function of the temperature. Below 60°F. this reaction is very slow, and so long 
as this temperature persists, there is comparatively slight formation of the binding or cement- 
ing substance, though it may be formed later, after the temperature has risen. Furthermore, 
when water freezes it expands with a force of approximately 300 tons per sq. in. with a volume- 
tric increase of 8%, so that in frozen concrete there is a general disruption and dispersion 
of components, possibly during the setting process, and to such an extent that in concrete 
floors, due to lack of hydrostatic head, they rarely, if ever, become again fully consolidated, 
regardless of how fully the cement may later react with water. Concrete floors which have been 
frozen, therefore, are weak and scaly. Use heated aggregates, heated waXer, and proper protection 
when concreting in freezing weather. Avoid the use of salt. Its advantages are not commensu- 
rate with its disadvantages. 

6. Remedial Measures. — Obviously, the best remedial measures, so far as the generality 
of concrete floors is concerned, are the avoidance of improper procedures and the tmceasingly 
careful institution of proper ones. The foregoing list of the causes of defects is, therefore, a list 
of remedial measures, so far as floors yet to be laid are concerned. Where, however, recognized 
defects exist in floors in place, it is a matter of serious moment to effect their repair. 

6a. Retopping. — Where head room and goods or machinery in place will permit, 
removal of existing top down to sound concrete, with thorough chipping, roughening and 


cleanBtng of the exposed surface and the careful laying of a new top of proper quality is sor 
times the most advisable and in the end, the cheapest procedure. Details of bonding top c< 
\ti base have been previously given. 

6b. Chemical Hardeners. — Sodium or magnesium fluosilic&te is maHceted unt 
various trade names as a liquid hardener. A pronounced change is to^M noted in the appei 
ance of concrete so treated with a greater or less hardening of the surface and correspondi 
resistance to abrasion, according to the initial condition of the concrete. The use of fluosi 
cates, oriRinally marketed as "fluates" has been known for many years, with a comparative 
recent revival through aggressive selling campaigns since the extension of uses for concre 
buildings, with corresponding increase in the number of defective floors. 

Ar. Use of Oils. — Linseed oil, both boiled and raw, with and without additior 
and adulterants has been tried as a binder for dusting concrete floors, but has not proven adt 
quate for all uses. Its value may be gaged by the resistive and retentive values that bd oxidize' 
film of a like oil might be expected to possess when subjected to the same traffic or other condi 
tions which gave rise to the original complaint. 

China-wood oil, sometimes called Chinese wood-oil, or Tung oil, either alone or in com 
bination with linseed or other oils or resins has proven somewhat more effective than plaic 
Unseed. It is today in extensive use a^ 
a basis for various concrete floor, wall, 
and water-resisting paints and in proper 
combinations is very effective. But ex- 
cellent as are the properties of this oil 
it is not necessarily an effective remedy 
for all dusting floors. The conditions 
attendant on each individual case 
should be minutely studied and the 
chances of success estimated, rather 
than attempting the haphazard appli- 
I cation of any palliative, proprietary or 
public, on the chance that the result 
desired will be secured. 
Fio. 8.— Paint film on i!on<:rc(e floor. (Miiaiu(i*d 20 diuni.) Sd. Floor Coatings and 

Paints. — It is oftentimes desirable for 
reasons of sanitation or of appearance, on good as well as on dusting floore, to use a surface 
coat of paint. A number of excellent floor paints are on the market which can be had in a 
variety of colors, the surface obtained being hard and sufficiently resistant to abrasion for 
most purposes and capable of withstanding the action of water almost perfectly. Ilie man- 
ner in which such a floor paint penetrates into the body of the concret«, leaving a hard re- 
sistant film at the top, is shown in the micrc^raph of Fig. 3. 


7. Structural Functions. — Concrete floors and concrete sidewalks have similarity of func- 
tioning as abrasion and impact-resisting surfaces and dissimilarity of functioning in that side- 
walks have solid bedding and low concentrated load and are rarely called upon to act as beams. 
There is further dissimilarity in that the perfection of surface of a concrete floor is not required 
of the concrete sidewalk, since the latter is in the open where slight surface disintegratioos do 
no harm and are not noticeable. 

8. Essential Qualities. — The aame general conditions governing the manufactureof concrete 
floors apply to the manufacture of concrete sidewalks since their requisites are held in common. 
A concrete sidewalk is also subject to the same character of disintegrations as is a concrete 

'" 263 

ion i»3 

all »t«, i 

■^ firry off ***^ «[ itote '1'**'^ . t lo '^"^ " , . ;« o-purat"' '"' 


ture greatly prolongs the period of plasticity, leaving water uncombined and free to form 
disrupting ice. 

Even with heated aggregates, therefore, and later surface protection, great care must be 
exercised to be sure that minute local disruptions (Fig. 4) due to frost have not made the con- 
crete, particularly at the top, ready for further disintegrations, either through attrition or by 
other actions. The best and safest plan is not to lay concrete sidewalks in freesing weather. 

9/. Protecting Sidewalks in Hot Weather. — In view of what has been said in 
Art. 5 under "Floors" with regard to the effects of floating and troweling in bringing water to 
the surface and remembering further the requirements of cement in regard to quantities of 
water necessary for slow hydration and chemical interactions, it is readily to be understood 
that rapid evaporation in hot, dry weather will cause hair cracks, through a too rapid removal 
of water, leaving behind and directly at the surface such of its products aa may be in solution, 
with further weakening of the concrete through deprivation of the remaining cement of its 
necessary reacting water. 

To guard against such happenings, as soon as possible after finishing, the pavement should 
be covered with a moist protecting canvas supported above the surface by a frame, or by a 
layer of moist sand, either canvas or sand being kept moist for several days. Only by observing 
precautions such as these can success be obtained. 

9^. Special Surface Finishes. — ^The same special finishes such as carborundum, 
or iron, suggested for concrete floors (Art. 4) are advocated for concrete pavements by their 
sponsors. The same remarks apply equally to both uses, as does a restatement of the fact 
that no additive so far brought out is capable per se of overcoming defects in manufacture and 
that none is superior in all-around qualities to good clean quartz of proper grading. 

10. Vault-light Pavements. — Basement areas under sidewalks in cities are a valuable asset 
but it is necessary that they be cheaply lighted. This desirable end is accomplished through 
the use of a combination of round or square lenses of thick glass set in cement mortar, formed 
into pavement blocks of requisite size, reinforced with steel rods and supported on steel or 
concrete girders. A template is used in spacing the lenses, the mortar being troweled in place 
as in usual cement mortar work. From the standpoint of considerations previously presented, 
this type of sidewalk should and does prove very isnduring, by reason of the large area of resist- 
ant glass exposed to abrasion and the lessened area of cement mortar, but care should be taken 
to have the cement {>ortion as carefully proportioned, mixed, and laid, as it would be where the 
entire wearing surface is of cement mortar. Joints between adjacent blocks and adjoining 
constructions are sealed with a waterproofing compound. Shattered vault light sidewalks 
are frequently seen, this shattering being due to the wheels of heavily laden hand trucks, or to 
improper working or use of the cement mortar, with later disintegration through frost or 
other actions, with loosening of lenses and permitting of continued spalling of their edges 
until the light-transmitting properties of the glass are impaired, or the lens broken com- 
pletely. These diflliculties, however, do not reflect upon the value of this construction 
properly executed. 

11. Concrete Citrtiing. — Concrete in curb and gutter construction has met with much 
favor. It proves generally excellent provided it is properiy underdrained. Integral curb and 
gutter blocks are made in lengths having definite cross-joints but no longitudinal joints, as such, 
by the action of frost or vegetation, would invite separation by frost or by the action of pene- 
trant vegetation. 

12. Sommaiy. — Concrete sidewalks, no less than other constructions, are structures of 
concrete; and to be enduring, they must be good concretes. Furthermore, they must be 
particularly protected against natural disintegrating forces, chief among which is frost. This 
requires good subdrainage as a prime requisite; and of almost equal importance, a dense struc- 
ture resistant to the penetration of water. To secure this, rigid observance of the principles 
of mixing good concrete must be insisted upon. The man with "20 years' experience" may 
only be one who persisU} in the least advisable procedures. 



IS. Structural Functions. — Concrete sidewalks and concrete roadway pavements are 
simOar in that they are impact and abrasion-resisting surfaces bedded on a continuous sub- 
base; and they are dissimilar in the perfection of the surface required and in the degree of 
impact and abrasion which they must sustain. 

14. Essential Qualities. — Ck>ncrete roads, no less than concrete floors, are, first of all, 
concretes, so that their character is governed by 4he laws basically controlling the making of 
good concrete. Particularly must provision be made against disintegration through weather- 
ing, through the heavy impact of shod hoofs and tires of loaded vehicles, and through the 
landing action of fast motor traffic. The best means for this purpose at present known are: 
(1) proper subdrainage; and (2) the securing of high-density concrete having adequate 
cementation of adequate quantities of rock products in the wearing surface, through care in 
selection and proper proportioning of materials; through adequate mixing, careful placing, and 
proper curing. 

16. One-course and Two-course Pavements. — The tendency at the present time is towanl 
the use of one-course, rather than two-course roadway pavements. A certain roughness of 
surface is very desirable to prevent slipping; and in a pavement with the concrete properly 
proportioned and mixed there is a requisite amount of rock material in the surface to withstand 
abrasion, with the added advantage that the slab is monolithic, without separation planes, as 
is so often the case where one course is laid upon another which has already set. The general 
method of finish is, however, similar screeding, floating, and troweling being done in the usual 

16. The Making of Concrete Roadways. 

16a. Porous Subbase. — Inasmuch as a concrete roadway pavement should not 
he called upon to sustain beam stresses, a continuous unyielding bed must be provided, else 
cracking and unequal settlements will occur. Such a solid bedment of the concrete slab requires 
special provision against upward heaving by frost action, which necessitates either a porous 
subbase of a depth sufficient (together with side drains) to clear the ground of water below 
frost line, or else one having a sufficiently yielding nature to permit local ground eruptions 
without extension of the disturbance to the concrete slab above it, though this latter is almost 
impossible to secure. 

It is regretable that so little attention is paid to this important feature of subdrainage. 
The majority of concrete roadways are placed directly on rolled ground, often without pretense 
of drainage and in some cases, with even a pronounced dish toward the center. To counteract 
the effect of settlement or frost action encouraged by this initial fault, reinforcement is embed- 
ded in the slab, Ho of 1% of steel being the Joint Conference recommendation, but twice to 
three times that amount being actually necessary to secure a modicum of assurance. Concrete 
roadways should be capable of enduring and rendering splendid service without reinforcement. 
Actually, even the reinforcement too often proves inadequate to the unnecessary task imposed 
on it. 

166. Proportioning and Selecting of Materials. — Arbitrary proportions are 
generally used in pavement work, regardless of possible benefits obtainable by better grading. 
For one-course work, average specifications call for 1:2:3 concrete; and for two-course work, 
1:2)4 : 5 concrete in the base with a topping of 1 : 2 mortar, although a concrete, using fine stone 
instead of sand requiring even less cement, would be preferable. 

The principles of proportioning and selection of materials for concrete roadways as laid 
down by the Aggregate Committee of the 1914 National Conference are as follows : 

1. For fine aggregate, uae only sand or other fine aggregate free from very fine particles, and which has been 
■etiially tested by mechanical analysis, and for the tensile strength of standard mortar. 

2. Use coarse-grained sands or hard stone screenings with dust removed. 




3. Use sand or other fine aggregate that is absolutely clean. 

4. For coarse aggregate, use hard stone, such as granite, trap, gravel, or hard limestone. 

6. If bank gravel or crushed stone is used, always remove the sand or screenings and remix in the proper 

If local conditions prevent following any of these rules, adopt some other material than concrete for your 

More detailed requirements for fine aggregate are: 

The sise of the fine aggregate shall be such that the grains will pass when dry a screen having 34 -in. openinjp>. 
In the field a H-in. mesh or in some cases a \*i-\n. mesh screen may be used for this separation. 

Not more than 10% of the grains below the U-in. sise shall pass a sieve having 50 meshes to the linear inch, 
and not more than 2 % shall pass a screen having lOD meshes to the linear inch. This is an exceptionally coarae 
sand, but eoane sand is a necessity for a durable pavement. 

16c. Joints. — Expansion joints must be provided to prevent cracking due to 
temperature changes. These should be provided at linear intervals of from 30 to 50 ft. depend- 
ing upon the climate of the region in which the pavement is situated. Joints should also be 
placed at changes in grade, and longitudinally between curbs. The usual joint is from y^ to 
^ in. wide and the National Conference recommends a preformed plastic filler. A variety 
of special compounds for this purpose arc on the market and also metal-and^lastic inserts 
intended to reduce the surface exposure of joints to a minimum. 


f%rosnt Oif Mufei aivin^ majciTTHini slieiiutli 

Fio. 5. — Proper consistency of concrete for road work.* 

16d. Curing. — Difficulties that have arisen in the use of concrete pavements 
have forced recognition of some usually neglected though well-known principles of concrete 
making. Among these is protection of the pavement for a period of time against drying in 
warm weather and frost in cold weather. For the purpose first named, earth dams permitting 
flooding the pavement with water have been used with success; also protecting canvases, or 
layers of sand, or earth, or sawdust have been used each being kept moist. In the winter time 
protection against frost has been obtained by the use of hay or straw and sometimes manure. 
The use of the latter, however, is dangerous, inasmuch as manure not only discolors the concrete, 
hut may bring about dtsintegration through penetration and decomposition of oiganic acids. 

16e. Consistency. — Observation indicates a general tendency to mix pavement 
concretes too wet (see Fig. 5). The proper consistency is plastic, permitting compacting and 
molding, with surface finishing, but without runoff, or separation of ingredients. Overwct 
concretes in roadway pavements cannot be expected to have better qualities than the stme 
character of mix possesses in other types of structures. 

1 D. A. Absams: Bulletin PorClaiMl Caaient 



By Ai>£IS£Rt p. Mills ^ 


1. S t reng t h in General. — The strength of mortar and concrete made with a given cement 
is dependent primarily upon: (1) the inherent strength of its aggregates, particularly the large 
aggregates; (2) the proportion of cement per unit of volume of mortar or concrete; (3) the 
degree of compactness or density of the mortar or concrete; and (4) the time afforded subsequent 
to final deposition in molds or forms. 

(1). With a given aggr^ate (sand, or sand and broken stone or gravel) the mortar or con- 
crete strength will, other things being equal, increase with increased proportion of cement in 
the mixture. 

(2). With a given proportion of cement in the mixture the mortar or concrete strength will, 
other things being equal, increase with increased density of the mixture. 

(3). Strength normally increases with age for an indefinite period. 

The first of these rules does not apply in comparing mortars or concretes which have been 
made with different aggr^ates, with different cements, with different proportions of water 
used in gaging, or with different methods of mixing or placing the mixture, nor in comparing 
mortars or concretes which have cured under different conditions. 

The density of the mixture secured will depend primarily upon the gradation in sizes of 
the aggregate, but may also be affected by the manner and extent of manipulation of the mate- 
rial in mixing and placing, and by varying the proportion of water used in gaging. 

The second rule fails as a basis of comparison when different cements are used, when the 
a gg re ga tes possess a different mineral character or are to a differing degree contaminated with 
mineral or organic impurities, and may not apply when methods or conditions of making, 
placing, and curing of the mixture vary. 

The third rule is independent of most other factors except that certain circumstances 
such as the proportion of water used in gaging and the conditions of curing nxay affect the rate 
of increase of strength with age. An apparent falling off or retrogression in tensile strength 
after from 3 to 6 months is commonl3' noted in tests of neat cement, and less commonly in 
tests of mortars. A less-marked retrogression in compressive strength is also frequently ob- 
served in tests of neat cement, and a sUght retrogression in compressive strength of mortars 
and concretes is not infrequently observed after long periods. 

A number of empirical formulas have been derived by various experimentors which attempt 
to express a definite mathematical relation between strength of mortar and concrete and the 
absolute volume of cement and sand, or cement, sand, and coarse aggregate, in the mixture. 
Owing to the fact that the effect of many factors, some of which are mentioned above, cannot 
he taken into account by such formulas, their application is restricted to certain classes of 
laboratory investigations which do not involve many of the variables encountered in the use of 
similar materials on construction work. 

2. Laboratory Tests, Their Use and Significance. — The value of cement mortars and 
concretes as structural materials depends primarily upon their mechanical strength and dura- 
it Pnrfcnor of M>t<tri>li. Cornell Univenity. Author, "Materiak of Coutniotion.'* 



bility when hardened. The conditions encountered in practical use, however, are necessarily 
Ko variable as to exchide the possibility of the establishment of standards based directly upon 
practical experience. 

The only established American standards for mortars are those for tensile strength (see 
Appendix A). These standards merely fix values of tensile strength, determined under labora- 
tory conditions, which experience has shown may be expected of cements and sands found 
satisfactory in practical construction work, in order that inferiority in any particular mortar 
materials may be detected by deviation from such standards. In other words, knowing that 
good concrete or mortar sands will in laboratory tests show a tensile strength not inferior to 
that of standard sand mortars, it is assumed that any sand which exhibits like tensile properties 
in the laboratory will not fail to satisfactorily meet the conditions of structural use. 

The conditions encountered by a mortar in a structure are not duplicated in the laboratory, 
but the laboratory method is so standardized that the external conditions of the test may be 
duplicated elsewhere or at a different time. 

Two factors operate to lessen the importance of laboratory tests of mortars as an indication 
of suitability of the material for construction uses: (1) the closeness of the relation between the 
results of tensile tests and the qualities which a mortar in a structure will be called upon to 
show may properiy be considered open to question, and (2) laboratory tests of this class of 
material cannot be made with any great degree of precision, and the results may be very 
much in error if the work is not performed under proper conditions by a skilled operator. 

Mortars are never used structurally in such a way that they will be depended upon to 
carry tensile stress, while they are commonly used to carry compressive stress. From this 
circumstance it may be argued that a compressive test in the laboratory will afford a more 
direct indication of the structural qualities of the material than does the tensile test. The 
compressive test is less easily made than the tensile test, however, and calls for more elaborate, 
more expensive, and less portable equipment. It is not easy to establish the relationship 
between laboratory test results and structural qualities of mortars, but the tensile test has been 
made so much more generally than the compressive test, that the average man has no experience 
by which he may judge the value of the latter, while he has observed that materials which pass 
the tensile test infrequently prove unsatisfactory in a structure. Whether the compressive 
test will fail any less frequently as an indication of unsuitability remains to be conclusively 
shown, but the somewhat inadequate data available seem to point toward this conclusion. 
It is undoubtedly a fact that certain natural impurities in sands, as well as certain classes of 
material sometimes intentionally added to mortars for special purposes such as decreasing 
permeability or altering the appearance, reveal their injurious character to a much more 
marked extent in compressive tests than they do in tensile tests. 

The results of laboratory tests of mortars are affected by a number of factors, not all 
of which are readily subject to control. It is never possible to determine precisely the relative 
qualities of mortar materials tested in different laboratories or by different operators. The 
atmospheric condition of the laboratory, with respect to both temperature and humidity, is 
one important factor which ought not to be subject to variation, but is so nevertheless. The 
most important consideration, however, is the fact that very slight variations in the detail 
methods of manipulation of the materials in making test specimens affect the test results to 
so great an extent that a very close check between the results of two different operators is 

An experienced operator may be able to check his own results within say 5%, but a second 
equally-experienced operator who can also check his own results thus closely, may not be able 
to check the first man withm less than 16 or 20%. Each man has developed invariable methods 
of manipulation, but the methods of the two men will never be identicaL This fact need not 
invalidate comparative tests, however, and the results of tests of standard mortar and mortar 
made by the same experienced operator with a commercial sand substituted for the standard 
•and should be truly comparable. 


Bearing in nuad the cauadeniians >bow dianund. it nnr fae(«ncjnd«<d: ^r (hat hbora- 
tory tceeptaoBt te«U of mortu- nmst not be roBsidend ta Aow the absolutr stMngUi vhk-h 
n»y be dcTcU^ml in » guarttue. but mertjy to indk*)^ tbr tippmamxW rdalive vahtr of tbr 
proposed mfttemls mud rnktcnak vrba^' eahabilny has becti pror«d; mod t,2' that Uhoratofy 
testa must nrver be minteted to Mn-oraP vho hai not h*d s lat^ (spenence in ""t-ing this 
partk-uUr kind of lest vitb the mdvuiuife ol « fuBr fqumiMl iBbontorr. The testing of 
roDcrete materiak is not the job far a novice, and tb« avera^ firid laboratory i$ not a fit plan- 
to do the wMk. 

S. Heat, Mortv, and Coacrrtc Sti^sEth Ca^^ared. — A MunpahsoD of thr strniflh of 
oeinent, cement mortar, and caocrne can raih- be made wbra the many variable SmetOK which 
inflnence the resolte of (est« l)aTP been eliminated so far as if practically paeeihle. Tbis mcana 
that only tests made with idaitical matcnalB imder the same ansfHccs are truly compatablc. 

Ihe strengths of the cement, mortar, and concrete mixtures shown graphically by the diagrams 
of Figs. 1 and 2 are based upon one series of tests made by the Techoologic Branch of the U. S. 
Geological Survey at the Structural Materials Testing lAboratories formerly located at St, 
Louis, Mo. The complete report of these tests is contained in BulUtin 344 of the U. S. Geolog- 
ical Survey and Technologic Paper 2 of the U. S. Bureau of Standards. 

The diagrams for neat cement and 1:3 standard sand mortar are averages of three tests 
of each mixture for each of nine separate brands of cement. For the commercial sand mortars 
and all of the concrete mixtures, a blend of these nine cements was used. The diagmms aver- 
age the results of three tests of each mortar and 18 to 21 tests of each concrete. (The irregu- 
larities shown by the diagrams for the commercial sand mortars would doubtless not appear 
if they represented the average of a laiger number of tests.) The same commercial sand was 
used in the mortare and all of the concretes. It is described as Merrimac River sand and is 
composed of flint grains having comparatively smooth surfaces. It has a fairly well-graded 
composition, its void cont«nt is not particularly high, but it is finer than is desirable. The four 
coaiae aggregates are typical well-graded aggregates of the classes indicated. It should not 
he concluded from these tests that granite or gravel aggregate can be depended upon to excel 
hmestone in all cases. Very alight differences in two apparently aimilsr materials of llic Humo 
class will often make a great difFerencc in their value as concrete uggrcgnto. 




It is not intended that the diagrams of Figs. 1 and 2 shall form a basis for definite con- 
clusions concerning the relative strengths of various cement mixtures, but only to show in a 
general way the results of tests of typical materials in various mixtures. The value of any 
given material can be determined only by tests of its qualities regardless of the qualities other 
materials of the same type may show. The more important of the many factors which affect 
mortar and concrete strength are considered below. 







■ aWI 



iflnrv - 






iifltt - 




MOM - 




c — 


M 9aot - 





^ MQC - 



~ TOOA — 

/ p >vwv 


" 6000 - 


S VMfl > 





1 • ' 1 ~ T 


•Qaooq - 
















































flODO "I 















ta n 








Age of specimens 

FiQ. 2. — ReUtion between comprenive strenKth of neat cement, cement mortars, and concretes 

(9 brands of cement). 

4. Aggregates of Mortar and Concrete. — Testing of the cement used in all important 
concrete structiires has been common practice for many years, but the importance of the 
quality of the aggregate, in its relation to the quality of mortar and concrete of which it forms 
a constituent, has not yet come to be adequately appreciated by the majority of engineers, 
architects, and contractors. While cement of good quality is essential to the -making of good 
concrete, its manufacture has today been standardized to such an extent by exacting speci- 
fications that in the average case it is actually safer to assume, without tests, that the cement 
is satisfactory than to assume that the aggregate materials most readily available may properiy 
be used without careful experimental determination of their quality. 

The principal requisites for concrete aggregates are structural strength and durability, 
a proper gradation of sizes of particles, and cleanliness or freedom from deleterious matter. 

The unsuitability of a weak, soft, or porous material is quite obvious, but a well-graded 
limestone aggregate may make better concrete than a harder granite aggregate whose void 
>tent is high, and a coating of matter partly or wholly of organic origin upon the particles 


of the best-graded granite aggregate obtainable may cause the concrete in which it is used to 
have exceptionally poor qualities. 

The physical testing of aggregates is, unfortunately, not controlled by any generally- 
accepted specifications, but methods are at the present time undergoing standardization by the 
technical committees of the most interested National engineering societies. Some of the 
largest engineering oiiganizations, as well as State, Federal, and municipal public service 
commiasions, employing large quantities of concrete, make a regular practice of subjecting 
all aggregate materials used to a systematic physical examination. The variability of aggregate 
materials available in different localities, and even of an aggregate from a single source of 
limited extent, magnifies the importance of tests not only in choosing the most suitable aggre* 
gates from those available for construction work in any given locality, but also in certifsring 
the quality of all shipments of that aggregate to the job. This entails a considerable expense 
for the testing of materials which possess a very low intrinsic value. It is an expense which 
may be justified, however, by the direct benefit gained by a thorough knowledge of how avail- 
able materials may best be utilized. An unsatisfactory material may be greatly improved by 
washing, perhaps, or by screening and readmixture of the different grades in different propor- 
tions, or the judicious mixture of two available materials may be found to yield a material 
greatly superior to either one alone. 

It is very desirable, also, that the proportions of the mixture of cement and fine and 
coarse aggregate be not rigidly specified, but only that the physical properties of the result- 
ing concrete shall be up to a fixed standard of strength or, in some cases, density or im- 
permeability. This will often mean that when the local materials are inferior, a concrete of 
the required quality may be attained either by using the local materials in a rather rich 
mixture, or by importing better aggregate from a distance, using a leaner mixture. Thus 
the relative costs of the additional cement used with the local materials, or the freight charges 
on imported aggregate will be the factor which deteimines the choice. 

For tests, specifications, and properties of aggregates, see chapter on "Aggregates'' in 
Sect. 1. 

5. Effect of Mineral Character of Aggregates. — The structural strength and durability 
of concrete aggregates is dependent upon the mineral character of the rock from which it is 
derived. In the case of coarse aggregate of artificially crushed stone, the original qualities of 
the rock have obviously not been altered. When the rock has been broken down into gravel 
or sand through the operation of natural agencies, the structural qualities of the individual 
particles of the material will still be identical with those of the parent rock except for possible 
changes effected by chemical agencies. 

The principal classes of rocks from which concrete aggregates are derived are granites, 
trap-rocks, limestones, and sandstones. Granite is an igneous Vock of variable structure 
and texture, whose principal mineral constituents are quartz and feldspar with varying 
amounts of mica, hornblende, etc. The structural qualities of granites vary greatly but 
granites as a class rank among the hardest, strongest and most durable stones. The term 
trap-rode is commonly used to include basalt, diabase, and a number of other igneous rocks 
possessing similar chemical and physical properties. The principal mineral constituents 
of most of these rocks are pyroxene and feldspar. They are commonly rather fine-grained, 
hard, tough, and durable. lAmestones contain carbonate of lime, calcite, or carbonate of 
lime together with the double carbonate of lime and magnesia, dolomite, as the essential 
constituent. Sand and clay are common impurities and some varieties contain large 
amounts of shells and other fossils. Limestones vary greatly in structure, strength, hard- 
ness, and durability. Some of the limestones are superior in structural qualities to some 
of the granites, but the average limestone is inferior to the average granite or the average 
trap-rock as concrete aggregate. Sandstones consist of grains of varying sizes, chiefly 
quartz, bound together by silica or iron oxide or, less frequently, by lime carbonate or 
clay. The structural qualities, strength, hardness, and durability of sandstones vary 




greatly according to the texture, and the class of the binder. A siticious binder exceb 
my other in all respects; an iron oxide binder is usually, though not slwaya, superior to 
one of lime carbonate; and sandstones having a clay binder &re in all respects least valuable 
as concrete aggregate. 

Fig. 3, which is based upon tests made at the U. S. Bureau of Standards ^Ttek. Paper 58), 
shows the great variability in strength exhibited by concretes made with different clawee of 
coarse aggregate aud with different aggregates of the same general class. The proportions 
were 1 :2 :4 in all cases, and the same cement, a blend of nine standard brands of Portland 
cement, was used throughout. Two river sands of similar character and somewhat similar 
granular analysis were used in making the specimens of each coarse aggregate. It will be noted 
that, except in the case of the granite concretes which included only four different aggregates, 
the range in strength of concretes with different aggregates of the same class is often more than 
100% of the average strength. 

All sands are derived from rocks which have been broken down or disint^prated 
through the operation of purely-physical agencies, without change of mineral identity, 
or which, in addition to disintegration, have been more or less decomposed by chemical 









.i " 






















1 ,"■*!:•-' — Jl 

Fid. 3. — Compuativa ■ttanithi of conertte with ti 

>• typa of m^grtt^tt. 

agencies involving the formation of new compounds. The principal disintegrating agen- 
cies are temperature changes, which are operative because of the unequal expansion and 
contraction of the component minerals and because of frost action, and abrasion caused 
by the flow of water, glacial action, or by winds. Chemical decomposition is accom- 
plished through the solvent power of water, facilitated often by the presence of various 
chemically-active substances, acids, etc., carried by the water. 

Quarts is the mineral which makes up the bulk of the particles of most sands. Thin 
is due to the fact that only the harder constituents of rocks survive as sand, and quarts is 
not only a very common constituent of rocks, but is also very hard and resists chemical 
decomposition. The fact that quarts is the principal constituent of a sand does not 
insure its suitability for concrete, however. Comparatively small amounts of certain 
minerals like mica or even feldspar or hornblende, or very small amounts of organic impuri- 
ties will render a quarts sand altogether unfit for use. Sandstone is a common source of 
quarts sand, and the quality of the sand will depend upon the character of the binder of 
the original rock, since the individual particles of sand arc made up of still smaller parti- 
cles of quarts hound together by silica, iron oxide, lime carbonate, or clay according lo 
the class of the rork. Sands are seldom derived from trap-rock or granite directly, 


though sand beds may often be largely made up of the constituent minerals of tbcn; 
rocka. Pyroxene and horablende &ra complex silicates possessing a degree of hardness, 
strength, and durability slightJy inferior to that of quarts. Hornblende particularly has 
inferior weathering qualities. Feldspars are essentially silicates of alumina with potash, 
■oda, or lime. They are considerably less strong and durable thaii quarts. Mica is a 
very objectionable constituent of sands for concrete. It is soft, has low strength, particu- 
lariy in shear, and its laminated structure promotes the percolation of water. Its surface 
is also of such a character that the bond secured by cement is very poor. Limestones do 
not serve as a source of concrete sands although calcite and dolomite may occur in sands 
derived from sandstones having a calcium carbonate binder. Limestone screenings or 
crusher dust are also sometimes used as fine aggregate though the concrete made there- 
with is usually inferior in strength to that made with an average sand, and it is also apt 
to be, or will in time become, more permeable, owing to the solubility of calcite and dolomite. 

Sand depoaita being rarely of a residual character, but usually deposited by stream or 
glacial action, and being also of sueh a character that the percolation of surface waters 
through the beds is very easy, 
the material is often contami' , 
naled by matter, much of it ' 
of organic origin, carried in [ 
suspension by water. Thus | 
the coating of the grains by j 
such substances as tannic acid i 
is frequently encountered. ^ 
The effect of such impurities i 
is extremely detrimental and 1 
the difliculty with which their .< 
presence may be detected in- 
rreaaes the importance of care- ~ 
ful teats of concrete sands. 

S. Effect at 9tape and Size * 
of Afgregate*. — Specifications 

frequently call forashaip sand, ^„g in days 

U.. one composed of rough Fio. 4»,-E(fKto!.i«of «dd upon l,«,trenBth 1:1 morOr. 

angular particles, in spite of (Twu of R. F. D«vu.) 

the fact that in many localities 

river or beach sands having somewhat rounded particles are the only ones obtainable, and 
have been used with perfect satisfaction. The shape of the particles is chiefly important in 
so far as it affects the void percentage of the material, and rough, irregular particles do not 
romparct better than rounded particles. On tbe contrary, rounded particles which afford 
BO opportunity for bridging will compact into the densest mass. In some instances tbe 
surface facets of angular particles may afford a better adherence for cement than the 
roctnded suri'aces of water worn particles, but this factor is leas influential than is 
tbe density of the mass, and mortars and concretes made with aggregates composed of 
worn, roonded particles are not inferior in strength to those made with "sharp" aggregate, 
and very often excel the latter in strength. The same consideratiotks apply in the cose of cnubod 
stone m. gravel as coarse aggregate. Concrete ol excellent quality or very inferior quality 
may be made with either class of material, the mineral character of the particles and ths 
grwiatioa td aiies being much more influential factors than the shape of the partit^tes of 
•ggngate. Tl»: leqoiiement tA sharpneflB is based upon all erroneous idea of the positive 
advantage gained; it olten works hardship, or injury, or is unenforcible, and should be 
ofDilted from q)eeifications. 

The |M«vailing sixe ot the particles of aggregates M more important than tbe shape, but 


far 1m8 important than the gradatioa of eiies because o[ the direct relation of the latter to 
denaitjr and strength. A comparatively coarse sand is always preferable to fine sand. It 
has 8 smaller surface to be coated with cement per unit of volume and therefore requires 
less cement to produce a mortar of a given strength. It is less difhcult to fill its interstices 
with cement than in the case of fine sand, and a denser mass is usually secured with the 

same proportion of cement. 
E A composition of coarse sand 

* and finer material which will 

B servo to fill the voids in the 

S coarse material will usually 

■- excel either a very coarse or 

£ a very fine aand alone aad 

e will lead to economy of ce- 

t mcnt. This is particularly 

^ true when permeabiUty is im- 

'I portant, a mortar made with 

(* a combination of coaree aad 

fine sand, or one in which the 

'" "a™ In'Sy. "° »"» "' l>"'i'l" •" ««" 

. graded from coarse to fine, 

Fia. 4&. — ElTect oluK of uod upon tciuileattencth 1:3 morUi. . . , . . ,, 

(Twu of R. P. D»vu.) bemg less permeable than one 

made with either excluaively 
coarse or exclusively fine material. 

■ The relation between size of sand and tensile strength of mortar is shown by Figs, ia, 
46, and 4c for 1:1,1: 2, and 1 : 3 mortars, respectitrely. These diagrams are based upon 
tests made in the laboratories of the College of Civil Engineering, Cornell University, by R. P. 
Davis (" Materials of Constnio 
tion," by A. P. Mills, pp. 152- 
153). The various sands used 
were prepared artificially by 
separating a natural beach sand 
of nearly pure quartz into eight 
sizes or combinations of sizes. 
It is shown that the sand which 
passes the 20-mc8li sieve and is 
retained on the 30-mesh sieve 
produces the strongest mortar 
for all mixtures and at all ex- 
cept the early ages. The sand 
of all sizes finer than that pass- 
ing the lO-mesh sieve ranks 
second for all mixtures, the 
blend of equal amounts of lO- 

20 and 30-40 sand ranks third, 7 88 W 180 mo 

10-20 sand ranks fourth, and *9» '" *^ 

all finer siiea of aand rank Fi"- ■«■;— Kff«i 
lower in the order of their fine- 

7. Relation Between Densit7 and Strength.— The term density is employed referring to 
mortars and concretes, meaning the ratio of the sum of the absolute volumes or absolutely 
solid substance of the individual constituents contained in a measured unit volume of mitrtar 
or concrete to the measured unit volume of the materiala combined in the form of mortar or 

Sec 6-71 



concrete, water being neglected as an individual constituent. In other words the density is the 
solidity ratio, the ratio of the volume of solids to the volume of the mass of mortar or concrete. 
Many experimental studies have shown that the strength of mortars and concretes is 
directly proportional to the density of the mixture. The density of the mixture is dependent 
partly upon the thoroughness of mixing, the amount of water used in gaging, etc., but is pri- 
marily dependent upon the gradation of sizes of the aggregates. Aggregates in which the 
relative amount of particles of different sizes is such that the particles of one size just suffice 














, '^^4l-» 






7 3 


■ •. 




ff 1* 





C 300 


& « 















MOO a» MHO ao tw 

.650 AM ATO JBflO J690 300 710 720 

.P«nsity of rnor+ar 

730 740 .750 7M .770 7S0 790 

Fko. 5a. — Rriation of "density" or solidity ratio of 1:3 mortar to tensile strength. Age, 13 weeks. 

to fiU the voids of the next larger size will have a minimum void space, and will therefore re- 
quire a minimum proportion of cement to secure a product of given strength. The ideal con- 
crete of maximum density would be made with aggregates of this character, and, in addition, 
would be so proportioned that the mortar just suffices to coat the particles and fill the voids of 
the coarse aggregate, and the cement just suffices to coat the particles of sand and fill its voids. 
This ideal can of course only be approximated in practice because the larger particles of each 








1 § 1 1 i § 








44 ** 













to ■ 




■" 1 



















390 AOO M iCO OO MO JbSO MO A70 jUO MO TOO .710 .TtO J30 740 750 .160 770 


Fig. 56. — RdatioD of "deDsity" or solidity ratio to compressive strength of 1:3 mortars. Age, 13 weeks. 

aggregate will not closely approach each other. Owing to the wedging action of the smaller 
particles the larger stones and grains of sand are forced apart so that the density of the mixture 
is certain to be less than that theoretically possible. A slight excess of mortar over that theo- 
retically required is usually beneficial to density and strength. 

The relation of density of 1 : 3 mortars to tensile strength is shown by Fig. 5a which com- 
prises tests of 157 different natural sands made by the U. S. Bureau of Standards (TecA. Pawf 



[Sec 5-S 

68). The numbers on the diagram indicat« the order of fineness of the Bands, No, 1 being the 
couaest, and No. 157 the finest. The corresponding relation of density and compreaahre 
strength (or the same mortars is shown by Fig. 56. Note that Figs. 5a and 56 show that 
the density of mortars is in general inveisely proportional to the 6neness of the sand used, the 
finer the sand, the lower the density. 

w> - 

tZ - " " - _ "-:'■" 

iZ - - " 

iz - --- .^- ' - - 

iZ- ^^ 

IVu ^ 

liw • '' 

\Z "^1.^'^ ■ 


Fid. 6.— Rebtlon betimn dcrwily knd n 

Afc.4 WMki 

The relation of density to compressive strength of 1:2:4 concretes made with various 
aggregates is shown by Fig. 6. The diagram is based upon tests made by the U. 8. Bur«au of 
Standards (Tech. Paper 58, Tables 23-26), The data are not extensive enough to definitely 
fix the relation sought l>ecsUBe of the inevitable wide variation in test results due to variable 

Fid. 7.— Effect a[ m 

ID laniite itraicth o( I : 3 ibuubird Mul m 

factors other than density. That the relation ia a direct one, approximately that shown by 
the straight line upon the diagram, is, however, a justifiable conclusion. 

B. Effect of Micft, Clay, and Loam in Aggregates.— The occurrence of mica, clay, and 
■n in aggregates has been explained in connection with the coDsideration of mineral 


I ompooition in Art. 5. The very detrimental efiect of mica upon the strength of 1 : 3 
standard sand mortar ia shown by Fig. 7. The diagrams are based upon testa made by W. 
N. Willis (.Eng. New*, vol. 54, p. 145). The loes in strength amounU to IS to 25% with 
2H% of roica in the sand, 25 to 45% with 5% of mica, and becomes still more marked as 
the proportion of mica is increased. Mr. Willis also observed that increasing the proportion 
iif mica increasnl the voids in the sand from 37% with no mica, to 67% with 20% of mica. 
The weight was at the same time lowered 20%, and the amount of water required in gaging 
the mortar was 3 times that required ia gaging 
niiirtani free from mica. 

The efFect of clay in sands is dependent upon 
its state of subdivision and the uniformity with 
Mhich it is distributed through the sand. In most 
laboratory teats the addition of clay in moderate 
amounts has been found to be beneficial, l^^pical 

results of laboratory teats are exhibited by Fig. 8 ■- >«rtanti^ of ciay imw^ 

which ia derived from tests made bv P. L, Roman Fio. s.— eitke of cUy uMn urtnsth of I ^ 3 

(Eng. A C<mi., vol. 43, p. 403). With the materials X' m'i3S'L"he iZ-'SoVy.r" """"" "''""" 
here used the maximura increase in strength due 

to clay additions was observed to be about 20%, and was secured with additions of 10 to 15% 
of flay. Similar tests made by L. T. B. Southwick and G. A. Wellman {,Eng. Ree., vol. 63, 
p. 332) show that maximum strengths of l:l^i, 1 iS, 1: 4)^, and 1 ;6 mortar mixtures are 
secured with 3%, 10%, 15%, and 20% of clay, respectively. These laboratory results do not 
prove that similar percentages of clay will be beneficial or harmless in natural clayey sands. 
The manner of distribution and degree of fineness of the clay in concrete sands will be the de- 
termining factors, and the amount permissible will not usually approach the above limits. 
Lumps of clay do not become broken up in concrete mixing, and should be carefully excluded 
from aggregates. 

Loam, in the usual acceptance of the term, is earth which is made up of vegetable mold 
together with clay or sand or both. It is extremely injurious to mortars and concretes because 
of its content of organic matter. Fig. 9, derived 
from the series of tests of F. L. Roman above re- 
„. ferred to, shows the effect of loam and organic 

h matter in sand upcm strength. From these testa 

o it appears that 5% of this loam (about 1.5% 

^ organic matter in the sand) reduces the strength 

g of the mortar about 20%, and other amounts are 

£ nearly proportionally detrimental. Organic 

> , matter naturally occurring in sands is frequently 

J , found detrimental to an even greater extent than 

^. _ , .. J is indicated by these tests, wherein the organic 

ii., 9-— EBeotoforfm^^lMinuponsirensthof j:3iojj^ in the shape of fine powder was mixed with 
the sand. Oi^anic matter not infrequently covers 
sand particles with a film which is not easily perceptible, but which tremendously retards the 
normal rate of hardening and gaining strength. An investigation made by Sanford E. Thomp- 
son (TroTu. Am. Soc. C. E., vol. 66) led to the conclusion that organic matter constituting over 
10% of the silt and at the same time over 0.1% of the sand is distinctly injurious. 

9. Effect of Consistency. — The important relation of consistency to the strength of mortar 
and concrete is shown by Figs. 10 and 11. The amount of water is expressed as a percentage 
of the total dry weight of aggregates and includes any moisture carried by the sand or coarec 
aggr^ate in its natural condition. These diagrams are based upon tests made in the labora- 
tories of the Sheffield Scientific School under the direction of Prof. Barney [Eng. and Cont., 
vol. 42, p. 244). 



These teste show that a quite definit« percentage of water is required to produce & mortar 
or concrete of maximum strength nith given materials, for the particular materials uaed in 
this case the ina\iniuni 1 :2 murtar strength was nttaincd with about lfi.6% of water, and tbe 
maximum strength of 1:2:4 concrete with about 8.4% of water. For other materiala or 
other mixes these consistencies for maximum strength will not remain the same, but for any 

ftrc»nto9« of watw bawd on Wal dry yii«(jW of 
Pro. 10. — ES«t of eaiuuttnc7 upon itrencth of I : Z mottar. 

mixture of given materiale there is a critical consistency which will be productive of higher 
Htrcngth than any drier or wetter consistency. This tact is particularly important in view of 
the rommon practice of using extremely wet concrete mixes in order to be able to deliver the 
i-onTete on the work cheaply and expeditiously by the use of chuting devices between muter 
and furms. The fact should be understood that such wet mixtures may be used only with a 

Fta. 1 1 . — EScct of eonuteDCT upon (trcDCth a[ 1 : 1 : 1 oonmta- 

considerable sai-rilice in strength of Ihc < un< rete pla< ed. Un the other hand, these t«t8 indicate 
that nothing is gained l>y making a concrete so dry that it must be rammed or tamped in place 
in:itead uf being pudtlled. Conr'retm containing 8 to 9% uf water are of a sufficiently musby 
consistency to be readily puddled, but from 12 to 15% uf water, or even more, is commocily 
used to produce the fluid consistency flesirable for chut« or spout dcliveiy. 


For hAimf III effects Irom the use of excess water and for suggested proceduree, see chapter 
on "WKier** in Sect. 1. 

111. Comprc B WTe aad Tenale Streagflifl Comp arw L — The i^mpreasive and tensile strengths 
of the same mortar mixture may be contrasted by a comparison of the diagrams of Figs. 1 
and 2, pagn 217 and 218. A direct comparison for 1:3 standard sand mort&r is afforded 
b>~ Fig. 12 which is baaed upon testsof seven brandsof cement made in the Btructural Materials 
I^bo iat oryfonneriyma i nta ine datSt.Louis, Mo.,bytbeU.S.Geol(^icalSurv^(U.6.Geol.Surv. 

Ag* ir\ days 

to Uuile atrencth oH : 3 aUodanl Portlnad 

huU. 331). The specimens used were standard tension briquettes and 2-in. compression cubes. 
1 1 is ebown by the diagram that the average value of the ratio of compreBsive strength to tensile 
strength is far from being constant as the age increases because of the relatively more rapid 
rate of gain in tensile strength during the first fen' weeks and the very sUght gain or actual 
retrogTeoBion which characterises the tenwle strength after the first C months. The average 

value of the ratio for all cements is about 6 between the 1-month and the 6-month periods, 
but the individual brands of cement show variations of from 15 to 40% from this average. 

A similar series of tests, made under the same auspices, with a blend of the above 7 brands 
of cement and 22 commen ial sands in 1 : 3 mortars has been made the basis of the diagrams of 
Fig. 13. Because of the very wide variations shown by the 22 different mortars, only the 
average for all the mortars, the average (or the 5 mortars showing the highest value of the 



[Sec. 6-1 

ratio and the average for the 5 mortars showing the lowest value of the ratio are plotted. The 
same variation in the ratio of compressive to tensile strength with age shown by standard sand 
mortars is exhibited by the commercial sand mortars. 

An inspection of Figs. 12 and 13 w^ill show that the tensile strength of mortars is not more 
than a very approximate indication of the probable compressive strength of similar mortars 
with the same cement, and even then the age, the mixture, and the sand used are souroes of 
variation which must be taken account of. 

The tensile strength of concrete is a property of little importance because, being low in 
comparison with the compressive strength, concrete is practically never designed to carry 
tensile stress. When concrete is used in situations involving tensile stress it is more economical 

Tensile and Compressive Strenqths of 


Character of coarse 
aggregate and mix 

Age, montha (approx.) 



(lb. per sq. in.) 



(lb. per sq. in.) 

tensile strength 



compressive strength 



Limestone 1 :2:4 6 








Average. . . . 





Sandstone 1:2:4... 















Average. . . . 





Sandstone 1 -.2^ :5... 

6 ' 2 















1 1,889 







Average. . . . 







to ttae steel reinforcement than to use the very large sections which would be required if the 
roncrete were depended upon to cany tension. 

An indication of the comparative strength of concrete in tension and compression is afforded 
by the table shown on page 228w These data were derived in tests made by the writer in the 
laboratories of the College of Civil Engineering, Cornell University. The concrete was 
mixed and the specimens molded in the field. 

Note that the values of the ratio of tensile to compressive strength in this table would 
have been somewhat lower had the specimens been tested in compression at the same age they 
were in tension. 

11. Strength of Plain Concrete Columns. — The strength of plain concrete columns, as 
determined by teste of laboratory specimens whose dimensions are comparable with those of 
columns used in structures, is usually not less than' 75 nor more than 90% of the strength of 
cubes of the same concrete, the column length not exceeding 10 to 12 diameters. 

A number of series of teste of plain concrete columns are tebulated below and on page 
230. The strength of cubes of similar concrete is indicated where data from comparable teste 
are available. 

Strenqth of Plain Concbete Colgtmns 
Watertown Arsenal Teste ^ 

Mixture and 





character of ' 





coarse aggregate 


(lb. per 
sq. in.) 




1 : 1 Mortar 





1 :2 Morter 





1:2 Morter 

6 . 




1:3 Morter 





1 :3 Morter 





1 :4 Morter 

- 6 




1 :4 Morter 





1 :5 Morter 





1 :5 Morter 





1:1:2 (Pebbles) 





1:1:2 (Pebbles) 





1:1:2 (Trap rock) 





1:2:3 (Pebbles) 





1:2:4 (Pebbles) 





1:2:4 (Pebbles) 





1:2:4 (Trap rock) 





1:2:4 (Trap rock) 





1:2:5 (Pebbles) 





1:3:6 (Pebbles) 





1:3:6 (Pebbles) 





1 :3:6 (Trap rock) 





1:2:4 (Cmders) 





1:3:6 (Cinders) 





» "TeiU of Metals/' 1904, 1905. 



[Sec 5-12 

University of Ilunois Tests* 

col. strength 





strength of 


(lb. per sq. in.) 






strength of 


:ib. per sq. in.) 









cube strength 






4 g 



































12 in. eyl. 

12 in. cyl. 
12 X 12 
12 X 12 
12 X 12 
12 X 12 
12 X 12 
12 X 12 

12 in cyl. 

12 in. cyl. 

12 in. eyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyi. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 

12 in. cyl. 


University of Wisconsin Tests* 







12 X 12 
12 X 12 
12 X 12 
12 X 12 


I BtiiU. 10 and 20 of the Univ. of 111. Eng. Ezpar. Sution. 

s Data from tests of cubes made at ages which do not correspond even approximately to the age of the column 
made from the same oooerete have been omitted. 
* Bull. 300. 

18. Bffect of Method of Mixing. — ^The method of mixing mortars and concretes may 
vary with respect to: (1) amount of water used; (2) duration of mixing operation; and (3) 
detail method of manipulation. The effect of v^ariation in the amount of water used is con- 
sidered in Art. 9. The effect of the duration of the mixing operation is shown by Fig. 14 
which is based upon tests by Prof. Scofield of Purdue University (Eng, and Cont., Jan. 17, 1915). 
All of the concrete was mixed in a Chicago Cube Mixer of 2H-cu. ft. capacity, run at the rate 
of 26 revolutions per min. These concretes are all much stronger than the average com- 
mercial concrete but this fact does not affect the significance of the test results. It appears 
that with the particular materials used a very decided advantage is gained by operating this 
mixer much longer than the usual period of mixing. The actual time of mixing most advan- 
tageous to the quality of concrete produced under given conditions will probably vary greatly 


with different materiala and with different mixers. It is very probable, however, that the 
ftv«age roncret« used in every day practice would be ironaiderably improved in quality, if it 
were mixed tor a longer period. Thb b certainly true of roni'retes which are turned out at a 
rate of a batch per minute as is sometimes the case. A distinct advantage ia gained by mixing 
beyond the pomt which produces a batch of even color. The mass becomes more viscous; 
there is less danger of separation of fine and coarse material; for agiven water content it appears 
to possess a wetter consistency and flows better in transporting by chute and in depositing in 
the forms; and it forma a concrete of greater density, less permeability, and greater strength. 

It cannot be said that machine-mbiing will invariably produce better concrete than hand- 
iDixing, but for all except the smallest work it is less expensive and is, therefore, generally pre- 
ferred. Hand-mixing is more opt lo be severely slighted thnn maihine-mixing because of the 
heavy labor involved and Ihe comparatively long time required. 

Total time of mixing in minutes 

Pla. 14.— Effect o( timo of miiing upon ttrcngth of concrele. 

In comparative testa of concretes made with the same materials, weighed and molded by 
employees ot the laboratory of the U. S. Bureau of Standards, but mixed in the field by three 
different contractors, each being permitted to use his own methods of mixing, both hand and 
machine, all conditions being the same except the actual mixing of the materials, variations 
<A as much as 70% in compressive strisngth were obtained (Tech. Paper 58, U. S. Bureau of 

13. Effect of Uethod of Placing. — The importance of the effect of the methods of manipula- 
ting and molding laboratory specimens of mortar upon the qualitiea of the specimens shown 
by tests has been discussed in Art. 2. The same factors arc operative in the case of molding 
laboratory specimens of concrete, and their disturbing effect becomes even more pronounced 
when the work is done in the field. Experimental data arc lacking which might show the extent 
of the effect of variations in molding methods, but an indication of the importance of this fac- 
tor is afforded by any scriea of tests of mortar or concrete specimens made from the same batch 
of material and stored and tested in an identical manner. A number of such apparently iden- 
tical specimens may perhaps vary in strength less than 5% it made by an experienced operalor, 
but ■ second equally-experienced operator using the same materials and the same general 




methods will often obtain results which are not within 20% of those of the first man. 1 
deviation must be attributed primarily to slight differences in detail methods of molding' 

When special methods of handling the materials are considered, only very scanty comi>£ 
tive data are available. The following tests reported by R. E. Goodwin of the Matex- 
Testing Division of the New York Public Service Commission indicate that concrete placed 
mass may possess greater strength than when it is cast in molds of the size ordinarily \xi 
(Eng. New8y Feb. 18, 1915). Several pieces of concrete were cut from existing subway etr 
tures at places designated before the concrete was placed in the forms. While the concr 
was being poured in the forms, samples from the same batches were cast in molds. A port i 
of the specimens thus molded were stored in moist sand on the work, while others were stoi 
in the laboratory moist room. In addition similar specimens were made from the same ma. 
rials brought from the work and mixed and molded in the laboratory. The pieces of conert 
taken from the work were cut from various portions of 12-in. walls at the age of 2 mont 
and after being rough-dressed were polished smooth to exact dimensions. All tests were ma. 
at the age of 00 days; the concrete was a 1 : 2 : 4 mixture ; and all specimens were 6 by 6 by 12 i 

CoMPREssrvE Strength of Field and Tebt-b ample Concrete 

Concrete made on the work 
(All in one line are from the same batch) 

Concrete made in the laboratory 
(Samples of same materiak from the work usee 


cut from 

12-in. wall 

poured in 
molds and 
stored on 
the work 

poured in 

molds and 
stored in 

moist room 

of batch 


made in 


and stored 

on the work 


made in 


and stored 

in moist room 

of batch 

(4) 3,096 

(4) 2,410 
(4) 2,415 
(4) 2,760 

(4) 2,880 

(4) 2,870 
(4) 1,720 

(3) 2,060 

(4) 1,776 

very wet 

very wet 

(4) 2,136 
(4) 2,225 
(4) 1,980 
(4) 1,900 

(4) 2,080 
(4) 1,930 
(4) 1,705 
(4) 1,910 


(2) 2,685 
(4) 2,020 


(16) 2,670 

(10) 2,480 

(15) 2,110 

(16) 2,060 

(16) 1,910 

(Figures in parentheses indicate number of specimens averaged for each result.) 

The quality of concrete deposited under water is usually considered to be decidedly in- 
ferior to that of concrete placed under normal conditions where water is not encountered, 
^rhis is doubtless true if the material is permitted to fall freely through the water, or if the cir* 
cumstances are such that the formation of laitance is facilitated. That first quality concrete 
can be made in subaqueous construction was shown, however, during the construction of the 
Detroit River Tunnel. In this case concrete of 1:3:6 mix was deposited at a depth of 60 
to 80 ft. below the water surface through 12-in. tremies. Test cores cut from the tremle- 
deposited concrete by a 6-in. shot drill showed a compressive strength of from 2740 to 4000 lb. 
per sq. in. 1 year after deposition. Other specimens in the shape of roughly-cut 6-in. cubes of 
1:2:4 tremie-deposited concrete developed a strength of from 1800 to 3040 lb. per sq. in., 
and it was believed that their strength was impaired by the operation of cutting (Trans. Am. 
Soc. C. E., vol. 74, p. 338). The matter of the pressure under which this concrete was deposited 
probably has some bearing upon the quality, for in this case the hydrostatic pressure at the 
bottom of a tremie tube was about 30 lb. per sq. in. 

14. Effect of Regaging. — The Joint Committee on Concrete and Reinforced Concreto 
recommends that ''the remixing of mortar or concrete that has partly set should not be per- 
mitted" (Proc. Am. Soc. C. E.. Dec, 1916, p. 1673), and most engineers specify that mortar or 




•iifiTi^te shall be used within 1 hr. or even yi hr. after it is gaged. It is undoubtedly generally 
drisable on construction work to adhere to the practice of not permitting regaginig and it is 
jartieulaiiy important that concrete which has stood undisturbed for some time be not per- 
mitted to get into the form in 

Effect of Regaging upon Strength of Mortars — Age 

4 Months 

Office of Public Roads Tests^ 

t^ uMiiilastic condition, but 
'hf harmful effect of regaging 
t: often less pronounced than 
> commonly believed, and 
'leeptions to the general rule 
my under certain circum- 
•".aoces properly be made. 

The data on this page 
•how that the strength of Port- 
iaod-cement mortars is not in- 
jinnidy affected by allowing 
thfm to stand for periods of 
iram 1 to 3 hr., and then re- 
Ptnng and molding. In fact 
the delayed treatment ap- 
pears slightly beneficial owing 
probably to the increased 
unount of working given the 
loaterial. This effect is one 
tbat has been shown with singular unanimity by a considerable number of experimenters using 
lU classes of cement. One fact brought out particularly by the tests of Mr. Sabin is that the 
Qiaterial regaged should not merely be remixed, but should have sufficient water added so 
tbat the original consistency will be restored after regaging. 



made up 

into briquettes 


after mixing 

Mortar broken up 

after initial aet and 

made into briquettes. 

Water added to 

restore normal 


Mortar broken up 

after final set and 

made into briquettes. 

Water added to 

restore normal 


Tensile strength in lb. per sq. in. 





Initial set 
Final set 

1 hr. 42 min. 
7 hr. 15 min. 

Effect of Regaging upon Tensile Strength of 1:2 Mortar — ^Age 1 Year 

Tests of L. C. Sabin< 



1 hr.. 


Stood 3 hr., 


each hour 

and then 


Stood 3 hr., 

regaged each hour 

with water added 

to restore original 

oonsistenoy and 

then molded 

Stood 5 hr., 


each hour 

and then 


Stood 5 hr., 

regaged each hour 

with water added 

to restore original 

consistency and 

then molded 




and then 





and then 


No water added 
579 I 565 

Water a 

dded to 

m regagmg 

restore ori 

ginal consistency 


• • ■ • 




Natural cements and quickngetting Portland cements appear to be less capable of showing 
^^ undiminished strength after regaging than do normal or slow-setting Portlands. 

The most pronounced efifect of regaging of mortars and concretes is in the direction of 
retarding the set and delaying the hardening, thus reducing the strength at early periods. 
Candlot ("Ciments et Chaux Hydrauliques, " 1898, p. 358) found that mortars regaged after 
attaining their final set all required 8 to 10 hr. to set regardless of the rapidity or slowness with 
^^c\i the mortar originally set. This effect of regaging alone will often be a suflBcient cause 

* U. 8. Department of Agriculture, BiUl, 235. 
''* Cement and Concrete," p. 253. 



[Sec. 6-15 

for prohibiting regagcd mortar or concrete on construction work requiring an early assumption 
of load. 

Candlot also found that regaging had a very detrimental effect upon adhesive strength of 
mortars to stone, the loss being often 50%. Sabin (''Cement and Concrete," p. 290) also 
found that regaging was detrimental to adhesive strength of mortar to stone, the effect being 
more pronounced with rich mortars. Earnest McCuUough {Eng, News, Jan. 11, 1906) found 
that regagcd mortars showed a loss in power to adhere to old mortar or concrete, but found that 
the addition of 10 to 12% of lime to the regaged mortar produced a material whose adhesive 
strength considerably excelled that of mortar placed when freshly mixed. 

16. Effect of Curing Conditions. — The principal variations in curing conditions which 
affect the process of hardening and gaining strength of mortars and concretes are: (1) varia- 
tions of moisture conditions, and (2) variations of temperature. 

The effect of different conditions of exposure to moisture is shown by the following data 
based upon tests made at the U. S. Bureau of Standards {Tech. Paper 58). The test specimens 
were 8 by 16-in. cylinders. 

Effect of Variation of Moisture Condition in Curing Period 

Tests of Bureau of Standards' 

Mix, clans of ccincrete. and curing conditions 

Comprps^ive strength (lb. per aq. in.) 

1 week 

4 weeks 

13 weeks 26 weeks 52 weeki 

1 : 6 gravel (quaking) : In damp closet entire period. 

• • • 





1:6 gravel (quaking): 4 weeks in damp closet, 
then removed. 

■ ■ • 





1:2:4 trap rock (quaking): Immersed immedi- 
ately after molding. 

• • • 


3,570 -h 

4,094 + 


24 hr. in damn closet, then immersed f 

• • • 


3,978 + 


4,247 + 

8 weeks in damp closet, then immersed 

■ • • 




1:2:4 gravel (mushy) : Sprinkled daily for 1 week, 
then stored indoors in dry room. 




4 weeks in damp room, then placed in open, exposed 
to weather. 

• • • 



1:2:4 gravel (quaking) : In damp closet entin period 
Open air, exposed to weather entire period 

« • • 


These tests indicate : (1) that concrete specimens cured in the moist air of the damp closet 
until tested become somewhat stronger than ones inunersed in water after a longer or shorter 
period in the damp closet, or ones immersed immediately after molding. (This may be due in 
part at least to the fact that the immersed specimens were tested while still holding much more 
water than the damp closet specimens.) (2) Specimens cured in ttie damp closet are con- 
siderably stronger than ones cured in the comparatively dry air of the laboratoiyv even though 

> Tteh. Paptr 58. 




the l&tter were ipnnUed d&ily for the first week. (3) Specimens gained strength in the open 
air exposed to the weather to a considerably greater extent than did others cured indoora in a 
comparatively dry room, but not to aa great an extent as ones stored in the damp closet. 

The rdation between the mean temperature encountered during the curing period and the 

I of Fig. 

This diagram \ 

Erti^ngth of 1:2:4 concrete is shown by the diagra 
structed by A. B. McDaniel as a summary of 
the tcsulta of testa made at the Engineering Ejc- 
periment Station of the TJniveraity of Illinois IBuU. 
81). The ctHicrete used was a 1:2:4 mixture by 
weight (1:2.2:3.6 by volume), the coarse ance- 
g«te being crushed limestone. A portion of the 
specimens were 6 by 6-in. cylinders, some were 
6-in. cubes and the balance were 8 by 16-in. 
cylinden. All values were reduced to an equiva- 
lent value for 8 by l&-in. cylindere, however. 
Ten setA of specimens were made, and each set was 
subjected to a different mean temperature through- 
out the period of curing. The mean temperatures 
employed were 26.5% 27.1°, 30°, 34.7°, 35.5°, 
48.5% 68°, 71.8°, 72.8°, and 95.6° respectively. 
The marked effect of low temperatures in at least 
delaying, if not permanently preventing, the 
hardening process is excellently shown by the 

diagram of Fig. 15. The diagram represents the results obtained with only one mixture of 
one class of materials, but the relative effect of various temperatures on other concretes 
may be expected to at least approximate the relation found for this concrete, and the dia- 
gram should furnish sugge»' 
Effkct op CrntNG 1 : 4 Mortar under Steam Pressukb 
(8 X 16-in. cylindere) 
Testa of Bureau of Staodaids' 


unterwl durioc the iwiin 

tive information useful ii 
tinmting the strength of con- 
crete cured at abnormal tem- 


Certain classes of con- 
crete products such as tiles 
and blocks are subjected to 
an accelerated hardening 
treatment by the use of steam. 
The effect of such treatment 
upon strength is shown by 
data given on this page from 
tests of the U. S. Bureau of 
Standards [Tech. Paper 58). 
The results show that up 
to 80 lb. per sq. in. gage prea- 
Bure, steam has an accelerat- 
ing action upon the harden- 
ing of cement mortar, and 
that the compressive strength increases with the pressure as weU as with the time of exposure 
io steam. A compressive strength considerably {in some cases over 100%) in excess of that 
obtained normally after aging for 6 weeks may be obtained in 2 days by using steam under 
pressure for curing. Furthermore, the eteam permanently aecelerat«s the hardening of the 




c™^....™.. 1 

*■ ■" ' 






















































concrete which subsequently mcreases in compressive strength with age upon exposure to the 

It was noted that the steam-cured concrete was more uniform in appearance and lighter 
in color than normally-aged mortar from the same materials. These tests were made upon 
Portland-cement mortars, but the same conclusions were found to apply to concretes. Tlie 
mortar or concrete should obtain an initial set before exposure to the steam treatment. 

16. Bffect of Freezing. — ^The effect of low temperatures in delaying or permanently pre- 
venting the hardening of mortar and concrete has been shown by Fig. 15. In the event of the 
temperature being close to the freezing point of water from 4 to 8 times as long a period 
is required to obtain a final set as is required at normal room temperatures. If water in mortar 
or concrete freezes before the cement has set, it is not available for the chemical action of setting 
and hardening and hence the concrete or mortar will not set at all until the ice melts. These 
facts must be borne in mind when removing forms from concrete placed during cold weather. 
If the temperature hovers above the freezing point for some time after concrete is deposited, 
there is a possibility of the water drying out before the greatly delayed setting has taken place. 
If, however, the concrete has begun to set before the temperature drops considerably below the 
freezing point, the expansion of the water in solidifying produces an expansive force in excess 
of the cohesive strength of the green concrete. This action results in a destruction of the bond 
and crumbling of the concrete when the ice melts. If the temperature does not fall more than 
a very few degrees below freezing, the result may simply be the further delaying of the set 
without appreciable injury. 

Two factors operate to lessen the injurious effect of freezing upon mortar and concrete: 
(1) concrete is a rather poor heat conductor, the outer portion therefore serving as an insulation 
for the bulk of the material and preventing an injurious low^ering of the temperature in the 
interior of the mass; and (2) the chemical action of setting and hardening of cement being an 
endo thermic reaction, the heat evolved serves to raise the temperature of the material and 
so offsets to a degree the loss of heat by radiation and conduction. Experimental data secured 
during the construction of the Arrow Rock Dam and the Kensico Dam (see Art. 42) indicate 
that mass concrete shows a rise in internal temperature of from 20 to 40°F. above the initial 
temperature within a period of from 15 to 30 days. 

Serious injury is often suffered by concrete which encounters temperatures considerably 
below the freezing point within the first few hours after placing, but this injury is usually con- 
fined to the outermost portion of the work and seldom penetrates more than an inch or two of 
depth. A very frequent form of injury is a scaling off of a very thin crust of rich material 
which has been flushed to the surface in finishing the work. 

Numerous experimental studies of the effect of frost action on mortars have been made and 
have led to somew^hat conflicting conclusions, but practically all of these have involved the use 
of such small specimens, briquettes, 2-in. cubes, etc., that the condition of exposure is compar- 
able only to that of the outermost surface layer of concrete. 

Authorities are quite in accord in prescribing that '* concrete should not be mixed or de- 
posited at a freezing temperature, unless special precautions are taken to avoid the. use of 
materials containing frost or covered with ice crystals, and to provide means to prevent the 
concrete from freezing after being placed in position and until it has thoroughly hardened." 

17. Effect of Salts. — Common salt (NaCl) is frequently used as an ingredient of the mixes 
of concrete or mortar which must be placed in cold weather. Its primary effect is the lowering 
of the temperature at which water will freeze. Approximately 1 % of salt in the mixing water 
lowers the freezing point l^F. Calcium chloride (CaClt) is also used to a lesser extent to 8en*e 
the same purpose. 

The effect of common salt and calcium chloride upon the strength of a 1 :2 :4 limestone 
concrete is shown by the diagrams of Fig. 16 which are derived from tests made by H. £. 
Pulver and S. E. Johnson of the University of Wisconsin {The Wisconsin Engineer, October, 
1013). Hie specimens were 4-in. cubes, and the tests were made in duplicate, one series of 


speeimenfl being cured indoors at normal temperatures (60°-75°FO, the other cured out of 
doora or ID a refrigerator at temperatures below 33'F. The amounts of the salts used are 
ezpretBed as perrentagea by weight of the mixing water. 

Tlie testa show that common salt used alone is quite injurious to the strength of concrete 
cured at normal temperatures, the loss being roughly proportional to the amoimt of salt used. 
With concrete cured at temperatures below freeiing, however, it facilitates the hardening 
proceaa. The tests show an increase of strength for the addition of N'aCI up to 12%, after 
wbich there is a decrease. Comjnon salt retards the setting of concrete to a considerable degree. 

Calcium chloride used alone is beneScial to the strength of concrete, whether cured at 
normal temperatures or below freeting, up to about 4% CaCIt, at which point the maximum 

'^ P Bii< i i t aj i NaU t^TcenTuga nai.1 ^ Peixentage NaCI 

Fio. IS.— Eff«t of ulu upoD ilRDctb o( I : 2 : 4 concret*. 

ipceiraens cund >t Donni! tciiipenlun> <«> to 7S° FX 

ipcamciia Eumt Bllow tEmpntum (tx]oir32* P.). 

strcBfth ia obtained. This maximum strength, however, in the case of the cold-cured specimens 
is only abont one-half the maximum strength of the cold-cured specimena having 12% of com- 
mon salt. Calchim chloride accelerates the setting of concrete. 

With concretes cured at low temperatures the best effect was obtained with a mixture of 
2% of CaCl, and 9% (rf NaCI. This mixture gave about the same strength aa the cold-cured 
concrete having 12% of NaCI alone, and was not as detrimental to the strength of normally 
cured concrete aa the NaCI alone seemed to be. 


18. Effect of Hjdrated Lime and Wsterproofing Compaunds. — The addition of hydrated 
lime in email percentages has not a vety marked effect upon the strength of laboratory specimens 
of mortar and concrete. Fig. 17, which is based upon testa made by Prof. Harry Gardner 
of the University of Kansas {Bng. See, vol. 64, p. 3C9), shows the effect of varying percentages 
by weight of hydrated lime upon the tensile strength of 1 : 3 standard sand mortals. Tbe 
replacement of tbe cement by hydrated lime appears to be slightly beneficial to Htrength up to 
about 15%, except in the case of ieste made at tbe ages of 3 and 7 days wherein tbe effect was 
generally detrimental in proportion to the amount of lime used. Other tests, notably those of 
H. S. Spackman {Concrete-Cement Age, vol. 4, p. 112 and Eng. Ree., vol. 69, p. 25) have shown 
that small amounts of hydrated lime aometimea appear to atTert tbe strength favorably, some- 
times unfavorably. The most pronounced efTect of hydrated lime added to mortars and con- 
cretes is its producing a more plastic, better- working material. The fat, viscous mortar 
produced spreads better under the trowel, and in the case of concrete the presence of a small 
amount of lime hydrate tends toward the production of a mixture of greater uniformity by 
prevention of the separation of fine and coarse 
materials. This fact may constitute a distinct ad- 
vantage in the case of the average construction 
job which would not be noted under the ideal 
conditions of mixing and molding in tbe laboratory. 
The effect of a large number of commercial 
waterproofing compounds upon the tensile and 
compressive strengths of mortars in 1 : 4, 1 : 6, and 
1:8 mixtures has been investigated by the V. S. 
Bureau of Standards {Tech. Paper 3). Figs. I8a 
and 18b show the results of tests of the effect of 15. 
such compounds upon tbe compressive strength of 
J 1:4 mortar. These results are typical of the re- 

sults obtwned with all mortar mixtures, and the 
results obtained in tensile testa do not differ 
greatly from those obtained in compression. The 
proprietary compounds are, of course, not identi- 
fied by name, but they are classified as follows: 
p. , No. 27 is dolomitic hydrated lime; No, 28 is a 

t, - T— , , . , i' : solid chemically active filler designed to form an 

Fio. 17. — Effect of hydrmlcd lime npon taoila , i , ,. . . - i ,.11 .. «„.».. 

■tnofth at 1:3 auDdKnl merur. insoluble lime resmate Void filler; Nee. 29 to 36 

inclusive are water-repelling solid substances con- 
sisting essentially of stearic acid with soda and potash or Ume, designed to form an insoluble 
lime soap; No. 37 is cement containing a water-repelling substance; Nos. 38 to 40 inclusive 
are chemically active liquid fillers designed to fill the Voids with either tar, insoluble lime eili- 

19. Effect of Sea Water Dsed in Gaging. — Tbe use of sea water to gage cement mortars 
and concretes is almost invariably forbidden by specifications for work done in localities where 
the use of sea water might be convenient. It has not been conclusively shown, however, that 
the use of sea water instead of fresh water has a particularly harmful effect. Meaara. Taylor 
and Thompson ("Concrete, Plain and Reinforced," 1816, p. 106) found by a very limited number 
of tests of 1 1 2 : 4 concrete cubes that there was no appreciable difference in strength of specimens 
gaged with sea water and other specimens gaged with fresh water. 

Results obtained in comparative tensile tests of mortar briquettes made by Cloyd M. 
Chapman are shown by Figs. 19.a, 19,6 and 19,c (Enff. New*, vol. 63, p. 291). These tests 
were made upon three sets of specimens: series (A) sperimens were gaged with fresh water and 
cured in fresh water; series (B) specimens were gaged with fresh water and cured in sea water; 
wd scriea (C) epecimcns were gaged with sea water and cured in sea watet. The epecimcna 


were tensile briquettes and the sea-water curing was done in tanks in the laboratory using a 
frequently changed bath of sea water. These tests indicate that the use of sea water for gaging 

Age of test pieces ir 





— ?'''j2s5 














Ag« cif t«st pieces In weeks 
euEth of wkterproDfsd mortan. (One part Portlsad cemi 

Flo. IB. — EHeet upon Uiu 
Iratb nl«(— cund in Imh »* 
vith (ca w«tei — cured in *«■ ir 

iii not pulJcuUrly detrimental t« the tensile strength of mortars except in very lean mixtures. 
The condition of curing of these specimens wsa probably not as severe a test as actual immet- 


sion in moving sea water would have beco. Ou the other hand, the section of the specimens 
wss so small that any superficial or surface effect of the sea water would appear to have an 
injurious effect much greater than that suffered by concrete of large bulk exposed in sea wal«r. 
SO. Effect of Oils Uaed in Gaging.—The use of certain classes of mineral residual oils in 
gaging mortals and concretes with the object of dampproofing them or reducing their pennea- 

I Kudual oils upon compnwve Binngth o( 
[TcaU o( Locnti W>]J« Psie.) 

bility lends importance to the consideration of the effect of such oils upon the strength of the 
mortars or concretes in which they are used. Figs. 20 and 21 present the results of teste 
made by Logan Waller Page, Director of the Office of Public Roads (IT 8. Dcpt. of Agriculture, 
Bvil. 230)> It appears from these tests that the use of mineral oils up to from S to 10% of 

Fia. 31. — EBHt at muwnl nndiul oib opon ihimjii ■Innctli of soncrata. (TaU of Locui WmU« P*cf.) 

the cement lowers the comprcasive strength to a moderate degree only but that larger amounts 
may be very injurious. 

Other tests made by Arthur Taylor and Thomas Sanborn {Trant. Am. Sac. C. B., vol. 
76, p. lOM) using Weatcm aspbaltic oils showed a more marked falling off in strength of mortar* 

S«c. S-211 



Ihon was observed in Page's teats. At 28 and 50 days the eompivBivc Htruigllis of S-in. cubes 
uf 1 :3 mort&T nude by the incorpoTstioD of Weatctn aOa were u fiJIows: 
Mftnf oQa used for vatioua 

• AsPHAi-nc Oii3 UPON St«enoth or 1 : 
Compressive Straigtb 


SOdM^ 1 

lib, |w 1 ReUtivr 








Boikr fud 






Boiler fuel 






Boiler fuel 






Boiler fuel 






Richmond fuel 




Road on No. 6 




Liquid asphalt 




cammercial purposes contain animal 

oils, v^etable oils or admixtures 

of these. Such oils have been found 

to be capable of not only weakes- 

ing cement mortars and concretes, 

but actually to disintegrate con- 
crete in some cases, the effect bnng 

most pronounced in the early stages 

of setting and hardening (see testa 

of James C. Hain, Eng. Neutt, 

March 16, 1905). 

21. ElfiMt of X-ahance.— I^- 

tance is a whitish substance whkh 

is washed out of concrete and sub- 

nequently deposited as a ecum when 

there is an excess of water uaed in 

mixing (see chapter on "Water" in 

Sect, 1), or when concrete is depos- 
ited in water, or when water collects 

in pools on the surface of freshly laid concrete. The taitance consists of the finest flocculeut 

matter in the cement blether with some silt and clay from the aggregates. Its occurrence 

is explained by the formation of amorphous hydrates in the eaiiy stages of the setting of cement. 

The coinpoeition of laitance is practically identical with that of cement, but it hardens only 

very slowly and never acquires much strength. As a consequence, if not removed by water 
and brushes or by a steam jet, it forms a distinct 
plane of weakness between successive layers of 

fllO concrete. The washing out of a portion of the 

finest part of the cement means the loss by the 
* "^ concrete of just so much of its most valuable 

% lU constituent, because it is the impalpably fine 

£ portion of the cement which is most active in 

£ binding together the inert particles of the aggre- 

g, 10 gate. This same material alone does not develop 

f^_ great coheeiveness, however. A familiar example 

of this fact is afforded by the relative behavior 
I *> of very finely ground cements and cements of 

I eo only average fioeness in neat and mortar mixtures. 

. The cement of average fineness will greatly excel 

) in neat strength, but the cement which is ground 
Aga fntirval In ioys gj, fingjy that the proportion of impalpably fine 

_mile ■treiiBth nf material is veiy large will form mortara of greatly 

superior strength. 

S2. Rate of Increase In Hortar Strength 
— KetrogresHOn. — The relation between the early test strength and the subsequent gain in 
strength is shown by Figs. 22 and 23 which are based upon the tests of the former Structural 
Materiab Laboratory of the U. S. Geological Survey in St. Louis (U. 8. Geol. Surv. BuU. 
331). The rate of gain in both tensile and compressive strength for these 1 : 3 mortars (made 
with seven typical Portland cements) is shown to be approximately inversely proportional at 


all ages to the strength at 7 days, those mortars which show the lowest b 
strength at 7 days maintaining the best rate of gain in strength at all a; 


e or compreeaive 

FlQ. 23. — Rkta o[ in 

A^ interval in dayi 

cault lbs nitrate of SO to 70 tesUJ 

The following table indicates the relation between the eariy strength of mortar and sub- 
sequent retrogression in strength as determined by the Structural Materials Laboratory Teats 
above quoted : 

Retroorebsion in Stbenotb of 1:3 Standabd Mortaks 
Tests of Structural Materials Laboratory 




Strtnstli ■! 7 ilmyi 

(lb. p«T n. i" ) 

7 and IS 



M ind 

ISO d.y. 




Below 200 



200- 250 



250- 300 



300- 350 



Below 800 



800- DOO 





Above 1,500 





%%, Tnuuren* Stteuitli.— The transverse strength of granular brittle materials Lke 
mortan and eoncrrtes is best expressed by the Uodvltu of Rvpture. The moduhis of rupture 




is the ai^Mucnt stresB in the extreme fiber of a transvose test specimen under the load which 
produces rupture. For specimens of rectangular section of breadth b and height hj loaded 
centrally €m a span 2, the breaking load being W, the modulus of rupture is computed by the 

Modulus of rupture «= —j 

The extreme fiber stress thus computed is not the actual fiber stress because the formula 
involves the inaccurate assumption that the material deforms elastically for all stresses up to 
rupture. The comparative relations between results are not affected by this inaccuracy of 
the formula, however, when the tests compared are made upon specimens of similar material, 
because the computed values of the modulus of rupture are very nearly proix>rtional to the 
actual stresses. 

Since the extreme fiber stresses on the tension side and on the compression side of a beam 
of homogeneous material are equal, and the tensile strength of mortar or concrete is only a small 
fraction of the compressive strength, the transverse strength of mortar or concrete is almost 
wh(^y dependent upon the tensile strength. The modulus of rupture found in transverse 
tests will invariably be considerably in excess of the tensile strength, however, because the 
computed stress in the extreme fiber considerably exceeds the actual stress. 

The data on this page constitute a summary of a portion of an extensive series of tests of 
transverse strength of mortars and concretes made by Wm. B. Fuller. (*' Concrete, Plain and 
Reinforced " by Taylor and Thompson, p. 334.) The tests were made upon specimens 6 by 6 in. 

in section, supported on spans of 30 and 60 in. 
crushed trap-rock aggregate were used 
throughout the series of tests. The speci- 
mois were broken at the age of 1 month. 
84. Shearing Strength. — The shear- 
ing strength of mortars and concretes 
possesses great significance because com- 
pressive failure of short compressive speci- 
mens or structural members is usually 
failure by shearing on a diagonal plane, 
and because shearing stresses are impor- 
tant considerations in all cases of concrete 
beams. It is very difficult, however, to 
make experimental determinations of 
pure shearing strength because most 
methods and devices which may be used 
to make shearing tests involve either a 
cutting action, bearing pressures, or beam 
stresses. The data on the shearing strength 
of mortars given on page 244 are derived 
from tests made by Feret. The speci- 
mens used were prisms, 2 by 2 cm. in sec- 
tion, subjected to single shear, the condi- 
tions being such that beam stresses prob- 
ably affected the results considerably. 
Specimens were tested after 5 months 

One brand of cement and the same sand and 

Transverse Strength of Mortars and 


Tests of William B. Fuller 

by wdcht. 

by Tolume. 

Modulus ci rupture 
(lb. per SQ. in.) 

cement: sand: 

cement: sand: 



age of 6 





































:1.17: 2.06 





:1.17: 3.09 





:2.34: 4.12 





:2.34: 5.17 





:3.51: 5.17 





:3.51: 6.21 





4.68: 8.25 












Shearing Stbenqth of Cement Mortars 
Tests of R. Feret^ 

Approximate pro- 
portions by weight 

Ultimate strength (lb. 

per sq. in.) 

Ratio of 
shear to 

Character of sand 

^fc^ • ■ ^^ • v^^^ ^ ^**^ ^r ^ m^^^f^f^^^^ 







Very coarse granite sand 





















































Mediumnsized very shelly sand. . . 1 






1 ^ 







1 1 
1 1 






1 1 
1 -i 






1 1 

1 *■ 







1 *■ 






























Very fine silicious sand 1 






■ ^^ ^m _ y ^V^H^V^P ^i^ v^r ^v^P ^ ^>^ ^» ^"^ ^i^^^^^ ^i^^^^' ^ ^ ^fc*^ ^••^•^ »• • ^^ 










































1 i 0.3 






£qual parts of coarse medium and 1 






fine ground quartzitc 1 











20-31-me8h ground quartzitc. . . 1 




Neat Portland cement 

1 ' 





A ^ ^^V>B ■r V ^^ • • ■ W^* • ^ • ^0 ^^ • • • ^^ •.• V****** •• 

The data at the top of page 245, showing the results of shearing tests, are derived from tests 
made at the Massachusetts Institute of Technology under the direction of Prof. C. M. Spofford. 
The specimens were cylinders 5 in. in diameter and 15^2 in* long. The ends were securely 

> OaU Uken from "Concrete. Plain and Reinforced" by Taylor and Thompson, p. 136. 1900 Edition. 

Sec. S-24] 



clamped in cylindrical bearings and the load was applied along the middle third of the length 
by a semi-cylindrical block. The final failure appeared to be by true shear. 

The following data are 
taken from tests made at 
the University of Illinois 
Engineering Experiment 
Station under the direction 
of Prof. A. N. Talbot {Bvll. 
8). Two methods of testing 
were used. In the first a 
6-in. hole was punched in a 
concrete plate or block; in 
the second, a short beam 4 
by 4 in. in cross-section was 
securely clamped at the ends 
and the middle third of the 
length loaded. Three forms 
of specimens were used in 
the punching tests: (1) plain 
concrete plate; (2) recessed 

Shearing Strength of Concrete 
Summary of Massachusetts Institute of Technology Tests* 




Shearing strength 
(lb. per sq. in.) 



(lb. persq. in.) 

5 by 13^i-in. 


Ratio of 
shear to 


























Shearing Strength of Concrete 
Summary of University of Illinois Tests 


Form of specimens 

Method of 



(lb. per 

sq. in.) 

Compressive strength 
(lb. per sq. in.) 

Ratio of shear 
to compression 













Plain plate. 

Plain plate 

Recessed block. 

Recessed block. 

Reinforced recessed block 

Reinforced recessed block 
Restrained beam 

Restrained beam 



damp sand 

damp sand 





damp sand 








damp sand 

damp sand 





damp sand 






damp sand 
damp sand 





damp sand 





damp sand 
damp sand 





damp sand 













> Taken from Bull. S, Uni. of 111. Eng. Exp. Sta. 




Adhesioh or New to Ou> Concrete 

Transverse StreniEth of Joints — ^Tests of Hector St. 

Georige Robinson 

Method emplojred to 
• bond 

Computed tea- 


in ea t f em e 


(lb. per 

■q- in.) 

Solid specimens with no 



Surface (molded against 
rough board) merely 


Surface roughened with a 
chisel, cleaned and 


Surface roughened, 
cleaned, and thoroughly 
coated with neat cement 


Surface treated with hy- 
drochloric acid, washed, 
brushed, and wetted 













bond. % 







filled with additional fresh concrete. All specimens 

concrete block; (3) recessed oonciete 
block reinforced outside of the area 
subjected to the direct action of the 

S6. Adhesive StrengOu— The ad- 
hesion of mortars to various building 
materials is a matter <rf much impor- 
tance which has, however, been insufli- 
ciently investigated. Fig. 24 presents 
the results of tests made by General 
£. S. Wheeler (Report Cliief of Engi- 
neers, U. S. A., 1895, p. 3019, and 1896 
pp. 2799, 2834). Discs of the material 
concerned, 1 by 1 in. square and \i in. 
thick, were prepared and inserted in 
the center of briquette molds. The 
molds were subsequently filled with 
mortar and the specimens were tested 
in the usual manner in tension. Mr. 
Wheeler found that a consistency wet- 
ter than that which gives a manimnm 
tensile strength is required to give a 
maximum adhesive strength of mortar 
to stone, even though the surface of 
the stone be saturated with moisture. 
Irr^^larities of the surface of stone or 
brick appeared not to affect adhesive 
strength, but a dirty surface, or in- 
sufficient moistening of the surface 
greatly reduced adhesion. 

26a. Adhesion to Con- 
crete Previously Placed. — ^The adhe- 
sion of concrete to old work of the 
same character constitutes an impor- 
tant problem in many classes of con- 
struction work, but few experimental 
determinations of the bond between 
new and old work have been made. 
The data on this page have been derived 
from a series of tests made by Hector 
St. George Robinson in 1912 {Proe. In* 
stitute of Civil Engineers, vol. 189, p. 
310). The specimens used were prisms 
of 1:2:4 concrete 30 in. long and 4 
by 4 in. in section. One set were solid 
prisms. The remainder were made by 
placing a stop-board in the mold 8 in. 
from one end and allowing the con- 
crete molded in this end to harden for 
7 days before the stop-board was r^ 
moved and the balance of the mold 
were tested after further hardening for 


28 days. Four difierent treatments accorded the face of the old concrete to improve the 
bond »re enumerated. The testa were made by rigidly clamping the 8-in. portion of 
earh beam in a bxed support and loading the cantilever beam at a point 20 in. from the sup- 
port. The relative strengths of the various iointa thua tested woa determined by computing 
the extreme fiber stress on the tension side of the joint (modulus of rupture). TTie ap- 
parent tensile strength thus computed is much higher than the actual tensile stress, but the 
rdative efficiencies of the various methods of securing a bond are nevertheless shown. 

Fta. 24.— AdbMivc rtrength or Portland remenl mortir. 1 pirt ptment: 1 part cnuhed qnvtl (No*. 20-30). 
(Whedcr, Report of Chief o[ Enc'ri, 1893.) 

Mb. Adhesion or Bond to Steel.— See Art. 2, Sect. 6. 
S6. Strvngtta of natural Cement Hortar and Concrete. — The production and use of natural 
cement in the United States has declined so rapidly since 1899, when the amount produced 
reached its maximum of nearly I0,CO0,0CO bbl. per year, and greatly exceeded the output of 
Portland cement, that the present importance of natural cement as a material of engmeering 
construction is almost negligible in comparison with that of Portland cement. The reSAons 
for the great decline in importance of natural cement are briefly : (1) the great improvement in 
quality and lowering of cost cf Portland cement; (2) the tnfcrJDrity of the average natural ce- 



Pio. 3Sa. — Tnuile atreDgth of ulunl cemcDt— svernce of 10 bnndi. (SoUn.) 

ment to the average Portland cement in structural qualities; (3) the great variability in quality 
ahown by natural cements owing to the lack of close control of the manufacturing process; 
and (4) a general distrust of natural cement among engineers and others which is often alone 
responsible for its use being forbidden by spec ificalions. 

Natural cement, mortars, and concretes vary greatly in strength owing to a great varia- 
bility in both composition and constitution ot the <ement. "This variation is found not only 
in comparing cemmts from different localities, but even in comparing samples taken at different 
times from the output of any one locality. The on^ general statements that may be made 



ISw. S-27 

ronceraing their strength is that natural cements rarely show more than half the tensile strength 
of Portland cements of the same age, and their compressive strength rwely exceeds tine- 
third that of Portland cement in similar mixtures " (" Materials of Construction," by A. P. 
Mills). The digrams of Figs. 25a and 25b average the results obtained from tensile 
tests of mortals of ten representative brands of natural cement made by L. C. Sabin (" Report 
Chief of Engineers," 1895, p. '2937), and compressive tests of eight to nine brands of natural 
cement made by Clifford Richardson {Briekbuiider, vol. 6, p. 253). 

Very few data are avadable showing the strength of natural cement concretes. Tests 
made at the Watcrtown Arsenal in 1899 show the following strengths of 12-in. cubes of 1 : 2 : * 

e brand of typical natural c 

t ("Tests of Metals," I8«9 

(lb. per'i. in.) <lb- per'IHi. in.) 

(lb '^?^ in.) 

CompresBive strength at 3 months 

Compressive strength at 14 months 

400 263 
914 585 


Sabin ("Cement and Concrete," p. 314) quotes the following tests made by A. W. Dow for 
the Engineer Commissioner of the District of Columbia in 1897. The specimens were 12-in. 
cubes of 1:2:6 concrete, made with six different coarse aggregates and t«sted at the age of 
12 months. Comparative tests of a Portland-cement concrete made with the same aggregati^ 
with the same proportions are also reported. 

(lb. ^^. in.) 

(lb. per .q. Id.) 

(lb. pinj'in,) 

Compressive strength of 1 : 2 : 6 




Compressive strength of 1 : 2 : 6 

The standard specifications of the American Society for Testing Materials (A. 8. T. M. 
Standards, 1916) require that the minimum tensile strength of 1 :3 natural cement mortar 
made with standard Ottawa sand shall be: 

7 days (1 day in moist ai 
28 days (1 day in moist ai 

, 6 days in water)— 60 lb. per sq. in. 
, 27 days in water)— 126 lb. per sq. in 

17. Strength of Cinder Concrete.— The compressive strength of a number of mixee of 
concrete made with anthracite coal cinders and sui different Portland ccmenta is shown by 




the f<dlowmg summanes of two smes of tests made at the Watertown Arsenal C'Tests of 
Metals," 1896, 1903, and 1904). The specimens of each test series were 12-4n. cubes, and the 
average vahies given are means of from two to four tests. 

Strength of Cinder Concrete 
Watertown Arsenal Tests — ^Tests Made in 1898 


Brmnd of 


(lb. per sq. in.) 


of mixture 

1 month 

3 montha 































































Tests Made in 1903 and 

1904 (One Brand of Cement) 

Comprewive strength (lb. per sq. in.) 

5 weeks 

32 weeks 

1 year, 16 weeks 




Ave. of 3 



Ave. of 2 



Ave. of 4 

1:2 :4 
1:3 :6 










More recent tests of cinder concrete, the results of which should be indicative of the 
range of quality of the cinder concrete used in building construction, are summarized in the 
table on page 250. These tests constitute a portion of a study of " Cinder Concrete Floor Con- 
structiaii" by Harold Perrine and by George E. Strehan {Traru, Am. 8oc. C. £,, vol. 79^ 
p. 523). The qiedmens were 8 by 16-in. cylinders made by competent men with laboratory 
training, but the material waa taken from that going into the floors of various structures 
then in process of construction in New York City. The samples were taken without advance 



Perrine and Strchan Tests 

— '»— 1 



.—th.| 1,^ 

1:2:5 coiitinuou»-mbter concrete, lor-gnule cindcra. 








1:2:6 band-mbted concrete, good rindere 

notice been given the contractor, and the apecimenB, after being molded on the job, ' 

tested in the Columbia Univeruty Laboratory. 

Stbznoth of Cinder Concrete 
Structural Materiala Laboratories' Tests 













Testa of 1:2:4 ctoder concrete 
made at the Structural Materiala 
Testing LaborBtories at St. Louis in 
1909 are summariied on this pa^ 
{Tech. Paper % U. S. Bureau tA 
Standards). Teste of 21 cylindera 8 
by 16 in. are averaged. 

38. WoTking Stresses.— For work- 
ing stresses recommended by the Joint 
Committee, see Appendix B. 


». StreM-atnin Cnrres for Hortan and Concretes.— Typical strees-etrain cutvm fo 
a number of classes of 1 :2 :4 conrrete at the age of 1 year are presented by Fig. 26, Thes 
flurvea average the results of tests of 21 specimens (8 by 16-in. cylinders) for each elate of con 

« for conereta, {Cunrei u 

Ace 13 mo.) 

Crete, made at the Structurml Materials Laboratories at St. Louis (U. 8. Bureau of Standard* 
TecA. Paper 2). At earlier ages tests of these same concretes resulted in Btrem-etrain curves 
closely resembling these l-year-teat curves except that the slope of the curves is less, and the 
curvatun greater, at the earlier teat periods. 




Stre^HStnin curves for mortars exhibit the same i hararteristice as do these currea tot 

SO. Yield Point.— -Mortars and concretes are not perfectly elastic materials for any range 
of loading, there being a slight decrease in the proportion of stress to strain as the stress in- 
craues, and a slight pennsnent set for very low stresses. Even the first portion of the stress- 
strain curve is, therefore, not a perfectly straight line, but for purposes of curve plotting it is 
sensibly so up to a certain point. This point where deviation from a practically straight-line 
relation between stress and strain is first perceived is designated the yield point. Beyond the 
yield point the slope of the curve decreases at an increasingly rapid rate, and the permanent 
set increases correspondingly. 

The following table gives values of the yield point in tests of mortars of three proportions 
made by the Bureau of Standards (Tech, Paper 58). 

Yield Foist op Mortars 

by voliun* 



TleM point 
Ob. pa *i. Id.) 

' elJlidty 
lib. pa >q. in.) 


Yipld point 
ab. per iiq. iTt.) 





Ob. p«r M. in-l 










Hie yield points of various classes of 1 :2 :4 and 1 :3 :6 concretes at ages up to 1 year 
are shown by Ilg. 27 which is based upon teste of the Bureau of Standards {Tech. Paper 58). 

Fin. 27.— Yield point otI:Z:lu<ll:3:S toaetvUa. 

VaJuea of ultimate compressive strength for the same concretes are shown for purposes of com- 
parison. These teats indicate that the yield point of the average concrete is in the neighborhood 
of three-tenths of the compressive strength at 1 month and about four-tenths of the compressive 
strength at 1 year. 

SI. Hodnlns of Elasticity. — ^The modulus of elasticity of an dastic material is the quotient 
obtained by dividing unit stress by the corresponding unit deformation or strain, the limit 
of elastic behavior not being eiccecded. In American practice the unit of measurement is 
POUd4s per square infh. Since mortars and concretes are not perfectly elastic materials, the 
deformation not bearing a constant relation tu the stress for any range of loading, the quotient 



[S«c. 5-32 

of stress divided by strain will vary, decreasing as the stress increases. Properly speaking an 
inelastic material has no modulus of elasticity, but practice has sanctioned the use of the term 
in connection with mortals and concretes, meaning the quotient of any small stress increment by 
the corresponding strain increment. Hie value of E thus computed is, therefore, the slope of a 
short chord of the stress-strain curve, or if the stress increment be very small, E at any point on 
the stress-strain curve is represented by the tangent to that curve. Within the limits of work- 
ing stresses for mortars or con- 
£1 ^; cretes the value of B changes 

St J- only very sUghtly, and its initial 

J k value may, therefore, properly be 

^ . M xiaed for purpoaea of design. Thi* 

.E C initial value of E may most con- 

'■t i" veniently be determined by not- 

i i ing the slope of the tangent to 

% -f the first portion of the stress^trsin 

i 9 The values of the initial mod- 

4 ^ ulus of elasticity for the thn^ 

' £ mortar mixtures mentioned in the 

last preceding chapter are gi^'en 
Ags In wssks in the above table. Values of E 

Bad 1 ; 3 : 6 concrcta. found for the various concretes 
also mentioned above are given in 
Fig. 28. Little may be said by way of generalisation concerning E for concretes. It increases 
with age and with the richness of the mix, but varies greatly with different classes of aggre- 
gate mat«riab and with different aggregates of the same general class. E for cinder concrete 
appeals to be something less than one-half the average value found with rock concretes. 

The design values of E recommcoded by the Joint Committee are given in Appendix B. 

Fid. 23.— Moduliu of dutlcili 

S2. Co«fflci«Dt of Expamioii. — Mortars and concretes expand as the temperstui« is raised 
and contract as the temperature is lowered. The coefficients of linear expansion per degree 

Fahrenheit for a series of mortars 

and crushed st«ne concretes tested 
under the writer's direction in the 
Isborstorics of the College of Civil 
Engineering, Cornell University, in 
1916 are listed in the Uble on this 
page. All of the specimens were 
molded in the shape of bricks 8 in. 
long, 4 in. wide, and 2 in. thick. 
Heating was done in a specially-con- 
structed resistance type of electric 
furnace, and distortion was meas- 
ured hy a special extensometer 
actuated by fused quarts contact 

bats extending through the furnace walls. Measurements wera made to the neuvst 
0.000,000,9 in. per in. of length of the specimen. The test results listed are averages of from 
two to ten tests of each class of material, the ages varying from I to 8 months. The range of 
temperatures employed was from about 70° to 212°F. 


— 1 


Coeffideit of 


CoeSwnt of 



1:2 :4 
1:3 :6 



Theoe testfl show tb&t the coefficient of expansioD of mortare and concretes increaseB with in- 
rrease of richnesa of the mix, but that the range of values between a very lean concrete and neat 
cement is compaistively short. An average concrete will have a temperature coefficient almoet 
exsrlly equal to that of the average steel. This fact is one of great importance in aU cases of 
reinforced concrete. 

SS. Moistnre Changes. — Mortars and concretes expand in volume if kept wet or immersed 
in water, and contract if exposed in air. Experiments made by Prof. A. H. White at the 
Univeraity trf Michigan (Proe. Am. 
tk>c. Test. Mat., toU. 11 and 14) -^ 
indicate that this property is not '^ | 
confined to the early hardening - £ 
period but is characteristic of roor- g:p 
tars and concretes even after 20 g-^ 
years in service. Fig. 29a shows ^-t 
the variations in length observed " o 
in two l-in. by l-in. by 4-in. bars g>, 
of 1 : 3 mortar tested by Prof, -g " 
White. These bars appear to S 
have suffered no impairment of the « 
ability to expand immersed and 

«,ntract when exposed to air even """'« '" V"^''* 

after a period of nearly 5 years. ^"' »« -Exp«u>ion .«d^™n.r«tioj. rfji^: 3 mortal, when ^lur- 
These particular specimens, when 

about 4 years old, show increases in length of about 0.05% in a 3 to 4-month period when 
placed in water after thorough drying out, and their contraction in air is scarcely leas rapid, 
t^iecimens of the same mixes made with other cements showed in a number of eases more ex- 
tensive volume changes than do the ones shown in Fig. 29a. One mortar stored in air con- 
tracted about 1.10% in the first 3 months. A specimen cut from the rich mortar top coat 
of a sidewalk which had been in service for 20 years expanded about 0.16% when stored in 



water for 2 years, and a portion from the gravel concrete base of the same walk expanded 
about 0.12% in the same period. 

The results of a number of tests of concrete made under various auspices are shown by Fig. 
296. Curves I, 11, and III show the changes in length of concretes tested by A. T. Goldbeek 
in the laboratory of the Office of Public Roads, U. S. Department of Agriculture (,Proc. Am. Soc. 
Test. Mat., vol. 11). The specimens were 8 in. square and 5 ft. long. Specimens I and II 


were stored in air; specimen III was wrapped in burlap which was kept moist continuously. 
Curves IV, V, VI, and VII are derived from tests made by Prof. H. C. Berry in the laboratories 
of the University of Pennsylvania {Proc. Am. Soc. Test. Mat., vol. 11). The specimens were 
in. square and the gaged length was 20 in. Specimens IV and V were stored in air while speci- 
mens VI and VII were stored in water. Curve VIII is the record of a test of a slab of concrete 
12 in. wide, 6 in. deep, and 10 ft. long between gage points, made by Prof. B. P. Fleming of the 
State University of Iowa. The slab was exposed in the air of the laboratory throughout the 
period of observation. Its behavior is notable in one respect in that a pronounced expansion 
was observed for the first 10 days, amounting to about 0.012%. After about 12 days it con- 
tracted continuously for about 75 days longer, at which time it shows a net contraction of about 
0.03%. Mr. Goldbeck states that some of his mixtures showed a tendency to expand in the 
early hardening period of a few days but the amount of expansion was very slight. 

The tests of Fig. 29& indicate that the changes in volume of concretes when exposed in 
either a wet or a dry situation are rather variable with different cements and aggregates. It 
appears, however, that a concrete which dries out in the air may be expected to contract from 
0.02 to 0.05%, and when immersed in water may expand at least half this amount. If the 
concrete in a structure is so restrained that it is not free to expand or contract, it is possible 
therefore that stresses amounting to from 400 to 1000 lb. per sq. in. in tension may occur, E 
being considered to be 2,000,000 lb. per sq. in. This means that the tensile strength of concrete 
is exceeded and the concrete will, and commonly does, crack. DifRculty caused by the expan- 
sion of concrete in a damp or wet situation is not so commonly encoimtered, and the stresses 
introduced will never cause compressive failure. They may, however, cause a buckling action 
in the case of continuous surfaces of large extent. 



84. Fire-resistance Properties. — Concrete ranks highly as a fire-resistant and fireproofing 
material principally because it possesses a low rate of heat conductivity and has a low coeffi- 
cient of expansion practically equal to that of steel, in addition to being incombustible. Other 
masonry materials like some of the natural stones and terra-cotta are no less incombustible 
than concrete, but are inferior to the latter as a fireproofing material because they possess either 
greater conductivity or a higher coefficient of expansion. 

Tests of the conductivity of concretes made by Prof. Ira H. Woolson {Proc. Am. Soc. Test. 
Mat., vols. 5, 6, and 7) led to the conclusions: " That all concretes " — stone gravel, and cinder — 
"have a very low thermal-conductivity, and herein lies their ability to resist fire. That when 
the surface of a mass of concrete is exposed for hours to a high heat, the temperature of the 
concrete 1 in. or less beneath the surface will be several hundred degrees below the outside. 
That a point 2 in. beneath the surface would stand an outside temperature of 1500**F. for 2 hr. 
with a rise of only 500^ to 700^ and points with 3 in. or more of protection would scarcely be 
heated above the boiling point of water." 

The low thermal-conductivity of concrete is partly due to its porosity, air spaces affording 
efficient protection against conduction, and partly due to the absorption of heat of vaporixa- 
tion by the water of combination in the set cement when the temperature of dehydration of the 
latter is reached. This dehydration probably begins at about 500°F. and is completed at about 
900^F. (8. B. Newberry, Cement, May, 1902, p. 95). The absorption of heat by the surface 
material in becoming itself dehydrated retards the dehydration of the underlying material. 
The surface concrete which is injured by heat, but which remains in place, affords protection 
for the material beneath, for it becomes a poorer conductor than the original concrete. The 
Joint Committee on Concrete and Rcinfoncd Concrete recommends that ** metal be protected 
by a minimum of 2 in. of concrete on girders and columns, 1}-^ in. on beams, and 1 in. on floor 

The experience gained in great conflagrations like the Baltimore fire, the San Francisco 


fire, the Edison plant fire, etc., has been that concrete exposed to intense heat for considerable 
periods becomes calcined to a depth of from ^. to 3i in. but shows no tendency to spall off 
except at exposed comers and edges (see reports cf Captain J. S. Sewell to the Chief of Engineers, 
U. S. A., and of Prof. Norton to the Insurance Engineering Experiment Station, on the Balti- 
more fire, Eng. News, March 24, and June 2, 1904; the report of S. A. Reed to the National Board 
of Fire Underwriters on the San Francisco fire, Eng. News, vol. 56, p. 137; the comments of 
various engineers upon the Edison plant fire, Eng. News, vol. 73, p. 38; and the report on the 
Edison fire of the National Fire Protective Association, obtainable in booklet form from the 
New York Board of Fire Underwriters). 

Prof. Norton and others have concluded that there is little difference in the action of fire 
on stone concrete and cinder concrete. 

96. Weathering Qualities. — The principal agencies affecting the durability of concretes 
and mortars which are classed as weathering agencies are changes of atmospheric temperature, 
wind and rain, and changes of atmospheric moisture. The expansion and contraction of mor- 
tars and concretes subjected to variations of temperature and moisture conditions are respon- 
sible for practically all failures of these materials under conditions of exposure to the weather. 
Either temperature effects or moisture effects may be alone operative, or both effects may be 
combined. Temperature stresses caused by the shrinkage of continuous large surface areas 
are particularly apt to cause cracking. This cannot be wholly prevented, but cracks can be 
made less harmful by the use of steel reinforcement so placed that a multitude of small cracks, 
which do not open up much, replace a few large and deep cracks. In the average situation the 
introduction of dangerous stresses caused by a tendency to expand or contract is more apt to 
be due to moisture changes than to temperature changes, because the volumetric changes in 
the latter case are less marked. The expansion and contraction of rich mortars and concretes 
is considerably more extensive than that of leaner mixes when the moisture condition varies, 
and the same thing appears to be true to a lesser extent when the temperature varies. This 
circumstance is responsible for the difficulty often encountered in causing a surface coating of 
comparatively rich material plastered upon a leaner base material to adhere permanently if 
the bond between the two is at all defective. The surface material tends to expand and con- 
tract more than the underlying material not only because it is richer in cement, but also be- 
cause it protects the underlying material from as extensive temperature and moisture changes 
as it itself experiences. The result is the introduction of excessive stresses in the surface 
material, the opening up of tension cracks, or buckling due to compressive stress, and the 
ultimate spalling off of the surface layer. The principal preventive measures which may be 
adopted are the use of as lean a surface coat as is practicable, the use of as thin a plaster coating 
as possible thus favoring the formation of many small cracks rather than a few large ones, 
and the adoption of all measures tending to make a strong bond between jLhe two classes of 
material. An excessive amount of troweling of surfaces is to be avoided because of the flush- 
ing to the surface of a film of nearly neat cement which will tend to x>eel off in some cases. 

36. Abrasive Resistance. — The abrasive resistance of mortars '^is primarily of importance 
in the determination of the best mortar for use in the top coat of concrete floors, walks, and pave- 
ments. Resistance to abrasion will always be dependent not only upon the cement, as regards 
the tenacity with which it clings to the sand grains, which will be largely dependent upon its 
fineness and its lime content, but also upon the hardness of the sand used. Abrasion either 
wean away the cement and the sand grains, or it pulls the sand grains out of the cement matrix. 

''With soft sand particles the resistance to abrasion with a given cement decreases con- 
stantly as the percentage of sard is increased. With hard sand grains the abrasive resistance 
increases as the proportion of sand increases, until the volume of cement becomes relatively 
too small to bind the sand grains together thoroughly. This limit is found to be reached when 
the mortar contains not more than two parts of snnd to one of cement." ('* Materials of Con- 
struction," by A. P. Mills.) The abrasive resistance of concretes is dependent almost wholly 
upon that of the mortar made by its cement and fine aggregate. The coarse aggregate is prac- 


tically always forced back from the surface of concrete exposed to abrasion. If it is expoiied. 
the same considerations of relative hardness of the stone particles and proportion of the mix 
applicable in the case of mortars apply to the concrete. 

87. Action of Sea Water. — The behavior of concrete in sea water is a problem which 
has occupied much of the attention of engineers for many years. The question has often 
been discussed, and many attempts have been made to determine experimentally the exai't 
action of sea water upon concretei and the causes of that action. The amount of accurate in- 
formation available is rather meager, however, and the results of experimental investigations 
are inconclusive and often contradictory. Many concrete structures in sea water out of tbc 
range of frost action have remained intact and uninjured for many years. Others have been 
seriously disintegrated, particularly between high and low tide levels. The disintegration is 
evidently often due in part to frost action, but chemical action is frequently indicated by the 
softening of the mortar, and the complete disintegration of mortar and concrete specimens 
by subjection to the action of sea water at normal temperatures in the laboratory has been 
accomplished. The exact nature of the chemical action involved cannot be definitely stated. 
It is commonly believed, however, that the magnesium sulphate in the sea water is the most 
injurious constituent, and that the magnesium chloride and calcium chloride are somewhat 
less active. The magnesium sulphate attacks the lime in the cement, also the alumina, form- 
ing large and rapidly growing crystals of hydrated magnesia and calcium sulpho-aluminatc. 
Both magnesium chloride and sodium chloride attack the silicates of the cement. 

The chemical action is accompanied by various physical phenomena. Sometimes the maas 
swells, cracks, and gradually falls apart; sometimes the mortar softens and becomes disinte- 
grated leaving the coarse aggregate exposed and finally permitting it to fall away; and occa- 
sionally a crust forms on the surface which later cracks off. 

An important symposium of European investigators' studies of the problem of concrete 
in sea water is afforded by the several papers of Chapter XVII of the Proceedings of the Sixth 
Congress of the International Association for Testing Materials held in New York in 1012. 

The conclusions arrived at from a study of important German and Scandinavian tests 
have been expressed, in part, as follows (Concrete and Conetrudional Engineering, January, 1910) : 

1. Good Portland cements such as are now on the European market, are very resistant to the action of 
water. A marked difference in the behavior of cements of slightly different composition has not bee n found, ex- 
cept that a high proportion of aluminates tends to cause disintesration. 

2. In a dense mortar, the chemical action is confined to an outer layer of small depth, further action being 
checked by the slowness of diffusion. A porous mortar, by admitting salt water to the interior, is apt to erack by 
expansion owing to chemical change. 

3. The main agency in the destruction of mortar and concrete in marine embankments, harbor works, groynes, 
etc., is not chemical action, but the alternations of saturation, drying in the sun, freesing, etc., due to the alternate 
exposure and covering by the rise and fall of the tide. 

4. The denser the mortar the better (1 cement :3 sand is too poor). An admixture of fine sand with the 
ordinary sand increases the closeness of the mixture. A well-graded aggregate would be advantageous for the 
same reason. 

6. The addition of finely ground silica or trass to the cement before mixing is |>ossibIy advantageous in the 
case of weaker mortars. It is very doubtful whether anything is gained by adding trass to the richer mortars. 

6. The destructive action of the sea being mainly physical and mechanical, and not chemical, tests by mere 
immersion in still sea water are of very little value in determining the behavior of concrete in marine engineering 
works. A mixture which disintegrates under this test is certainly useless, but a mixture which pssies the test 
may disintegrate under the more stringent conditions of practical use. 

7. As long a period as is practicable should be allowed for the hardening of e<merete blocks before placing 
in the sea. 

8. The bdiavior of test specimens for the first 12 months is very irregular, and definite eonduaons can only 
be drawn from the results of long-period tests. 

The most notable American investigations of the subject are one made by the Bureau 
of Standards {Tech. Paper 12) and one begim in 1908 by the Aberthaw Ck>n8truction Co. in 
cooperation with the United States Navy Department (reported in a pamphlet issued by the 
Aberthaw Construction Co. in 1914; also in Eng, Rec,, March 21, 1914). Few general conclu- 



^■' -n.^ BBT br dnn fiom the molts of the first 5 veAis* ohservatioiis of the Aberthaw tests. 
"Hlzs is paitnlAifT tme aiiee European experience has shown that the first indieations of 
:c jaTv to BUT ronmtes appear only after horn 5 to 10 3reais' exposure, and in some cases only 

30 reais. The most notable indkatioos afforded by the Aberthaw tests after 
o vvais ri iH wiH r in a latitude mvotving wide variations of temperature and frequent freeiing 

!• C utwte which is aHemate^ inuneised and exposed as the tide rises and falls is most 
to B jorr. 

2. lean Mixtm c B (1 :3 :6) are very much more subject to attack than rich mixtures (1 : 1 
2 . and Bcdiom m ixtur es (1:2:4) are more TulneraMe than rich ones. 

3u Concteie mixed with a plastae or even a very wet consistency appears to withstand sea 
1^ ^XfT attack better than concrete of a drr consistencT. 

4. The retative immmiity from attack by sea water of concretes made with cements of 
\ arious daasesy inchidinK a low-iron cement, high- low- and average-alumina cements, an iron- 
'Te ccflKBt. and a sfaig IVsitland cement, has not been condoaively estabttshed. 

The Joint Committee makes the following recommendations for concrete placed in sea 

be i p ^ H H T^iffTWfH. 

it so as to aenire the 
old aad new vork fllioald be 
oiUilit is thoroosUy bard and imperrioaa. 

miaad, and plawd so as to 

ab«mM be cmrtivBy 

Ode demity: the eonrrrto 

water-ti^t: and the eoDcrcte 

qI Aflkafi. — ^The effect of alkali on concrete is a problem resembling in many 
r e sp e cts that of the action of sea water on concrete. The problem is of especial interest in 
f-onnectaon with concrete constniction in the arid regions of the West, where soluble salts are 
present in the sofl to an extent not usually found dsewhere. 

The prindpal sahs encountered in alkali waters usually include: magnesium sulphate, 
i-3dciam sulphate and sodium sulphate, magnesium chloride, sodium chloride, and potassium 
f-hloride, together with carbonates of magnesium, sodium, and potassium. Of these the 
•4ilphate8 appear to be most active in causing disintegiation of concrete; the chlorides also are 
active, while the carbonates ai^xar to be without effect. 

The attempts at an explanation of the manner of attack of these salts upon concrete 
have hitherto encountered the same difficulty found in the case of sea water — an unsatisfactory 
knowledge of the constitution of cement. From the physical point of view the action exactly 
resembles the action of frost except that it is more rapid. There exists, apparently, a disruptive 
force which quickly destro3rs the bond and causes disintegration. This action appeals to 
proceed most lapidly in the parts of a structure subjected to alteinate wetting with alkali 
water and drying in the air. In porous concrete the action proceeds much more rapidly than in 
dense concrete, where, indeed, it may make no progress at alL 

As in the case of the injurious action of sea water on concrete, instances of failure caused 
by alkali waters are merely isolated ones, presenting an interesting field for study, but not 
ronstitnting a very serious menace to the future of ccmcrete construction in the arid regions 
of the West. The remedy in the present state of our knowledge is, as in the case of marine 
structures, a matter of the possible physical precautions only — the securing of the densest 
possible concrete, thus preventing injury by the exclusion of the salt-bearing wateia. 

39. Actioo of AcidSy (Mis, and Sewage. — ^The Joint Committee Repoit makes the following 
statements eonceming the effect of acids and oils upon concrete: 

bat do not affect 

b ni dencd is affected a inn e cia blj only by 
re. that eoatain acids, nay 
<-oncrete that m tbocoo^Uy hardened. 

Concrete is nnnffected by sach mineral oils as petroleom and ordinary engine oils. Oik which contain fntty 
setds prodoee injnrioas effects, forauns contpoandB vith the fime which may result in a diBntesrmtion of the con- 
rreCe in aontnct with 




The use of concrete sewer pipes has led to considerable study of the effect of sewage and 
sewage gases upon concrete. Sidney H. Chambers concluded from an investigation reported 
before the Concrete Institute (Great Britain) in 1910: 

Th«t the gases in solution in sewmge and those expelled from it, arising from its deoomposation, do act in- 
juriously upon Portland-oement concrete, notwithstanding the fact that the concrete is constituted of sound and 
good materials, when the following conditions prevail: (1) A high degree of putrescence of the sewage; (2) a mois- 
tened surface, which held or absorbed the putrid gsses; (3) the presence of a free air supply. Further, that in the 
absence of one or the other of the above-enumerated factors little danger from erosion need be feared. 

Rudolph Hering is responsible for the following statements concerning the effect of thf* 
acids in sewage upon concrete (quoted from a report to the President of the Borough of Brook- 
lyn in 1908 by Gustave Kaufman in Proceedings of the National Association of Cement Users, 
vol. 8, 1912, p. 725). 

Portland cement used for the manufacture of concrete pipes is attacked by certain strong adds, such aa 
sulphuric acid, which converts the carbonate into sulphate of Ume, which is comparatively soft and easily eroded. 
Therefore cement pipe cannot be used where strong acids are known to enter the sewers. 

The acid question should be viewed in a reasonable light. When the dilution of sewage is sufficient th« 
discharge of a small amount of even strong acid will not cause objectionable effects, as evidenced by European 
cities where the use of concrete sewers is almost exclusive in some cities, ss Paris and Vienna. In England concrete 
sewers are also very common. 

The greasy substance which is usually found to coat the perimeter of a sewer under the water line tends 
to protect the cement from the action of acids to some extent. 

Over 400 miles of concrete sewer pipe laid in the City of Brooklyn during a period of over 
50 years are giving eminent satisfaction. 

40. Electrolysis in Concrete. — ^Experience and laboratory tests have shown that under 
certain conditions concrete may be seriously damaged by electroljrtic action caused by the 
flow of electric current between the concrete and iron or steel embedded therein. The phe- 
nomena assumes importance under certain conditions of use of reinforced concrete, also the use 
of concrete foundations and footings in which the bases of columns of buildings, bridges, and 
elevated railway structures are embedded. A very important laboratory and field study of 
the entire problem has been made by the U. S. Bureau of Standards, and the conclusions arrived 
at in this investigation constitute the authority for the statements made here {Tech. Paper 
18y Bureau of Standards). 

The electrolytic effect differs according to the direction of flow of the current. If elec- 
trically positive iron or steel Ls in contact with concrete the iron will become corroded provided 
the concrete is moist or wet and the potential gradient is high enough to heat the junction to a 
temperature not below about 45*'C. (113T.). The minimum potential gradient found effective 
in causing corrosion in moist concrete was about 60 volts per ft. Iron when corroded expands 
to about 2.2 times its original volume and causes mechanical pressure found in some cases to 
reach values as high as 4700 lb. per sq. in. This causes cracking of the concrete. The passivity 
of iron below 45^C. is due chiefly to the inhibiting effect of Ca(OH)» in the concrete, and on 
this account old concrete in which the Ca(OH)t has been largely carbonated is probably more 
susceptible to electroljrsis than new concrete in the same moisture condition. For air-dried 
concrete a much higher potential gradient is required to produce the temperature at which 
corrosion becomes dangerous than for moist concrete. Under actual conditions, therefore, 
corrosion from stray currents may be expected only under special or extreme conditions. 
Normal wet concrete carrying current also increases its resistance a hundredfold in the course 
of a few weeks, owing partly to the precipitation of CaCOi which fills up the pores. This 
further lessens danger of trouble. 

Electrolysis of the concrete in contact with negative iron is manifested in a different way. 
The concrete near the cathode becomes softened, beginning at the cathode surface, and extend- 
ing to a depth of H ^' o^ more. This softening practically completely destroys the bond 
between iron or steel and concrete. While the anode effect becomes serious in normal concrete 
only on comparatively high voltages, decreases much more rapidly than the voltage, and almost 


disappears at voltages likely to be encountered in practice, the cathode effect develops at all 
voltages, the rate being roughly proportional to the voltage. The softening effect is due to 
the gradual concentration of alkalies, Na and K, near the cathode, these finally becoming strong 
enough to attack the cement. The cathode effect is wholly limited to the vicinity of the cath- 
ode, the strength of the mass of the concrete not being affected. 

Salt or calcium chloride, even in very small amounts (a fraction of 1%), multiplies the 
rate of corrosion of iron at the anode many hundredfold because it increases the conductivity 
of wet concrete, destroys the passivity of iron at ordinary temperatures, and prevents the 
increase in resistance with flow of current by preventing the precipitation of CaCOi. Salt 
should therefore not be used in structures subject to electrolytic action, and special considera- 
tion should be given to the possibility of electrolytic action in the cases of all concretes exposed 
to sea water or salt brine. 

The danger of electrolysis of rcinforced-concrete structures through the operation of 
stray currents has been overestimated. Certain precautions are necessary under special 
conditions, but there is no cause for serious alarm. Non-reinforced-concrete structures are 
practically immune from injury by electrolysis. 

The precautions to be adopted in the special cases where electrolysis is to be feared include 
avoidance of grounds in direct-current circuits in buildings; providing insulating joints in 
pipe lines which enter buildings outside the walls; completely isolating the buildings by in- 
sulating joints in pipe lines which enter the building and also continue on beyond; providing 
a copper cable shunt around the building if the potential drop is large; isolating from the con- 
crete lead-covered cables entering the building; and interconnecting all metal work in the 
building if practicable, but without connecting this metal work to ground plates or to pipe 
lines outside of the insulating joints. 

41. Effect of Manure. — Manure is occasionally used to cover up fresh concrete in freezing 
weather, not only because it is a poor conductor of heat when rather dry, but also because its 
decomposition is a source of heat. Experience has shown (see Eng, Newa^ vol. 49, pp. 11, 104, 
126, 127, and 175; also Journal of the New England Waietworke Association^ vol. 22, p. 242) 
that manure not only discolors the work, but that it also has a marked disintegrating effect if 
placed in contact with freshly placed concrete. The injury is especially pronounced if rain 
wets the manure during the early hardening period and carries the uric acid into the concrete. 
The use of manure as a preventive of freezing of fresh concrete may be considered permissible 
only if the work is covered first by a material which will be sufficiently impermeable to prevent 
the seepage of acid into the concrete. 

Concrete which has once thoroughly hardened appears not to be susceptible to injury 
by contact with manure except that it is in some cases somewhat discolored. 


42. Else of Temperature in Setting. — The chemical combination of water with Portland 
cement is an endothermic reaction, the heat evolved being sufficient to materially raise the 
temperature of mortars and concretes during the period of setting and hardening. The total 
rise in temperature, the rate of increase, and the time interval before the maximum temperature 
is reached are all variable, depending upon the character of the cement used, the proportions of 
the mixture, the size of specimen or bulk of material involved (in so far as this determines the 
distance of the point at which temperatures are measured from any exposed surface of the 
material), external temperature conditions, the amount of water used in mixing, etc. 

The results of observations of temperatures acquired at the center of 12-in. cubes of cements 
and mortars tested at the Watertown Arsenal ("Tests of Metals," 1901, p. 493), are shown by 
Fig. 30o. With these specimens the maximum temperature was attained in from 12 to 18 


hr. except in the case of natural cements which reached their maximum temperature much ear- 
lier. The heating effect is less than one-half as great with 1 : 1 mortar as with neat cement, aod 
is quite small with leaner mixtures. 

Radiation has a great deal to do with the temperatures observed at the centers of these 
comparatively small specimena, aa ia shown by comparisoQ of the curvea for mortars with Pig. 
306 which shows the rise in t«mperature of a 1 :2^ : 6 concrete, the temperature having been 
measured in the midst of a very large mass of concrete with the thermometer covered by a 

TliTM \n hours 

depth lA from 1 to 9 ft. of concrete as indicated. Fig. 30b constitutes a portion of a report 
upon an investigation of the temperature changes in mass concrete made during the construc- 
tion of the Arrowrock Dam by the United States Reclamation Service near Boiae, Idaho ("Tem- 
perature Changes in Masa Ctmctete" by Cbaries H. Paul and A. B. Mayhew, TVnn*. Am. Soc. 




C. E.y vol. 79, p. 1225, 1915). A summary of the results obtained in these tests is given in the 
following table. 

Temperature Changes in Mass Concrete 
Arrowrock Dam Tests 

Proportions oi mix (parts by Tolume) 

to nearest 
face (ft.) 

Depth of 

concrete over 



Rise in 


in degrees 




in degrees 


highest tem- 
perature, hr. 

i^mnd esment 















































































* Thermometer bedded in fresh concrete in shallow trench dug in concrete 11 days old. 

Similar tests made during the construction of the Kensico Dam at Valhalla, N. Y., by 
George T. Seabury, are reported to have established that (Proc, Am. Soc. C. E., vol. 29, p. 1247- 

With the 1:3:6 concrete used, thermometers being inserted as soon as the concrete was placed and immedi- 
ately eoTered to as great a depth as possible, the rise in temperature was uniformly about 40*'F. 

The maximum concrete temperature reached was often well above lOO^F. in summer, in one instance, 118.5^. 
This maximum was usually reached in about 15 days. 

The rate of increase in temperature was about 1® per hr. for 4 or 5 hr. gradually increased to S^-IO^ per hr. 
daring the period of final set of the cement, and then dropped suddenly to about \i^ per hr. for a considerable 
time. A total rise in temperature of from 25® to 30®F. often occurred subsequent to the period of final setting. 

48. Porosity. — The porosity of a mortar or concrete is expressed by the percentage of void 
space (space filled by air or uncombined water) in terms of the total volume. It is determined 
experimentally by subtracting from the total apparent volume the volume of solid matter and 
dividing by the total apparent volume. The total apparent volume may be determined by 
direct measurement or by determination of the volume of water it displaces, absorption being 
prevented by a waterproof coating of grease or varnish. The volume of solid matter is deter- 
mined by weighing the specimen dry in air, and subsequently in water after the pores of the 
material have been thoroughly impregnated with water. The difference in these weights 
divided by the weight of a unit volume of water is the volume of solid matter. Extreme accu- 
racy in determining porosity is not possible because of the difficulty encountered in completely 
fiilling all the voids in the material. 

The porosity of mortars and concretes is principally dependent upon the consistency of the 
mixture, and the granularmetric composition of the aggregates. Plastic or wet consistencies 
will in general produce mortars having less void space, and therefore lower porosity, than dry 
consistencies, and a well-graded aggregate will form less porous mortar than one whose particles 
are not well graded in size. A concrete will usually be considerably less porous than a mortar 
because of its proportion of comparatively non-porous coarse aggregate, and a fine-«and aggre- 
gate will in general produce more porous mortar than a coarse sand because of the larger amount 
of wat^r required in gaging the latter. 

The porosity of mortars will usually be found within the limits of 15% and 30%, an average 
1 : 2 mortar showing 20 to 25% porosity. Concretes show from about 12 to about 20% porosity, 


the lower figure applying to concretes having a relatively large proportion of coarse agg^regate 
and the higher figure to concretes having a relatively low proportion of coarse aggregate. 

44. Permeability and Absorptive Properties. — ^The permeability of mortar or concrete is a 
measure of the rate at which water under a given pressure will pass through a given thickness 
of the material. The absorptive properties of a mortar or concrete constitute a measure of the 
rate at which moisture will be absorbed when the material is exposed in damp situations or 
covered with water under negligibly small heads. 

Permeability is an important consideration where water-tightness of walls, etc., is re- 
quired and percolation of water is not admissible. 

Absorptive properties of a mortar determine its value as a dampproofing coat, particularly 
in the event of its use as a plaster over metal lath, which must be protected to prevent corrosion. 
In view of the disintegrating effect of expansion and contraction of mortars used as a plaster, 
etc., the moisture content (which largely affects this expansion and contraction) should not be 
greatly variable. Thus the least absorptive mortar will be most durable, up to the limit 
reached when the cement content is relatively so high that the expansion and contraction is 
disproportionately increased. 

The determining of precise information concerning each of these properties is dependent 
upon a standardization of methods of conducting tests. Such standard methods have not 
yet been adopted, and it is therefore impossible to quote data as to the absolute permeability 
or absorptive power of mortars. 

Tests to determine the relative permeability and absorptive power of mortars were made 
at the Structural Materials Laboratory as St. Louis in 1909, and are reported in Technologic 
Paper 3 of the Bureau of Standards. Owing to the small number of tests made and certain 
unsatisfactory features of the testing method employed, only a few general conclusions will 
be drawn from the report of these tests. (1) Permeability decreases rapidly for all mixtures 
with increase in age of the specimens when tested; (2) permeability decreases considerably 
with the continuation of the flow; (3) permeability increases with the leanness of the mixture, 
the dryness of the mixture, and increased coarseness of the sand. 

Absorption was found to be dependent upon the same factors: it decreased with the ago 
of the mortar as a rule, but not as rapidly as did the permeability (especially with the leaner 
mixtures); it decreased but slightly with increased richness of the mixtures; and the wetter 
mixtures were slightly less absorptive than the dryer mixtures. 

The permeability of mortars and concrete is closely related to the porosity, but the relation- 
ship is not alwasrs direct, and is by no means constant, since the continuity and sise of the pores 
determines permeability more than does the actual percentage of voids. 

The Joint Committee Report says concerning the permeability and the waterproofing of 
concrete : 

Many expedients have been resorted to for rendering concrete impervious to water. Experience shows, 
however, that when concrete or mortar is proportioned to obtain the greatest practicable density and is mixed to 
the proper consistency, the resulting mortar or concrete is impervious under moderate pressure. 

On the other hand, concrete of dry consistency is more or less pervious to water, and though compounds of 
various kinds have been mixed with the concrete or applied as a wash to the surface, in an effort to offset this 
defect, these eipedients have generally been disappointing, for the reason that many of these compounds have at 
best but temporary value, and in time lose their power of imparting impermeability to the concrete. 

In the case of subwayv, long retaining walls and reservoirs, provided the concrete itself is impervious, cracks 
may be so reduced, by horisontal and vertical reinforcement properly proportioned and located, that they will 
be too minute to permit leakage, or will be dosed by infiltration of silt. 

Asphaltic or coal-tar preparations applied either as a mastic or as a coating on felt or cloth fabric, are used 
for waterproofing, and should be proof against injury by liquids or gases. 

For retaimog and similar walls in direct contact with the earth, the application of one or two coatings of 
hot ooal-tar pitch, foOowing a painting with a thin wash of ooal-tar dissolved in benaol, to the thoroughly dried 
surface of concrete is an efficient method of preventing the penetration of moisture from the earth. 

46. Protection of Embedded Steel From Corrosion. — ^The Joint Committee Report says 
concerning the corrosion of metal reinforcement in concrete: 




Teats and ezperienoe indieate Uutt rt ec l suffidMitly embedded in cood oonerete is w^ protected ecainat 
eomwon, no matter whether loeated above or bdow water level. It is recommended that such protection be not 
leas than 1 in. in thickoeaa. If the eoncrete is porous so as to be readily permeable by water, as when concrete is 
laid with a very dry constancy, the metal may corrode on account of the presence of moisture and air. 

The historic tests made by Prof. C. L. Norton for the Insurance Engineering Station in 
Boston in 1902 led to the following conclusions the validity of which has never been disproved by 
either tests or experience {Eng. News^ vol. 48, p. 333) : 

1. Neat Portland cement, even in thin layers, is an effective preventive of rusting. 

2. Concretes, to be effective in preventing rust, must be dense and without voids or cracks. They should 
be mixed quite wet where applied to the metal. 

3. The corrosion found in cinder concrete is mainly due to the iron oxide, or rust, in the cinder, and not 
to the Bulphur. 

4. Cinder concrete, if free from voids and wdl rammed when wet, is about as effective as stone concrete in 
protecting steel. 

Further tests made by Prof. Norton in 1903 showed conclusively that steel reinforcement, 
corroded before being embedded in concrete, does not corrode further, provided only that it 
has a continuous unbroken coating of concrete. This fact is important since it is almost impos- 
sible to prevent exposure of reinforcing steel on construction work to the elements, and a large 
proportion of such steel is therefore corroded when placed in the work. 

Experience has abundantly shown that if concrete be mixed sufficiently wet so that it will 
flow about the reinforcement with only a moderate amount of puddling, the thin film of rich 
mortar which coats the steel affords a perfect preventive of corrosion. 

46. Weight of Mortar and Concrete. — The weight of mortars and concretes varies with the 
proportions of the mixture, the consistency used in mixing, and the character and granular- 
metric composition of the aggregates. William B. Fuller found the following range of weights 
of mortars of various proportions made with the same sand and cement : 

ProDOrtions of mixture 








Ave. weight Gb. per cu. ft.) 

The Bureau of Standards found the following relation between weight of gravel concrete 
and the proportions of mixture {Tech. Paper 58) : 

Proportions of mixture 

Ave. weight flb. per cu. ft.) 







The same series of tests of the Bureau of Standards indicate the following relations to hold 
between weight and consistency: 

The cinder concrete appears to 
be slightly heavier the wetter the 
mixture, but all of the rock con- 
cretes are slightly heavier the drier 
the mixture. No universally appli- 
cable relation between weight of 
concrete and the class of aggregate 
used is shown by tests. Different 
gravel concretes, for instance, will 
^ow greater variations in weight 
than the difference in average weight of gravel concretes and granite concretes. For pur- 
poses of design the weight of any class of stone concrete may be assumed to be 150 lb. per 
cu. ft., whfle cinder concrete may be assumed to weigh 115 lb. per cu. ft. 

Proportions by 
Coarse aggregate 

Weight per cubic foot 

Watery or 


Soft, mushy 

Stiff quaking 











1. Advantages of Gombming Concrete and SteeL — Steel can be put into a form to resist 
a given tensile stress much more cheaply than to resist an equal amount of compressive stress. 
This comes from the fact that the solid bar is well adapted to take tensile stresses, while for 
compressive stresses the steel must be made into forms of more extended cross-section in order 
to provide suflScient lateral rigidity. Other facts to be noted are the lack of durability of 
9teel in many locations and its failure to stand up under a high heat. 

Concrete, on the other hand, cannot be used in tension except to a very limited extent, 
but its compressive strength is sufficiently high to be of structural importance. It is also a good 
fireproof material and has great durability. In addition, concrete is a cheaper material than 
steel, can readily be obtained in almost any locality, and tests and the results of observations 
show that it thoroughly protects embedded steel from corrosion. 

From the above considerations it follows that the advantages to be gained by using con- 
crete reinforced with steel instead of either material separately will vary with different types 
of structures. In structural members subjected to both tension and compression, as in all 
forms of beams, the proper combination of the two materials meets with the bdst success. 
Steel rods embedded in the lower side of the beam carry the tensile stresses while the compressive 
stresses are carried by the concrete. Here the steel is used in its cheapest form and the con- 
struction may be made strong, economical, and very durable. In compressive members of 
appreciable length, such as columns, a combination of the two materials is also quite advanta- 
geous, although to a varying degree depending upon whether the reinforcing steel is used in 
the form of small rods or as structural-steel shapes. 

8. Bond Between Concrete and SteeL — Most of the tests on bond have been made by 
embedding a short reinforcing bar in a block of concrete and pulling it out in a testing machine. 
In the kind of tests referred to, the concrete is in compression and conditions do not correspond 
to those ordinarily encountered in beams and slabs. Pull- 
out tests of this character, however, have been found to be 
of valuable aid in determining actual values for bond. This 
conclusion was reached through an extended series of ex- 
periments at the University of Illinois during the period 
1909-1912; a series which include both pull-out tests and Fig. i. 

beam tests, as described further on in this article. 

In a series of experiments made at the University of Wisconsin, test beams were arranged 
as shown in Fig. 1, the reinforcing bars being embedded only a short distance from each end, 
leaving the middle portion exposed. The stress in the rods were computed from the observed 
deformations. The beam was prevented from failing in the early stages of the tests, by an 
upper set of auxiliary rods. Failure finally occurred by the pulling out of the lower rods, as 
intended. Pull-out tests were also made with concrete cylinders and rods arranged as shown 
in Fig. 2. Cylinders designated as (a) were tested in the ordinary way with their upper surfaces 
bedded against the lower face of the pulling head. In cylinders (b) tension was applied to 
both upper and lower rods, bringing tension also in the concrete. The principal conclusions 
arrived at were as follows (the word alaiic is used in connection with beams progressively 
tested to failure under loads gradually applied, and the word repeated occurs in connection 
with those beams subjected to repeated loadings) :^ 

t Bull. 5. Tol. 5, Univeraity of Wisconsin, 





The ttatio bond between 1:2:4 concrete and plain round steel rode increaaee with age at leaet up to 6 moDths. 
About 80% of the 6 monthe' bond strength is developed in 1 month. 

Owing to the variation in the results of individual tests and the difference between laboratory and practical 
working conditions, it does not seem as though the maximum static bond between concrete of the class used and 
plain rods less than H in. diameter should be assumed greater than 250 lb. per sq. in. or for the rods or larger sise 
200 lb. per iq. in. 

The method of making bond tests by pulling a rod from a cylinder of concrete in such a manner that th« 
concrete around the rod is compressed gives results which are neither of quantitative nor qualitative value. Th^ 
results obtained are dependent largely upon the compressive stress acting on the head of the cylinder. Cylinder 
tests in which the rod and concrete are both subjected to a tensile stress give results more in accord with the bond 
values obtained from beam tests. 

The static bond between the class of concrete employed and corrugated bars is about twice as great as that 
which can be developed with plain round rods of about the same sise. The static bond between concrete and rusted 
rods is very much greater than that obtained where plain round rods are used. 

From the tests under repeated loadings it seems evident that 50 to 60 % of the static bond between concrete 
and plain round rods may be repeated a large number of times without failure in bond; that 00 to 70% are the 
corresponding figures for corrugated bars. Under repeated loadings the bond between concrete and rusted round 
rods is considerably greater than that between concrete and plain round rods. 

Considering the severity of the tests made there seems to be no valid reason for believing that the bond be- 
tween concrete and plain round rods will be destroyed under repeated loadings, providing a proper working value is 
used. Such a value for concrete of the class used in these tests should not be over 50 lb. per sq. in. 


In tests at the University of Illinois, attention was given to 
obtaining accurate measurement of slip of bar through the concrete 
as the loading progressed, in both the ordinary pull-out tests an d in 
tests on beams. In the beams of this series the concrete was not cut 
away from the rods as in the tests at the UnivenBity of Wisconsin. 

In the pull-out tests the amount of movement o{ the free end 
of the embedded bar was measured by means of an Ames gage. 
In the beam tests, the Ames gage was used to measure center deflec- 
tions and the movement of the ends of the reinforcing bans. In 
many of the tests, observations were also made on the amount of 
slip of the reinforcing bar with respect to the adjacent concrete at 
several points along the length of the beam. 

The concrete blocks used in the pull-out tests were usually 8 in. 
in diameter and 8 in. long, with the bar embedded axially. The 
beams tested were 8 by 12 in. in section with an effective depth of 10 
in. The span length was generally 6 ft. All beams were tested 
with two symmetrical loads, generally at the one-third points of the 
span. With the exception of six tests, the longitudinal reinforce- 
ment consisted of a single bar of large diameter placed horizontally throughout the length of 
the beam. The principal results of these tests and the conclusions reached were as follows:^ 



t 1 




t 1 

% 1 


1 i • 

* * 


i i 

■ ■ 




'c \ 

• • 

■ * 



1 if 




* If- > 


Fio. 2. 

PoU-out TesU 

Bond between concrete and steel may be divided into two principal elements, adhesive resistance and sliding 
resistance. The source of adhesive resistance is not known, but its presence is a matter of univeraal experience 
with materials of the nature of m<ntar and concrete. Sliding resistance arises from inequalities of the surface of 
the bar and irregularities of its section and alignment together with the corresponding conformations in the con- 
crete. The adhesive resistance must be overcome before sliding resistance comes into action. In other words, 
the two elemeots of bond resistance are not effective at the same time at a given point. Many evidences of the 
tests indicate that adhesive resistance is much the more important element of bond resistance. 

Belatioa of Bond Stress to Slip of Bar as Load Increases. — Pull-out tests with plain ban show that • con- 
siderable bond stress is devdoped before a measurable slip is produced. Slip of bar begins as soon ss the adhesive 
resistance is overcome. After the adhesive resistance is overcome, a further slip without an opportunity of rest 
is accompanied by a rapidly-increasing bond stress until a maximum bond resistance is reached at a definite amount 
of sUp (see Fig. 3). 

BmU. 71. Engineering Experiment Sutioo, University of Illinois. 




The true rdation of slip of bar to bond stress can best be studied by considering the action of a bar over a 
very abort section of the embedded length. The difficulties arising from secondary stresses made it impracticable 
to conduct tests on bars embedded very short lengths. The desired results (Fig. 3) were obtained by varying 
the forma of the specimens in such a way that the effect of different combinations of dimensions could be studied. 

Pull-oat tests with plain bars of the same sixe embedded different lengths furnish data which suggest the 
values of bond reaistanoe over a very short length of embedment, or indicate values of bond resistance which are 
iniependent of the length of embedment. Tests with bars of different sixe which were embedded a distance 
proportionAl to their diameters give the true relation when the effect of size of bar is eliminated. Two series of 
tivts of this Idnd on plain round bars of ordinary mill surface gave almost identical values for bond resistance after 
eliminating the effect of length of embedment and sise of bar, and we may consider that these values represent the 
stresses which were developed in turn over each unit of area of the embedded bar as it was withdrawn by a load 
applied by the method used in these tests. These tests showed that for concrete of the kind used (a 1 : 2 : 4 mix, 
storad in d&mp s»nd and tested at the age of about 60 days) the first measurable slip of bar came at a bond stress 
of about 260 lb. per sq. in., and that the maximum bond resistance reached an average value of 440 lb. per sq. in. 
If we conclude that adhesive resistance was overcome 

St the fizst measurable slip, it will be seen that the ad- 
hesive resistance was about 60% of the maximum bond 
resistanoe. This ratio did not vary much for a wide 
range of mixes, ages, sixe of bar, condition of storage, 

Widing resistance reached its maximum value 
for plain bars of ordinary mill surface at a slip of 
about 0.01 in. The constancy in the amount of slip 
corresponding to the maximum bond resistance for a 
wide range of mixes, ages, sixes of bar, conditions of 
•torages, etc., is a noteworthy feature of the tests. 
With farther slip the sliding resistance decreased 
slowly at first, then more rapidly, until with a slip of 
0. 1 in. the bond reaistanoe was about one-half its maxi- 
mum valoe. 

Bond Rssistaacs in Terms of Compressive 
Strsflgtli ol Coocrets. — ^Pull-out tests with plain round 
bars show end slip to begin st an average bond stress 
equal to about one-sixth the compressive strength of 
64n. cubes from the same concrete; the maximum bond 
resistanos is equal to about one-fourth the compres- 
sive strength of 6-in. cubes. These values were about 
the same for a wide range of mixes, ages, and conditions 
of storage. In terms of the compressive strength of 8 
by 16-in. concrete cylinders these values would be about 
13 <^for first end slip and 19% for the maximum bond 

Distribatioa of Bond Stress Along a Bar. — The 
tests indicate that bond stress is not uniformly dis- 
tributed along a bar embedded any considerable length 




1 — 










9 t 













.m h 


f tYUir 



Cmbedmenf variabfe. 
.♦ li'in. plain rounds^ 

Embedment yariabte. 
o Plain rounds. Diamehw 





edmenf" mar/ah 

1 1 






l-Z-4 concrm 


ge about- Zm 
1 f 







J02 X)3 

Slip of Bar -Inched 

and havinc the load applied at one end. Slip of bar Fxo. 3. — Relation of bond to slip of bar as load increases, 
begins first at the point where the bar enters the con- 
crete, and the bond stress must be greater here than elsewhere until a sufficient slip has occurred to develop the 
fway»Hiwm boud rcKstance at this point. Slip of bar begins last at the free end of the bar. After slip becomes 
general, there is an approximate equality of bond stress throughout the embedded length. 

Vsriatioa of Bond Resistance with Size, Shape, and Condition of Surface .of Bar. — The maximum bond 
resistance was not materially different for bars of different diameters. 

Rusted bars gave bond reastances about 15 % higher than similar bars with ordinary mill surface. 

The tests with flat bars showed wide variations of bond resistance and were not conclusive. Square bars 
gave valufes of unit stress about 75 % of those obtained with plain round bars. 

T-bais gare lower unit bond resistance than plain round bars, but gave about double the bond resiBtancc 
per anit of length that was found for the plain round bars of the same sectional area. 

With polished bars the bond resistance is due almost entirely to adhesion between the concrete and steel. 
Sameroos tests with polished bars embedded in 1: 2: 4 concrete and tested at 60 dajrs indicated a maximum bond 
resistaoee of about 160 lb. per sq. in., or about 60 % of the bond resistance of bars of ordinary surface at small 
amounts of slip. 

Adhesive resistance must be destroyed, sliding resistance largely overcome, and the concrete ahead of the 
projectioas must undergo an appreciable compressive deformation before the projections on a deformed bar 
beeoms sffsetive in taking bond stress. The tesU indicate that the projections do not materially assist in re- 
sisting a foree tending to withdraw the bar until a slip hss occurred spproximating that corresponding to the 


mazimura sliding raauitancc of plain ban. As slip conttmies a larger and larger portion of the bond stress is taken 
by direct bearing of the projections on the concrete ahead. 

In determining the comparative merits of deformed bars, the bar which longest resists beginning ol slip 
should be rated highest, other considerations being equal. 

The oonorete cylinders of the pull-out specimens with deformed bars were reinforced against bursting or 
splitting, because it was desired to study the load-slip relation through a wide range of values. In only a few tcata 
was the marimum bond resistance reached at an end slip less than 0.1 in. It should be recognised that, in gencraJ. 
the bond stresses reported for deformed bars at end slip of 0.05 and 0.1 in. could not have been developed with 
bars embedded in unreinforoed blocks. These high values of bond resistance must not be considered as available 
under the usual conditions of bond action in reinf orced-concrete members. In the tests in which the blocks were 
not reinforced, evidence of splitting of the blocks was found at end slips of 0.02 to 0.05 in. 

The normal components of the bearing str e s se s developed by the projections on a deformed bar may pro- 
duce very destructive bursting stresses in the surrounding concrete. The bearing stress between the projections 
and the concrete in the tests with certain tsrpee of commercial deformed bars was computed to be from 5800 to 
14,000 lb. per sq. in. at the highest bond stresses considered in these tests. The large slip and the high bearing 
st r ess e s developed in the later stages of the tests show the absurdity of seriously considering the extremely high 
values that are usually reported to be the true bond resistance of many tjrpes of deformed bars. 

Round bars with standard V-shaped threads gave much higher bond resistance at low slips than the com- 
mercial deformed bars. The average bond resistance at an end slip of 0.001 in. was 612 lb. per sq. in. The maxi- 
mum bond resistance was 745 lb. per sq. in. These were the only deformed bar tests in which failure came by 
shearing the surrounding concrete. 

The 1-in. twisted square bars gave a bond resistance per unit of surface at an end slip of 0.001 in., only 88 r; 
of that for the plain rounds. Following an end slip of about 0.01 in., these bars showed a decided decrease in bond 
resistance, and a slip of 5 to 10 times this amount was required to cause the bond resistance to regain its firat 
maximum value. After this, the bond resistance gradually rose as the bar was withdrawn. Some of the bars were 
withdrawn 2 or 3 in. before the highest resistance was reached. The apparent bond stresses at these slips were 
very high; but. of course, such stresses and slips could not be developed in a structure and could not hare been 
developed in the tests had the blocks not been reinforced against bursting. Such values sm entirely meaninglcas 
under any rational interpretation of the tests. 

Aachofinf of Reinforcing Ban. — The tests with plain round bars anchored by means of nuts and with washers, 
only showed that the entire bar must slip an appreciable amount before these forms of anchorage come into action. 
Anchorages of the dimensions used in these tests did not become effectire until the bar had slipped an amount 
corre s ponding to the in«.»imiii« bond resistance of plain bars. With further movement the apparent bond re- 
sistanoe was high, but was accompanied by excessive bearing stresses on the concrete. 

The load-slip relation for bars anchored by means of hooks and bends was not determined.^ 

Iflflneace of M athod of Caiiag Concrete. — Tests on specimens atomd under different conditions indicate that 
concrete stored in damp sand may be expected to give alxMit the same bond resistance and compressive rwristance 
as that stored in water. Water-stored specimens gave values of maximum bond resistance higher in each instance 
than the air-stmed specimens; the increase for water storage ranged from 10 to 45%. The difference seemed tf> 
inc r ease with age. The presence of water not only did not injure the bond for ages up to 3 years, but it was an 
important factor in producing conditions which resulted in high bond resistances. However, it was found that 
specimens tested with the concrete in a saturated condition gave lower values for bond than those which had been 
allowed to dry out before testing. The bars in specimens which had been immersed in water as long as 3^ years 
showed no signs of rust m other deterioration. 

laftaence of Freezing of Concrete. — Specimens made outdoors in freeaing weather, where they probably 
frose and thawed several times during the period of setting and hardening, were almost devoid of bond strength. 

loiloence of Age and Mix ef Concrete. — Pull-out tests made at early ages gave surprisingly high values of 
bond resistance. Plain bars embedded in 1 : 2 : 4 concrete and tested at 2 dajrs did not show e^ slip of bar until a 
bond stress of 75 lb. per sq. in. was developed. Bond resistance increases most rapidly with age during the firwt 
month. The richer mixes show a more rapid increase than the leaner ones. The tests on concrete at ages of 
over 1 year showed that the bond resistance of specimens stored in a damp place may be expected ultimatriy to 
reach a value as much as twice that developed at 60 days. 

The load-slip relation of leaner and richer mixes was similar to that for 1:2:4 concrete. For a wide range of 
mixes the bond reoUitanoe was nearly proportional to the amount of cement used. This relation did not obtain 
in a mix from which the coarse aggregate had been omitted. 

Bffect ef Coalimed end Re p e a ted Load. — ^When the application of load was continued over a conaiderabie 
period of time or when the load was released and reapplied, the usual relation of slip of bar to bond resists nee 
oonsiderabiy modified. The few tests which were made indicate that the bond stress corre sp onding to 
of slip i* the highest stress which can be maintained permanently or be reapplied indefinitely without failure of 

> Other tests have shown that a semicircular hook of 4 times the diameter of the bar and well embedded 

in concrete may be assumed to develop the elastic limit of the steel without exceeding the bearing strength of 

the eonerete. The curved ends should consist of bends through 180 deg. with a short length of straight rod 

beyond the bend. A short cross rod aids greatly in distributing the bearing stress in the conerete. Short 

ve hooks upon the ends of bars are not of great value. 





Xffsct of Coocnto Ssttiiic JSadmt Pnnm. — Bond resistanee of piain ban is greatly iDcrMsed if the eon* 
m mvmd to eet under preaeure. With a preesure of 100 lb. per aq. in. on the fresh oonerete for 5 days after 
the marinmm bood resistanee was increased 92 % over that of similar bars in concrete which had set 
Hie gjit m Ur density of the ooncrete and its more intimate contact with the bar seems to be 
for the inereaaed bond resistanoe. light pressures gave an appreciable increase in bond resistance, 
ban the effect of pressure was slight. 
As nii^t have been expected, the ecmpressive resistance of ooncrete setting under pressure was increased in 
ratio as the bond resistanofc At the age of 80 days the initial modulus of elasticity in compression 
wfaiidi set under a pressure of 100 lb. per sq. in. was about 37 % higher and the compressive strength 
by about 73 % over that of eonerete which had set without pressure. The density of the concrete, 
by the unit weights* was increased about 4 % by a pressure of 100 lb. per sq. in. on the fresh eon* 
increase in strength and doisity was relatively greater for the low than for the high pressures. A 
ttaned for 1 day, or until the ooncrete had taken its final set and hardening had begun, seems to have 
the same effect in inenasing the strength and elastic pr(^>erties of the concrete as when the pressure 
tinned for a much longer period. 

Beam Tests 

below. All 

eomputed values for bond stresses in the 6-ft. beams in the scries of 1911 and 1912 were as given 
were of 1 : 2 : 4 concrete, tested at 2 to 8 months by loads applied at the one-third points of the 
are given in pounds per square inch. 
In the beams reinforced with 
plaitt bars end slip begins at 67% of 
the maiiin nm bond resistanee; for 
the eomgated rounds this ratio is 
51 %, and for the twisted squares, 

Tbe band unit resistance in 
beams icsnf arced with plain square 
bars, computed on the superficial 
area of the bar, was about 75% of 
that for •■***'*■• beams reinforced 
with plain round ban of similar sise. 
reinforced with twisted 

of tests 

First end 
slip of bar 

End slip 

of 0.001 



1 and IV^-in. plain round. . . 
%^«in. plain round 





9ft-in. clain rountl 

1-in. olain souare 

1-in. twisted square 

iH-in. corrugated round 

square ban gave values at small slips about 85 % of those found for plain rounds. At the maximum load, the 
bood-unit stress with the twisted ban was 90 % of that with plain round ban of similar sise. 

In the beams reinforced with IH-in. corrugated rounds, slip of the end of the bar was observed at about the 
as in the fdain ban of companble sise. At an end slip of 0.001 in., the corrugated ban gave a 
about 6% higher and at the maximum load, about 30% higher than the plain rounds, 
in which the longitudinal reinforcement consisted of three or four ban smaller than those used in 
most ai the tests gave bond stresses which, according to the usual method of computation, were about 70% of 
the atresaea obtained in the beams reinforced with a single bar of large sise. It seems probable that the lower 
eompoted bond stresses in these tests are due to erron in the assumptions made as to the distribution of bond 
stress and not to actual differences of bond resistance in the ban of different sise. 

The tests on beams with the loads placed in different positions with respect to the span gave little variation 
ta bond usistanfii'i during the early stages of the tests. The maximum bond resbtanoes increased rapidly as the 
load iviproached the supports. These tests indicate that the variation in the maximum bond stresses must be due 
to the presence of other than normal beam action. 

Ttie bond stresses developed in the beam tests indicate that with beams of the same cross-section the bond 
stresses are distributed in the same way during the early stages of the test in beams varying widely in sxmui length 
and loadtng. During the later stages of the test, the distribution of bond stress seems to depend largely upon 
the eooditions of stress in the ooncrete through the region of the span where beam bond stresses an high. The 
distribution of bood stresses in beams of different cros s se ction appanntly varies with the relative dimensions of 
the beam and the reinf oreing ban. 

In the reinforoed-ooncrete beams it was found that very small amounts of slip at the ends of the bar repre- 
sented critical conditions of bond stress. For beams failing in bond the load at an end slip of 0.001 in. was 89 to 
94 *^ of the maximum load found in beams reinf oroed with plain ban, and 79 % of the maximum load for similar 
reinforoed with corrugated ban. As soon ss slip of bar became general, other conditions were introduced 
aoon caused the failure of the beam. 

The bood stecsses devdoped in a retnforoed-conerate beam by a load applied as in these tests varies widely 
over the region in whidi b^m bond stresses are present. High bond str e s ses are developed just outside the load 
paints at comparatively low loads. The load which first developed a bond stress nearly equal to the maximum 
bood roristaooe in the region of beam bond stresses produced a stress near the support which was not mfjre than 
about 15 to 40 % ol the maximum bond reststanoe. As the load is increased, tlie region of high bond stress is thrown 

the supfMxt, and at the same time the bond stress over the region just outside the load point 


beoomefl tteadily smaUer. Thia indioates a piecemeal development of the maximum bond atreM ai the load a* 
increased. The actual bond etresses in certain tests varied from lees than one-half to more than twice the averaicv* 
bond resistance computed in the usual manner. 

Slip of bar in a reinforoed-oonorete beam has a marked influence in increasing the center deflection during 
the later stages of loading. 

The comparison of the bond stresses developed in beams and in pull-out specimens from the same nmtcrial* 
is of interest. 8uoh a comparison should be made for similar amounts of slip. In the pull-out teste the mazimum 
bond resiBtance came at a slip of about 0.01 in. for plain bars. The mean bond resbtance for the deformed hmi> 
tested was not materiallv different from that of the plain bars until a slip of about 0.01 in. was developed ; with a 
continuation of slip the projections came into action and with much larger slip high bond stresses were developed 
The beam tests showed that about 79 to 94 % of the mazimum bond resistance was being developed when the bar 
had slipped 0.001 in. at the free end; hence the bond stress developed at an end slip of 0.001 in. was used as a baais 
of the principal comparisons in the pull-out tests. However, it is recognised that, under certain conditions, the 
stresses developed at larger amounts of slip may have an important bearing on the effective bond resistance of 
the bar. 

The puU'Out tests and beam tests gave nearly identical bond stresses for similar amounts of slip in many 
groups of tests, but it seems that tlus was the result of a certain accidental combination of dimensions in the two 
forms of specimens and did not indicate that the computed stresses in the beams were the correct stresses. How- 
ever, it is believed that a properly designed pull-out test does give the correct value of bond reststanoe, and givetf 
values which probably closely represent the bond stresses which actually exist in a beam or other member as 
slipping is produced from point to point along the bar. The relative position of the bar during molding may be 
expected to influence the values of bond resiBtance found in the tests. 

A working bond stress equal to 4 % of the compresnve strength of the concrete tested in the form ol 8 by 16- 
in. cylindefs at the age of 28 days (equivalent to 80 lb. per sq. in. in concrete having a compressive strength of 
2000 lb. per sq. in.) is as high a stress as should be used. This stress is equivalent to about one-third that causing 
first slip of bar and one-fifth of the maximum bond resistance of plain round bars as determined from pull-out tests. 
The use of deformed bars of proper design may be expected to guard against local defideneies in bond resistance 
due to poor workmanship and their presence may propwly be considered as an additional safeguard against ulti- 
mate failure by bond. However, it does not seem wise to place the working bond stress for deformed bars higher than 
that used for plain bars. 

3. Length of Embedment of Reinforcing Bars to Provide for Bond. — ^Let /• be the working 
tensile strength of the steel, A, the area of bar, o the circumference of bar, d the diameter or 
thickness of bar, u the working unit bond strength, and x the required length of embedment 
(or grip) for the above values of /• and u. Then, to develop the strength of the steel, using 
either round or square bars, 

i. IT* - 

xau = Aj, ^-r'J* 

4. Ratio of the Moduli of Elasticity. — ^Let /« » unit stress in steel, /e — unit stress in con- 
crete, Et » modulus of elasticity of steel, and Ec » modulus of elasticity of concrete. Since 
the modulus of elasticity of a material is the ratio of stress to deformation, it follows that, for 
equal deformations, the stresses in. the steel and concrete will be as their moduli of elasticity. 

/- E. 
fc " E. 

This ratio uf the moduli is generally denoted by the letter n, or 

/• ^ nfc 

The equation just given shows that if the stress in either the steel or concrete of a concrete 
column is known, the stress in the other material can be found, and this relation is made use of 
in the derivation of column formulas. Fig. 26, Sect. 5, page 250, shows that the modulus of 
elasticity of concrete in compression is less for the greater loads, and hence the value of n is 
greater. Thus, it is plain that with increasing loads in concrete columns the steel receives a 
greater proportionate stress, the variation in the amount carried by the steel depending on the 
variation in the value of n. In order to take account of the fact that under increasing loads the 


steel receives an mcieasmg proportion, it is desirable to use a value of n in the computations 
for design somewhat larger than that which is obtained by taking a value of Eg corresponding 
to working loads on small prisms (about 10). A value of 15 for n may well be used for the ordi- 
nary 1:2:4 mix. 

In concrete beams, experiments show that the tension which remains in the concrete just 
below the neutral axis, and properly not allowed for in the derivation of the beam formulas, 
has its effect in the position of the neutral axis and the strength of the beam. It is found that 
a value of 15 for n is not too large for calculations of strength of beams, assuming the ordinary 
1:2:4 mix, although great accuracy in this respect is not necessary. This value of 15 for 
n is the one most generally used, but a value of 12 is also frequently employed. The value of 
15 corresponds to a value of Ee of 2,000,000 which is somewhat low as determined by compressive 

For the proper values of n to use for other mixtures see recommendations of the Joint 
Committee, Appendix B. 

Comparatively few tests have been made on the elasticity of concrete in tension, but these 
s«eein to indicate that for small stresses, it is practically the same as in compression, although 
probably slightly less. 

S. Behavior of Reinforced Concrete Under Tension. — Early tests indicated that the ulti- 
mate stretch of reinforced concrete in tension is as much as 10 times that of plain concrete, 
but such results were due to the fact that it was found extremely difficult to determine just when 
the concrete begins to crack. Cracks do not become noticeable, even on very close examination, 
until a stretching occurs corresponding to a tensile stress much beyond the ultimate tensile 
strength of the concrete. The steel causes a uniform elongating of the concrete so that the 
cracks which open up are very small and remain invisible for some time. 

A method of detecting minute cracks in the tensile side of beams was accidentally discovered 
in 1901-02 in some experiments made at the University of Wisconsin. It was found that when 
beams were hardened in water and only partially dried before testing, very fine hair-cracks be- 
came noticeable at a moderate load. Before these cracks occurred, however, dark wet lines 
appeared across the beam, and it was observed that each of these lines was later followed by a 
very fine crack. These water-marks were proven to be incipient cracks by the sawing out of 
a strip of concrete along the outer part of the beam. Careful measurements of extension showed 
that these streaks, or water-marks, occurred at practically the same deformation at which the 
concrete ruptured when not reinforced. This same phenomenon has since been observed by 
many careful experimenters, and the fact is now generally established that concrete, reinforced 
with steel, does not elongate under tensile stress to any greater extent before cracking than plain 

A reinforced-concrete beam for working loads is usually more heavily stressed on the ten- 
sion side than the ultimate tensile strength of plain concrete — enough steel being usually em- 
bedded near the lower face to permit the full allowable compressive strength of the concrete 
to be utilized. The presence, then, of the cracks above referred to, accurring long before a 
reinforced-concrete he&m has obtained its working load, must seriously affect the tensile strength 
of the concrete. The moment formulas now in most general use for the design of reinforced- 
concrete beams neglect entirely the tensile strength of the concrete. 

Experiments have shown that concrete when well placed and mixed somewhat wet, com- 
pletely protects the steel in the tensile side of a beam from corrosion, even when the unit stress 
in the steel somewhat exceeds the elastic limit. 

S. Shrinkage and Temperature Stresses. — ^In reinforced-concrete structures which are 
free to contract and expand, the stresses occurring from temperature changes and from shrink- 
age in hardening are due wholly to the mutual action of the steel and concrete. Of the stresses 
produced from these two causes, those which result from hardening are the greater, but experi- 
ments show that even these are not sufficient to be of practical importance. In regard to the 


temperature stresses, they are negligible by reason of the nearly equal rates of expansion of the 
two materials. 

On the other hand, if reinforced-concrete structures are restrained by outside forces, or if 
they are of such dimensions that they cannot be considered as sufficiently well bonded to act 
as a unit — such as long retaining walls — ^then the stresses resulting are much greater, and the 
tensile strength of the concrete will be reached (this will occur with a drop in temperature some- 
where between 10 and 20°F.)f thus producing cracks, called contraction cracks. To prevent 
plainly noticeable cracks due to shrinkage and lowering of the temperature, all exposed surfaces 
should be reinforced with about 0.3 of 1 % of steel, based on the cross-section of the concrete. 
This b less than the amount required theoretically, but experience shows this amount to give 
very satisfactory results where the foundations are stable. If the structure is fixed in two 
directions, the reinforcement must be placed accordingly. The above percentage of steel 
should be figured for an area of crossHsection of maximum thickness of about 12 in. 

No amount of reinforcement can entirely prevent contraction cracks. The steel can, 
however, if of small diameter and placed close to the surface, force the cracks to take place at 
such frequent intervals that the required deformation occurs without any one crack becoming 
large. No cracks will open up to be plainly noticeable until the steel is stressed beyond its 
elastic limit. The amount of steel should be such, then, that without being stressed beyond its 
elastic limit, it will withstand the tensile stress resulting from the maximum fall of temperature 
(usuaUy considered to be 50**) in the steel itself plus the tensile stress necessary to crack the con- 
crete. A high elastic-limit steel is thus advantageous. 

The size and spacing of the cracks will also depend upon the bond strength of the rein- 
forcing rods. The distance between cracks in any given case will be the length required to 
develop a bond strength equal to the tensile strength of the concrete. Thus, bars with ir- 
regular surfaces which provide a mechanical bond with the concrete are in general more effective 
than smooth bars. 

7. Weight of Reinforced Concrete. — Reinforcing steel in the usual proportions adds from 
3 to 5 lb. to the weight of plain concrete per cubic foot. The weight of plain concrete for the 
various kinds of aggregate may be found on page 263. Reinforced concrete is usually assumed 
as 150 lb. per cu. ft. in making computations for design. 



L Forces to be Resisted. — As expressed by the Joint Committee the forces to be resisted 
are those due to: 

1. The dead load, which includes the weight of the structure and fixed loads and forces. 

2. The live loadj or the loads and forces which are variable. The dynamic effect of the 
live load will often require consideration. Allowance for the latter is preferably made by a 
proportionate increase in either the live load or the live-load stresses. The working stresses 
recommended (see Appendix B) are intended to apply to the equivalent static stresses thus 

2. Distrilmtioii of Stress in Homogeneous Beams. — The following statements and for- 
mulas are in accordance with the theory of homogeneous beams : 

1. At any cross-section the internal forces, or stresses, may be resolved into normal and 
tangential components. The components normal to the section are stresses of tension and 
compression, while the tangential components add together and form a stress known as the 
resisting shear. 

2. The shear at any cross-section is borne by the tangential stresses in that section. The 
moment at any section is borne by the component stresses normal to that 


3. The neutral axis passes through the center of gravity of the cross- 

4. The intensity of stress normal to the section increases directly with 
the distance from the neutral axis and is a maximum at the extreme fiber 
(Fig. 1). The intensity of this stress at any given point in the cross-sec- Fig. 1. 
tion is given bv the formula 


in which / » fiber stress at distance y from neutral axis. 

M = external bending moment at section in inch-pounds. 
y = distance in inches from neutral axis to any fiber. 
/ s moment of inertia of the cross-section about the neutral axis. 

5. The general formula which gives the longitudinal shear per square inch {v) at any 

desired point in the cross-section is 


in which V — total shear at the section in pounds. 

Q = statical moment about the neutral axis of that portion of the cross-section lying 

either above or below (depending upon whether the point in question is above 

or below the neutral axis) an axis drawn through the point in question parallel 

to the neutral axis. 

/ = moment of inertia of the cross-section about the neutral axis. 

6' •= width of beam at the given point. 




[Sec 7-2 

In the above formula, by the term statical moment is meant the product of the area mentioned 
by the distance between its center of gravity and the neutral axis. For example, the longi- 
tudinal shearing intensity at a point c in a rectangular beam, Fig. 2, may be expressed as 



V — 


I Areo 

portion,- A* 

Fio. 2. 

For rectangular beams and all beams of uniform width, the largest value of v for any given sec- 
tion will occur at the neutral axis since the statical moment 
Q has its maximum value for a point on this axis, and 6 is 

6. If a beam is of constant cross-section throughout, the 
maximum values of / and v will occur at the section where 
M and V respectively have maximum values. 

7. In addition to the longitudinal or horizontal shear at 
any point there coexists a vertical shear and the intensity of 
this vertical shear is equal to the intensity of the horizontal 

8. The intensity of the shear at the top and bottom of a beam is zero and the intensity 
of shear (horizontal and vertical) along a vertical cross-section for a rectangular beam varies 
as the ordinates to a parabola, as shown graphically in Fig. 3. The maximum value occurs 

3 V 
at the neutral axis and is % the average intensity, or ^ - r^* 

9. At the neutral plane there exists a tension and compression at angles of 45 deg. to the 
horizontal, and the intensity of these forces is equal to that of the shear. 

10. At the end of a simply supported beam where the shear is a 
maximum and the bending moment a minimum, the stresses lie prac- 
tically at 45 deg. to the horizontal throughout the entire depth of 

11. At the section of maximum moment, the shear is zero and the 
stresses are horizontal. 

12. If / represents the intensity of horizontal fiber stress and v the intensity of vertical 
or horizontal shearing stress at any point in a beam, the intensity of the inclined stress will be 
given by the formula 


Fio. 3. 

and the direction of this stress by the formula 



where K is the angle of the stress with the horizontal. 

13. At any given point maximum compressive 
stress and maximum tensile stress make an angle of 
00 deg. with each other. 

14. The directions of the maximum stresses for 
a simply supported beam uniformly loaded are as 
given in Fig. 4. The general direction of the stresses 
in a beam with any given loading may be determined 
by means of the formulas for t and K given above. 

15. The common theory of flexure gives the unit stress correctly at the important section 
of maximum moment and also for the extreme fibers in other sections, since at these points 
the shear is zero. Where the shear is not zero an inclined stress is the result and the flexure 
formula gives only the horizontal component of this stress — ^namely, the fiber stress. 

Ufic« of maxMimii tarwdn 
..... uns» of mosunum 

Pio. 4. 

Sm. 7-3] 



FxQ. 5. 

8. Assumptioiis in Theory of Flexure for Homogeneotts Beams. — ^The two main assump- 
tions in the common theoiy of flexure are: 

1. If, when a beam is not loaded, a plane cross-eection be made, this crossHsection will still 
be a plane after the load is put on and bending takes place (Navier's h3rpothesis). 

2. The stress is proportional to the deformation — namely, to the 
amount of elongation or compression per unit of length (Hooke's Law). 

From the first assumption it follows that the unit deformations of the 
fibers at any section of a beam are proportional to their distance from 
the neutral axis. By means of the second assumption the important 
principle is established that the unit stresses in the fibers are also pro- 
portional to the distances of the fibers from the neutral axis. 

4. Plain Concrete Beams. — ^The first assumption in the common 
theory of flexure, as given in the preceding article, may be applied directly 
to plain concrete and also to reinforced-concrete beams. Careful measurements seem to show 
some deviation from a plane, but in general this assumption seems to be warranted. From 
this fact it follows (as stated above) that deformations of the fibers are proportional to the 
distances of the fibers from the neutral axis. OS in Fig. 5 is the stress-deformation diagram 
^^__^_^__^^^___.^_^_^ for concrete in compression with the deformations repre- 
^ \ X- sented vertically. The curve OT is the stress-deforma- 

'^^i^: ^^^^-:rrsrr:z^::.-:r,r::S^!!^S(^ tion diagram for concrete in tension. For working loads 
♦ ' the curves OS and OT do not vary materially from straight 

lines and the unit stresses in the fibers at any section of a 
plain concrete beam may thus be assumed to vary directly 
as the deformations and consequently as the distances of the fibers from the neutral axis. 
Hence, the conmion flexure formula for homogeneous beams applies when the loads are work- 
ing loads. For ultimate loads, however, ths formula does not strictly apply. 

A plain concrete beam will fail by cracks opening up along the uneven lines which are 
shown in Fig. 4 on account of the low strength of concrete in tension. If concrete were only 
stronger in tension, then the plain concrete beam might be of 
some structural value. In order to offset this disadvantage of 
plain concrete, steel is used. 

6. Purpose and Location of Steel Reinforcement. — Steel 
reinforcement should have the general directions shown in Fig. 
6 in order to take the tension in the beam and prevent the 

cracks starting along the lines indicated. Fig. 7 is the simplest method of reinforcement 
and quite often used for light loads. In beams highly stressed, curved or inclined reinforce- 
ment is needed, in addition to the horizontal rods. The most common method is to use several 
bars for the horizontal reinforcement and then to bend up some of these at an angle of from 30 
to 45 deg. as they approach the end of the beam and where they are not needed to resist bending 

stresses. The concrete is depended upon to take care of 
the compressive and pure shearing stresses, its resistance 
to such stresses being large. 

6. Tensile Stress Lines in Reinforced-concrete 
Beams. — Lines of maximum tension in the concrete of re- 
inforced-concrete beams are considerably inclined imme- 
diately above the line of the steel. The inclination of 
these lines is greater, the greater the shear, and the less 
the horizontal tension. The inclination, therefore, increases toward the end of the beam. At 
points nearer the neutral plane, the horizontal tensile stresses become less and the inclined ten- 
sion approaches the value of the shearing stress, while its inclination approaches 45 deg. Fig. 
8 is an attempt to represent roughly the general direction of the inclined tensile stresses in a 
simply supported beam uniformly loaded and with horizontal reinforcement. 

Fia. 7. 

Vtr y HW« fnyqn m 
on occount c€ concrete 

Fig. 8. 



(Sec. 7-7 

7. Flexure Foimulas for Reinforced-concrete Beams. — A great many varietiee of flexure 
formulaa have been proposed from time to time to be used in the design of reinforced-concrete 
beams. As might be expected, many of the earlier formulas considered the concrete to carry 
its share of the tension which we know now cannot be done with safety. Only two classes of 
flexure formulas are at the present time in practical use. In each of these classes, tension in 
the concrete is neglected and a plane section before bending, is assumed to be a plane after 
bending takes place. 

The formulas almost universally used and made standard by the Joint Committee 
relate to working stresses and safe loads, and are based on the straight-line theory of stress 
distribution. The other formulas referred to above relate to ultimate strength and ultimate 
loads and the stress-deformation curve for concrete in compression is assumed to be a full 
parabola. Ultimate4oad formulas are used to such a limited extent that only a few pages of 
this handbook are devoted to their consideration — ^namely. Arts. 10 and 1 1. 

8. Assumptions in Flexure Calculations. — The following assumptions are made in deriving 
the flexure formulas: (1) the adhesion of concrete to steel is perfect within the elastic limit of 

the steel; (2) no initial stresses are con- 
sidered in either the concrete or the steel 
due to contraction or expansion; (3) the 
applied forces are parallel to each other 
and perpendicular to the neutral surface 
of the beam before bending; (4) sectional 
planes before bending remain plane sur- 
faces after bending within the elastic limit 
of the steel ; (5) no tension exists in the con- 
crete; (6) modulus of elasticity of concrete 
is constant. 

9. Flexure Formulas for Working Loads — Straight-line Theory. — The unit stress in the 
steel is within the elastic limit, and the unit stresses in the concrete at the given section of the 
beam are considered to vary as the ordinates to a straight line (see Fig. 9). Tension in the 
concrete is neglected. The formulas follow^ (see Notation, Appendix D): 

Fxo. 9. 

* The formulas may be derived as foUown: 

Total compreaeive resutanoe — total tenule resistance, or 

>4A*W - A J. 

From the assumption that deformations vary as the distances of the fibera from the natural axis and 
stress proportional to deformation 


which reduces to 

/• — /«» — r—* or /e - 

SJbd E^{1 - k) 

n(l - *) 

,or A; 



The total resisting moment of the beam is the sum of the moments of the total compressive st 
the total tensile stresses about the neutral ads, xx 

M - HkdiH/ckbd) + rf(l - k)AJ, 
- J4M*6d« + A,Ad{l - k) 

EliminatinK k between equations (a) and (b), the following formula fur steel ratio results 



and of 



Introduetng the value of /• from equation (6) into equation (a), we have 

H**M - ^.fi( 1 - i) - 

y%i^ - pbnil - k) 

S«:. 7-^1 BEAMS AND SLABS 277 




k — V2pn '\- (pn)^ pn- - .- 

A. \i Uk 
P " M /. /7. , , \ 2/. 
/. W. / 


-HMi(M«), or M'=|^' or /« = l^i 


-pU(M*), or M'=py^-- or /. " -^ j^ 

/-"^•^ or -/•*,. 
A: 7t(l — k) 


show that for a eiven ratio of -'» p and A; remain t 

of beams.. The formula for Me gives the resisting moment when the maximum allowable value 
of /« is introduced as the limiting factor and the formula for M, gives the resisting moment 
when the maximum allowable value of f, is the limiting factor. The lesser of these two re- 
sisting moments, when proper working values are assigned to /« and /., is the safe resisting 
moment of the beam in question. 

Unlike steel beams, reinforced-concrete beams require a preliminary formula to be solved 
before the formula for resisting moment may be employed. Solving this preliminary formula 
locates the position of the neutral axis which is in the same position only for beams of a given 
percentage of steel reinforcement. 

The method of procedure in flexure formulas is to determine the vertical section of the beam 
where the moment is a maximum and apply the formulas at that section. Either formula for 
p, containing the values of fe and /«, determines the amount of steel reinforcement which is 
needed to cause the beam to be of equal strength in tension and compression. The formulas 
for resisting moment determine the bending moment which a beam will safely withstand (for an 
existing structure) or the size of the beam needed to resist a given bending moment (for a pro- 
posed structure). 

If a beam is over-reinforced, its resisting moment depends on Me, and if under-reinforced 
on Mg. 

If it is desired to find the fiber stresses in concrete and steel of a given beam, the formulas 

/, = — — -and/r = T-T— (or/e = — 7— I should be used, where M is the external bending 
A^jd kjhd* \ k / 

moment in each case. For a given external 3f , either M* — or W = may be used to de- 
Mi Pfj 

from whieh 

* - \^2pn + ipn)* - pn 
Substituting the value of A^fa from (a) into (c), we get 

Mc - H/e*(l - H t) W* 


Me - HfJejbd^ 

Sabetituting the value of /« from (a) into (e), and remembering that A« — pbd, 

M, - pfjbd* 
Equation (a) may be solved to give 

/c - — , or p - ^^ 



[Sec. 7-7 

7. Flexure Formtilas for Reinforced-concrete Beams. — A great many varieties of flexure 
formulas have been proposed from time to time to be used in the design of reinforced-concrete 
beams. As might be expected, many of the earlier formulas considered the concrete, to carry 
its share of the tension which we know now cannot be done with safety. Only two classes of 
flexure formulas are at the present time in practical use. In each of these classes, tension in 
the concrete is neglected and a plane section before bending, is assumed to be a plane after 
bending takes place. 

The formulas almost universally used and made standard by the Joint Committee 
relate to working stresses and safe loads, and are based on the straight-line theory of stress 
distribution. The other formulas referred to above relate to ultimate strength and ultimate 
loads and the stress-deformation curve for concrete in compression is assumed to be a full 
parabola. Ultimate-load formulas are used to such a limited extent that only a few pages of 
this handbook are devoted to their consideration — ^namely, Arts. 10 and 11. 

8. Assumptions in Flexure Calculations. — The following assumptions are made in deriving 
the flexure formulas: (1) the adhesion of concrete to steel is perfect within the elastic limit of 

the steel; (2) no initial stresses are con- 
sidered in either the concrete or the steel 
due to contraction or expansion; (3) the 
applied forces are parallel to each other 
and perpendicular to the neutral surface 
of the beam before bending; (4) sectional 
planes before bending remain plane sur- 
faces after bending within the elastic limit 
of the steel; (5) no tension exists in the con- 
crete; (6) modulus of elasticity of concrete 
is constant. 

9. Flexure Formulas for Working Loads — Straight-line Theory. — ^The unit stress in the 
steel is within the elastic limit, and the unit stresses in the concrete at the given section of the 
beam are considered to vary as the ordinates to a straight line (see Fig. 9). Tension in the 
concrete is neglected. The formulas follow* (see Notaiiony Appendix D): 

Fko. 9. 

* The formulas may be derived ua follows: 

Total compressive resistanoe — total tensile resistance, or 

W.kbd - AJ. 

From the assumption that deformations vary as the distances of the fibers from the natural axis and 
stress proportional to deformation 


which reduces to 

EjBd E^a - k) 
f» ■• fen — r — • or fe ^ ZTT^TTx* or * - 

»(1 - *)' 

1 + 


The total reabting moment of the beam is the sum of the moments of the total compressive st 
the total tensile stresses about the neutral axis, or 

.V - H W(HMW) + d(l - k)A^, 
- M/ck^bd* + AJMl - *) 

Eliminating k between equations (a) and (6), the following formula for steel ratio results 





Introdueing the value of /• from equation (6) into equation (a), we have 

\^hd - -4.n( 1 - A ) - 

>»*«6 - p6fi(l - *) ^ 

tr.i^t>n .1, - ^ ,,"■ 

.1 principle 

lOe- Thus, 

'jlosa in each 


termine cross-eection, when the p used is obtained from the formula p »■- : or from 

p = — -» m which k = — • 

2/. ,+A 


IixuBTRATiVE Pboblbm. — What will be the resisting moment (M) for a beam whose breadth (6) is 8 in. with 
a distance from the center of the reinforcement to the compression surface ((0 of 12 in., the area of steel section being 
0.96 sq. in.? Assume n > 15; fe ■■ 650 lb. per sq. in.; and /• ■■ 16,000 lb. per sq. in. 

From (1) 

From (4) 
From (5) 

A, _ 0.96 ^^, 

P " W "" (8) (12) " °-°^ 

* - \/(2)(0.01)(15) -f- (0.01)*(15)* - (0.01)(16) - 0.418 
j - 0.861 

Me - H(650) (0.418) (0.861)(8) (12)* - 134,700 in .-lb. 

Mm - (0.01) (16,000) (0.861) (8) (12)* - 158.700 in.-lb. 

Me is the lesser of the two resisting moments and hence controls in the design. 

iLLUSTBATnrx Pbobleii. — Assume the beam of the preceding problem to be 14 in. deep and subjected to a 
bending moment of 130,000 in.-lb. Compute the maximum unit stresses in the steel and concrete. 

From (1) 

From (4) 

From (5) 

A, 0.96 nnnim 

** " 6d "■ (8)(14) " O™**® 

k - V'(2) (0.0086) (16) + (0.0086) '(IS)' - (0.0086) (15) - 0.395 
j - 0.868 

130,000 " \^) (0.395) (0.868) (8) (14)« 
fe "■ 480 lb. per sq. in. 

130.000 - (0.0086)(/,)(0.868)(8)(14)* 
/• - 11,100 lb. per sq. in. 

Illustratxvjb Problsm. — A beam is to be designed to withstand a bending moment of 300,(X)0 in.-lb. and to 
have equal strength in tension and compression. A 1 : 2 : 4 concrete will be used with Bg ■• 2.000,000 and fe •■ 600 
lb. per sq. in. The pull in the steel Ib to be limited to 14,000 lb. per sq. in. Its modulus of dastidty Bt is 30,000,000. 

B. _ /. 70 

From (1) and (2) 

k - 

1 + 

* " 14000 — " °'^^ *"** ' " 0.870 

From (3) 

( 600) (0.39 1) 

^ " (21)(14.00b) 


Either (4) or (5) may now be used in determining 6 and d since the amount of steel to be employed will eause 
simultaneous majdmum working stresses. 

From (5) 

^ 300,000 

*" (0.008i)(14.000)(0.870) ^"^ 

Many different values of 6 and d will satisfy the last equation. If b is taken as 10 in., then 

d« - ^J5^ - 293, or d - 17M in. 


A. - (0.0084) (10) (17.25) - 1.45 sq. in. 
If 194 ia. is allowed between the teanon sttrfaee ol the ooncrete and the oonter of the Btaei« the entire depth of the 
beam should be 19 in. 

Sec. 7-10] 



10. Flexure Formulas for Ultimate Loads. — The stress-deformation curve for concrete in 
compression is assumed to be a full parabola. Experiments show this to be very nearly the 
case for ultimate loads. The amount of reinforcement is considered as sufficient to develop 
the full compressive strength of the concrete without stressing the steel beyond its yield point. 
Failure under such conditions (Fig. 10) will occur by crushing the concrete. 


Avtroge compreMive 
streftft * § "^ 

lotat. oompr«o«iv« 

Fia. 10. 

The formulas follow (see Notation, Appendix D) : 




j ^1 -Hk. 

% ^2 fek 

3 ' /. 

^(-^ + 1^ 
/. \2nfc ^ V 

Me - Hfckj(bd^), or M« = 



M. = pfj(hd^)y or 6d« = 







In the above formulas, /, — elastic limit of the steel, and fe = ultimate compressive strength 
of concrete. 

When using the above formulas, it should be remembered that the amount of steel in the beam 
is assumed as sufficient to cause the ultimate resisting moment to be due to the concrete. Thus, 
the resisting moment of the beam may be figured by using the formula for Me. If an amount 
of steel is used such that the ultimate strength of the concrete and the elastic limit of the steel 
would be reached simultaneously, either Me or M, may be used to determine the ultimate 
resisting moment. If a less amoimt of steel is used than the amount just mentioned, the condi- 
tions of the assumption do not hold, and the formulas given above cannot be \ised. When this 
happens the ultimate moment may be figured by means of formulas based on a parabolic varia- 
tion of compression in the concrete and applicable for any load up to the ultimate. The para- 
bola for such a case is not a full one and the formulas are cumbersome to use and not at all 
fitted for practical use. 

The formulas for ultimate loads, however, can readily be employed to design a beam for 
equal strength in tension and compression. The method is to first find the required amount 

M M 

of steel. Then either W* = —. — —-orbcP =---.may be used to determine the size of beam 


Ii*LU8TRATivB Problsu. — A beam is to be designed to have equal strength in tension and compression and 
to safely withstand a bending moment of 150,000 in.-lb., the ultimate compressive strength of the concrete being 
taken at 2000 lb. p«r in. and the elastic limit of the steel at 40,000 lb. per sq. in. Assume n >■ 16. 



ISec. 7-n 





1 + - 
* ^ 30 

J - 1 - H* - 0.775 

P - 

2 0.598 


3 20 

With a factor of safety of 4, the ultimate bending moment is 600,000 in.-lb. and 



With 6-8 in., then 


( Ji) (2000) (0.598) (0.776) 



121.5, or d - 11 in. 
A. - (0.02) (8) (11) - 1.76 sq. in. 

11. Flexure Formulas for Working Loads and for Ultimate Loads Compared. — ^Formulas 
for ultimate loads are open to the objection that when a factor of safety is applied which will 
bring the stress in the concrete to about a good working stress, the stress in the steel becomes 
unduly low from a standpoint of economy. A factor of safety of 3 or 4, as is usually taken, 
leaves a high stress in the concrete with the stress in the steel far below what is usually consid- 
ered a safe stress. Beams designed by the ultimate load formulas will generally be of smaller 
cross-sectional dimensions than when the straight-line formulas are employed ; but, on the other 
hand, a larger amount of steel is required. Practically identical results will be obtained by the 
two classes of formulas if about 15% lower compressive stress is permitted in the concrete by 
ihe ultimate load formulas than by the formulas based on the straight-line theory. There 
seems to be no good reason why the simple formulas based on the theory of straight-line stress 
variation should not be used for purposes of design, safe working stresses being employed. 

12. Lengths of Simply-supported Beams. — The span length for beams and slabs simply 
supported should be taken as the distance from center to center of supports, but need not be 
taken to exceed the clear span plus the depth of beam or slab. 

13. Shearing Stresses. — In Fig. 11 is shown a small portion of a concrete beam, so short 
that no appreciable portion of the load on the beam acts directly upon it. The opposing total 

compressive forces are denoted by C and C; and the tension 
in the steel on each face by T' and T, The tension in the 
concrete may be neglected. Let V be the total shear on 
this small portion of the beam. From conditions of equili- 
brium, C ^ T' and C — T. The total horizontal shearing 
stress upon a horizontal section immediately above the steel 
is T' — 7, and if b denotes the breadth of the beam and v 
y the imit shear (horizontal or vertical) at any point between 
the neutral axis and the steel, then 

Fia. 11. 

F — 


r - T 

The various couples acting upon the element produce equilibrium ; hence 

Vx = (7" - T)jd 


Substituting this value in equation (1) there results 

which is the value of shear intensity at any point between the neutnd axis and the steel. 

Sec.T-14| BEAMS AND SLABS 281 

The value of j for working loads varies within narrow limits and v will change but slightly 
if the difierent values of j are inserted in equation (2). The average value of j for beams in 
ordinary coDBtruction is J^. Using this value, equation (2) reduces to 

7 bd 





Shearing stress is the same at all points between the neutral axis and the steel, and above 
the neutral axis it follows the parabolic law. Fig. 12 represents the distribution of shearing 
strees on a vertical cross-section assuming no tension in the concrete. 

The longitudinal tension in tbe concrete near the end of beam modifies the distribution 
of the shear, increasing the shearing stress somewhat at the neutral axis and decreasing it at 
the level of the reinforcement. Equation (3), however, gives results 
which are sufficiently accurate and are derived for beams having the 
horicontal bars straight throughout. When any web reinforcement is - 
used, the distribution and the amount of the shearing stresses at the end 
of a simply supported beam are materially different frum the foregoing. 
Tlie analysis of the stresses becomes more complex and a determination 
of their value impracticable. Even here, however, the above formula 
serves a useful purpose. It ia found that shear is the chief factor in the 
failure of a beam by diagonal tension and either formula (2) or formula (3) may be used in de- 
sign if properly controlled by the results of experiments. 

Failure by the actual shearing of the concrete in a beam is not a likely occurrence under 
any conditions as the shearing strength of concrete ia at least one-half the crushing strength. 

14, Uethods of Strengthening Beams Against Failure in Diagonal Tension. — The in- 
tensity of the diagonal tensile stress at any point in a beam depends upon the shear and hori- 
lontal tension in the concrete, with shear as the chief factor. The percentage of horizontal 
reinforcement must also be considered, since the amount of steel employed affects the hori- 
zontal deformation and consequently the tension in the 
, concrete. Thus beams may be strengthened against 
failure in diagonal tension by keeping the horizontal 
tension small tiirough the use of considerable horizontal 
steel at pomts of heavy shear, by avoiding heavy shear- 
ing stresses, and by providing some type of web rein- 
forcement. A low unit working stress in whatever 
type of web reinforcement is employed is also much to 
be preferred. 

The most unfavorable part of a beam as regards 
diagonal tension is at points of excessive shear com- 
bined with considerable bending moment. A sufficient 
number of reinforeing rods should be extended horizon- 
PiQ, 13 tolly to the ends of the beam to provide for bending 

with low unit stresses in the steel. In small beams, ver- 
tical stirrups looped about the horizontal rods may be employed throughout for web reinforee- 
ment but in large beams under heavy shearing stresses, both stirrups and bent rods should be 
used. The stirrups in large beams should be securely fastened to the longitudinal rods in 
such a way as to prevent slipping of bar past the stirrup. Inclined web members may also 
be used in place of vertical stirrups if securely attached to the horizontal rods. Vertical 
stirrups may be made in various forms, as indicated in Fig. 13. 

U. Bloment and Diagoaal-tension To stB— General.' — When a beam begins to fail by 
yielding of the steel at, or near, the section of maximum bending moment, any further load 

< For detailed Inslmcnl oF this lubjert, lee "Concrete. Plain and Rrin[or<;nl" by Tituik and Thouphin 
(1916 F^ditiun). 




[Sec. 7-15 

rapidly increases the deformation, large cracks open up in the concrete on the tension side, the 
neutral axis rises on this account, and the ultimate failure soon occurs by the crushing of the 
concrete. A steel tension failure is found to occur when the amount of steel used is less than 
the amount determined by theoretical formulas which makes the beam of equal strength in 
tension and compression. This result agrees, then, with what is expected. Likewise it is 
found that with a larger amount of steel than is theoretically required, the yield point of steel 
is not reached and the beam fails directly by crushing of the concrete. Beams with no web 

reinforcement and with the existence of large shearing and moment stresses, 
cF fail by inclined cracks opening up in the concrete, thus substantiating to a 
considerable degree the theoretical deductions regarding the internal stresses in 

The residts of breaking tests on reinforced beams with different percent- 
ages of steel reinforcement compare well with the results derived from the- 
oretical formulas. Considering the nature of the material, the calculations by 
the two assumptions of stress variation are found to agree sufficiently with 
the experimental results to justify their use in problems of design. 

Another method of testing reinforced-concrete beams is by the use of ex- 
tensometers to measure distortions, so that the deformation of the steel and of 
the extreme fiber of the concrete may be calculated and the neutral axis determined. In 
making beam tests it is customaiy to place equal loads at points dividing the length into 
three equal parts. The advantage of this arrangement lies in the fact that the bending 
moment is practically uniform between the loads and, if measuring devices are attached, the 
deformations of the fibers at the top and at the bottom may be easily determined. If in Tig. 
14 the deformations be aa' and W, the neutral axis O is located by connecting 
a' and b' with a straight line, intersecting ab at 0. 

In moment calculations, the position of the neutral axis is of prime im- 
portance and once this is known, the actual strength may be determined with 
little uncertainty. The formula for k shows that the position of the neutral 

axis depends only upon the percentage of steel employed and upon the ratio 


•g or n. The value of Eg is the only value in the formula which is uncertain. 

It might be well to take the value as determined by the ordinary compression ^ ^ ^ ^ 
test for use in theoretical formulas, but closer results can be obtained from ^ ^Q* 15- 
these formulas if the value of n is taken so that for average conditions the neu- 
tral axis is in as nearly as possible the same position theoretically and experimentally. The 
reason that Ee, as thus determined, will give better results at working loads, is due to the effect 
of the remaining tension in the concrete below the neutral axis — a stress which is properly not 
allowed for in the resisting moment. 

In making the experiments above described, it was observed that the neutral axis raised 
as the loading increased, k being approximately j^ at working loads. It was also noted that 

for the lower loads the neutral asus as deter- 
mined from the theoretical formulas is more 
uncertain and generaUy lower in the beam 
than for the higher loads. This is undoubtedly 
- ■ - due to the relatively large influence of the ten- 

I I sile strength of the concrete in such cases. This 

Fio. 16. i^ ^^ ^^c neutral axis as the load increases is 

shown in Fig. 15. Consider aibi to be the plane 
ab after a bending takes place just sufficient to bring the maximum tensile stress in the concrete 
to its ultimate value. When loads are applied which cause a greater bending moment, the 
concrete in tension becomes broken by fine cracks, and the steel takes a greater part of the 
tensile stress. The elongation at b now increases faster than at a, and the neutral axis rises 

Sec 7-15] BEAMS AND SLABS 283 

rapidly. When working loads are reached, the position of the neutral axis moves but little, 
and the steel takes all the tension. 

Fig. 16 illustrates a typical diagonal tension failure in beams reinforced with only horizontal 
bars. The initial crack forms at a and branches toward b. A little later the concrete begins to 
fail in a horizontal tension crack just above the rods, running from a toward the end of the beam. 
This horizontal crack is brought about by the new conditions which exist after the concrete 
has become cracked along the diagonal line and the normal diagonal tension has thus ceased 
to act. Sometimes this horizontal crack does not extend to the end of the beam — ^the final 
failure occurring either by the diagonal crack extending to the top of the beam or the horizontal 
rods pulling out. Thus final failure often occurs from stresses which are developed after initial 
failure has occurred. However, the initial failure and its cause is what is of importance in 

Tests show that it is possible to provide sufficient web reinforcement by means of stirrups 
and bent bars to develop the full strength of the beam whether governed by the crushing strength 
of the concrete or the elastic limit of the steel. It is found that part of the diagonal tension is 
taken by the concrete so that web reinforcement need not be designed to take all the diagonal 
tensile stresses. 

Vertical stirrups spaced a distance apart equal to, or greater than, the depth of beam help 
but little in preventing diagonal cracks between successive stirrups. They may prevent final 
failure, however, by preventing the extension of a crack horizontally along the reinforcing rods. 
Stirrups are found by tests to be most effective when spaced a distance apart equal to one-third 
the depth of beam. To give the best results they 

should be securely fastened to the longitudinal bars I I 

in the tension side of the beam. i | | | • | | | lT"i~M~rT" 

Tests in which curved and inclined rods were I j u\-^ j p . jjN j J >J j j 

used, but in which no rods continued straight for Li>*gtrj^<c 4-L-i-/-fc. X-iXU-KAuIvL.. 

the entire length of the beam, showed results very t t 

little better than for straight rods. ^la, 17. 

Vertical stirrups and bent rods combined are 
found by tests to give the very best results. Tests also seem to indicate that too much re- 
liance should not be placed upon one or two bent rods. For this reason, even if one or two 
rods are bent up properly to take the diagonal tension, it would be good design to consider this 
rod, or rods, as not taking any diagonal tensile stress and to provide a thorough web reinforce- 
ment by means of stirrups. 

Tests show that bent bars may be inclined at any angle between 30 and 45 deg. without 
the beam showing any marked difference in strength. Beams having sharp bends in the rein- 
forcing bars are found to have less strength than beams with bars having circular bends of a 
radius about 12 diameters. 

Fig. 17 represents the conditions which developed in the test of a beam. The cracks are 
numbered in the order of their appearance, final failure occurring at crack No. 4 and being due 
to inadequate web reinforcement. The stirrups were stressed beyond their yield point. 

It appears from tests of beams in which bent rods were employed with a good anchorage at 
their ends, that the anchorage is quite advantageous in increasing web resistance. This form 
of construction is also found to be an insurance against failure at low loads through defective 
concrete or insufficient bond. 

Hooks at the ends of the horizontal tensile bars prevent slipping of the bars in the concrete 
and are found to increase the strength of the beam materially. 

The results of experiments show that the ultimate compressive strength of concrete in a 
beam is at least equal to its crushing strength as determined by tests on cubes hardened under 
similar conditions; also, that the yield point of the steel should be regarded as ultimate strength 
as far as reinforced beams are concerned. When the steel reaches its yield point, the beam de- 
flects, and failure soon occurs by the crushing of the concrete. 


16. Bond Stress. — The tension in the horizontal steel near the lower surface of a reinforced- 
concrete beam is a maximum near the center of beam and decreases each way toward the end. 
The difference in the tension between any two points is transmitted to the concrete by the bond 
between the steel and the concrete. 

A formula for bond may be derived for beams in which the reinforcement is horisontai 
or straight throughout. The total shearing stress per linear inch between the steel and the 
concrete, considering a length of beam equal to x, is 

T "T 


From Fig. 11 

Vx = (r - T)jd 


T -T y 

X jd 

(bond stress per linear inch) 

• F . . 
and the bond stress per square inch of the surface of the steel bars is -^ divided by the sum 

in inches of the circumference of the bars at the given vertical cross-section. If ii — unit 
bond stress, and So the total circumference of all bars in a beam at the given section, then 


The above formula shows that theoretically the bond stress is a simple function of the 
shear and varies with the shear. Thus, shear diagrams may be used to represent the variation 
of bond stress along a beam. When using the above formula, the average value oi j ^ "^ 
may be taken. 

If we consider simply supported beams, tests on rectangular and T-beams loaded at the 
quarter points show that when stirrups are used the beam is stiffened and the bond stress along 
the horizontal rods near the end of beam is somewhat reduced. A reason for this may be 
shown ip the fact that, after the concrete begins to crack from diagonal tension, the stirrups 
aid in carrying part of the tensile stress which results from the bending moment then existing 
at the line of the diagonal crack; the stress in the horizontal rods at the end of beam is thus 
reduced and likewise the liability of failure through bond. A greater reduction of the bond 
stress has been found to exist when the web reinforcement is provided by means of bent rods 
and stirrups. The reduction becomes considerable when about one-half of all the rods are bent 
up, provided, however, that a sufficient number of rods be thus employed. Results seem to 
indicate that no reduction should be considered in design unless the number of rods bent 
be greater than two or three and that the bends be made at least at two points at each end of 
beam. Tests show that for conditions especially favorable, an average of fiO% more bond stress 
may safely be allowed on the horizontal rods at the end of beam than would be considered 
safe by the above formula. No allowance should be made when only stirrups are employed 
for the web reinforcement. 

It has been f oiud in the testing of simply supported T-beams with steel straight throughout 
and loaded at the one-third points, that the bond stress along the horizontal steel is affected by 
the presence of tensile stresses in the concrete and that this bond stress is usually a maximum 
just outside the load points. The observed bond stress at these points was in some cases as 
much 80 50% greater than the computed stress. 

In beams reinforced for diagonal tension the bond stresses along the horizontal bars are 
not distributed as uniformly as in beams having the reinforcement horizontal or straight through- 
out. The bond stresses are found to be concentrated at and near stirrups and at and near 
points of bending of the longitudinal rods. Such concentration of bond stress causes local slip 
of the longitudinal bars unless the web reinforcement is well distributed along the beam. If a 
stirrup is not rigidly attached to the horizontal rods and local slip of ban occuns, the effective- 
ness of the stirrup is somewhat impaired. 

Sec. 7-17) 



The bond stress in continuous beams is treated in Art. 39. 

The bond strength of vertical (or inclined) stirrups may be insufficient to develop the re- 
quired strength of the stirrups with respect to tension. This possibility must also be investi- 
gated in the design of beams having web reinforcement in the form of bent rods. Tests show 
that it is safe to assume that the stress in a stirrup or bent-up bar may be transferred to the 
concrete above a point 0.6 the depth of beam from the upper surface. In most cases it is found 
that stirrups must have hooked ends. 

For illustrative problem, see page 298. 

17. Web Reinforcement In General. — Inclined web reinforcement may be separate mem- 
bers firmly connected with the horizontal reinforcement to prevent slipping, or some of the 
horizontal bars may be bent up near the ends of the beam where they are not needed to resist 
bending. The vertical reinforcement may be used separately or in combination with inclined 
reinf orcementy depending upon the preference of the designer and upon the amoimt of diagonal 
tension to be provided for. Vertical stirrups should be looped around the horizontal bars and 
in important beams should also be firmly secured to these bars by wiring or otherwise. Stir- 
rups should usually be looped or hooked at the top in order to prevent slipping due to insuffi- 
cient bond (see Art. 19). 

The proportioning of web reinforcement cannot be done with any degree of exactness since 
very little experimental work has been performed along this line. However, rough determina- 
tions of what is required may be obtained on rational grounds. The only information from 
tests is the value of the maximum shearing stress which measures diagonal tension failure — (1) 
for beams with horizontal bars only, and (2) for beams having an effective system of web rein- 
forcement. Also, tests on beams, with and without web reinforcement, show that when rein- 
forcement is provided for diagonal tension, the concrete may be assumed 
to carry its full value of the shear and the steel the remainder. It is 
generally conceded as safe practice in the designing of beams to use 
only two-thirds of the external vertical shear in making calculations of 
the stresses to be taken by the web reinforcement. 

Consider now Fig. 18, in which V represents the average total shear 
over the portion s of the beam. Let v' represent average unit horizon- 
tal shear on any plane below the neutral axis. Then (see Art. 13) 


The total shear over any such horizontal plane is i/hs] whence 

,. Vs 

vbS = -TT 


The function of stirrups, either vertical or inclined, is to resist by their tensile strength 
that portion of the above shearing stress which is not carried by the concrete. 

Assume a vertical stirrup to be placed at the section A-A, and to oppose the shear over the 
portion of the beam. The total stress in the stirrup is A*/, (in a U-shaped stirrup, A» is the sum 
of the areas of the two legs), and it is produced by that part of the total shear over the horizontal 
plane hs not taken by the concrete. Assuming the steel to take two-thirds of the total shear, 

2 Vs 

— ■ ■ 




Fia. 18. 

v' = 

A J. -g 












(vertical stirrups) 





.-|-^^ (3. 

For inclined members and bent-up bars, the lines on a beam representing the direction in 
which the diagonal tensile cracks are likely to occur, are crossed more times per unit of length 
for a given horizontal spacing than would be the case if vertical stirrups were employed; that is, 
a given amount of inclined steel is much more effective in taking diagonal tension than the same 
amount of vertical steel. It may be assumed that the allowable stress in the inclined bars is 

approximately . /Lo and the required area of steel, assuming the steel to take two-thirds of 

the shear, is 

_ 2 0.7(7.) 

^ 3 fjd ^^^ 


* 2 0.7F ^^^ 

Since tests show that bent bars may be inclined at any angle between 30 and 45 deg. without 
a beam showing any marked difference in strength (see Art. 15), the Joint Committee recom- 
mends that the longitudinal spacing of vertical stirrups should not exceed one-half (H) the depth 
of beam and that of inclined members and bent-up bars should not exceed three-fourths (^) 
of the depth of beam. 

18. Region Where No Web Reinforcement is Required. — ^There is a region near the center 
of most beams in which the shear does not exceed that permissible for plain concrete. In this 
region no shear reinforcement is required. The distance from one support to a point beyond 
which no stirrups are required may be found as follows for a uniformly loaded beam. 

I « span of beam in feet. 

w » uniform load in pounds per foot. 

Xi s distance in feet from left support beyond which no stirrups are required. 

Vi s unit working shear for plain concrete. 

Vi « total working shear, producing unit shear of Vi. 
From equation (2) on page 280, it is obvious that, where Vi » v, no stirrups are required. 
At this point 



"' bjd 



— WXi 

2 ■ 


or, in terms of v at the end of beam, 

Xi =» 
When V at the end of beam equals 3pi, then 

Xi « 

w -?) 


This derivation applies only to a static uniformly distributed load over the whole span. 

Suppose A lO-ft. beftm (b -• 10 in, and d i- 20 in.) is uniformly loaded with a static load of 2900 lb. per ft 
and anume ti "■ 40 lb. per aq. in. according to recommendation of Jmnt Committee for 2000-lb. concrete. Abo 

aaeumejd ■> 7ftd. Then 

10 (40)(10 )(17 5) 
Jfi - 2 ~ 2900 ■ 

When designing for floor systems it is often more proper to consider the uniform load tu 

DiAnitAU t 





Spacing of U-StiTups io irchjs 

S,„,„, „ U-B, 

Spacing of W-Strrrup» m InchM 

Spocl-^ of W-5.1 











S|ncing St U-Himip* In *chw $(»cl--3 of U-Shrrupi fn 

Spacing of W-Stirrup* *n inc*i«» Spockng a< W-Sfirrup» In 


I I 

I I 

(pacing at U'Stirivpi in Inch** Spacing cf U-Wirrupt In 

Spacing of W-Stirrvpt in lneh«» Spacing at W-Stlrru- 



Spoemg of W-Sftrrvp* In inchH Spacrti^ of W-St*"/* m 


i- 1 

Ivadl of W-SHttv*, tn .KtM S»w^ •* W-Stwrwf* >n 



I I 

I 1 

I I 

i I 

Svocfng of U-S*-'rFiip« In 
Spoclng of W-SHmrp. In 

i I 

Spacing sT W-Mimip* In InehH SpOCInf or W-MliTup* « IncH* 

K load. In this case the ehear at the center of the span ia not zero, but is equal at 
n to -^ ' where w' is the live load per foot. When plotting the ehear diagram this value 
should be plotted at the center of the beam and the total end shear at the aupport. The varia* 
lion in shear between theae two points may be safely aaaumcd to be a straight line. The shear 
to be carried by the concrete being known, the point in the beam beyond which no shear rcin- 
rorcement is required may be quickly located. 

19. Vertical Stimips. — The required total area of cross-section of a vertical stirrup may 
1>e determined by the formula (see page 285) 

. 2 V* 

^- 3 /jd 

assuming the web reinforcement to carry two-thirds of the tot«! shear. (For U-shaped stirrup, 

A, is the sum of the areas of the two legs.) With a given size of stirrup, this formula may be 

solved to give the spacing required, or 

3 A4.jd 


The value of V should be taken at the section where the spacing is desired. This spaciog 
formula may be solved directly by means of DiaEraroa 1 to VIII inclusive for three siiea of 

If the shear diagram is drawn for any given beam, it ia convenient to use the above fonnuin 
in the form 



and, for varioua even-inch epacings (<) of the atimipa, to solve for the corresponding totftl 
external shears. At the point where the ordinate to the shear diagram scales a computed total 
shonr, there the spacing may be made the even inch used for » in the formula. 

Tests have shown that little or no value is derived from stirrups spaced a distance apart 
equal to, or greater than, d (see page 283). A practical limit suggested by the Joint 0>m- 
mittee is one-half the depth. 

In restrained beams the firat stirrup should be placed no farther than one-half the mini- 
mum spacing from the edge of support and, in beams simply supported, the first stirrup should 
be placed not farther than one-half the minimum spacing from the center of support. 

The variation of shear intensity along a uniformly loaded beam ia shown in Fig. 19. Ute 
area A BCD represents the total stress to be taken 
by the stirrups at each end of beam. The ordi- 
nate AB represents two-thirds of the shear at the 
support per 1-in. length of beam. 

Some attention must be paid to the diameter 
of stirrup which it will be possible to employ in any 
given case. The diameter should not be so small 
that the stirrups will be placed too close together 
for convenience in construction, yet not so far 
apart that the limiting value i^id is exceeded. 
Fio. 19. But, in addition to such consideration, the bond 

strength of the stirrup must be inveot^aled if the 
stirrups are not made with hooked ends, since the danger of slipping determines the maxi- 
mum diameter which may be employed. 

The distribution of bond stresses devdoped on the surface of the stirrups is indeterminate. 
Evidently it must not be expected that tension will be transferred through bond to the con- 
crete until the compression area of the beam is reached, or until a point but little below is 
reached. Experiments show that it is safe to assume the grip of a stirrup to be O.fl the depth 
of beam. Using notation on page 270. 

/^. - 0.6 dot! 

But, for round or square stirrups, letting i - maximum diameter of stimip. 


If each end of the stirrup is bent into a prong or book, tben stimipa (ri larger diamel«r may b« 
used than is indicated by the above formula. Testa show that if hooka with a semicircular 

* Diaimna dmUar to Ihoae br PauiK S. ButLXW In In*. Ntm, Ovt. 11, lOIS. 


bend of 4 diameteiB are w^ embedded in coDcrete, tbe stress in the bar wiQ reach the ebstir 
limit befcHe slipping takes place (see page 268). 

It is oonadered good practice to use ^e-in- stirraps with hooked mds for beams from 10 
to 25 in. deep^ *^-in. s ti rr up s for beams from 25 to 40 in. de^>, and ^ilnn. stiiraps for beams from 
40 to 60 in. de^. 

iLumnuLTiTK PwLCT. — ^A wtmpis supparted beam is S bgr 16 ia. in ctom wctian and th« MMton r««n- 

k 2 in. above the lower face of tbe beam. Spaa of the beam b S.5 ft. Cniform load of 1S00 lb. pw ft. 

If nimifieiy, the wfh is to be renforoad acainirt «tiagonal tanioii oaiiiK vertwal atixTopa. Allowable /, i* 10,000; 

« — 40; « « 80. 

V 7650 

•"^" (9)G4KU) -^^^P^wJ-w 

The allowafale shear a 40 lb. per eq. in., bcsnce eticTupe are neoeaBary. 

The diamftw of a stiiiup without any prone or hook ahoold not exceed 

If the atiriup e are to be bent at the iq>per end, 9i«4n. round bars may be ooaeidered eecure acainet aUppint- 
Stirrups are unneoeasary at a distance from support equal to 

'■-¥(>- s) ->•«>" 

The minimum spaeinc of stirrups (U-«hape) will occur at the supports, or 

3 2 (0.077) (10,000)(Ti)(14) ,_. 
• - 2 7650 " ^^^ *" 

The shear diacrmm for one-half the beam is shown in Fig. 20. For a 44n. stirrup spadnc 

F - I '(0077)(10.000)(%)O4) _ ^^ ,^ 

The point where V « 7090 lb. is easily found by scalinc to the shear diagram. For a 5-in. stirrup spaoinc V ^ ^% 
(7000) - 5670 lb. For a 6-in. spscing V - H (7090) - 4730 lb.. Ptr. Time can he Mived by iisinc Dlaffram I 
in findinc values of V directly. 


xJ _^-5hear Dioaram 




Fza. 20. FiQ. 21. 

20. Method of Placing Stirmps from the Moment Diagram. — It is a well-known principle 
of mechanics that the difference in moment between any two points along a beam is equal to 
the average total shear over the distance between the points multiplied by that distance. Thus, 
from Fig. 21, 

Vs = (Ma - Mc) 

y^ iMA' Mc) ^j^ 

For loads concentrated at points along a beam this law is not strictly true, unless in each 
case the concentration occurs at a point midway between the transverse sections chosen; Init 
in the case of concentrated loadings, by beams cast against girders in concrete construction, 
aiul evai by loadings on slabs transmitted to the floor beams, the concentration may not bo 
sharply defined, and there is no determinate law of shear variation over such a region. More- 
over, as this discussion will show later, the distance t is relatively small wher(» shear ' 




Within tbc linuta of Actual conditioDa in reinfoiced-conciete coDstiuctioii, thefcfoic, thp 
above Btaumeat may be considered very appn>iiinBt« to the truth, for the beun loaded witb 
concentrated loada. 

Substituting the above value for (' in equation <1) on page 285, we have 
(.M4 - Mc) 3 AJ^ 
* "2 * » 

Mi = (Ma - Mc) = i.5AJ^ (2< 

in which Mt is the increment of moment between the ends of the portion of the beam to be re- 
inforced with the Btinup. 

If the Htirnip is inclined at an angle 9 with the horizontal, then 

Jf. -i 


When — 45 deg. (or, accurately enough, any angle between 30 and 45 deg.) 

M, = 2.lAJ^d 14 1 

Consider the portion of the beam shown in Fig. 22a, loaded in such a manner as to pro- 
duce the moment curve OF. It is desired to reinforce the portion shown with vertical stirrups, 
keeping in mind the principles just laid down. A certain stimip has been adopted which, for 
ttiis paiticular beam, gives the value of Mt from equation (2) equal to the vertical distance 

shown in Fig. 22a. The first increment intercepts the portion On of the curve OP; the second, 
mn, and so on let one of these intercepts, as mn, be projected on the beam, thereby defining 
the area ABDC on the diagram of the beam. This area is the portion of the beam over which 
the adapted stirrup will exactly carry the shear. The length of the portion is seen to vary as 
the shear varies along the beam. Since the stirrup is required to carry the shear fur this 
portion of the beam, it should be placed through the center of gravity of the shear area for this 
portion. Likewise, each other portion of the beam defined by the projection of M, would 
have a stirrup through the center of gravity of its shear area. 

To eliminate the feature of having to locate this center of gravity, the following method is 
proposed: Lay off as the firat value H Mi (Fig. 226). Let all other spaces be equal to JIf , as 
before. 'Hieae increments have tit', n', etc., for points of intersection on the moment curve. 
Let theae pobts be projected on the beam. E^h projection will thus determine the position 
of a stirrup. This scheme gives very closely the same results as before, the stirrup being 
placed slightly nearer the support than when placed through the center of gravity of the shear 
area for the portion of the beam; this error, however, is well inside the accuracy of placing this 
part of the reinforcing. 

Diagrams IX to Xll inclusive are plotted for a ready solution of equations (2) and (4) 
l^ven above. Their uae is explained in an illustrative problem on page 295. 



■puvd-qsui ^a tfnatnem u| diuj|4* jq^ lutuausut (iiiioiu 6uipuaQ 

uJMq JO ijiMp •Aip^M] 




A timple meUiod of constructing the p&rftbola or moment diagram for unifoim loading is 
aa follows: I^et AO, Fig. 23, repreeeot the base of the parabola, with a middle ordinate of 4.5 
kt D. Divide the baae into any desired number of equal parts, as for instance, six. Number 
these pointa from each end beginning with eero. Divide the middle ordinate by the product 

" 0.5. This constant, if multiplied by the product of 
iple, the ordinate at 

of the numbers at that point, bs . ' „ 

the pair numbeis at any point givee the ordinate at that point. For 
C i» 2 X 4 X 0.5 - 4.0. If an ordinate is de- 
sired at a point between the equal divisions, as 
X, the fractional part of the division may be 
expressed for the point from each way. At X 
the distance from A is 2.6 units, and from 0, 
3.4 uniU. The ordinate at X is 2.6 X 3.4 X 
O.S •• 4.42. If the middle ordinate does not 
fall at an even division, as would be the case if 

an odd number of units were used, the fractional values for the mid-point would be used the 
same as the full values in the above case. 

Iu.D>T**Trra PamLaic. — DaMrmliM tba *!» wid Bpccins of vertlu] itirrups for iheiir niaforoemant in a 
beuB lowlad u ihan in Fi(. 24a. 

Pot Uw lUtio unilorm lowl of 1000 lb. per [t., the moment at the center Is 

■(_• (i00Q)(ai>'(i2) 

- 062.000 in,-lb. 

■ (1000) (7) (12) - 8*0,000 in.-ib. 

J third of the beam. The moment curt 

. the 



d to form 

the curve < 

The totel end ehear i* (1000) 1211 y, + 10,000 - 
20.S001b. Juat outaide the third point the shoir is20,SOO 
- (1000) (7) - 13.500 tb. Juit inside the third point it is 
13.S00 - 10,000 - 3900 lb. At the center the ghev is kto. 
The shear diagram ia shown in Fig. 24c. 

In Beiure the foUowinc valuea wiU be adopted: /, - 
10.000, /, - S50; B - 15. It is found that for equal 
strength in tension and compreauon a depth (d) of 28 in. is 
required when b - 20 in. Also; - 0.8TS. 

Tbe allowable shear to be carried by the concrete 
slone will be taken a* 30 lb. per mi- in. The eoncrele will 
then carry (30)(2D){2e) - 15.000 lb. TLia value i> plotted 
aa point a on the shear diacrsm. Fig. 24c; and the dittance 
from the support to point a ia the distance along the beam 
where shesr reinforcement ia required. 


vable ■ 

lOXNM lb. 

to scleet a aiia of atirrup that will not be eo email that 
the apadnc will be too amall for convenient construction, 
nor yet large enough to cause the apacdng to be greater 
than )M. a limit recommended by the Joint Committee. 
To aeeompUah thia. let ■ vertical Una be projected from a. 
Fig. 24c, to the moment curve, aa at b. 1 


Fio. 34. 

vertical projaetioi 

thia case 200.000 in.-lb.) . With this 
IK. enter Disgram IX at the right and 
ive to tbe left until tbe value of ; - O.S75ia reached; then 
ital Una from d - 26 in. Thia point ia found to give a 
p at«« equal to 0.58 •«. in. 

Four ^i-ln. round r«di (In a oomUnad ara* of O.OOt sq. in., and by sntetiog Diagram IX at d - 36 in.. 
~ " sq. In., thenee vertloally to / • 0.B7S and OnaUr to the right. Jtf - 205,000 in.- 



[Sec 7-21 

Basmmiic m in Fig. 226, ^(205.000) is Iftid off at the base of th« moment curve, m dc, Fig. 24b, aiter 
which equal increments of 205,000 are laid off until the point b is passed. Through the points thus fixed, hori- 
sontal lines are drawn to intersect the moment curve. From each intersection a vertical line projected to the 
beam locates a stirrup. 

21. Bent Bars and Vertical Stirrups for Web Reinforcement. — The ends of the horisontal 
bars in a reinforced-concrete beam may usually be bent up to assist vertical stirrups in pre- 
venting diagonal tension failure. Although in some cases these bars may be found theoretically 
to take all the diagonal tensile stresses not taken by the concrete, vertical stirrups are alwa^'s 
desirable, as shown by tests. 

Plain rods bent up to provide web reinforcement often lack sufficient bond strength to 
render them fully effective. Where bent up at a considerable angle they should be turned again 
horizontally and extend some distance along the upper part of the beam, as shown in Fig. 25. 
In heavy construction the ends of all bars should be bent into a hook. The most convenient 
method of using reinforcement is to bend up two rods at a time and make all the bars inclined 
at the same angle with the horizontal. The bars bent should theoretically be such as to keep 
the center of gravity of the beam cross-section in the line drawn vertically through the center 

of the section. An exception occurs to the bending of 
two rods at a time, in the case of an odd number of 
horizontal rods. Here, one of the bends may consist 
of either one or three rods. 

If bent rods are not required to provide for di- 
agonal tension, then the horizontal rods may be dis- 
pensed with at the points, beyond which they are not 

.^ eBewjs-. j needed to provide for tension due to bending. This 

i V J i7"^.>. - I method of stopping off the horizontal rods is not de- 

sirable, however, as the bond in the concrete near the 

gsi X ^ ^ ^\K 1^ middle of the beam is not as good as would be the 

I - ~» — —.»—.—— =..r«^:-^^^^ ^^^g^ ^g^P ^YiQ end where the moments are smaller. 

T l^-v-s- -H Also, when a bar is discontinued, the stress in those 

PiQ, 25. which remain is immediately increased tending still 

further to impair the bond between the steel and the 
concrete. This is true whether or not a hook is employed on the discontinued rods. With 
bent up rods a better condition exists. The horizontal components of the upturned bars act 
with the bars unbent in taking the tension due to bending, and so in general the tension in the 
horizontal rods decreases somewhat more gradually toward the end of beam as it should. 

At the bend in a horizontal rod, the unit stress in the concrete may become excessive if the 
bend is too abrupt. Tests indicate that the strength of beams with bars having sharp bends is 
less than for beams with bars having a radius of bend equal to about 12 diameters. 

The bent rods, if of the same diameter, should be so arranged that each rod will take an 
equal part of the diagonal tension — ^that is, if they can be bent in this way and still provide 
satisfactorily for the horizontal tension. The points where the bent rods should cross the 
neutral axis of the beam may be found by any of the methods given for determining the spacing 
of vertical stirrups, remembering that a rod bent at 45 deg. (or, as tests show, at any angle be- 
tween 30 and 45 deg.)^ is ^ » 1.43 times as effective per unit area as a vertical stirrup. If 

the rods cannot be bent as desired, then vertical stirrups must be used to provide for some of the 
diagonal tension, and it would be preferable in all cases to use stirrups even in that part of the 
beam where the bars are bent up. 

Consider uniform loading and let A BCD in Fig. 25 be the diagonal tension area. Assume 
that four rods may be bent near the end of beam and assume also that the first two bars cannot 
be bent up nearer the center of beam than the point /T, Fig. 25. The area cdef should be made 

> See p. 286. 

Sec. 7-221 BEAMS AND SLABS 297 

equal to 1.43 times the allowable tensile stress in the two No. 2 bars; likewise area abed should 
represent 1.43 times the allowable tensile stress in the No. 1 bars. (The vertical rt should theo- 
retically pass through the center of gravity of the area and, if d^ is made ^2 suid tej^2 0fd€, this 
¥rill be practically accomplished. The error is not serious U dtia made equal to te.) If the No. 
1 bars cannot be bent up so that the areas abed and cdef have the line cd in common, then a ver- 
tical stirrup, or stirrups, will be needed to take care of the space between these areas . If the areas 
overlap, the No. 1 bars may be bent up nearer the end of beam if no other condition governs. 
In any case the distance a should not be greater than ^d — ^that is, the maximum spacing at 
which inclined web reinforcement can be considered effective. Stirrups will at least be needed 
to the left of ab and to the right of ef. The area taken care of by a vertical stirrup is equal to 
its tensile value. 

The Joint Committee recommends that bent bars be considered as adding to diagonal 
tension resistance for a horizontal distance from the point of bending equal to ^d. The Joint 
Committee also recommends for the case where stirrups are needed in combination with bent- 
up bars, that the stresses in the stirrups be determined by finding the amount of the total shear 
which may be allowed by reason of the bent-up bars, and substracting this shear from the total 
external vertical shear. Two-thirds of the remainder is recommended as the shear to be con- 
sidered as carried by the stirrups. 

The bond strength of inclined bars must be investigated. This strength should be pro- 
vided in the upper portion of the beam. As with vertical stirrups, it is arbitrarily assumed that 
no stress is transmitted from the steel to the concrete below a point which is 0.6d below the upper 
surface of the beam. 

Assume that the stress in an inclined bar is its working stress. This gives the maximum 
condition. Using the notation of page 270 and V for length 

Vou = AJm 
and for round or square bars, 

lu =-r 


V — /- diameters 

If the allowable /« — 10,000 lb. per sq. in. and the allowable u = 80 lb. per sq. in., then V 
(as shown at No. 2 bars in Fig. 25) should equal 31 diameters by the above formula. If V = 
16,000 lb. per sq. in., Z' = 60 diameters. 

The value of hooks on the ends of bars is discussed on page 268. 

See illustrative problem on page 298. 

22. Points Where Horizontal Reinforcement May be Bent. — For uniformly loaded beams 
the bending-moment curve is a parabola as shown in Fig. 26. 

Let a — area of bars required at the center of beam, 
as » area of bars to be bent. 

Pt = 100 — = % of total steel that may be bent up. 

M = malum moment '^' 

xt s distance from support to point where bars may be bent. 
Mg^ » moment at dbtance xt from support. 


M a 

3f«j a — at 
Substituting values of M and M,^, and solving for xs, 



[Sec. 7-22 

Fig. 26,' based on the above formula, indicates the points at which rods may be bent up for 
three types of beams with the maximum bending moments specified. To illustrate the use of 

the diagram, assume that a beam designed for 3f « -jj^ requires 3.5 sq. in. of steel at the 

center. To find the point where 40% of the steel may be bent up and still leave suflicient steel 

to carry the tension, trace horizontally from the 40% mark at the right to the curve M « — 

and then vertically to the lower margin where 0.22/ is read. 







' I 00 




























• • 





























^ OA 

















u in 



. .. 

















50 •£ 






0.1 0.E 0.3 

Location of section in terms of span 




If it is desired to bend up a number of rods two or more at a time, then z\ should be deter- 
mined for each bend. After this is done, the remaining horijsontal bars should be secure 
against slipping. 

For concentrated and unsjrmmetrical loading, the maximum moments at various sections 
will need to be determined, in order to ascertain the points where the horijsontal bars may 

J be bent up. From these maximum moments obtain 

the required area of horizontal rods at ^e different 
points (1, 2, 3, and 4, Fig. 27). Plot a curve to scale, 
as shown. Thus, ah represents the area required at the 
point a. On the center ordinate lay off the required 
areas of the rods, and draw horizontals as shown. The 
rods may be bent up where these horizontals cut the 
curve but it would be better, however, to cany them a short distance beyond the theoretical 

Illustkatitb PKOBLmM. — Daaicn the left end of a simply sapporied beftm to spaa 15 ft. and to •opport tbo 
loede shown in Fig. 28 with eqiud strength in tension and eompression. Allowable /c ■• 660: « ■> 80: * * 40 
withoat web reinforeement and 130 with an effeetire web rainfoteeinaBt. 

> Diagram taken from an article in Eng. Ifevt, Aug. 19, 1018, by Kabl D. ScawaaDKinEB. 

Fia. 27. 

Sm. T-221 BSAMS and SLABS 

TIm nutioB* *n renMly loiuul uid mn i 
load don tha itu»t pamm Uirouib tb* *>li» h 
tba onUotm IomI <d 1000 lb. par ft. 

M - (2S,50a)(S)(lZ) - (33,000)(4)(13) - l^«.000 In.- 
Il ■■ fauBd that ■ dcptb (d) of 28 in. ii nquind whao ft - 16 In. 

It ooouii at tha ec 

wd aB)(!t)us) 

- (lfl)(2S)(0.OOTD - S.tSaq. In. 

TDimd rod) - 3.S3 aq. in. Bond for 


Thua. ■( thii and of beam lottr rod* mi 

Tba eanonta iril] ba foand to Uke 

wilt taka <40)(1S) - MO lb. per Un. in. 

can of any diaconal Mnaion betn 

an the « 

no.nti.ted load* 


diavxul tanaioE 

) at the 


-cS ■•"«'"'■ >*"'■"" 

atad load, it I* 

S-"""-- '" 

Tba total diacon^ tamioa to ba takan by tba wab n- 
ioffvetmeat ia npnaeoted by a trapeioid (Fif . 2B) . 
tba paraUel (idea of obkh at* » (1040) - fWO lb. and 
M (SSOj - S90 lb. and tba lcn(th 4 ft. Heooa, toUl 
atraat Id tbia part of baam to ba takan by tba eon- 
oeta aad by the wab 
fl90+Mq ^ 

< 12 

30.720 lb. 
be bent, tbeir compara- 


(4)(0.M18)<1S,000)(1.43) - 40.400 lb. 


whether the taoaile ■ 

BOW ba made to aae 
the bottom of the beam 
It bars. lU. It abowa 
the beDdios-moment curve plotted to icalei and the 
pointa wheta the baja an^y ba bant up aje datarminad 
by the method deacribad on pa^e 208. !t ia clear 
that tba ban oannot be bent up aa deaind to provide 
thorouchly for diafonal teoaion. The point* wbare 
tba bata ai« aetoally bent up an about 2 in. beyond 
tba theoretieai pointa aa detenninad by moment. 

U eaah bant-up bar ia aaaumed to take diaci>- 
nal lanaion to the left only of ita point of beDtUnf. 
tbaDthearaaKAA^eiamaiDauDpTDvidad lor. BUrmpa 
win ba provided to taka diaconal tanaion betwaan 
tbe point t, where the line be produoed maata the 
iwutnl line, and tba adjacent load. Only ooaatlmip 
Ui Uad to alao plaoa atinupa at tba podtiooa lodloatad 
inad baiB abould ba M diaueten. or 3TH in. 


28. Transverse Spacing of Reinforcement. — ^The shearing stress along ab. Fig. 30, should 
equal the amount of stress transmitted by bond along bed. If bond and shearing strengths 

were equal, ab should equal bed, and the clear space between bars should be -^—n — diameters 

» 1.57 diameters. But shearing strength here employed is controlled by the diagonal tension. 
For » 90 lb. per sq. in. and u » 80, a6 should be at least 1.40 diameters; and for v » 120 lb. 
per sq. in. and u = 80, a6 should be at least 1.05 diameters. There is likely to be more or less 
tension in the concrete surrounding the bars, and, besides, since the concrete is liot easily placed 
between the rods, it may have a lower strength in that vicinity. A clear spacing of IH to 
2 diameters is advisable unless it is determined by computation that the bond stress is very 
much lower than the allowable. In the above discussion plain bars only have been considered. 
Deformed bars, if stressed to their full bond value, should be spaced farther apart than plain 

The Joint Committee recommends that the lateral spacing of parallel bars should not be less 

than 3 diameters, center to center, and that the distance from the side of the beam to the center 

of the nearest bar should not be less than 2 diameters. In order that concreto 

may be readily placed between bars and also give sufficient concrete on the sides of 

the beam for fire protection, it is also advisable to require that the spacing of rods 

•'•4k-# ^ ^^^ ^^^ than 1 in. in the clear (if the maximum size of aggregate does not ex- 

^^ ■ ceed 1 in.) and that 1}^ in. in the clear be also considered the minimum distance 

Fio. 30. of the rods from the sides of the beam. Thus, the least width of beam should be 

the greater of the two values determined from the following formulas: 

6 = [3(n - 1) + 41di 
h = at{n - 1) + ndi + 3 

in which h « least width of beam in inches, 
di » thickness of the rods in inches. 

n s maximum number of rods which occurs in a horizontal layer. 
Qg — maximum size of aggregate in inches. 

For an aggregate with a maximum size of 1 in., the width of beam for all rods greater than 
^^ in. in diameter will ordinarily be governed by the first formula and for ^-in. rods and less, 
by the second formula. 

Where two or more layers of rods are used, the rods should be so placed as to permit the 
mortar to run between them. The Joint Committee specifies a limiting clear space of 1 in. and 
does not recommend the use of more than two layers '^ unless the layers are tied together by 
adequate metal connections, particularly at and near points where bars are bent up or bent 
down." The Joint Committee also advises that ''where more than one layer is used, at least 
all bars above the lower layer should be bent up and anchored beyond the edge of the support." 

84. Depth of Concrete Below Rods. — ^Tests show that a 2-in. thickness of concrete is 
necessary to thoroughly protect embedded steel from the direct action of flames. Flat slabs 
are found to be affected to a less depth than projecting members such as beams and 
columns. The Joint Committee suggests that "the metal in girders and columns be protected 
by a minimum of 2 in. of concrete; that the metal in beams be protected by a minimum of 1 y^ 
in. of concrete; and that the metal in floor slabs be protected by a minimum of 1 in. of 

The following depths of concrete below the center of steel may ordinarily be employed 
except where conditions are unusually severe. 


D<ilHh to steel (rf) ^^'^'f^JSl"*" 

3Ji in. and under ?i in. 

Between 3^ in. and 4?^ in 1 in. 

4>i in. and over 1 ^ in. 

S«c 7-25] BEAMS AXD SLABS 301 

Bbaxs akd Gibders 

Depth to steel (rf) '^iiiw rtid***' 

10 in. and under 1 in. 

Between 10 in. and 20 in IH >i^* 

20 in. and over 2 in. 

26. Ratio of Length to Depth of Beam for Equal Streni^ in Moment and Shear. — With 
given working stresses in concrete and steel, there is a definite ratio of length to depth of beam 
which will give equal strength in moment and shear. First, consider beams simply supported. 
For a single concentrated load at the center of span 

d vi 

in which vi » allowable shearing stress and /« » working stress in steel. For a uniformly- 
distributed load 

d vi 
For beams loaded with equal loads at the third points, 

J ^^ 

d vi 

Taking for example, vi = iO lb. per sq. in., /. - 16,000, fe « 050, n « 15, and, using an 

average value of J4 forj, we have the following ratios for -.• 

For concentrated load at center of span -j -^ 6. 16 

For uniformly distributed load -i -> 12.32 

For equal loads at the third points . ■> 0.24 

It should be clear that the strength of beams of greater relative length than obtained by the 
formulas will be determined by their moment of resistance, while that of shorter beams by their 
shearing resistance. 

In the case of continuous beams the above formulas will apply if I is taken as the length 
between points of inflection. It is often convenient to know the extreme limit in design. The 
Joint Committee recommends 120 lb. per sq. in. for the shearing strength of concrete when 
adequately reinforced against diagonal tension. This is a low figure but is adopted in order to 
prevent any likelihood of cracks opening up in the concrete. Suppose then, it is required to 

know the minimum value of -j for a given beam, uniformly loaded. From the formula for 

uniform load, using the working stresses given above, 

I (4) (16 ,000) (0.0077) 

d ^ 120" " "^ 

At the same time that the ratio of length to depth is being investigated for moment and 
shear, there are other conditions which must be considered. For instance, the ratio of length to 
breadth of beam should not exceed a value of about 25 if the beam is not supported laterally. 
The reason for this is found in the fact that the upper part of the beam u a column, and to 
prevent additional stress due to side bending the length should not exceed about 25 times the 
width. On the other hand, the best-shaped beam is one in which b lies between }4d and !^d. 
In any given case, to satisfy all requirements and arrive at a satisfactory design^ two or three 
trials may be required. 


86. Economical Proportions of Rectangular Beams. — Without taking the cost of web 
reinforcement into consideration, it can be shown mathematically that the cost of a rectangular 
reinforced-concrete beam to resist a given bending moment and be of equal strength in tension 
and compression, varies inversely with the depth, directly with the square root of the breadth, 
and directly with the cube root of the ratio of breadth to depth. 

The breadth and depth of a rectangular beam to be of equal strength in tension and com- 
pression may be found by means of the formulas (see page 277) 

M..?f hd'.^ A.-VM 

If b or d, or the ratio -\ is decided upon, the proportions and steel area of a beam to resist a 

given bending moment are definitely determined. Thus for a fixed breadth, fixed depth, or 
fixed ratio of breadth to depth, the cost of a beam will vary with the working stresses employed 
since values of k, j, and p depend wholly on values of /e, /■, and n. 
Where the depth is fixed, it is found that, if the ratio 

cost of steel per unit volume 
cost of concrete per unit volume 

does not exceed a value of 60 to 80 (60 a common value), no economy results from using/, greater 
than 16,000 lb. per sq. in. when /« = 600 to 700 lb. per sq. in., or from using /• greater than 
12,000 lb. per sq. in. when /« — 400 to 500 lb. per sq. in. Somewhat higher values than these 
may be economically used for /« when the breadth of beam is fixed. In both cases cost de- 
creases as /e increases. When the ratio ^ is fixed, no economy results from using /« greater 

than 16,000 to 18,000 when /« » 400 to 600, or from using /« greater than 14,000 when 
/« » 700. In this case cost increases as /« increases. 

In the case where the crossHsection of beam is determined by shear, the maximum depth 
theoretically permissible is that for which hd is just large enough to carry the shear. With a 
beam designed for moment alone, the cost decreases as the depth increases, but the area of the 
cross-section becomes less. A point must be reached when the beam will be of just the required 
strength in moment and shear (see Art. 25). The question which now arises is whether or not 
a still greater depth will result in greater economy. The quantity hd must now remain constant 
for the greater depths. But hd\ on the other hand, is increased and the concrete stress (/«) 
decreased. A smaller value for/e permits the use of a smaller percentage of steel, and the cost 
is still further reduced. Thus it should be clear that the proportions of a beam will not be 
determined by shear excepting as to minimum cross-section — an increase in depth always result- 
ing in a gain in economy. It should be noted in this connection, however, that although deep 
beams are economical of concrete, the wooden forms cost more than they do for shallow beams. 

27. Rectangular Beams with Steel in Top and Bottom. — Compressive stresses are usually 
carried by concrete more economically than by steel. It is sometimes desirable, however, to 
place steel in the compression as well as in the tension side of the beam. When a rectangular 
beam is limited as to size, double reinforcement is sometimes the result, and in such cases the 
value of the steel reinforcement on the compressive side needs to be known. The effectiveness 
of steel in compression has sometimes been questioned, but the results of tests indicate that 
the steel does its share of the work. 

The Joint Committee recommends that ''the reinforcing bars for compression in beams 
should be straight and should be 2 diameters in the clear from the surface of the concrete. For 
the positive bending moment, such reinforcement should not exceed 1% of the area of the 




The formulas which are used in the design of double-reinforced rectangular beams are 
derived by means of the same fundamental principles as for beams with single reinforcement. 
In deriving the following formulas the compression in the concrete is assumed to follow the 
linear law and the tension in it b neglected; the formulas then apply to working conditions only. 






— ^ 

Fio. 31. 

Let A* » crossHsectional area of compressive reinforcement (Fig. 31). 

d' » distance from the compressive face of the beam to the center of the compressive 

p' » ratio of cross-section of steel in compression to cross-section of beam above the 


tensile steel — 


/'• « compressive unit stress in steel. 

k = ^2n ( P + P' ^') + w*(p + pV - n(p -h p') 

k = 


1 + 




k* H-2p'n( A; - ^') 

/. = 

AJd pjbd* 

t - f'^ 
-'* nil - k) 

k - 



1 -fc 

. _ Md - k) 
J. - - -^ 

M. - MV.pi 

d « 







The cases which may be met with in practice, with the method of solution in each instance 
indicated, are as follows: 

1. To determine fiber stresses. 

Compute p, p', and j* 

Solve for k from formula (1) and j from formula (2). 
Substitute value of j in formula (3) and value of k in formulas (4) and (5). 
Solve directly for fiber stresses. 
2. To determine moment of resistance. 

Compute p, p', and -r- 


Solve for A; and j. 

Substitute value of A; in formula (6) and find value of /« when the maximum allowable 

value of /e is substituted. 
Solve formula (7) for 3f « using either the value of /« determined from formula (6) or 

the allowable value of /., whichever is the lesser. 

3. To determine d for a given h and given values of A,, —„ /«, and /«. 

(Trial method. Best shown by use of diagrams. See illustrative problem, page 347.) 

4. To determine p and p', or only p' (see Art. 27o). 

Formulas for shear, bond, and web reinforcement are the same for double-reinforced beams 
as for beams with tensile steel only. When using formulas for shear and bond stress along 
horizontal tension rods of beams double-reinforced, an average value of j = 0.85 may be taken. 

27a. Formulas for Determining Percentages of Steel in Double-reinforced 
Rectangular Beams. ^ — The formulas given below are based on the fundamental fact that for 
any given values of /« and /., k has exactly the same value regardless of the shape or type of 
beam. This single value for all beams is expressed by the formula 

1 + -•'; 


It follows from this that if steel is added to the section without changing the extreme fiber 
stresses, this added tensional and compressive steel must form a balanced couple whose stresses 
conform to the stresses already in the section. 

Let pi » steel ratio for the beam without compressive steel. 
Ps » steel ratio for the added tensional steel. 

P = Pi + P». 

p' = steel ratio for compressive steel. 
Ml » moment of the beam without compressive steel. 
Aft — moment of the added steel couple. 


1 + 4 

P.«g (2. 

Mx «/.pi(l -5)W« (3; 

Aft ^ M - Afi (4; 


Pt = / ;r/T (5 » 


p = Pi -h Pj («) 

P' - P« ^ " J, (7) 

^ d 

(See page 348 for illustrative problem.) 

88. Deflection of Rectangular Beams. — Fig. 32 gives the Keneral form of 4i deflection dia- 
gram for a reinforced-concrcte beam. The portion AB shows the deflection beC9i)i ^^® ^^^' 
rrete has begun to fail in tension near the center of the beam, BC shows the defle^^^aqp during 

< From ihMUi by Il<MBirr 9. Bbako submitted to crmduate school of rnivcndly of Rjfrrgi^'% {laitDU fulfiU- 
ment of the requirements for the Master *a Degree. 

Sec. 7-28a] 



a second or readjusting stage, and CD the deflection with the steel near the center of beam 
carrying practically all the tension. 

88a. Maney's Method. ' — The deflection of a reinf orced-concrete beam of whatever 
shape may be determined by the formula 

D == c^ (ec + e,) 


D — maximum deflection (if desired in inches, the units specified be- 
low should be used). 
{ — span (inches). 
d » depth of the beam to the center of the steel (inches). 

«e = unit deformation in extreme fiber for the concrete = is-- 

f, = unit deformation in extreme fiber for the steel = 





Fia. 32. 

c = - in which 

Ci =» the numerical coefficient in the formula for deflection of homogeneous beams, 

D =^ Ci ~pT* depending on the loading and on how the ends are supported. 

Ct » the numerical coefficient in the formula for bending moment, M = Ctwl*, 

for a simple beam loaded at center, c » H2 or 0.0833 

uniformly loaded, c = ^g or 0.1041 
loaded at the third points, c = ^%i6 or 0.1065 
for a beam with fixed ends, loaded at center, c = J^4 or 0.0416 

uniformly loaded, c = H2 or 0.0313 
loaded at the third points, c = ^44 or 0.0347 

886. Tumeaure and Maurer's Method.' — Turneaure and Maurer recommend that 
8 to 10 be used for n in the formulas which they have derived, and which are given below. They 
also state that the formulas presented are the result of modifying the deflection formulas for 
homogeneous beams in accordance with the following assumptions: 

1. The representative or mean section has a depth equal to the distance from the top of 
the beam to the center of the steel. 

2. It sustains tension as well as compression, both following the linear law. 

3. The proper mean modulus of elasticity of the concrete equals the average or secant 
modulus up to the working compressive stress. 

4. The allowance for steel in computing the moment of inertia of the mean section should 
be based on the amount of steel in the mid-sections, since stirrups and bent-up rods do not 
affect stiffness materially for working loads. 

The following are the deflection formulas for rectangular reinforced concrete beams: 

n = <^» Y^' " 
E\ ' bd^ ' a 

a = H[k^ -h (1 - ky + 3np(l - k^] 

_ 1 + 2np 
2 + 2np 



From equations (2) and (3), the value of a for any values of p and n may be computed, and then 
the deflection from equation (1). The notation employed in the above formulas is as follows: 

* See paper by G. A. Manbt, presented before the seventeenth annual meeting of the American Society for 
Testing Materials. 

* "Principles of Reinforced Concrete Construction" 2d Edition, p. 116. 


D » maximum deflection (if desired in inches, the units specified below should be used). 

h » breadth of the beam (inches). 

d » depth of the beam to the center of the steel (inches). 
W *- total load (poimds). 

/ «s span (inches). 

p s steel ratio. 
Eg s modulus of elasticity of the reinforcing steel (pounds per square inch). 

n — ratio of the moduli of elasticity of steel and conc^te. 

a » a numerical coefficient depending on p and n. 

X; — proportionate depth of the neutral axis. 

Ci » the numerical coefficient in the formula for deflection of homogeneous beams, 

Cx -wj^ depending on the loading and support. For example, 

for a cantilever loaded at the end, c 

for a cantilever uniformly loaded, c 

for a simple beam loaded at center, c 

for a simple beam uniformly loaded, c 

for a beam with fixed ends, load at the center, c 

for a beam with fixed ends, uniformly loaded, c 

= >i84 

The following are the deflection formulas for reinforced concrete T-beams (referred to later) : 

E/ hd* ' fi 

,b' b'/i\ , t 

in which ^ is a coefficient depending upon the steel ratio and n, and other symbols as before. 

29. Slabs. — ^A reinforced-concrete slab should be figured in the same manner as a rectangu- 
lar beam, the bending moment being usually computed for a width of slab equal to 1 ft. The 
ratio of steel in a slab is most readily found by dividing the cross-section of one bar by the 
area between the centers of two adjacent bars, this area being the spacing of the bara multi- 
plied by the depth of steel below the top of slab. 

Slab bars should not be placed too far apart to properly take stress directly nor yet should 
they be spaced so close that the concrete cannot be properly placed between them. The main 
tensile reinforcement should not be spaced farther apart than 2^ times the thickness of the slab. 
The minimum limit should be about the same as in beams. 

Shearing faUures are not usually important in slabs, but in special cases of heavy loading 
the same care should be used as in the design of large beams. > 

89a. Moments in Continiioiis Slabs. — For uniformly loaded spans, fuDy 
continuous over two or more intermediate supports, a moment of Hi^^^ i^^^y ^ ^^^^ hoih in the 
centers of all spans and over all supports, for both dead and live loads. For slabs continuous 
for two spans only, with ends restrained, the bending moment both at the center support and 

near the middle of span should be taken as—' For very unequal spans or spans of imusual 

length, the moments should be computed more accurately. 

896. Provision for Negative Moment in Continiunui Slabs. — Slabs having 
spans of any appreciable length should be reinforced against negative moment. This may be 
done by bending up a part of the rods in the spans on each side of a support and extoiding each 

> See p. 344 for illuiirstiTe probleme otiiig diesrains. For flat eleb floors, eee eliapter in Beet. 1 1 . 

Sm. 7-20cr BEAMS AND SLABS 307 

aet of bent rods along the top of beam into the adjoining span. The bend in the bars should 
be near the }i points in the span, and usually at an angle of 30 deg. with the hoiicontal. Too 
sharp an angle may tend to crack the slab. 

When placing slab reinforcement in long spans, a top reinforcement at least to the third 
point will be desirable. In ordinary spans the steel should at least be lapped a sufficient dis- 
tance over supports to provide adequate bond strength, and the steel should be bent up far 
enough from the support to provide properly for negative moment. 

29c. Floor Slabs Supported Along Four Sides. — When a floor panel is square, 
or nearly so, the slab may advantageously be reinforced in both directions. Exact analysis 
of stresses in such a case is impossible, but some important facts have been brought out by ap- 
proximate solutions for uniform loading. The theory applied in such an analysis depends upon 
the fact that the load at any point on the slab is distributed to the two systems of reinforcing 
bars at that point, in proportion to the stifiFness of the beam elements lying in those directions. 

The following recommendations are from the report of the Joint Committee : 

Floor aUbs having the supports extending alone the four sides should be designed and reinforoed as con- 
tinaous over the supx>orts. If the length of the slab exceeds IH times its width, the entire load should be carried by 
tranaveiBe reinforcement. 

F<v uniformly distributed loads on square slabs, one-half the live and dead load may be used in the cal- 
culations of moment to be reeisted in each direction. For oblong slabs, the length of which is not greater than 1^ 
times their width, the moment to be resisted by the transverse reinforcement may be found by using a prop<»tion of 
the live and dead load equal to that given by the formula 

r - - - 0.5 


where I * length and b ■■ breadth of slab. The longitudinal reinforcement should then be proportioned to carry 
the remainder of the load. 

In placing reinforcement in such slabs account may well be taken of the fact that the bending moment is 
greater near the center of the slab than near the edges. For this purpose two-thirds of the previously calculated 
moments may be assumed as carried by the center half of the slab and one-third by the outside quarters. 

29d. Cross-reinforcement in Slabs. — When the length of a floor panel is large 
compared to its breadth, the longitudinal reinforcement (that is, reinforcement parallel with 
the length) is of little value in carrying loads, but a small amount is nevertheless generally 
desirable in preventing shrinkage and temperature cracks and in binding the entire structure 
together. It is more important for wide beam spacing than when the beams are closely spaced. 
The amount of steel to use is usually selected somewhat arbitrarily, and }^-in. or ?^-in. rods 
spaced 18 to 24 in. apart is common practice. The top of the slab over a girder should be re- 
inforced transversely not only for stiffening the girder, but also to provide for the negative 
bending moment produced with the bending of the slab at right angles to the direction of the 
principal slab steel. 


80. T-Beams in Floor Construction. — When a slab and beam (or girder) are built at the same 
time and thoroughly tied together by means of stirrups, bent-up rods, and cross-slab reinforce- 
ment, a part of the slab may be considered to act with the upper part of the beam in compression. 
This form of beam is called a T-beam, and the extra amount of concrete in the compressive 
part of such a beam makes possible a considerable saving over the rectangular form. The 
thickness of the flange is fixed by the thickness of slab required to support its load, but the 
width of slab which can be taken as effective flange width must be selected somewhat arbitrarily. 

81. Tests of T-beams. — ^T-beams are found to fail under essentially the same condi- 
tions aa rectangular beams, and the same general principles apply. Tests show that the maxi- 
mum load carried can be materially increased by placing cross bars in the top of slab and 
by adding fillets between the flange and the beam. Cross bars are found to be actually needed 
to insure T-beam action but fillets are not required in ordinary cases. 



[Sec. 7-32 

T-beams with projections of flange on each side of web of 10.5 times the thickness of the 
slab have been found to carry a load only 5% larger than beams with projections of 6.8 times 
the thickness of the slab. No appreciable difference was found in the latter beams between the 
deformations at the edge of flange and the deformations in the flange at the web. 

32. Flange Width. — The Joint Committee has recommended the following rules for deter- 
mining flange width : 

(a) It shall not exceed one-fourth of the span length of the beam. * 

(fr) Its overhanging width on either side of the web shall not exceed 6 times the thickness of the slab. 

Beams in which the T-form b used only for the purpose of providing additional compression area of concrete 

should preferably have a width of flange not more than 3 times the width of the stem and a thickness of flange 

not less than one-third of the depth of the beam. 

83. Bonding of Web and Flange. — The web and flange of a T-beam can be considered well 
tied together when slab reinforcement crosses the beam and when the web reinforcement extends 
well up into the slab. The bonding should be especially well looked after near the end of beam, 
and this is generally accomplished by means of the bent rods and stirrups brought up as high 
as possible, in addition to the slab reinforcement (as mentioned) acting at right angles to the 

length of the beam. Along the center of the 
beam the differential stresses between the 
beam and slab are not large, but it is better 
to. insert vertical stirrups extending up into 
the slab at occasional intervals since shrinkage 
of the concrete is apt to part the slab from the 
beam if there is not some means to hold the 
two together mechanically. The thinner the 
sections, the more thorough should be the 

34. Flexure Formulas. — With a T-beam 
it is necessary to distinguish two cases; namely, (1) the neutral axis in the flange, and (2) 
the neutral axis in the web. 

Case I. The Neutral Axis in the Flange, — All formulas for "moment calculations" of 

rectangular beams apply to this case. It should be remembered, however, that h of the for- 

A» A, 
mulas denotes flange width, not web width, and p (the steel ratio) is r^* not ^rn (Fig. 33). 

Case 11, The Neutral Axis in the Web, — The amount of compression in the web (aaaOj 
Fig. 33) is commonly small compared with that in the flange, and is generally neglected. The 
formulas to use, assuming a straight-line variation of stress and neglecting the compression in 
the web, are: 


Croas sectton 

Stress Diogram 

Fia. 33. 

k = 


kd = 

k « 

2n<i A, -f bt* 
2nA, + 26t 

pn + 

pn -h 
Skd -2t 


2kd - 
d -z 










Sec. 7-34a] BEAM^ AND SLABS 309 

f^^^JL (7) 

^' = ;^) . '^'^ 

/. = Mi_^) (9) 

Af. = /.^ jd (10) 

Approximate formtilas can also be obtained. From the stress diagram, Fig. 33, it is clear 
that the arm of the resisting couple is never as small as d — yit, and that the average unit 
compressive stress is never as small as K/e» except when the neutral axis is at the -top of the 
web. Using these limiting values as approximations for the true ones, 

Mc = ]4}M{d - MO (a) 

M. = AUd - MO, or A. = y^)(^^ ^,) (6) 

The errors involved in these approximations are on the side of safety. 

Formulas which take into account the compression in the stem are recommended where the 
flange is small compared to the stem. Such formulas may be found in the report of the Joint 
Committee, and are as follows: 

^ ^ nndA. -K6 ~ h')t^ ^ / nA, -K6 - b Qty _ nA^-^- (6 - b')t 

z = 


(kdt* - Ht*)h + [(W - 0* a + H<^kd - t)W 

t(2kd - t)b -h (fed - tyh' 
jd ^ d — z 

^* AJd 

- _ 2Mkd 

[(2kd - t)bt + (kd - iyh'\jd 

84a. Case n Formulas for Determining Dimensions and Steel Ratio for 
Given Workliig Stresses.^ — The following formulas are sometimes useful in the design of T- 
beams when the neutral axis is in the web : 

d ^ 2R, ^^^^ 

in which 



(Formula (12) should be solved by exact methods as the slide rule does not give satisfactory 

If the depth of beam is fixed by the headroom available, formula (13) gives directly the 
proper percentage of steel for given working stresses. 

> From thesis by Robbbt S. Bbabd submitted to graduate school of University of Kansas in partial fulfillment 
of the requirementa for the Master's Degree. 




86. Designing for Shear. — Since a T-beam will usually have ample strength in compression 
for any ordinary depth of beam likely to be selected, the design of the stem of the T, or the beam 
below the slab, is largely a question of providing sufficient concrete to take care of the shearing 
stresses and to give a good layout of the tension rods. The manner of providing reinforce- 
ment for shearing stresses in T-beams is similar to that described for rectangular beams. In 
T-beams, however, the reinforcement for shear should run well up into the slab in order to tie 
the beam and slab together. The shearing strength of a T-beam is about the same as that of a 
rectangular beam of the same depth and a width equal to the width of the stem of the T. 

86. General Proportions of T-beams. — T-beams should not be made too deep in propor- 
tion to the width of stem as such forms are relatively weak at the junction of stem and flange. 
The width should preferably be from one-third to one-half the depth in ordinary cases. For 
large beams the width may be made from one-third to one-fourth the depth. 

All re-entrant angles in concrete are points of weakness and such angles should, therefore, 
be avoided. 

87. Economical Considerations. — When a floor slab forms the flange of a T-beam, it Is 
possible to determine economical proportions for the stem. 

Ck>nsider a portion of a rectangular beam one unit in length. 

Let c 


cost of concrete per unit volume. 

ratio of cost of steel to cost of concrete per unit volume. 

cost of beam per unit length. 

depth of beam below slab. 


« e\h'd* 



Md' + Ht) 


using the approximate formula (6) on page 309. 

When d' is fixed by the headroom available, the cost will be a minimum when V is made as 
small as possible, and its value will then be determined by the shearing stress or by the space 
required for the rods. The expression also shows that the cost will decrease with increased 
values of /«, and that with a fixed value of h'd' the cost decreases with increase in depth. If the 
value of b' is assumed as fixed, then there is a definite value of d which will give minimum cost. 
The following expression has been deduced from the preceding equation and will give the value 
of d for minimum cost when the value of b' is fixed: 



From this expression the best depths for various assumed widths may readily be determined and 
the desirable proportions finally selected. 

The following table is convenient in determining values of r: 

88. Conditions Met widi in De- 
sign of T-beams. — ^In practice the de- 
sign of T-beams will take one of the 
following forms with method to be 
followed in each case indicated: 

1. To find moment of resistance 
or fiber stresses. 
The values of k and j may be 
found from equations (3) and 
(6), or from equations (2), (4) 
and (5), on page 308, and then 
the values of the fiber strcsBcs 
from equations (7) and (8), or 
the moment of resistance from 

Cost of 


cts. per lb. 

Coat of conerete. doUara per cubic yard 

















• • 










equations (0) and (10). When the moment of resistance depends upon the concrete, 

equation (9) is useful in determining the value of /« to use in equation (10). To obtain 

this value, the maximum allowable value of /« should be inserted in equation (9). (If 

the value of A; is found to be less than j, then the problem falls under Case I and the 

formulas for rectangular beams apply.) 

2. To design a T-beam in which the flange forms a portion of a floor slab. 

Depth and width of stem of beam should be selected with reference to shearing strength, 
space for necessary rods, and other considerations. The depth having been selected, 
the amount of steel may be approximately determined by equation (h). The amoimt 
of steel being known, the value of j may be determined by equation (6). The value 
of k should also be foimd from equation (2) or (3) in order to ascertain if the beam 
falls under Case I or Case II. The stress in the concrete, corresponding to the allow- 
able working stress in the steel, is then found from equation (8). 

3. To find the minimum depth for a single-reinforced T-beam. 

The value of ^ or d, may be found from equation (12) for given working stresses, and 

the amount of steel from equation (13). The width of stem should then be selected 
with reference to shearing strength, etc. 

4. To design a T-beam not connected with a floor system. 

First method: 

First, select suitable proportions for the web. A flange thickness is then assumed 

such as to give satisfactory proportions between i and d. The value of ^ is then 

known and k and j can be determined from equations (1), (4), and (5). The 
area of steel and the breadth of flange are then found from equations (10) and 
(11) respectively. 
Second method: 

For any assumed thickness and width of flange, the depth of beam ia&y be de- 
termined by equation (12) and the percentage of steel from equation (13). The 
width of stem should then be selected with reference to shearing strength, etc. 
(When making approximate computations for shear or bond stress along the horizontal 
tension rods, an average value of j » ^ may be assumed, as for rectangular beams.) 

89. Design of a Continuous T-beam at the Supports. — Negative bending moment exists 
at the supports of continuous beams and tensile steel must be placed in the top of beams over 
supports to prevent cracks opening up at these points. For the usual case of equal spans and 
indefinite live load, the common method of providing for this negative moment is by bending 
up one-half of the rods on each side and extending each set over the supports into the adjoin- 
ing span. The remaining lower horizontal rods in each span are carried horizontally through 
the supporting columns. 

In a design of continuous T-beams at the supports it should be noted that the flange is 
under tension, that the stress in the concrete is negligible above the neutral axis and that 
a rectangular section may be considered at such points. The method of design is thus similar 
to the design of a double-reinforced rectangular beam at the center of span with the exception 
that the compressive and tensile stresses about the neutral axis are inverted (see page 302). 
Since a T-beam in the center of span becomes a rectangular beam over supports, the stress 
in the tensile steel at the support will generally be greater in ordinary designing than the corre- 
sponding stress at the center of beam; that is, this stress will be greater if half the rods are bent 
up on each side and lap over the support. For this reason, then, when selecting the steel at 
the center of span, a little more than the required amount at that point should be chosen. It 
should be noticed, however, that the column has some strengthening action at the support 
and it will not be necessary to keep too closely to the allowable stress. 



[Sec. 7-32 

T-beams with projections of flange on each side of web of 10.5 times the thickness of the 
slab have been found to carry a load only 5% larger than beams with projections of 6.8 times 
the thickness of the slab. No appreciable difference was found in the latter beams between the 
deformations at the edge of flange and the deformations in the flange at the web. 

82. Flange Width. — The Joint Committee has recommended the following rules for deter- 
mining flange width: 

(a) It shall not exceed one-fourth of the span length of the beam. * 

(6) Its overhanging width on either side of the web shall not exceed 6 times the thickness of the slab. 

Beams in which the T-form is used only for the purpose of providing additional compression area of concrete 

should preferably have a width of flange not more than 3 times the width of the stem and a thickness of flange 

not less than one-third of the depth of the beam. 

33. Bonding of Web and Flange. — The web and flange of a T-beam can be considered well 
tied together when slab reinforcement crosses the beam and when the web reinforcement extends 
well up into the slab. The bonding should be especially well looked after near the end of beam, 
and this is generally accomplished by means of the bent rods and stirrups brought up as high 
as possible, in addition to the slab reinforcement (as mentioned) acting at right angles to the 

length of the beam. Along the center of the 
beam the differential stresses between the 
beam and slab are not large, but it is better 
to. insert vertical stirrups extending up into 
the slab at occasional intervals since shrinkage 
of the concrete is apt to part the slab from the 
beam if there is not some means to hold the 
two together mechanically. The thinner the 
sections, the more thorough should be the 

34. Flexure Formulas. — With a T-beam 
it is necessary to distinguish two cases; namely, (1) the neutral axis in the flange, and (2) 
the neutral axis in the web. 

Case I. The NetUral Axis in the Flange. — All formulas for "moment calculations" of 

rectangular beams apply to this case. It should be remembered, however, that b of the for- 

A, A, 
mulas denotes flange width, not web width, and p (the steel ratio) is r^* not r-A (Fig. 33). 

Case II. The Neutral Axis in the Web. — The amount of compression in the web {aaaa, 
Fig. 33) is commonly small compared with that in the flange, and is generally neglected. The 
formulas to use, assuming a straight-line variation of stress and neglecting the compression in 
the web, are: 

Cross section 

Stress Diogrom 

Fia. 33. 

k = 


1 -f 


k = 


2ndA, + bt^ 
2nA, -h 2bt 

z — 

pn + 
3kd - 2t 


2kd - 
jd = d — z 



J = 









6 -3:i 



flange is f*"'*^ -' j "•• lat ?"-;;:_ ^^ i. ■ -m, e — 

3ttf _ ' ir ^ 


beams when tbe 


XL r~:^-i. 


V * 

r» «: - - / - • 

p-(-) .-;'■- : ' : ^^^^ 

(Fonnula (12) should be srfved by exa'i iiie'ii'-.d- \xt -u* hii'u* ruW d'>*' nut give satisfactory 

If the depth of beam k fixed by tiit iieadr-ooiL a\a lafi.-. formula (IS) gives directly the 
proper percentage of sted for given -» orking rtr^Kset?. 

' From thens by Robkbt S- Bkaxd submitted to graduate bcLool of t mversity ^ •illroent 

of the reqairementa for the ICaater's Decree. 


A higher compressive stress may be allowed in the concrete at the supports than at the 
middle of span, because of the fact that the negative moment decreases very rapidly and only 
a short section is under maximum stress. Also, a slight excess of stress at this point does not 
in any way endanger the structure but merely increases somewhat the positive moment on the 

There are three methods of reducing the compressive stress in the concrete at the bottom 
of the beam over supports: (1) by increasing the amount of compressive steel in the bottom 
of the beam; (2) by increasing the area of compressive concrete, which may or may not require a 
flat haunch depending upon the width of the support ; and (3) by increasing the areas of both 
steel and concrete. 

The bond stress along the horizontal rods at the top of a continuous beam over supports 
may be found by the same formula as is employed for the tension rods at the end of a simply 
supported beam. However, if bent up rods are employed for web reinforcement and if these 
same rods are employed to take the tension over supports, the beam is greatly stiffened and the 
bond stress along the top rods is undoubtedly reduced appreciably below that given by the 
theoretical formula. This bond stress is affected by the amount of web reinforcement used in a 
somewhat similar manner to the way the bond stress is affected along the rods at the end of a 
simply supported beam (see page 284). 

In beams considered uniformly loaded, the rods which are bent should extend beyond the 
center of support to about the fourth point, or in beams of very definite live load to the third 
point (point of zero moment varies for different loadings), to provide thoroughly for negative 
moment, and this length should be increased if it is not sufficient to transfer to the concrete 
through bond, the greatest allowable tensile stress in the rods. 

If half of the rods from each span are used over the support, then half of the total number 
will extend to about the fourth point where the tension due to negative moment becomes zero. 
At this point the shear is only one>half of what it is at the end of span, if the beam is considered 
uniformly loaded. Since bond stress due to increment (or decrement) tension varies as shear, 
a sufficient number of rods are thus run out to the fourth point, and with the bent rods being 
added gradually to this number until all the rods are acting in this manner at the support, it 
is clear that this method of design is satisfactory even when the bond stress at the support 
is the maximum allowable. 

Rods should be bent in a position to take as much diagonal tension as possible, usually 
at an angle between 30 and 45 deg., and the points where the rods are no longer needed at the 
bottom of beam to resist tension may be found as explained on page 297. It is also necessar)' 
to determine where the rods over supports may be bent down. It will be on the safe side, 
and sufficiently accurate, to consider the curve for negative moment aa a straight line between 
the support and the point of zero moment at the fourth point. (For a very definite live load, 
zero moment should be assumed at the third point.) With this variation of the moment in 
mind, it is an easy matter to find where the rods may be bent down at the top of the beam. 
The designer must use his judgment in the matter, but this much may be said: if a bend 
cannot be made in a rod, as proposed, due to the controlling points- for bending at the top 
and bottom, a greater number of rods may be employed at the center of span in order to 
make this bending possible, and the design governed accordingly. It is evident from the 
above that it will not always be possible to place the rods so as to take all the diagonal ten- 
sion, in which case both stirrups and bent rods must be used. 

Another point to be noted in the design of a continuous beam at the supports is the bond 
stress of the compressive reinforcement. It can be shown that the bond stress per square inch 
for the tension and compression rods will be proportional to the product of the diameters by 
their distances from the neutral axis. Since the compressive steel will generally be nearer the 
neutral axis than the tensile steel, it follows that, if the compression bars are no larger in di- 
ameter than the tension bars, the bond stress per square inch will always be less than that of 
the tension bars. It is sufficient to consider simply the compressive stress in the steel and 

Sec. 7-40] BEAMS AND SLABS 313 

provide a sufficient length from this point to the end of the bar to transmit this stress. The 
working strength of the steel in compression cannot be reached without exceeding the com- 
pressive strength of the concrete in which it is embedded. CJonsequently, in common design 
it will be satisfactory to provide a lap beyond the support sufficient to take care of compress- 
ive stress in the steel equal to the maximum as determined by the concrete. 

40. T-beams with Steel in Top and Bottom. — The following formulas correspond to those 

for rectangular beams given on page 303. ^ — 3* 

*= Vi/_^ (1) 

p -fp' + 

A(2fc - A) - ^* (2fc - 2A) + 2p'n (k - j) (l - j) 

A(2fc - A) + 2p 



/. = ^, (4) 

/. - ■^>(i^ii> (6) 

M, = bd^f.pj (7) 

T-beuns.' A == i 

40a. Formulas for Detenniniag Percentages of Steel in Double-reinforced 


1 + 4 (^> 


J ^ 

6 - 6A +2A»-|- A^L-^ 



6 -3A 

Ml » /.?>,iW« (4) 

Mt^M-Mi (5) 


P2 7 :n^ (6) 



p = Pi + Pj (7) 

1 -fc 

P' = P» 

41. Deflection of T-beams. — Formulas are given in Art. 28. 

1 From thesis by Robert S. Beard submitted to graduate school of University of Kansas in partial fulfillment 
of the requirements for the Master's Degree. See p. 304 for notation, etc. ' 




42* Wedc»-itep64 Buiinf ' The aiudyM ol wedee-ah^Md 
the rtati g n of eoonterf orts and buttreaKs, and in the deaigii of cantifeTcr beama for oi^erliaii^iig 
aidewalka or roadway* on deck bridges. Formulaa follow for the general caae riiown m Fig. 34 : 

. ^ ppw eoa^t p%»coa*ft pn coe A 1 





, / coa«^. \ 
A. ^\ COB fit) fji / coa' fe \ 

/« W. "^ 7 
1^. - M/Jy (W«) coe» ^^ or W« 


iV« " Rf J (2Nf *) coa A, or btf* 

M;(coa« 0.) 

I or/e 


Pf J(COB fit) 


Bft \ 


, or/. 

Jb;(6rf«)(coa« fl.) 


Fso. 34. 

t \coa* fi,l n(l - t) 

When the compreasion aide of the beam is horixontal, coa fit becomea equal to unity. 

Likewise, when the tension side of the beam is 
horizontal, cos fit * 1. With both top and bot- 
tom faces of the beam horisontal, the above 
formulas reduce to those for a rectangular beam 
l... given in Art. 9. The above formulas ahould not 
*Xs be applied to beams having fi^ greater than 45 
deg. on account of approximations in the theory 
which depart more and more from accuracy aa 
fie increases. 

Let V\ represent the total ahear to be taken 
by concrete and web reinforcement, and let V 
represent total shear on section computed as for a rectangular beam. Then 

F| - F - ~ (tan ^. + tan A) 

fit and fit are to be taken as positive when they bear the same relation to the direction of V aa 
shown in Fig. 34, and are to be taken as negative when they bear the reverse relation. The 
formulas for shear, bond, stirrup spacing, etc. in rectangular beams apply to wedge-ahaped 
beams if F in formulas is replaced by Vi. 

48. Beama of Any Complex or Irregular Section. 

48a. Analytical Method*' — ^Long and cumbersome formulas for the analysia 
of complex sections can be avoided by a simple application of general methods of solution to 
the homogeneous transformed section. Such a general solution is here presented, based upon 
the standard notation and using the homogeneous section easily obtained by multiplying the 
steel areas by n, the ratio of the moduli of elasticity of steel and concrete. Thus an equivalent 
concrete section ia obtained, to which the ordinary principles of analysis for homogeneous 
beams are applied. An equivalent steel section could aa well be used if desired. 

Adopting the usual fundamental assumptions for the analysis of reinforoed^^oncrete 

> 8m Apptndia t of ** Earth PraHora, RaUdninc Walb and Bios" by 
"Brlda» Kngin— rini" by J. A. L. Wabdbix. 

* M«thod M ciTcn by J. R. Cimsl in Bmg. Am., M«r«b M, 1017. 

WtUAMU Caw. 8m alio tqL I of 

Sec 7-43a| 



beanie, based upon the atrughtrline theory, it can easily be Bhomi that an equivalent homo- 
geneous concrete aection will be obtained by multiplying the steel area by n. The neutral 
axis can then be located at the centroid, or center of gravity, of the transformed section. This 
is done moet conveniently by equating the statical moment of the equivalent area on one side 
to the statical moment of that on the other side. 

Knowing the position of the neutral axis, the moment of inertia of the transformed section 
with respect to this axis may be calculated, and the well-known fundamental formula M —— 
used to find the resisting momenta for a given section, or the fiber stresses produced by a given 
bending moment. 

Tit. 3S liva tha dimenuoiu ind noUtlon uMd (or the enut aDalyug of ■ rdnloToed-coDcntfl T-beun, in- 

nwtinc momrnt for this beam whrn A, - 4 iq. in., n - IS,/, - IS.OOOuid/, - eSO. the cDinpiwaive area U 
~ la with nipect to the 

illnded. lor eonvenieDoe. into the parte jhown in the G(ure. Equatini the 


(32 X 4)Ck - 2)+ i^- 60(24 - i,) 


ti - 7.31 and d - « - 19.69 

nl.-lSXiX (lB,6B)t . 


/.- (44X7.31.XH)- (32 X 3.31- X >^) 


- ^M2 

■ith 11 - IS, 

• r««i*tlni m 
the reetan^ 
.1. - 2.1S (q 

ia Ihua I,410,a00~-Elvtn b^ the iteel. 

Lin ehown in Fix. 30 lubjectAl to ■ pwitire 

idA' - l.eaaq.iD.tocamputathefibnatra 

The tranelormed ana al oomproaive ateel - IS X 1.69 - 
The tnulormed area or tenaile atml - IS X 2.26 - 

» (iMiiic tranalormed aeelinn) : 

-f- + 26.4(w - 2) - 33.8U2 - ») 


V - T.58 10. and <l - H - 14.42 in. 

TiMn the mome 

t o( lurtU will be: 

rd; - IS X 1.69 X <S.5S)' - 7S9 

nt. - 15 X 3.25 X (I4.42)> - 7018 

J, - 13 X (7,S8)' X >S - 1742 

I -I. + nI. + ni; - 9G49 



IS«. 7-436 

48&. Graphical M«tliod.^— Application of graphical metboda to the locatioa 
of the neutral axis and the determination of effective depth and resiating moment of reinforced- 
concrete beama is here proposed for cases of complex sections where the steel is distributeil 
as in beams reinforced for compression or where several layers of rods are used. It is assumed 
that graphical methods are familiar to the reader, and these methods will be illuBtrated both 
for investigating a given section and for designing a section to resist a given bending moment. 

TuFind Mi Ktatral Axim. 

UH»I matbod of findinc the a 
■lica. takin imrnU in order to ■ 
order to obtaia a bopuganeou* 

B probLem of fiodiDa the oeutnil ftiu atid reutina monu 
»d OD the comprtHioa aide, m in Flf. 37a. Applyiai i 
trf, the comprtodoa side of the beam is divided inlo tl 
[>r^ and euh glice u repreeented by ■ loree A,. A%, rti. 
reaor tbsstmllB multiplied by n - IS, The f om diaarmm 1 
■ rorcee" it then dnwD, F^. 37b, be^oaing irith nA, at a 


dof t! 

itiD( tha nmltBDt of tbcas fotM* ia 
■ectioD O of the Uue KO, wltb tbe lir» 
of ttua 


nultiplyiDC by tl 
nit'UVx 15. o 


ol the ilicea soiuidered. and 

that the ■tee) etraa in tend 

10.000/11 to loeata the itrtai Une PS. and 

ton* polyaoa, Ka. STd, in order to locate tb 

Knowini tha macnitud* of tha com pi 

dooely by the uaual formula*. 

in oancrele i* negleoted. 

Raiuint VoiiKnt.— Now la Fii, 37* the point P U nn 
lat found. RS a drawn to eautrnisBt icale to nptseDI thr 
a tbe concrete, 7(XI lb. pet aq. in., and the aliaiaht Hoc PS 

:lie valuaa at the tt^ting l<iT«a wrre obtained, alter notinc 
., whjab 11 leaa than tha allonrabJa valu*. abowinc that (be 
•iccnJ the allawable Talue. tbe line UV abould ba mad* 
irould lonm. The compnaBve (orcea ai* plotind in ih* 
It by the polyson, Fit. 37/ at X. 

tm, by lummation. to ba OTJOO lb. tbe rfleetliv depth ia 
97.200 X 17.3 - l,IS3,e00in.-lb. Thia baa bead ehackad 

nby V 

t. WoLn in Kna. Rtr. March 3 


oDoiplM MoUon*. M Fie- 38a, ou b« a 

le u ihown by Fi(i. 38b and 3: 

A-I39XI7S J-.ff -SM", 

A,-i99 X ITS xae-^jo" 

value lor the rwrnpceodon iteel ratio p*. The itreH line F7S ig Iben drawn in Fis. aoil by makinc R 
.hs allawsbln «onipre«ive >tre» in (he concrete (650) and UV equal to the aJlovabla strea in the steel i 
18,000) divided by n (IS), or 106«. The Intereection P then locata the neutral axis. 

Tba eomproinon aids la divided into a number of glices aa shown, and the areaa, Ai, . 
■ion Ble«l rsquirad, tba value o( nA, ia obtained from (ha polygon, Tig. 3M. and the foro 


thedoaacHaaYO. mA. is tiMaanfed off (31) and divided by » (15). givins 2.06 aq. In. for the area 
Tbe sted ratio p is then f rily eompvited. The reastias moment is obtained as before, mine the diacrams. Figs. 
aOd, a^c and 39/. bjr whieh a rfaiitlng moment of 902,000 in.-lb. is obtained, as ehovA. 

Now, if the vidth is increased and the atcel ratios are kepi eonstaat. the rtaiBtini m<Hnent will increase 
directly as the increase in width* Therefore, if the sssnmcd width be moltiplied by the ratio of the required moment 
to the mom«it obtained for this width, the reqoired width is found. This eomputation, giyen on the discrmm. 
shows that a 20-tn. width is neeeassry. The required areas of steel in eompresBon and tension ean then be com- 
poted ss giren, knowing the Taloes of the steel ratios p and p^. 

MtmetU of Inertia of CompUx Boom SoeHotu. — ¥i^ 40a shows the cross-section of a dooble-rcinforced 
concrete beam, the moment of inertia of which is desired. The compcession side of the beam is divided into anoall 
shoes, the srea of the steel beinc mohipfied by 15, and the aress of these slices are laid off in the foroe polysoD, 
Fig. 406. tocether with ail«, or 15 times the area of the tension steel. For conTenienee the pole p is taken 
with the pole distance B equal to 100 sq. in. to the scale at which the areas were laid out in the foroe polycon. 
From Rg. 406 the fnnicnlar polsnon F^ 40c ie drawn locating the neutral axis by the intersection O. It will 
be noted that all the string in this fonicular polygon are esrtended until they intersect the neutral axis, the axis 
about which / is desired, at points 1. 2. 3, etc. Now for conTenienee the jtole j/ is taken so that the pole distance 
H* equals 10 in. to the scale at whieh the sectum of the beam was drawn, and j/ is connected with points 1, 2, 3» 
etc. Now, paralld to these rays the eorrespondinc strinfs in the funicular polygon. Fig. 40d, are drawn, thus 
following Culmann's approximate method for finding the moment of inertia graphically. From this eonstnictioa 
weget/-£rxJ7'xF-100XlOX 9.25 - 9250 in.«. from which /« and /« can easily be obtained for any 
giren bending momeat by using the f ormulss 


44. Span Length for Beams and Slabs. — ^The Joint Committee recommends the following: 

The span length for beams and slabs simply supported should be taken as the distance from center to center 
of supports, but need not be taken to exceed the dear span plus the depth of beam or slab. For continuous or 
restrained beams built monolithlcally into supports, the span length may be taken ss the clear distance between 
the faces of supports. Brackets should not be considered as reducing the dear span in the sense here intended, 
except that when brackets which make an angle of 45 d^. or more with the axis of a restrained beam are built 
monolithieally with the beam, the span may be measured from the section where the combined depth of beam 
and bracket is at least one-third more than the depth of the beam. Maximum negatiye moments are to be oon- 
sidered ss eslstang at the end of the span ss here defined. . 

When the depth of a restrained beam is greater at its ends than at mid-span and the slope of the bottom 
of the beam at its ends makes an an^e of not more than 15 deg. with the direction of the axis of the beam at 
I, the span length may be measured from face to face of supports. 

46. Reconunendatkms of Jomt Committee as to Positive and Negative Moments. — In 
computing the positive and negative moments in beams and slabs continuous over several 
supports, due to uniformly distributed loads, the Joint Committee recommends the following 

(a) For floor slabs, the bending moments at center and at support should be taken at rj for both dead and 

live loads, where w repreeents the load per linear unit and I the span length. 

(6) For beams, the bending moment at eenter and at support for interior spans should be taken at r^ and 


for end spans it should be taken at r^ for center and interior support, for both dead and live loads. 

(c) In tho ease of beams and slabs eontinuous for two spans only, with their ends restrained, the bending 


moment both at the central support and near the middle of the span should be taken as j^ . 

(d) At the ends of continuous beams, the amount of negatiTe moment which wiO be devdoped in the beam 

will depend on the condition of restraint or fired nw, and this will depend on the form of constmotion us ed. la 

the ordinary ca ses a moment of ^g bimj be taken; for small hnsmn ronmng into heavy eolumns this should b« 


inereased, but not to exeeed rr- . 

Foi spans of unusual length, or for spans of materially unequal length, more exact calculations should b« 
made. Special consideration &■ also required in the case of concentrated loads. 

Evsn if the eenter of the span is designed for a grester bending moment than is called for by (o) or (6), the 
negative m 'ment at the support should not be taken as less than the values there given. 

4A. Theorem of Three If omenta. — By means of the theorem of three tnomenU, the momeots 

Sec. 7-46] 



at the supports of a continuous beam may be deduced. The theorem, mgiiming level supports, 
is in its general form as follows (Fig. 40 Ay: 

Mith + 2Af, \hh + Wi) + M, hit = - P« V/. iJh-k,*) - Pt Wi (3fc,-2*,«+fc,») (1) 

1 DvriTfttion of the tktortir. of Ihrte momenU ia aa follows: 

Let the ori^ of codrdinatee be B (Fic- •^)t with x meMured positively toward the left. Conaideriiig only 
the dflfomuttion due to the bending moment, and negleetins the deformation due to shearing forces, the equation 
of the elastic eimre is viven by 

^ H 

dx* ■ SJ 
in whieh M is the bending moment at any point z, y. If this expression is integrated once, there results an ex- 

;on of ^, the tHope of the tangent to the elastic curve at any point x. y. Thus the slope of the tangent at 



- ♦ 


f* Mdx 

dx Jb EI 

and f<ir the whole member the ehangein slope of the tangent becomes 

*B Mdx 





Fia. A, 

la Fig. B is shown a beam resting on several supports, all on 
the same leveL Let us consider the portion of the beam between A 
Mxtd B, by eatting it out dose to the support by the planes m and n. 

Tlw part of the beam cut out is shown in Fig. C. Each end has the same shear and moment acting upon it as it 
did in its original pontion, thus causing stresses throughout the portion AB identical with thoae acting before the 
out. The moment at any point L, to the left of the load P, is 





Ml - M,+ rvt" 




Fia. B. 






-jr* — 

Pig. C. 

The moment 

By taking 

at any point R to the right of the load is 

Afie - ifs + Ftf" - P(*" - U) 
moments first about A, and then about B, the values of Vt and V» are found to be 

ITt - Jlfa + PU 

V, - 

V, - 


ir« - jfs + p(i - w) 




It is apparent that the moment at any point in the beam may be found if the moment at the reaction can 
be f ooad. We will, therefore, pr oc e e d with the solution of a general case of a continuous girder with concentrated 

Conaider now a beam resting on several supports on the same level, and loaded as shown in Fig. D. The 
slope ci the tangent to the deflection curve at B, considering the portion of the beam to the left of JB, is ^: and 
that for the tangent to the deflection curve at B, considering the portion to the right of B, is ^ Binoe the de- 
of a beam must aeoessarily be continuous, MN is a strai^t line, and 

fh' -^ 




[Sec. 7-W 

This is the general equation of three moments for one concentrated load in each span. 
When there is more than one concentrated load in each span, the loading may be broken up into 
a series of cases similar to the above. For each set of loads the right-hand portion of equation 
(1) is solved, and the results finally combined, and placed equal to the left-hand portion of 
equation (1). 

The theorem may be applied to a beam with uniform loads. For the loading shown in 
Fig. 41 the theorem has the form 

4 4 



\ « ^"-4 -/,---4 ^^ ^"4 

t-ff ^» ^4 "9 

Fia. 40 A. 

For a constant moment of inertia and all spans equal, equation (1) reduces to 
Mi -h 43/, + Af4 = - PjZ(iki - iti») - P,2(2fc, - 3A;,* -f ik,>) 

When hi— kt ^ 0.5 this equation reduces to 

Mt + 4Mi + M4 = - 0.375P,Z - 0.875P,Z 

When in equation (2) the moment of inertia is a constant and all spans carry the same uniform 


Mrh + 2Mz(h + Iz) 4- MaU ^ - I W -h h*) 

and when /« = /j, 

Mt + 4Mt + A/« « - 




— I 

t -er~^-> 

i — <r-^i» 





Fig. 41. 



800 fh per ft 




1 < /r- >| 

Fra. 42. 

When using any of the foregoing formulas for continuous girders, it should be borne in 
mind that the supports are assumed to be on the same level throughout the process of loading. 

Illustrativb Pboblbm. — Compute the momenta at the supporta and the macnitude o