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

Full text of "Behavior of Mucks and Amorphous Peats as Embankment Foundations : Executive Summary"

IS I 



SCHOOL OF 
IMM CIVIL ENGINEERING 

INDIANA 

DEPARTMENT OF HIGHWAYS 



JOINT HIGHWAY RESEARCH PROJECT 
FHWA/IN/JHRP-87/2 - j 
Final Report 



BEHAVIOR OF MUCKS AND AMORPHOUS 
PEATS AS EMBANKMENT FOUNDATIONS 



Timothy Crowl and C. W. Lovell 



I 







PURDUE 



UNIVERSITY 



JOINT HIGHWAY RESEARCH PROJECT 
FHWA/IN/JHRP-87/2 - | 
Final Report 



BEHAVIOR OF MUCKS AND AMORPHOUS 
PEATS AS EMBANKMENT FOUNDATIONS 



Timothy Crowl and C. W. Lovell 



FINAL REPORT 

BEHAVIOR OF MUCKS AND AMORPHOUS 
PEATS AS EMBANKMENT FOUNDATIONS 



To: H. L. Michael, Director 

Joint Highway Research Project 

From: C. W. Lovell, Research Engineer 

Joint Highway Research Project 



May 26, 1987 
Project: C-36-5P 
File: 6-6-16 



Attached is a Final Report on the study, "Behavior of Mucks 
and Amorphous Peats as Embankment Foundations". The report is 
written by Timothy Crowl and C. W. Lovell. 

These materials of very low shear strength and very high 
compressibility are seldom used as embankment foundations. 
However, the research proposes a technique for successful use 
for low embankments (+10 ft), where loading is in stages. 

Stage height is controlled by suitable safety against bear- 
ing failure, lateral squeezing, embankment spreading, and rota- 
tional sliding. Strengths are determined at any stage of the 
construction by field vane shear measurements. Stage duration 
is determined by field pore pressure measurements. 

Settlement prediction is produced by: laboratory creep tests, 
the Gibson-Lo model, and previous correlations between laboratory 
and field settlement measurements. Surcharge loading is usually 
required before a pavement may be placed on the embankment. 

Examples of the proposed method are included, both with and 
without the inclusion of geotextiles. 

The report is submitted for review, comment and acceptance 
in fulfillment of the referenced study. 

Respectfully submitted, 



Co Lh Ji£ 

C. W. Lovell 
Research Engineer 



cc 



A 


G. 


Altschaeff 1 


R 


.A. 


Howden 


B 


K. 


Part r i dge 


J 


M. 


Bell 


M 


K. 


Hunter 


G 


T. 


Sat ter ly 


M 


E. 


Cant ral 1 


J 


P. 


Isenbarger 


C 


F. 


Schol er 


W 


F. 


Chen 


J 


F. 


Mc Laughli n 


K 


C. 


Sinha 


w 


L. 


Dolch 


K 


M. 


Mell inger 


C 


A. 


Venabl e 


R 


L. 


Eskew 


R 


D. 


Miles 


T 


D. 


White 


J 


D. 


Fr icker 


P 


L. 


Owens 


L 


E. 


Wood 


D 


E. 


Hancher 















Digitized by the Internet Archive 

in 2011 with funding from 

LYRASIS members and Sloan Foundation; Indiana Department of Transportation 



http://www.archive.org/details/behaviorofmucksaOOcrow 



FINAL REPORT 

BEHAVIOR OF MUCKS AND AMORPHOUS 
PEATS AS EMBANKMENT FOUNDATIONS 



by 



Timothy Crowl 
Graduate Instructor in Research 

and 

C. W. Lovell 
Research Engineer 



Joint Highway Research Project 

Project No. : C-36-5P 
File No. : 6-6-16 



Prepared for an Investigation 
Conducted by the 

Joint Highway Research Project 

Engineering Experiment Station 

Purdue University 



in cooperation with the 

Indiana Department of Highways 

and the 

U.S. Department of Transportation 
Federal Highway Administration 



The opinions, findings and concJusions expressed in this 
publication are those of the authors and not necessarily 
those of the Federal Highway Administration. 



Purdue University 

West Lafayette, Indiana 

May 26, 1987 



TECHNICAL REPORT STANDARD TITLE PAGE 



1. Report No. 

FHWA/IN/JHRP-87/2 



7. Government Acceliion No. 



3. Recipient'! Cotolog No. 



4. Title ond Subtitle 

BEHAVIOR OF MUCKS AND AMORPHOUS 
PEATS AS EMBANKMENT FOUNDATIONS 



5. Report Dote 



May 26, 1987 



6. Performing Or B oni lotion Cod 



7. Aufhorfl) 



Timothy Crowl and C. W. Lovell 



8. Performing Orgoni i alton Report No. 

JHRP-87/2 



9. Performing Organization Nome ond Addrete 

Joint Highway Research Project 
Civil Engineering Building 
Purdue University. 
West Lafayette, Indiana 47907 



10. Wore Unit No. 



11. Contract or Gront No. 



)2. Sponioring Agency Nome ond Addrel* 

Indiana Department of Highways 
State Office Building 
100 North Senate Avenue 
Indianapolis, Indiana 46204 



13. Type of Report ond Period Covered 

Final Report 



14. Sponioring Agency Code 



15. Supplementary Nol«» 

Prepared in cooperation with the U.S. Department of Transportation, 
Federal Highway Administration 



16. Aottrocl 

The construction of highway embankments over deposits of amorphous peat and 
muck is made difficult by the low shear strengths, high compressibilities, and 
excessive amounts of creep typically associated with soils of this nature. This 
report begins with a review of the compression behavior of these soils, including 
a method for predicting embankment settlements from the results of laboratory tests. 
A soil testing program is then developed for the determination of parameters 
required for embankment design and construction. Field vane shear tests are re- 
commended for the measurement of the undrained shear strength, and creep tests 
are recommended for calculation of the parameters required for settlement prediction. 
The report concludes with the presentation of a procedure for design and construc- 
tion of embankments over amorphous peats and mucks. The procedure relies upon 
the wse of stage loading, preloading, and in some instances geotextiles, to overcome 
the problems ordinarily encountered during construction over such soft soils. 
Design examples illustrating this procedure are provided. 



17. Key Wordf 

peat, muck, embankment, foundation, 
creep, settlement, stage construction, 
surcharge, geotextiles 



19. Security Cloeell. (ol this report) 

Unclassified ■ 



18. Distribution Slolement 



No restrictions. This document is 
available to the public through the 
National Technical Information Service, 
Springfield, VA 22161 



20. Security Cloi.lf. (of thli roge) 

Unclassified 



21- No. of Pogei 

126 



22. Price 



Form DOT F 1700.7 u-eil 



Ill 



ACKNOWLEDGEMENTS 



Special thanks are extended to Mr. David Frost for the advice and 
support provided during this project. The authors wish to express their 
gratitude to Mr. Barry Christopher for his assistance and recommendations, 
Thanks are also due to Mrs. Cathy Ralston and Mrs. Kathie Roth for their 
help in preparing this report, and for their support and encouragement. 

Financial support for this research was provided by the Indiana 
Department of Highways and the Federal Highway Administration. The 
research was administered through the Joint Highway Research Project, 
Indianapolis, Indiana. 



IV 



TABLE OF CONTENTS 

Page 

LIST OF TABLES vi 

LIST OF FIGURES vii 

LIST OF SYMBOLS AND ABBREVIATIONS ix 

HIGLIGHT SUMMARY xi 

CHAPTER I-INTRODUCTION 1 

Introduction 1 

Scope 3 

CHAPTER II-LITERATURE REVIEW 4 

Compression of Organic Materials 4 

Settlement Prediction 6 

Field Inconsistencies 6 

Prediction Model Inconsistencies 8 

Gibson-Lo Model 10 

Yield Envelope Concept 18 

Construction Techniques 23 

Stage Loading 23 

Berius 23 

Sand Drains 24 

Geot ex tiles 25 

CHAPTER III-TESTING PROGRAM 32 

Introduction 32 

Materials Studied 32 

Production of Samples for Laboratory Testing 33 

Creep Testing 38 

K Triaxial Testing 45 

Field Vane Shear Tests 53 

CHAPTER IV-PROCEDURE FOR DESIGN AND CONSTRUCTION 62 

Site Exploration 62 

Embankment Design 64 

Embankments without Geotextile Reinforcement 65 



Overall Bearing Capacity 66 

Lateral Squeeze 66 

Embankment Spreading 67 

Rotational Failure 70 

Geotextile Reinforced Embankments 71 

Overall Bearing Capacity 71 

Lateral Squeeze 71 

Rotational Failure 72 

Embankment Spreading 72 

Limit Fabric Deformation 73 

Stage Loading 75 

Preloading 77 

Field Observations 81 

Embankment Materials 82 

Construction Sequence 82 



CHAPTER V-CONCLUSIONS AND RECOMMENDATIONS 



86 



Conclusions .... 
Recommendations 



86 
87 



LIST OF REFERENCES, 



89 



APPENDICES 



Appendix 
Appendix 
Appendix 
Appendix 
Appendix 



93 

102 

106 

Design Examples 110 

Negative Numbers for Contact Prints 131 



Creep Test Results 

K Triaxial Test Check List 
Computer Programs 



LIST OF TABLES 



VI 



Table Page 

3.1 Soil Characteristics 34 

3.2 Values of C 44 

a 

e 

3.3 K Triaxial Test Results 54 

o 

3.4 Values of Undrained Shear Strength from 

Field Vane Shear Tests 57 



Vll 



LIST OF FIGURES 

Figure Page 

2.1 Primary Compressibility versus Stress Level. 

From Edil and Mochtar (1984) 13 

2.2 Secondary Compressibility versus Stress Level. 

From Edil and Mochtar (1984) 14 

2.3 Correction Curve for Laboratory Values of 

Secondary Compressibility. From Edil and 

Mochtar (1984) 15 

2.4 Dependency of Rate Factor for Secondary 

Compression on Average Strain Rate. From 

Edil and Mochtar (1984) 16 

2.5 Major Features of Yield Envelope. From 

Watson et al. (1984) 19 

2.6 Effective Stress Path during Multi-Stage 

Loading. From Watson et al. (1984) 22 

3.1 Slurry Consolidometer 37 

3.2 Slurry Consolidometer Adapted for 

Production of Creep Test Samples 39 

3.3 Strain versus Logarithm Time, Creep Test OL-6-2.... 41 

3.4 Strain versus Logarithm Time, Creep Test LR-6-1.... 42 

3.5 Calculation of C 43 

3.6 New K Triaxial Cap, Top View 49 

3.7 New K Triaxial Cap, Bottom View 50 

o r * 

3.8 K Triaxial Apparatus 51 

3.9 Loading Time-Rate Correction Factor versus 

Plasticity Index for Undrained Shear 

Strength as Measured by the Field Vane. 

From Bjerrum et al. (1972) 59 



Vlll 



Figure Page 

4.1 Description of Variables in Equation 4.3 68 

4.2 Embankment Spreading 69 

4.3 Pore Pressure Transducer Installation 

From Slope Indicator Company 76 

4.4 Plot of Log Strain Rate with Time from 

Laboratory Tests. From Lo , Bozozuk, 

and Law (1976) 79 

4.5 Inclinometer Type SGI. From Winterkorn 

and Fang (1975) 83 

4.6 Construction Sequence. From 

Barsvary et al. (1982) 84 

Appendix 
Figure 

Al Strain versus Logarithm Time, Creep Test OL-3-1 94 

A2 Strain versus Logarithm Time, Creep Test OL-6-1 95 

A3 Strain versus Logarithm Time, Creep Test OL-9-1 96 

A4 Strain versus Logarithm Time, Creep Test OL-9-2 97 

A5 Strain versus Logarithm Time, Creep Test LR-3-1 98 

A6 Strain versus Logarithm Time, Creep Test LR-6-2 99 

A7 Strain versus Logarithm Time, Creep test LR-9-1 100 

A8 Strain versus Logarithm Time, Creep Test LR-9-2 101 

Dl Embankment Configuration for Design Example Ill 

D2 Revised Embankment Configuration for 

Design Example 113 

D3 Plot of STABL Output for 10 Foot Embankment 116 

D4 Plot of Logarithm Rate Versus Time, 

Creep Test OL-9-2 117 

D5 Settlement Prediction for 10 Foot Embankment 120 

D6 Plot of STABL Output for Surcharged Embankment 123 

D7 Settlement Prediction for Surcharged Embankment 125 



IX 



LIST OF SYMBOLS AND ABBREVIATIONS 



ASTM 



field 



lab 



c 
c" 

E, 



'f r 



L 

1 



'all 



- American Society of Testing Materials 

- primary compressibility, also one half of 
deposit thickness 

- length of embankment (=1 ft for unit length) 

- secondary compressibility 

- field value of secondary compressibility 

- laboratory value of secondary compressibility 

- modified coefficient of secondary 
compressibility 

- undrained shear strength, see also s 

u 

- effective cohesion 

- minimum geotextile tensile modulus 

- minimum geotextile modulus to 
control rotational failure 

- coefficient of lateral earth pressure at rest 

- modulus of subgrade reaction 

- one half of embankment base width 

- lateral distance from crest to toe of embankment 

- bearing capacity factor 

- lateral earth pressure in cohesionless embankment 

- force resisting embankment spreading 

- ultimate bearing capacity 

- allowable pressure 



app 



f r 



t 

Ao 
e 



max 



Y 
\ 

b 

(-) 
v b ; lab 



'■b-'f ield 



sf 



applied pressure 

undrained shear strength, see also c 

required fabric strength 

required tensile strength of fabric 

time 

time of last strain reading 

applied stress increment 

strain 

last strain reading 

maximum strain in geotextile along 
embankment centerline 

unit weight of embankment soil 

rheological parameter in Gibson-Lo model 

rate factor for secondary compressibility 

laboratory value of 

rate factor for secondary compressibility 

field value of rate factor 
for secondary compressibility. 

effective angle of internal friction 

soil fabric friction angle 



xiR 



HIGHLIGHT SUMMARY 



The construction of highway embankments over deposits 
of amorphous peat and muck is made difficult by the low 
shear strengths, high compressibilities, and excessive 
amounts of creep typically associated with soils of this 
nature. This report begins with a review of the compression 
behavior of these soils, including a method for predicting 
embankment settlements from the results of laboratory tests. 
A soil testing program is then developed for the determina- 
tion of parameters required for embankment design and con- 
struction. Field vane shear tests are recommended for the 
measurement of the undrained shear strength, and creep tests 
are recommended for calculation of the parameters required 
for settlement prediction. The report concludes with the 
presentation of a procedure for design and construction of 
low embankments ( ± 10 ft) over amorphous peats and mucks. 
The procedure relies upon the use of stage loading, preload- 
ing, and in some instances geotextiles, to overcome the 
problems ordinarily encountered during construction over 
such soft soils. Design examples illustrating this pro- 
cedure are provided. 



CHAPTER I-INTRODUCTION 

INTRODUCTION 

A large number of deposits of amorphous peat and muck 
are located within the State of Indiana. Many difficulties 
are encountered when highway embankments are constructed 
over these soft soils. In the past, highway engineers have 
relocated roadways to avoid construction over peat or muck. 
In other instances, the organic material was excavated and 
replaced with a more suitable material. However, neither of 
these methods are economical by modern standards, forcing 
highway departments to develop more sophisticated methods 
which allow construction directly across deposits of such 
materials . 

Two characteristics associated with amorphous peats and 
mucks make them undesirable as materials for embankment 
foundations. Materials of this nature compress excessively 
when they are subjected to an applied load. A large portion 
of the compression is a result of the relatively high 
amounts of secondary compression associated with organic 
soils. These deformations occur over a long period of time, 
which compounds the problem. Deposits of these materials 
possess low preconsolidat ion pressures, so a large 



2R 



compression response is likely even at low stress levels. 
Amorphous peats and mucks are also characterized by very low 
shear strengths. Shear failures, which are both expensive 
and time consuming to renovate, can occur very easily either 
during or after construction. Deposits of amorphous peat 
and muck are highly variable, so that representative values 
of compressibility and shear strength are difficult to 
define. 

As a result of these typical characteristics, efforts 
to construct highway embankments over these materials have 
often resulted in poor performance in the form of excessive 
total settlements, large differential settlements, and shear 
failures. In addition, attempts to predict embankment set- 
tlements from the results of laboratory tests are often 
unsuccessful. 



It is the aim of this project to develop a procedure 
for the design and construction of low embankments (- 10 ft) 
over amorphous peat and muck. This procedure will include 
the use of a soil testing program to determine parameters 
required during embankment construction. In addition, using 
the method for settlement prediction presented in this 
report, it is hoped that more reliable settlement predic- 
tions may be achieved. 



SCOPE 

This report will begin with a review of existing 
literature concerning the topic. Included will be a discus- 
sion of selected highlights of previous work performed at 
Purdue University by Gruen (1983) and Joseph (1986). 
Emphasis will be placed on settlement prediction and con- 
struction techniques. 

Chapter 3 will describe the testing program developed 
for use during the design and construction of embankments 
over amorphous peat and muck. Sample preparation, testing 
procedures, and test results will be covered in this 
chapter. During this project, a major modification was 
required for the K triaxial apparatus, and a discussion of 
this design is included. 

The recommended design and construction procedure will 
be presented in Chapter 4. Material in this chapter will 
include a discussion of site exploration, implementation of 
the testing program, and subsequent embankment construction. 
In many instances on soft soils, geotextiles are used during 
embankment construction, and a discussion of the design of 
reinforced embankments is presented. Design examples will 
be included for unreinforced and geotextile reinforced 
embankments. This report will conclude with Chapter 5, 
which provides a summary of the content and recommendations 
for further research. 



CHAPTER II-LITERATURE REVIEW 

COMPRESSION OF ORGANIC MATERIALS 

To date, there has been a large amount of effort expen- 
ded to develop a better understanding of the compressibility 
characteristics of organic soils. Research has been perfor- 
med in both the laboratory and the field to predict the 
behavior of these materials under an applied load. The pri- 
mary goal of this research has been to increase the relia- 
bility of settlement estimates. 

Organic materials display four modes of deformation 
when they are loaded (Gruen, 1983): 

Instantaneous Strain: This mode occurs when the soil 
Is initially loaded as a result of the elastic deforma- 
tion of the soil mass. During this mode, gas that is 
trapped within the soil is also compressed or dissipa- 
ted. Instantaneous strain occurs very quickly. 

Primary Strain: Excess pore water pressures develop 
when a load is applied to the soil mass. During pri- 
mary strain, deformation is a result of the dissipation 
of these pressures as the water is expelled from the 
pores. This strain mode occurs relatively quickly in 



most instances, but accounts for a large amount of 
settlement . 

Secondary Strain: This strain mode begins during pri- 
mary strain and continues after excess pore water pres- 
sures have dissipated. The effective stresses between 
particles remain constant, implying that deformation is 
the result of creep. This strain mode continues for 
long periods of time, and is responsible for large 
amounts of settlement. 

Tertiary Strain: This mode has been observed only in 
laboratory consolidation tests. Edil & Simon-Gilles 
(1986) state that there is no known evidence of ter- 
tiary compression occurring in the field. Tertiary 
strain is indicated by an increase in the rate of creep 
greater than the rate of secondary strain. Tertiary 
strain must ultimately return to secondary strain. 

Dhowian & Edil (1980) state that at high consolidation 
pressures, the coefficients of secondary and tertiary 
compression approach the same value. This implies that 
secondary compression and tertiary compression occur 
simultaneously at high pressures. Dhowian & Edil offer 
no explanation for this mode of strain. 



SETTLEMENT PREDICTION 

There have been many attempts to predict the rates of 
settlement of embankments constructed over organic soils, 
with limited success. Inconsistencies between conditions in 
the laboratory and the field make accurate settlement 
prediction difficult. In addition, the models developed to 
make these predictions are approximate. 

Field Inconsistencies 

A number of discrepancies exist between laboratory and 
field conditions when these materials are loaded. Lefebvre 
et al. (1984) compared the results of laboratory tests and 
field performance for an embankment constructed on a fibrous 
peat. Two distinct trends were noted during this compar- 
ison. One trend indicated that the coefficient of secondary 
compression in the field was at least twice that exhibited 
under laboratory loading conditions. Also, the time for 
primary consolidation to occur in the field was less than 
that predicted using a one-dimensional theory of consolida- 
tion with the results from laboratory tests. 



Lefebvre et al. (1984) attributed the larger field 
values of secondary compression to a number of factors. 
During oedometer tests, only vertical deformations are 
allowed, while field deformations are not purely one- 
dimensional. Any lateral movements will Increase the amount 



of vertical deformation, and may be mistaken as secondary 
compression. They also suggest that variations of the water 
table within the peat and the embankment materials increased 
compression. Any decrease in the elevation of the water 
table was accompanied by a decrease in pore water pressure. 
As a result, the effective stresses in the deposit were 
increased, inducing additional settlement. Snow loads 
during winter months were also responsible for increased 
settlements. Lefebvre et al. contend that a portion of the 
vertical deformation assumed to be secondary compression was 
actually primary compression caused by increased effective 
s tresses . 

The f aster-than-predicted primary consolidation obser- 
ved in the field was in part due to the use of the Terzaghi 
theory in making the prediction. Terzaghi theory does not 
take into account any secondary compression that occurs 
during primary consolidation. The observed discrepancy was 
also due to radial consolidation which occurred In the 
field, but was not allowed in consolidation tests conducted 
in the laboratory. In fact, for fibrous peat, horizontal 
permeability actually became larger than vertical permeabil- 
ity as the peat compressed. Lefebvre et al. (1984) also 
attribute the difference between field and laboratory conso- 
lidation times to the variation of the water table and the 
resulting change in effective stresses. 



8 



Other sources of error are a result of poorer compac- 
tion around measurement instruments placed in the fill, pos- 
sibly affecting the observed settlements and pore pressures. 
In addition, the variability of the initial void ratio 
creates a range of initial compressibilities under low 
stresses. However, for loading greater than 20 kPa, the 
compressibilities were observed to converge to nearly the 
same value when an additional load was applied. 

In most instances, deposits of amorphous peat and muck 
are underlain by thick deposits of soft clay or marl. Weber 
(1969) observed the field performance of an embankment con- 
structed over a peat deposit. A layer of soft silty clay 
beneath the peat was observed to display long term compres- 
sion. Weber felt that this compression was significant 
enough to cause the poor correlation between settlement 
predictions and field measurements. 

Prediction Model Inconsistencies 



In a report by Gruen (1983), a review was made of exis- 
ting settlement prediction models for peats. None of the 
models considered account for settlement resulting from 
shear deformations. They state that shear deformations can 
comprise a large portion of settlement, so these predictive 
models provide approximate solutions at best. 



Of all the models treated by Gruen, the Gibson-Lo model 
was chosen as the most useful. This model was the easiest 
to use, yet It was found to be as accurate as other models 
reviewed. In order to determine the accuracy of predictions 
from this model, consolidation tests were performed in the 
laboratory. Using the results of these tests, settlement 
predictions were made using the Gibson-Lo model for consoli- 
dation tests performed at other stress levels. Gruen found 
that predictions were reasonably accurate for stress levels 
less than or equal to approximately two times the stress 
level for tests used to make the predictions. 



The Gibson-Lo model is a rheological model consisting 
of a Hookean spring in series with a Kelvin or Voigt 
element. The input parameters of this equation are strain- 
rate dependent. However, as Edil & Simon-Gilles (1986) dis- 
cussed, the field strain rate during secondary compression 
is often two to three orders of magnitude lower than the 
laboratory strain rate, due to the large difference in 
thickness between laboratory samples and field deposits. 
Therefore, the Gibson-Lo model can not be used directly to 
predict field performance from the results of laboratory 
tests. A correlation between parameters obtained from 
laboratory and field loading must be used to make accurate 
predictions of field performance. 



10 



Gibson-Lo Model 



The Gibson-Lo model provides a prediction of the one- 
dimensional compression of soils. This model is stated in 
the following equation from Edil & Simon-Gilles (1986): 



where : 



-(r)t 
e(t)=Ao[a+b(l-e )] 



e(t)=strain at time t 



(2.1) 



Ao=applied stress increment 

a=primary compressibility 

b=secondary compressibility 

T-=rate factor for secondary compressibility 
b 

A method presented by Lo , Bozozuk and Law (1976) allows 
for simple determination of the parameters a, b, and -r-. 
Using this method, the logarithm of strain rate is plotted 
versus time. The straight line portion of this curve 
corresponds to the time range of secondary compression. If 
the straight line is extended back to the y-axis, the 
parameters can be found by solving simultaneously the fol- 
lowing equations: 



line slope = 0.434 (t-) 



(2.2) 



y-intercept = log(Ao-A) 



(2.3) 



11 



a = ~ - b + be (2.4) 



where e ■ last strain reading 



t ■ time of last strain reading 



Edil & Mochtar (1984) recommend using linear regression 
during the time range corresponding to secondary compres- 
sion. This will tend to reduce the variation of the strain 
rate resulting from the use of unequal time intervals. 

The three parameters a, b, and t- obtained from labora- 
tory tests are somewhat dependent on the value of stress 
increment, final stress level, and the average strain rate. 
Stress increments less than approximately two times the 
stress level tested in the laboratory will cause little 
variation in the value of the parameters. However, this is 
true only for laboratory conditions. The parameters 
obtained from the analysis of field and laboratory perfor- 
mances are different as a result of the discrepancies 
between these conditions. 



During research conducted by Edil & Mochtar (1984), the 
laboratory and field behaviors of organic soils under 
loading were observed. The results were compared to deter- 
mine any existing relationships between the two conditions. 
From this comparison they were able to develop correlations 
between the model parameters for laboratory and field 



12 



performance. Figure 2.1 provides a curve of consolidation 
stress versus primary compressibility. Data points are from 
laboratory tests and from field observations as dis- 
tinguished in the figure. The figure indicates that the 
primary compressibilities in the field and the laboratory 
are comparable for the same stress level. Therefore, the 
laboratory value of the parameter "a" will compare with the 
field value when the variation of soil properties is con- 
sidered. In addition, the curve fitted through the plotted 
points can be used to correct the value of the parameter "a" 
when a prediction is desired at a different stress level. 

Figure 2.2 provides a curve of the secondary compressi- 
bility factor, "b", versus stress level. The points plotted 
are for peat data only. As illustrated in this figure, the 
field value of "b" is higher than the laboratory value at 

equivalent stress levels. Using Figure 2.2, a plot of 
b, 



field 
'lab 



versus consolidation stress was constructed as illus- 



trated in Figure 2.3. Once again, it should be noted that 
Figure 2.3 represents data from observations made on peat 
only. 



A plot of strain rate versus r- is provided in Figure 
2.4. This figure indicates that no correlation exists 
between <±) lab and <£) £leld . 



13 



10 



-2 



z 

CM 

E 

CD 



10" 



10 



-4 






x 



Portage 

O NoDiatvill* 
£ Diwi»> world 
° M«Ji»on Mo»qu« 
D MIOdlilOO 

Olin Avanu* 

C(OC»«r Resident* 

y Miooietoo wooawou 




1 

X Waupaca 
A Hhlnalandt* 

MKMMon 

8oit»ffl 
OLi CrotM 
Few Ow Lag 

Solid lymboli: Lab 
Gpan sym&a&FkcttJ 

■> 



J 



10 20 



50 



100 200 



500 1000 



CONSOLIDATION STRESS, CTlKPa) 



Figure 2.1 Primary Compressibility versus Stress 
Level. From Edil & Mochtar (1984). 



14 



Z 
E 

-Q 

> 

_J 
CD 

s 

LU 

cc 

0_ 



> 
DC 



O 
LU 
CO 



10- 



2 " 



10' 



10" 



- X 

A 
O 

o 

a 


"0 
o 

o 



10 



Open symbols: Field 
Solid symbols: Lab 




Wdupdca 

Rhinelander 

Middieton 

Soioerg 
La Crosse 
Fon du Lac 
Portage 
Nobiatvllle 
Disney World 
Madison Mosque 
i t 



20 



50 100 



200 



500 1000 



CONSOLIDATION STRESS, o=(kPa) 



Figure 2.2 Secondary Compressibility versus Stress 
Level. From Edil & Mochtar (1984). 



15 





10 




9 




8 


XI 

ffl 


7 


■a 


6 



4 - 



T r 



i r 



40 



80 



120 



160 



CONSOLIDATION STRESS, (f <kPa> 



Figure 2.3 Correction Curve for Laboratory Values 
of Secondary Compressibility. From 
Edil & Mochtar ( 1984) . 



16 



c 

1 



<75 

GO 



o 



o 

LU 
00 

g 

LL 

e 

o 

LU 



10 



10" 



10 



-3 



10" 



10 



10 



10" 



-6 



A RhineUndex 
O MIddleton 
O Solb«*g 
q Li Crosse 
t> Noblesvllle 

<3 Disney World 
oMidlion Mosque 
QMS * E 

Gngntno 
X Waupaca 
Fon Ou lac 
Portaga 



▼▼ 



- 





°/ 


1/ 


• 


- 
• 

Open symbols: Flald 









A 


D 


Solid Symbols Lab 
Middleton 


o/ 


a 


i 






a 

_L. 


Olln Avenue 
Crocker Residence 
Mlddiatoo woodwork 



io 8 10- 7 io -6 io-5 icr 4 io" 3 io" 2 



STRAIN RATE, (mil. "1) 



Figure 2.4 Dependency of Rate Factor for Secondary 
Compression on Average Strain Rate. 
From Edil & Mochtar (1984). 



17 



These figures can be used to correct the parameters 
obtained from laboratory tests for prediction of settlement 
in the field. Edil & Mochtar (1984) recommend performing 
the laboratory tests at stress levels that will be applied 
in the field to eliminate most of the effects of stress 
level. They also proposed increasing stress levels by 
amounts that will simulate field loading, rather than using 
a conventional load-increment ratio. 

Once laboratory consolidation tests are conducted in 
this manner, a curve of log strain rate versus time is con- 
structed and the appropriate parameters are determined. 
Using Figure 2.3, the value of "b" can be corrected. Figure 

2.4 can be used to obtain the r- (strain rate factor). If 

b 

the average field strain rate is not known from previous 
experience, it is recommended that a field strain rate two 
to three orders of magnitude smaller than that observed in 
the laboratory be assumed. Using these corrections, Edil & 
Mochtar (1984) and Edil & Simon-Gilles (1986) were able to 
predict quite accurately the settlement of peats when loaded 
in the field. 



One of the key assumptions in the development of the 
Gibson-Lo model is one-dimensional compression. During 
their study, Edil & Simon-Gilles (1986) noted that incli- 
nometers placed on the sites indicated very small amounts of 
lateral movement. In this case, the assumption was quite 
valid. 



18 



YIELD ENVELOPE CONCEPT 

In research performed by Joseph (1986), K triaxial 
tests were performed on thoroughly remolded samples of amor- 
phous peat and of muck. The results of these tests Indicate 
that a unique failure envelope in p'-q space exists for 
amorphous peats and mucks. The observed failure envelope 
began at the origin and developed concave upwards, indicat- 
ing that all strength was a result of friction, and that 
cohesion made no contribution to shear strength. The pres- 
ence of a unique failure envelope suggests that amorphous 
peats and mucks fit into the realm of classical soil mechan- 
ics, and behavior similar to soft clays can be expected. 

The yield envelope concept for soft clays was developed 
by Tavenas & Leroueil (1977). Joseph (1986) states that the 
yield envelope concept is valid for amorphous peats and 
mucks since the consolidation of these materials can be 
predicted by a generalized consolidation equation. 



Watson et al. (1984) provide a synopsis of the yield 
envelope concept. A diagram illustrating this concept is 
shown in Figure 2.5. The yield envelope is an envelope of 
stress states that separates the small strain response to 
loading from large strain response. The stress value at the 
point where the yield envelope intersects the K line is 
approximately equal to the preconsolidation pressure deter- 
mined in an oedometer test. 



19 



S| Cy (comprasslon) 



b 



o 



^ 



& 



f/ 



/A 

Larga strain 
stabla zona 




Small strain 
mata -stabla 
zona 



Small strain stabla zona 



YiaW anvaiopa 



Figure 2.5 Major Features of Yield Envelope. 
From Watson et al. (1984). 



20 

There are three important phases of soil response that 
can occur in this diagram, depending upon the effective 
stress state of the loaded soil. If the effective state of 
stress remains below the Mohr-Coulomb envelope and is con- 
tained within the yield envelope, the soil acts as an over- 
consolidated material. Upon loading there is a small amount 
of strain and excess pore pressures dissipate quickly. 

If the effective stress state is contained within the 
yield envelope, but is above the Mohr-Coulomb envelope, the 
soil is considered to be metastable. If an additional load 
is placed on the soil causing yield, the effective stress 
state will move along the Mohr-Coulomb line and strain 
softening will occur. Strain softening is ordinarily accom- 
panied by an increased amount of pore water pressures, total 
horizontal stresses, and lateral strains. 

The third type of soil response occurs when the effec- 
tive stress path is situated outside the yield envelope and 
below the Mohr-Coulomb line. When this condition exists, 
the soil behaves as a normally consolidated material. Large 
strains and porewater pressures develop upon loading. The 
excess pore water pressures dissipate rather slowly. 



Watson et al. (1984) suggest that in some instances 
construction of fills on soft soils should be performed 
using stage loads. Stage loaded construction allows for a 



21 

strength gain to occur in the foundation soil as a result of 
consolidation when the load is sustained. This strength 
gain increases stability when subsequent stage loads are 
applied . 

During embankment construction, the extent of lateral 
deformations should be minimized to reduce the amount of 
settlement. Watson et. al. (1984) state that by avoiding a 
stress state causing failure in the foundation material, 
lateral deformations can be controlled. They suggest con- 
structing the embankment in stages similar to that shown in 
Figure 2.6. Point E represents the initial state of stress. 
As the subsoil is loaded, the effective stress path moves 
along EA' and the total stress path moves along EA. 



If at this point construction to the final load is con- 
tinued, the effective stress path will move along EA'PR, and 
strain softening will occur. Large lateral deformations 
will develop as a result. However if construction is 
delayed, the excess pore water pressures will begin to dis- 
sipate and the effective stress path will move from A' to 
A'' in the figure. Once excess pore pressures have dissi- 
pated to point A'', the total load can be increased to total 
stress state B, which corresponds to effective stress state 
B' . In this instance, the effective stress path will cross 
the yield envelope and move into the large-strain-stable 
zone. Large consolidation settlements will occur in this 



22 




P. P' 



Figure 2.6 Effective Stress Path during Multi-Stage 
Loading. From Watson et al. (1984). 



23 

situation, but positioning of the stress state below the 
Mohr-Coulomb envelope will keep lateral displacements at a 
minimum . 

CONSTRUCTION TECHNIQUES 

Stage Loading 

Both Gruen & Lovell (1983) and Joseph (1986) recommend 
the use of stage loading when constructing embankments over 
amorphous peats and mucks. The low shear strengths associ- 
ated with these materials cause stability problems during 
construction. However, construction in stages, as discussed 
in the previous section, will allow strength gain while the 
load is held constant. In this manner the embankment can be 
constructed to its final height in a number of steps, and 
stability problems are avoided. In many cases on such soft 
materials, if the embankment were constructed to its final 
height in one step, the foundation materials would fail. 
Staged loading also allows for a large portion of settlement 
to occur during construction, reducing the amount occurring 
during the service life of the embankment. 

Berms 



Joseph (1986) discussed the use of berms during embank- 
ment construction. Berms are used to reduce lateral defor- 
mation and increase stability of embankments constructed on 



24 

soft soils. Raymond (1969) and Hollingshead & Raymond 
(1971) indicate that berms provide a useful means of reduc- 
ing undrained movements, therefore limiting the amount of 
shear deformations. Raymond (1969) suggests that successful 
use of berms requires that the berm width be 1 1/2 to 2 
times the depth of the peat and marl. 

Raymond (1969) emphasizes that the extent of the bene- 
fit provided by berms is greatly influenced by the sequence 
of construction. Less shearing stress develops if the 
embankment is constructed simultaneously from both outer 
edges inward to the centerline. This method was considered 
better than construction from the centerline outward or from 
one side to the other. Construction from the outer edges to 
the center has the advantage of trapping a developing mud 
wave in the center. In addition, if construction is per- 
formed in this manner, a berm failure is not likely to cause 
extensive damage beneath the central fill area. 

Sand Drains 



Joseph (1986) reviewed the use of sand drains to 
increase the rate of pore pressure dissipation during con- 
struction. From his review of literature on this topic, he 
concluded that sand drains do not provide much help in this 
manner. However, various researchers believe that sand 
drains have a beneficial effect on stability as a result of 



25 

pile action. Reduced wave action under traffic loads has 
also been observed when sand drains are in place, indicating 
they may sustain a portion of the load. Support may be pro- 
vided in a direct manner by the drains themselves, or in an 
indirect manner through arching. 

Geotextiles 

The use of geotextiles during construction of embank- 
ments over such soft materials is becoming increasingly 
popular. Research performed by various investigators indi- 
cates that geotextiles provide a number of positive effects 
when they are used during projects of this nature. 

Petrik et al. (1982) performed model tests to determine 
the behavior of a reinforced embankment. Two reinforcing 
materials were tested. One material was a polypropylene 
woven fabric, and the other was a brass sheet. The latter 
was included to determine the effects of a rigid reinforcing 
material . 



It was concluded from this research that the amount of 
horizontal deformation is significantly affected by the 
presence of reinforcement. However, the amount of vertical 
deformation that occurs is not substantially influenced by 
the presence of a geotextile. It was also concluded that 
both the bearing capacity and stability are enhanced through 
the use of fabrics. Tests performed on the embankment 



26 

models reinforced with brass were observed to result in 
higher bearing capacities, and to produce lesser lateral 
strain of the foundation material than those reinforced with 
the polypropylene fabric. These results imply that the 
stiffer reinforcement mobilized a higher strength in the 
subsoil . 

Hutchins (1982) discussed the behavior of a shallow 
embankment constructed over a deep black marsh muck with the 
use of geotextiles. A spunbonded polypropylene geotextile 
was placed on the muck surface, and a granular embankment 
fill was subsequently constructed. After construction, 
three types of plate load tests were performed at the site. 
The first type of test consisted of excavating a small hole 
in the embankment to the muck level. A plate load test was 
then performed directly on the fabric. The second type of 
test also consisted of excavating a hole to the muck level, 
however the geotextile was cut, and the plate load test was 
performed directly on the muck. The third type of test was 
performed on the in situ soil, away from the embankment. 

The results of this testing indicate that the effective 
bearing capacity under the geotextile increased by 39% at 
failure. The increase in bearing capacity was attributed to 
a modulus or membrane effect provided by the geotextile. 



During construction, a portion of the embankment was 
placed without the use of a geotextile. In this section the 



27 

contractor used twice the volume of sand to attain the same 
elevation as the section with the fabric. Hutchins (1982) 
theorized that the extra sand was used in the areas of local 
shear failure. The extra fill required in this section was 
a result of a lower modulus of subgrade reaction, K ,and an 

8 

irregular cross-section of the embankment. 

Hutchins (1982) cites bridging over weak areas as 
another advantage of the use of geotextiles. During site 
investigation, areas of low bearing capacity will be missed 
even during the most thorough investigations. Previously, 
engineers have designed embankments with a factor to account 
for these weak areas. However, this is not an economical 
design procedure. If geotextiles are placed in the embank- 
ment, they help to bridge the embankment over such areas, 
eliminating the need for overdesign for this criterion. 

Barsvary et al. (1982) also studied field performance 
of a highway embankment constructed over an amorphous granu- 
lar peat underlain by sands and a soft to stiff clay. A 
beneficial result of the use of geotextiles during this pro- 
ject was the formation of a barrier between the foundation 
and embankment materials. Barsvary et al. felt that the 
barrier reduced the problem of stability which is aggravated 
when the two materials intermix. 



To determine that the geotextile truly does provide an 
adequate barrier, excavations were made through the fill one 



28 

year after construction. There was no mixing of the 
subgrade with the embankment where geotextiles were used. 
However, in areas where no geotextiles were used, an irregu- 
lar interface developed between subgrade and fill materials 
and the two soil types were intermixed. 



During construction of the embankment over this peat, 
the first stage was completed without any sign of rotational 
failure. Barsvary et al. believe that the membrane effect 
provided by the geotextile prevented such a failure. 
Strains of 2 to 5 percent were observed in the transverse 
and longitudinal directions at this point. The fact that 
strains developed in the longitudinal direction implies that 
plane strain conditions may not be an accurate assumption 
under low embankments constructed on highly compressible 
soils. This is the result of inconsistencies of the founda- 
tion material such as soft areas and tree stumps. 

During the second stage of construction, the longitudi- 
nal strain was 8 percent at the center of the fill, while 
transverse strain approached failure. Some mud waves and 
tension cracks were observed to develop near the embankment 
toe, indicating that there was deformation in lateral shear. 
Although there was evidence of lateral deformation, there 
were no tension cracks or horizontal displacements observed 
in the embankment. The geotextile is credited with res- 
training the embankment and preventing any lateral spreading 
from occurring. 



29 

Hannon (1982) observed the performance of a test 
embankment constructed over San Francisco Bay Mud with 
geotextile reinforcement. The embankment was constructed to 
a height of 16 feet prior to settlement with a planned final 
height of approximately 10-12 feet. During one year, 
approximately 6 1/2 feet of settlement occurred. Wick 
drains were installed to accelerate consolidation. 

During construction of the embankment, high excess pore 
pressures developed, which the author felt were capable of 
causing a shear failure. Hannon (1982) believed that the 
geotextile was responsible for the successful construction 
of the embankment without incurring any failures. In an 
adjacent area, the embankment was constructed by end dumping 
without the use of sand wicks or fabrics. In this instance, 
construction resulted in the development of a large mudwave . 



As a result of loading from truck wheels during con- 
struction, approximately six inches of instantaneous 
compression were observed. The reaction was quite similar 
to that of a large waterbed. Hannon (1982) believed that 
the geotextile contributed to stability and kept the embank- 
ment intact. Over a three month period, three inches of 
lateral displacement were observed. However, in such a soft 
foundation material this is not considered to be a signifi- 
cant amount. 



30 

Research conducted by Boutrup & Holtz (1983) used the 
finite element method to analyze the effects of geotextiles 
on embankment behavior. A portion of their investigation 
focused on determining the benefits of placing geotextiles 
higher up in the embankment between lifts. A finite element 
analysis performed with three layers of fabric placed 
between lifts resulted in a maximum reduction in shear 
stresses of 13%. However, an analysis performed with one 
layer of fabric placed at the interface between the embank- 
ment and foundation materials resulted in an 11% reduction 
of shear stresses. It can be seen from this comparison that 
the benefit of multiple layers of geotextile is not signifi- 
cant. 

An analysis was also performed to determine the effects 
of the placement of two layers of geotextile between the 
embankment and the foundation. The results indicated an 18% 
reduction in maximum shear stress. One layer of geotextile 
posessing twice the modulus will produce the same results. 
Therefore, the use of this procedure is more effective than 
the use of multiple layers of geotextiles. 



Humphrey (1986) investigated the use of the cap soil 
behavior model in a plain strain finite element analysis. 
The cap model is a nonlinear elastic-plastic isotropic 
work-hardening plasticity model. A drawback of this model 
is the absence of a method to reliably determine the 



31 



required model parameters from conventional test results. 

Humphrey presents a method that allows for simple parameter 

determination. The main input soil parameters required are 

the compressibilities in virgin loading and 

unloading/reloading, the effective Mohr-Coulomb shear 

strength parameters (<(>' and c'), and the undrained shear 

strength ratio s /o '. 

u p 

A weakness of the analysis using the cap model is its 
inability to predict behavior when the principal stresses 
rotate 90 degrees. As a result of this, Humphrey recommends 
using finite element analysis only for an estimation of 
forces in the geotextile when the foundation fails, or to 
make a comparison between reinforced and unreinforced 
embankment behavior. Limit equilibrium analysis should be 
used during design to determine the factor of safety against 
various failure modes. For a more indepth discussion of the 
design of reinforced embankments, the reader should consult 
both Humphrey (1986) and Boutrup & Holtz (1983). 



32 



CHAPTER III-TESTING PROGRAM 

INTRODUCTION 

A critically important portion of this research inclu- 
des the development and implementation of a soil testing 
program to determine various parameters required for design 
and construction of embankments over amorphous peats and 
mucks. This chapter will provide a discussion of the 
materials tested during the project and give a description 
of their behavior. Also covered will be the preparation of 
samples for laboratory testing, as well as procedures and 
results of creep tests, K triaxial tests, and field vane 
shear tests. During K triaxial testing, a major equipment 
modification was necessary, and a description of this dev- 
ice, as well as other required equipment not commonly found 
in geotechnical laboratories, will be provided. 

MATERIALS STUDIED 

The laboratory behavior of both an amorphous peat and a 
muck were studied during this research project. The amor- 
phous peat was obtained from a portion of the shoreline of 
Otterbein Lake in Benton County, Indiana. The muck was sam- 
pled from a depression along a portion of Lindberg Road in 
West Lafayette, Indiana. 



33 

Disturbed samples of these materials were obtained from 
these sites using shovels. Samples were placed in 10 gallon 
containers which were sealed with airtight lids. Sufficient 
water from the site was placed in the containers to keep the 
materials saturated in natural waters during storage. 

Laboratory tests were performed to determine the speci- 
fic gravity, organic content, liquid limit and plastic limit 
of each material, as described in ASTM Standard Specifica- 
tions D854-83, D2974-84, and D4318-84 respectively. A tabu- 
lation of these characteristics, as well as other values 
previously determined by Joseph (1986) appears as Table 3.1. 
Asterisks indicate values taken from Joseph. 

PRODUCTION OF SAMPLES FOR LABORATORY TESTING 



As stated in the previous section, the samples were 
obtained from disturbed sampling at the site. Samples were 
obtained by this means as a result of the decision to per- 
form tests on thoroughly remolded samples prepared from a 
slurry. This decision was based on the fact that it is very 
difficult to obtain an undisturbed sample of these materials 
suitable for triaxial testing. The low preconsolidation 
pressures typical of these materials makes them soft, and 
trimming of samples would result in a large amount of dis- 
turbance from handling. Also, for triaxial testing a mem- 
brane must be placed over the sample. This process would 



34 



Table 3.1 Soil Characteristics 



QUANTITY OTTERBEIN LAKE LINDBERG ROAD 

Specific Gravity 1.8-2.0 2.2-2.5 

Organic Content 55.3% 34.7% 

Liquid Limit ** 123.5% 

Plastic Limit ** 92.2% 

Water Content 365-465 * 130-140 * 

P H 6.75 * 6.5 * 

Initial Void Ratio 10-21 * 2.9-4.7 * 

Fiber Content Nil Nil 



* Indicates values taken from Joseph (1986) 

** Atterberg Limits could not be tested for this 

material as a result of the lack of cohesion. 



35 

seriously disturb the sample. Production of samples from a 
slurry has the advantage of allowing the sample to be formed 
inside a membrane for triaxial tests, which greatly reduces 
the amount of handling. 

In research conducted by Landva (1986) the validity of 
the use of remolded slurry samples for fibrous peats was 
investigated. When the results of tests performed on undis- 
turbed samples and remolded samples were compared, nearly 
identical shear and consolidation properties were observed. 
Landva (1986) concluded that the fabrics of undisturbed sam- 
ples and remolded samples were quite similar after consoli- 
dation under initial loads. 



Before producing a slurry, the appropriate water con- 
tent must be decided upon. Krizek & Sheeran (1970) present 
factors to consider before choosing a water content. They 
state that a high water content slurry has the advantages of 
being easily deaired and more easily placed in the consoli- 
dometer. In addition, the method of placement in the conso- 
lidometer has less influence on the fabric of the sample. 
However, the disadvantages of a high water content are 
segregation of soil particles in the slurry and a longer 
amount of time required to consolidate the sample. Krizek & 
Sheeran (1970) recommend determining the best water content 
by trial and error, using 1.5 to 2 times the liquid limit as 
a starting point. 



36 

During this project, a water content of approximately 
1.5 times the liquid limit was chosen for the muck. Atter- 
berg limits could not be tested for the amorphous peat as a 
result of the lack of cohesion. Therefore, the minimum 
water content providing a workable slurry was chosen by 
trial and error. 

Once the slurry water content was determined, slurry 
production could begin. The appropriate amounts of soil and 
distilled water were obtained and placed in a blender. The 
blades of the blender were covered with a few layers of 
masking tape to prevent damage to the individual soil parti- 
cles. When the mixture appeared to be uniform, the slurry 
was removed from the blender and placed in a deairing cham- 
ber for 24 hours or until the level of the slurry was not 
observed to decrease when the vacuum was released. After 
deairing, the slurry was poured into the consolidometer and 
the appropriate consolidation pressure was applied. 



The consolidometer present in the laboratory is shown 
in Figure 3.1. This equipment allows samples to be consoli- 
dated within a membrane. Originally this consolidometer was 
designed for use in a loading frame where the consolidation 
pressure was applied through a load cell. However, the 
loading frame present in the laboratory was being used quite 
steadily for another project. To overcome this, an inexpen- 
sive clamp was constructed to hold the piston in place. 



37 




"TPf ^ 




Figure 3.1 Slurry Conso 1 idome t er . 



38 

Since consolidation loads required to produce samples were 
low, dead weights were placed directly on the piston to 
eliminate the need for a loading frame and load cell. 

There was no existing equipment in the laboratory for 
production of samples for creep tests. A modification of 
the consolidometer was developed to allow consolidation of 
the slurry to occur in the oedometer ring to reduce handling 
of the sample. A photograph of this modification is shown 
in Figure 3.2. 

CREEP TESTING 

Samples were produced from a slurry in the oedometer 
ring at pressures of 0.7 psi, 3.0 psi and 6.0 psi. Once 
samples were consolidated, the excess material was trimmed 
away to form a sample with a thickness of one inch. The 
weights of the oedometer ring and sample were recorded for 
each test. The oedometer cell was then assembled and placed 
on the loading frame. To eliminate the effect of tempera- 
ture variation on the creep test results, the loading frame 
was surrounded by insulated panels. Samples were backpres- 
sure saturated for approximately 48 hours, beginning at 10 
psi and increasing the pressure to 40 psi in increments of 
10 psi. 

Samples were then reloaded to the consolidation pres- 
sure at which they were formed in the consolidometer. Once 



39 






Figure 3.2 



Slurry Consol idometer Adapted for 
Production of Creep Test Samples. 



40 

secondary compression was observed, the loads on the samples 
consolidated to 0.7 psi, 3.0 psi, and 6.0 psi were increased 
to 3.0, 6.0, and 9.0 psi respectively. Two tests were also 
conducted by increasing the load from 0.7 psi to 9.0 psi to 
simulate construction of the entire embankment. Loading was 
sustained on each sample until sufficient data of secondary 
compression had been collected for determination of the 
modified coefficient of secondary compression, C . 

Typical test results for each material are provided in 
plots of strain versus log time in Figures 3.3 and 3.4. The 
remaining test results are provided in Appendix A. Test 
materials and consolidation pressures are signified in the 
following identification scheme. Samples from Otterbein 
Lake are represented by OL and samples from Lindberg Road 
are identified by LR. The number following these symbols 
indicates the final consolidation stress of the test. The 
second number indicates the number of the test performed at 
that stress level. Therefore, 0L-6-1 identifies the first 
test performed on samples from Otterbein Lake consolidated 
at 6 psi. Tests OL-9-2 and LR-9-2 represent tests on sam- 
ples formed at 0.7 psi and consolidated at 9.0 psi. 



The values of C were calculated for each test as 

illustrated in Figure 3.5. The resulting values are shown 
in Table 3.2. It should be noted that tertiary compression 



41 








CN 






1 






■ 


' 




1 

O 






4-1 


<* 
. o 






• 




H 


















a. 


■ 




01 


' 




0) 




r\ 


u 


n 
. o 


to 


01 


i m 


0> 


B 


i 


t> 


•H 




J 


H 




c 






••-4 


s 


. 


E 


4-1 
•H 


«« 


Qj 


to 


. e 


E 


00 




••4 


o 




1- 


J 


, 




CO 






a 

GO 






l-i 
0) 


■ 




> 


. o 






*4 




c 






•H 






(0 






M 


' 




4-> 






W 



f 2 



o 


o 


o 


o 
o 


E 


o 
m 


e 


CD 


N 


o 


•4 


o 


. 





o 


o 


1 


i 



o 
m 

in 

o 



o 


o 


o 


o 


O 


o 


R 


o 


in 


O 


o 


in 


m 


O 


in 


C) 


(V 


r* 


o 


N 


o 


*4 


n 


in 


o 


o 


-J 


•4 


.4 



CO 

n 

<u 
u 

D 

oo 

•rH 



uiejis 



42 




i 

PES 



H 

a. 

a) 

U 

u 



B 



6 



i-i 

nj 

oo 
o 

-4 

w 

3 

> 



CD 



O 
O 



O 
O 



o 
o 
o 

in 

w 

a 
i 



o 
o 


a 

O 


o 
o 




o 
a 


K 


w 


ri 


o 


o 
1 


o 
1 


o 

1 



o 
a 

s 



o 
o 

o 



u 

O 

60 



43 











in 
















<£> 
















cc 














oo 


.-i 














.H 
















o 


II 














• 








in 








o 


o 






oo 








II 












(l) 




r~» 


ff 


O 






rH 




a 




nH 


iH 


c 






II 




u 




• 

o 


1 


c 

t 

a 






-P 









+j 


1 


o 
o 


II 




8 1 




c 


u 


8 1 


o 


T 


n 


r» 











<: 


n 




H 


x> 




< 




•H 
03 




i— i 

< 


O 


2 


o 

• 

o 


X) 

• 






e-l 






1 


^H 














o 


3 


ii 


II 


II 


II 


B f*l 




1 


i-H 


U) 


(J 


+J 


0) 


• H 




_J 


(0 


a 


<a 




a 


/^ o 







u 


u 




< 


u 


1 o 
II 
a) 




















< 





ft 
. o 



u 





/> 




o 


V) 


<4H 


•• 


* 


o 




*» 


c 




3 


o 




C 


•H 




•«4 


4-1 




£ 


tu 




%• 


3 


a 

•4 


E 


tj 

i-l 
CO 




■•4 


U 




H 


• 

to 

01 


• 





M 

P 


•4 




00 



[n 



E 



K 



Table 3.2 Values of C 



TEST NO, 



44 



OL-3-1 
OL-6-1 
OL-6-2 
OL-9-1 
0L-9-2 
LR-3-1 
LR-6-1 
LR-6-2 
LR-9-1 
LR-9-2 



0.0070 
0.0184 
0.0167 
0.0298 
0.0171 
0.0086 
0.0083 
0.0105 
.0247 
0.0135 



45 

was observed in tests performed on the peat samples from 
Otterbein Lake. This was expected as this material was a 
peat. As previously mentioned, Edil & Simon-Gilles (1986) 
state that although tertiary compression has been observed 
in the laboratory, they have no evidence of such occurrence 
in the field. However, it is possible that it does occur 
under field loading conditions, but is masked by other set- 
tlement phenomena occurring at the same time. 



K TRIAXIAL TESTING 
o 



The second portion of the testing program consisted of 
performing K triaxial tests. Gruen & Lovell (1983) state 
that construction of embankments over peat deposits results 
in deformations resembling those in axial compression. K 
conditions were chosen for a number of reasons. For a long 
narrow loading such as an embankment, deformations are 
assumed to be negligible in the direction of the embankment 
axis. Deformation would therefore only occur in the direc- 
tion perpendicular to the embankment centerline. It was felt 

that K conditions would be a much better approximation of 
o 

this behavior than isotropic conditions. Also, according to 
Lambe & Whitman (1979), if two samples are consolidated to 
the same vertical effective stress, one isot Topically and 

one under K conditions, the K consolidated sample will 

o ' o 

possess a lower undrained shear strength. Therefore, K 
conditions were chosen as they provide a conservative 
estimate of the undrained shear strength. 



46 

It is the aim of the construction method to be 

developed later in this report to construct embankments over 

these materials with reduced lateral deformations. This is 

essential as the Gibson-Lo Model for settlement prediction 

does not account for shear deformations. K conditions 

o 

therefore represent most closely the anticipated loading of 
the subsoil. 

The principle of K triaxial testing requires the 
prevention of change in cell water volume during consolida- 
tion, preventing lateral expansion of the sample. This is 

accomplished with the use of equipment developed by Cam- 
panella & Vaid (1972). They developed equipment utilizing a 
Bellofram piston with the same diameter as the soil sample. 
During consolidation the cell valve is closed to keep the 
volume of cell water constant. As the sample consolidates, 
the sample volume lost from compression is replaced by the 
piston as it advances into the cell. 



Problems were encountered with the use of this equip- 
ment on soft materials such as amorphous peats and mucks. 
The Bellofram piston in the existing equipment provided only 
1 -r inches of stroke, which was not sufficient to consoli- 
date and shear samples. In an attempt to alleviate this 
problem, the sample height was reduced from 6 inches to 5 
inches, and the pedestal was elevated one inch to accomodate 
this change. The modification was attempted on the basis 



47 



that for the same percentage of strain required to complete 
the tests a lesser amount of vertical deformation would be 
required. This adjustment did not provide the necessary 
amount of correction. 

Since Bellofram does not manufacture a 2 ■=■ inch diame- 
ter piston with sufficient stroke, a new piece of equipment 
had to be developed. The solution was to construct a system 
using a linear bearing to allow movement of a 2 ■=■ inch diam- 
eter piston to replace the need for the Bellofram. The 
design is similar to that of an oedometer cell. As with the 
oedometer, an O-ring is placed beneath the bearing on the 
inside of the cylinder to prevent leakage of water into the 
bearing . 



The weight of a solid steel piston would result in 
excessive loads being placed on the samples. It was there- 
fore decided to use a hollow aluminum piston to reduce the 
weight. The aluminum tube was anodized to harden it suffi- 
ciently to prevent damage as it moved through the linear 
bearing. However, when the anodized aluminum was run 
through the bearing, the piston jammed and streaks developed 
in the piston walls. It was later determined that even 
though the piston surface was anodized, the aluminum beneath 
the surface was still soft, allowing indentations to develop 
which jammed the system. 



48 

At this point there was no option but to use a case- 
hardened steel piston. To counteract the excessive piston 
weight, a counterbalance system was developed. Photographs 
of the redesigned linear bearing system are shown in Figures 
3.6 and 3.7, while the counterbalance system is detailed in 
Figure 3.8. 

Once the counterbalance system was constructed, the 
weight required to balance the piston had to be determined. 
It was not necessary to counterbalance the entire piston 
weight as there was some friction in the system at the 
interface between the piston and the O-ring. The piston was 
also partially supported by the buoyant force of the water. 
The counterbalance weight was determined by attaching con- 
tainers of water to the cable pulley system. Equal amounts 
of water were added to each container until the piston was 
no longer observed to descend under its own weight. 



During the first test performed with this equipment, a 
large amount of resistance was observed when the cell pres- 
sure was raised to large values. It is hypothesized that 
the high cell pressures exerted on the O-ring caused it to 
tighten around the piston as a result of the Poisson's 
ratio. To reduce friction, the lubricant used on the piston 
was changed from silicon oil to automotive grease. Lubri- 
Matic Multi-Service Lubricating Grease, available at Sear's 
Automotive Departments, was chosen for its high shear 



49 




Figure 3.6 New K Triaxial Cap, Top View, 



50 




Figure 3.7 New K Q Triaxial Cap, Bottom View 



51 




Figure 3.8 K Triaxial Apparatus 



52 

stability and water resistance. The increased resistance was 
also caused in part by uplift forces, as the diameter of the 
top platen was slightly less than that of the piston. The 
piston was then calibrated for resistance at high pressures, 
which was subsequently determined to be approximately 1 psi. 

Samples were produced from a slurry in the consolidome- 
ter at a pressure of 3.0 psi. After the samples were conso- 
lidated, the excess material was trimmed to form a sample 
with a height of 6 inches. The samples were then carefully 
mounted on the base pedestal. Once the plastic former was 
removed from the samples, great care was taken not to 
vibrate the apparatus, as the vibrations were observed to 
cause the soft samples to slough. 



After the samples were installed, backpressure satura- 
tion began. Cell pressure was constantly kept 1 psi higher 
than backpressure to prevent ballooning of the membrane. 
Using the control panel, cell pressure and backpressure 
could be increased simultaneously, but the axial pressure on 
the piston had to be increased separately. To avoid damag- 
ing the sample with large stress differences, the cell 
pressure/backpressure and axial pressure were increased in 
steps of 1 psi. If larger increments were used, the sample 
was observed to deform as a result of differences between 
axial and radial pressures. The samples were backpressure 
saturated to 80 psi. A high cell pressure was desired to 



53 



reduce the compressibility of water under additional 
stresses. This is necessary to keep the cell volume con- 
stant during K consolidation. 

Once the samples were saturated, they were consolidated 

under K conditions. Samples were consolidated at twice 
o r 

their preconsolidation pressures to reduce the effects of 
disturbance from handling. Therefore, samples formed from 
the slurry at 3.0 psi were consolidated at 6.0 psi. The 
samples were allowed to consolidate until secondary compres- 
sion was observed. At this point, the samples were loaded 
axially until shear failure occurred. A step by step listing 
of the procedure for K triaxial tests including installa- 
tion, saturation, consolidation, and axial loading is pro- 
vided in Appendix B. The results of K triaxial testing are 
presented in Table 3.3. 

FIELD VANE SHEAR TESTS 



When the results of K triaxial tests were used in 

o 

design analyses, a number of problems developed. Based on 

the conclusion of Joseph (1986) that a linear strength line 

exists from to 30 psi, it was assumed that the value of 

s /o would be constant. However, it was later observed 
u p ' 

that Joseph's testing program was performed at stress levels 
equal to or greater than 15 psi, and that the linear 
strength line was an extrapolation. Thus, the assumption of 



Table 3.3. K Triaxial Test Results 
o 



54 



TEST NO- 



s (psi) 
u 



S-OL-6-1 
S-OL-6-2 
S-OL-6-3 



0.384 
0.408 
0.321 



2.21 
2.18 
2 .76 



55 



a constant value of s /o ' initially made during this pro- 

u p J or 

ject was not sound. 



The only method of finding the shear strength at lower 
stress levels was to perform the tests at the preconsolida- 
tion pressures to be encountered in the field for direct 
measurement of the undrained shear strength. However, the 
low stress levels often associated with deposits of this 
nature made triaxial testing at these levels very difficult. 
Samples formed at pressures required to simulate field con- 
ditions would be too soft, and would not be able to support 
themselves once the top platen was placed. 

The analysis of strength gain beneath the embankment as 
a result of consolidation was found to be cumbersome. 
Approximations had to be used to determine the extent of 
consolidation beneath the embankment. More difficulty was 
encountered when attempts were made to find the extent of 
consolidation adjacent to the embankment. A large amount of 
strength gain adjacent to the embankment is a result of hor- 
izontal consolidation, and this effect is difficult to esti- 
mate . 



In addition to technical problems, the recommendation 

of K triaxial testing was also impractical on the basis of 
o or 

economics. The equipment required for K triaxial tests, 
and the production of remolded samples, is not commonly 



' 56 

found in geotechnical engineering laboratories. These 
pieces of equipment would need to be custom made, and would 
be very costly. Also, a large amount of time is involved in 
the production of samples and the testing itself. 

In view of these facts, it was felt that the use of the 
field vane shear test to determine the undrained shear 
strength was much more practical. Field vane shear tests 
provide a more expedient method of data acquisition. By 
measuring directly the values of undrained shear strength 
both beneath and adjacent to the embankment, this method 
eliminates the need for assumptions regarding the effects of 
consolidation within the foundation. The equipment required 
for such tests is readily available and inexpensive. 

Field vane shear tests were subsequently performed at 
Otterbein Lake using the procedure outlined in ASTM Standard 
Specifications D2573-72. Tests were conducted using a 2y 
inch diameter vane. The values of undrained shear strength 
obtained from vane shear testing are presented in Table 3.4. 



According to Bjerrum (1972), the results of field vane 
shear tests are dependent upon the rate of loading, soil 
anisotropy, and progressive failure. When clays are loaded, 
the shear strength is observed to increase with the rate of 
loading. A clay which is failed within a few minutes can 
exhibit values of undrained shear strength considerably 



57 



Table 3.4 Values of Undrained Shear Strength from 
Field Vane Shear Tests 



TEST NO. 



DEPTH (FT) 



SHEAR STRENGTH (PSF) 



1 
2 
3 
4 



1.5 
1.5 
1.5 
5.0 



384.0 
345.6 
371.2 
332.8 



58 



greater than the strength that would be mobilized over a 
longer period in the field. 

As a result of the rate effect, the factors of safety 
of failed embankments back-calculated from the results of 
field vane shear tests indicated that in general the vane 
test overestimates the actual strength. The amount of 
overes t imat ion increased with the plasticity index of the 
soil. In order to overcome this, Bjerrum et al. (1972) 
presented the correction illustrated in Figure 3.9. The 
results of field vane shear tests should be multiplied by 
the appropriate correction factor for use in design 
analyses . 

During this research, the results of these tests per- 
formed on the amorphous peat were not corrected. This 
material was non-plastic, resulting in a correction factor 
larger than one. It was felt that such a correction would 
be unconservat ive . 



The undrained shear strength of normally consolidated 
clays is dependent upon the direction in which shear occurs, 
indicating that the results of field vane shear tests are 
affected by soil anisotropy. However, when the results of 
triaxlal compression tests, triaxial extension tests, and 
direct shear tests were averaged and compared with the 
results of field vane shear tests, a reasonable agreement 
was observed. As a result, it was concluded that the vane 



59 



l.t 



1.0 



? 0.9 



J 0.8 



0.7 



£ 0.6 



0.5 



o 
o v 

\ S ( 





20 40 60 80 100 120 
Plasticity indei PI 



Figure 3.9. 



Loading Time-Rate Correction Factor 
versus Plasticity Index for Undrained 
Shear Strength as Measured by the 
Field Vane. From Bjerrum et al. (1972) 



60 



shear strength could be considered representative of the 
average shear strength along the slip surface. 

In some instances, soft clays possess stress-strain 
curves with a sharp peak, followed by a substantial loss of 

shear strength occurring after failure. When a clay fails, 
the peak shear strength will be mobilized simultaneously at 
all points only if the strains are uniform. However, this 
is not the ordinary response. 

The failure will initiate in the most severely stressed 
zones beneath the embankment, and gradually progress into 
the lesser stressed zones at the embankment sides. Once 
sliding occurs over the entire failure surface, the soil 
beneath the loaded area has been strained beyond the peak. 
If a strain-softening material exists, the results of field 
vane shear tests will overestimate the resistance of the 
deposit at failure. For such materials, vane shear testing 
should include measurement of the remolded strength, as 
described in ASTM Standard Specification D2573-72. 

These limitations should be considered when interpret- 
ing field vane shear tests for use in design analyses. It 
should also be noted that it was assumed that the behavior 
of amorphous peats and mucks is similar to that of soft 
clay, and that the use of field vane shear testing in the 
above manner is applicable to these materials as well. 



61 



Since vane shear test data in peats are scarce, further 
verification of this assumption is recommended. 



62R 



CHAPTER IV-PROCEDURE FOR DESIGN AND CONSTRUCTION 

This chapter provides a complete step-by-step procedure 
to accomplish the stated objective. Numerical examples are 
given in Appendix D. 

SITE EXPLORATION 



An important step in determining the behavior of 
embankments over amorphous peat or muck is the obtaining of 
reasonable soil parameters for analysis. However, before 
these parameters can be established, representative soil 
characteristics of the deposits must be determined. This is 
not an easy task when dealing with materials of this nature. 

Difficulty in finding representative characteristics of 
the deposits is the result of the variability typical of 
amorphous peats and mucks. In order to find the range of 
existing soil conditions at the site, a preliminary soil 
survey should be conducted. Disturbed samples may be taken 
at various depths using a hand auger or a power auger. Sam- 
ples obtained in this manner may be used for the determina- 
tion of water content, organic content, and specific grav- 
ity. As the compressibility and pre consol ida t ion pressure 
of the soil throughout the deposit are also needed, creep 
tests should be performed on undisturbed samples. Field 
vane shear tests should be conducted throughout the site at 
various depths to determine the undrained shear strength. 



63 

The most critical value to be measured during the site 
investigation is the undrained shear strength. The value 
that the designer selects must be conservative as a result 
of the low factors of safety used in design. For projects 
such as earth-dams, Terzaghi & Peck. (1967) recommend spacing 
borings at a maximum of 100 feet. The variability of depo- 
sits of amorphous peat and muck, requires more extensive 
testing. For the purposes of this thesis, a spacing of not 
greater than 25 feet along the embankment centerline is 
recommended. Tests should be performed near the surface, at 
mid-depth, and near the bottom of the deposit, in order to 
obtain sufficient information regarding the strength pro- 
file. 

The results of the preliminary investigation should be 
used to locate the poorest conditions at the site. If pos- 
sible, the roadway should be realigned to avoid this area. 

During site investigation, the depth of the amorphous 
peat or muck in the region of the proposed embankment must 
be determined. A procedure for estimating the thickness of 
peat deposits is provided in ASTM Standard Specification 
D4544-86. This procedure uses graduated steel rods of 9.5 ± 
1.0 mm diameter and 1.0 or 1.2 m length. The rods can be 
threaded together to allow use in deposits of any reasonable 
thickness. Testing involves pushing or driving the rod into 
the deposit until the resistance to penetration is observed 



64 

to increase sharply. This depth of increased resistance 
should be recorded as the deposit thickness. If sampling is 
desired, a piston-type sampling device as described in 
MacFarlane (1969) can be attached to the rod assembly. This 
method has a number of limitations, and the Standard Specif- 
ication should be consulted. 

As discussed in Chapter 2, the material underlying the 
amorphous peat or muck is often a soft clay or marl. This 
material can influence the behavior of the constructed 
embankment. These materials should be sampled as well to 
determine their effects on embankment behavior. It is 
advisable to continue sampling until a layer of adequate 
strength is reached. 

EMBANKMENT DESIGN 

Embankments constructed over soils of this nature can 
be designed with or without geotextiles, depending on the 
initial shear strength of the deposit. In some instances, 
geotextiles are necessary to allow construction to begin. 
As discussed in Chapter 2, geotextiles have been found to 
reduce the horizontal deformations of embankments, increase 
stability, bridge weak areas of the subsoil, and provide a 
barrier between embankment and foundation soils. This sec- 
tion will cover the design of embankments with or without 
geotextiles. Design examples for both procedures are 
included in Appendix D. 



65 
Embankments without Geotextile Reinforcement 

After the site investigation, the results of field vane 
shear tests provide a range of values of the undrained shear 
strength in the deposit. Rather than using an average value 
of the shear strength, a conservative value (in some cases 
the lowest measured value) should be used during design ana- 
lyses. The variability typical of these soils can result in 
a considerable amount of variation in shear strength, and 
the average value could be significantly greater than the 
measured lower values. 



The factor of safety used in this thesis for overall 
bearing capacity, rotational failure, and lateral squeeze is 
1.3. Attewell & Taylor (1984) state that for embankments 
constructed on a compressible foundation, a factor of safety 
on the order of 1.5 is ordinarily used during stability 
analysis. Values as low as 1.2 have been used when soil 
data and site conditions were well established. When 
analyzing stability of a preloaded embankment, Stamatopoulos 
& Kotzias (1985) state that a factor of safety in the range 
of 1.1 to 1.3 can be used, assuming that the correct input 
values have been used during analyses. Thus, although a 
value of 1.3 is used herein, when selecting a factor of 
safety, considerable judgement based on previous experience 
should be exercised. 



66 

Overall Bearing Capacity: The overall bearing capacity 
calculation is a simple one. This step is used to find an 
approximate value of the allowable height. For a strip 
loading on soils of this nature, the bearing capacity equa- 
tion reduces to 

q=cN c (4.1) 

where , 

q=ultimate pressure (psf) 

c=undrained shear strength (psf) 

N =bearing capacity factor determined from Vesic (1973) 
c 

The maximum allowable load providing a factor of safety 
of 1.3 should be calculated. Once the allowable load is 
known, the height of this load is found as 



H=- 



'all 



(4.2) 



where , 

q .^allowable pressure (psf) 

Y=unit weight of embankment soil (pcf) 

Lateral Squeeze: The weight of an embankment will 
tend to squeeze the foundation soil laterally. Jurgenson 
(1937) states that the force required to cause lateral 
squeeze of a soil between two rigid plates is equal to 



P=icBL 2 
a 



(4.3) 



67R 



whe re , 

P=total applied load (lb) 

a=one half of deposit thickness (ft) 

c=undrained shear strength (psf) 

B=length of embankment (=1 ft for unit length) 

L=one half of embankment base width (ft) 

A diagram illustrating these variables is provided in 
Figure 4.1. The total load, P, for the height of the 
embankment found in the previous step is then calculated for 
a unit length of embankment. From this, the required value 
of the undrained shear strength is 



Pa 
;req= BL 2 



(4.4) 



3V3ll 

The resulting factor of safety ( ) must be greater than 

req 

1.3. If this is not the case, the height of the first load 
may be decreased, or the geometry of the embankment can be 
adjusted to provide a longer base length . 



Embankment Spreading: The lateral earth pressure 
developed within the embankment, as shown in Figure 4.2, 
must be resisted by shearing stresses at the base. If suf- 



1. The authors' attention has been called to a "rule of 

thumb" which requires that c be greater than 1/3 of 

r e o 
the applied embankment stress. The authors are unable 

to identify the source of this rule. 



67Ra 



ficient resistance is not provided by the foundation, the 

embankment may become unstable. The lateral earth pressure, 

P , developed within a cohesionless embankment is 
a ' 



P =0.5YH 2 tan 2 (45~) 
a 2 



(4.5) 



68 




Pm 



\ 
\ 

\ 
s 
v. 






<*. « -ej^-n -» 



\ 
\ 
\ 
\ 
s 

s 



en 



c 
o 

•H 



3 

w 

c 

•H 

(0 

0) 



u 
efl 

> 



C 

o 



p. 

u 
u 
w 
<u 
Q 



n 

3 

•H 



s 
s 



69 




Figure 4.2 Embankment Spreading, 



70 



where, 

Y=unit weight of embankment soil (pcf) 

H=height of embankment (ft) 

^internal angle of friction of embankment soil 



The resistance, P , provided by the foundation soil is 



P =cl 
r 



(4.6) 
where , 

c=undrained shear strength (psf) 

l=lateral distance from crest to toe of embankment (ft) 



A factor of safety of 2 against embankment spreading is 
suggested for geotextile reinforced embankments (Fowler, 
1981) and has been adopted for unreinforced embankments as 
well. A calculated factor of safety less than this value 
will require the use of a lesser height of load. 

Rotational Failure: To investigate the stability of 
the embankment with respect to rotational failure, STABL4 
(Lovell, Sharma, & Carpenter, 1984) or STABL5 (Carpenter, 
1986) should be utilized. The stability analysis should be 
performed for the allowable embankment height found after 
the preceding analyses. If the stability analysis yields a 
factor of safety less than 1.3, another iteration should be 
performed using a lesser height of load. 



Resistance provided by the embankment material in 
unreinforced embankments may be included in the stability 



71 

analysis only if an overconsolidated or dessicated layer 
exists at the surface of the deposit. Otherwise, any 
lateral movements in the embankment can create tension 
cracks, sharply reducing the resistance within the embank- 
ment . 

Geotextile Reinforced Embankments 

If geotextiles are used during embankment construction, 
the allowable safe height of construction is increased as a 
result of the stabilizing action of the reinforcement. This 
section will discuss the design of geotextile reinforced 
embankments. The information in this section is based on a 
design manual by Christopher & Holtz (1985). The manual 
provides a more in-depth coverage of the topic, and is 
recommended reading when designing with geotextiles. 

Overall Bearing Capacity: The overall bearing capacity 
is calculated in the same manner as for the unreinforced 
embankments. Once again, the recommended factor of safety 
is 1.3. Once the allowable pressure is calculated, the safe 

height can be calculated. For geotextile reinforced embank- 

P 
ments, the average applied pressure can be estimated as -s-p, 

Z Li 

where P and L are as illustrated in Figure 4.1. 



Lateral Squeeze: Geotextiles have no influence on the 
extent of lateral squeeze. The required value of the 
undrained shear strength is therefore found in the same 



72 

manner as unreinf orced embankments. Embankments constructed 
with geotextiles require a factor of safety of 1.3 against 
lateral squeeze. 

Rotational Failure: Using STABL6 (Humphrey, 1986), a 
stability analysis should be performed for the calculated 
height of load. The value of the fabric strength required 
should be adjusted until the minimum factor of safety is 
1 .3. 

Embankment Spreading: When constructing embankments 
with geotextiles, the lateral earth pressures exerted by the 
fill are resisted by the reinforcement. If sufficient fric- 
tion is not developed between the embankment and the rein- 
forcement, or the foundation and the reinforcement, the 
embankment may become unstable. Instability may also occur 
if the foundation soils beneath the embankment can not 
resist the applied shear stress. 

These two failure modes dictate that the reinforcement 
must provide enough frictional resistance to prevent sliding 
along the interface. In addition, the tensile strength of 
the geotextile must be adequate to prevent rupture or tear- 
ing. The lateral earth pressure developed within a cohe- 
sionless embankment is given in Equation 4.5. The resisting 
force, P , provided by the geotextile is found as 



P =0.5YlHtan<}> , 
r T sf 



(4.7) 



73 



where , 

A = soil fabric friction angle 

sf 

l=lateral distance from crest to toe of embankment (ft) 
The value of 4> f is equal to 

<fi sf = tan~ 1 (4P a /YlH) (4.8) 

2 
The specified value of <t> c should be at least ■=• ( 4> . .) . 
v Y sf 3 T soil / 

A factor of safety against embankment spreading is 
found by dividing the resisting force by the actuating 
forces. A minimum factor of safety of 2 is recommended by 
Fowler (1981). 



The lateral earth pressures must be resisted by tension 
forces in the reinforcement. To prevent splitting or tear- 
ing, Fowler recommends a minimum factor of safety of 1.5. 
The resulting required fabric strength is 

T =1.5P (4.9) 

i a 

where T, equals fabric tension. 

Limit Fabric Deformation: The stresses required to 
resist lateral spreading are developed through strain in the 
geotextile. The modulus of the geotextile controls the 
amount of strain. The resulting distribution from lateral 
spreading is assumed to vary linearly from zero at the toe 
to its maximum value beneath the crest of the embankment. 



74 

This assumption is unconservat ive in view of the fact 
that a majority of geotextiles possess stress-strain curves 
that develop concave-upward, not linearly. A factor of 
safety equal to 1.5 should be used to determine the geotex- 
tile tensile modulus, E, . If the required modulus is calcu- 
lated from the tensile strength, T f , the factor of safety is 
included. The minimum geotextile tensile modulus, E , 
required is found as 

T 
E f ~~- (4.10) 

max 

where e is the maximum strain in percent expected in the 
max 

geotextile along the embankment centerline. 

Using the assumed linear strain distribution, the max- 
imum strain is two times the average strain beneath the 
embankment. A value of 5% average strain is recommended for 
design. The maximum strain would then be 10%, and the 
required fabric tensile modulus may be found as 

E f =10T f (4.11) 

The embankment will also deform until the required fabric 
strain develops to prevent a rotational stability failure. 
The actual behavior of the embankment in this condition is 
unknown, and assumptions outlined in Christopher & Holtz 
(1985) have been used. The resulting minimum required 
modulus to control a rotational failure is found as 

E fr=071T (4 ' 12) 



75 



where , 

T,. =required tensile strength of fabric 
f r 

E, =minimum fabric modulus 
f r 

STAGE LOADING 

As mentioned previously, the soft nature of amorphous 
peats and mucks often makes construction to the full height 
in one stage impossible, particularly if a surcharge is to 
be placed. Construction will therefore have to be performed 
in stages. Once the maximum first load, as calculated in 
the preceding analyses, has been applied, the foundation 
will begin to consolidate. The consolidation will result in 
a strength gain allowing further loads to be placed without 
inducing failure in the embankment foundation. 

To determine the duration of each stage load required 
for consolidation to occur, pore pressure transducers as 
shown in Figure 4.3 should be placed in the foundation. 
Once excess pore pressures induced by the previous loading 
have dissipated, no further strength gain will develop. 
Field vane shear tests should then be performed in the foun- 
dation beneath the embankment, and in areas adjacent to the 
embankment to determine the extent of the strength gain. 
Using the increased values of undrained shear strength, the 
aforementioned analyses should be performed to calculate the 
allowable height of the second stage load. This procedure 



76 



TO PORE PRESSURE 
TERMINAL 



OVERBURDEN 



POROUS 
FILTER 




3' MINIMUM 
DIAMETER 
BORING ^y 



PNEUMATIC 
TRANSDUCER 



3' SAND 



Figure 4.3 Pore Pressure Transducer Installation, 
From Slope Indicator Company. 



77 

of applying the load, allowing pore pressures to dissipate, 
measuring the increased shear strength, and placing subse- 
quent loads should continue until the final embankment 
height is reached. 

PRELOADING 

One of the problems associated with the construction of 
highway embankments over amorphous peat and muck is the 
large amount of secondary compression taking place over an 
extended period of time. To reduce the amount of settlement 
that occurs during the service life of an embankment, a sur- 
charge in excess of the final design embankment height 
should be placed. The necessary height of surcharge is 
found by first using the Gibson-Lo model to predict settle- 
ments induced by the design height of the embankment. As 
discussed in Chapter 2, the input parameters required for 
this model are obtained from the results of creep tests. In 
order to obtain the most accurate results, the creep tests 
should not be performed at conventional load increment 
ratios. Instead, they should be performed at stress levels 
simulating actual field loading. 



Creep testing begins by reconsolidating the samples at 
their preconsolidat ion pressure in the loading frame. Edil 
& Simon-Gilles (1986) recommend sustaining the load until 
deformation is reduced to 0.001 to 0.003 mm/day. At this 



78 



point, the next load is applied corresponding to the stress 
level induced by the design embankment height. The load 
should be sustained until enough data are collected to accu- 
rately calculate the values required for the Gibson-Lo 
model. For the materials tested during this project, a load 
duration of 10,000 minutes was found to be sufficient. 

Once creep tests are completed, a plot of log strain 
rate versus time, such as in Figure 4.4, should be con- 
structed. Then, using the method presented by Lo , Bozozuk, 

and Law (1976), the values of a, . , b. . , and (rO -, . are 

lab lab b lab 

found by using the values obtained from the Figure and solv- 
ing Equations 2.2 through 2.4 simultaneously. 



is discussed in Chapter 2, the values of a, . 

v ' lab 



and 



a,-., , , are approximately equal for similar stress levels, 
f ield j ^ 

The values of b, , and (r-) , , must be corrected to 
lab b lab 

corresponding field conditions. Figure 2.3 is used to find 

the value of b c , ,,. The value of (■^), J ,, can be deter- 
field b field 

mined from Figure 2.4. If the field strain rate is not 
known from previous experience, Edil & Mochtar (1984) recom- 
mend using a value two to three orders of magnitude smaller 
than that observed in the laboratory. 



It should be recognized that the recommended correla- 
tions in Figures 2.1 through 2.4 are best fit lines through 
data with a considerable amount of scatter, and thus these 



79 



a: 



u 



00 
O 




Slope = 0.434 £ 



Time ( 1 x IO 3 min) 



Figure 4.4 Plot of Log Strain Rate with Time 
from Laboratory Tests. From Lo , 
Bozozuk, and Law (1976). 



80 

correlations provide only an approximate relationship 
between laboratory and field performances. Their use can 
help improve predictions, however they still may not provide 
sufficient reliability, and they should be used with cau- 
tion. As a result, the use of laboratory test results for 
settlement prediction is still questionable. The most reli- 
able settlement predictions can be obtained by observing 
field performance for calculation of the Gibson-Lo model 
parameters . 

Using the corrected parameters, settlement prediction 
can now be conducted using the Gibson-Lo model. To facili- 
tate these calculations, two computer programs are provided 
in Appendix C. The first program, GIBSON. F, calculates the 
parameters of the Gibson-Lo model, and the second, 
PREDICT. F, provides a prediction of the strain within the 
deposit. The resulting settlement values are calculated by 
multiplying the strain values by the thickness of the depo- 
sit being analyzed. Both programs are written in FORTRAN for 
use on the IBM PC. The use of these programs for a specific 
case will be illustrated in the design examples of Appendix 
D. 



From the results of the settlement prediction, the 
amount of settlement expected within the service life of the 
embankment can be found. The objective of the surcharge is 
to induce that amount of settlement during the time required 



81 

for primary consolidation. To calculate the height of sur- 
charge required to accomplish this, the Gibson-Lo model 
should be used to predict the settlement induced by various 
heights of surcharge until the appropriate value is 
obtained. The results of creep tests simulating loading by 
the design embankment height may be used as long as the 
stress increase of the surcharge plus the embankment is less 
than twice that used during these tests, as concluded by 
Gruen (1983). Once the height of surcharge is determined, 
the preceding analyses presented regarding embankment design 
must be performed to ensure that the surcharge does not 
create any instabilities. 

FIELD OBSERVATIONS 



To aid in monitoring the behavior of the deposit of 
amorphous peat or muck when loaded, a number of field obser- 
vations should be made. The most obvious of these is a 
record of settlements along the embankment centerline. 
These measurements can be compared with the predicted set- 
tlements to check their accuracy. They can also be used to 
calculate the field strain rate of the deposit, to allow for 
correction of the rate factor for settlement prediction if 
required. Settlement measurements will also be used to 
determine when the required amount of settlement has 
occurred, allowing for removal of the surcharge. 



82 

Inclinometers should also be placed in the embankment 
site to measure any lateral movements of the embankment. A 
typical inclinometer, designed by the Swedish Geotechnlcal 
Institute, is illustrated in Figure 4.5. Data obtained from 
inclinometers should be interpreted carefully, as these soft 
materials can flow around the inclinometer. As mentioned 
previously, pore pressure transducers should be installed to 
observe the dissipation of excess pore pressures. All types 
of field instrumentation should be installed to provide 
redundancy. This will allow for any equipment that becomes 
inoperable or is disturbed during construction. 

EMBANKMENT MATERIALS 

Deposits of amorphous peat or muck are in low-lying 
areas and are very wet. Therefore, portions of the embank- 
ment will become saturated, particularly as settlement 
occurs. As a result of this, a well graded material pos- 
sessing a limited amount of fines should be chosen for con- 
struction above the water table. This will allow for 
embankment drainage and will reduce the effects of 
wetting/drying or freezing/ thawing . 

CONSTRUCTION SEQUENCE 



Barsvary et. al. (1984) present a sequence of construc- 
tion for embankments over soft subsoils. A diagram of their 
procedure is illustrated in Figure 4.6. Before actual 



83 



— Flexible eoupUnj 



Cudini ipnnt 




- Sirain gjugcjgJucd 
onlc^'ipnnj 



Figure 4.5 Inclinometer Type SGI. From 
Winterkorn & Fang (1975). 



84 




STAGE I 

1. Place working platform 

2. Place geotextile transverse to alignment 

3. Place 0.3 m granular and fold back geotextile 

4. Place and compact earth to anchor geotextile 

5. Place and compact embankment core 

STAGE II 

6. Place and compact earth to profile grade at 
edges 

7. Place and compact earth to profile grade at 
core 



Figure 4.6 



Construction Sequence. 
Barsvary et al. ( 1982) 



From 



85 



construction begins, they recommend placing a working plat- 
form on the foundation soil for construction mobility and 
easier placement of the geotextile. If geotextiles are to 
be used, they should be placed on the working platform, 
transverse to the alignment of the embankment. After plac- 
ing the embankment to a height of one foot, the geotextile 
should be folded back on top of this material as shown in 
the Figure. The geotextile should then be anchored by com- 
pacting earth above the folded region as in Step 4. The 
core of the embankment is then placed and compacted. Subse- 
quent lifts should then be constructed by placing and com- 
pacting the edges as shown in Step 6, followed by placement 
and compaction of the embankment core. Compaction lifts 
should be kept at about the same level, to aid compaction by 
lateral constraint. 



86R 



CHAPTER V-CONCLUSIONS AND RECOMMENDATIONS 

CONCLUSIONS 

This report has investigated the problems associated 
with the construction of low (± 10 ft) highway embankments 
over amorphous peats and mucks. A number of conclusions 
have been drawn as a result of this research: 

1. Based on previous work by Gruen (1983) and Joseph 
(1986), it is felt that the Gibson-Lo model is the best 
method of predicting the long-terra compression of organic 
soils . 

2. The use of relationships developed by Edil & Mochtar 
(1984) correlating the results of laboratory tests with 
field behavior will improve the results of settlement pred- 
ictions made with the Gibson-Lo model. However, these 
correlations are approximations and they should be used with 
caution. 



3. The use of K triaxial testing for the determination of 

o ° 

the undrained shear strength of a foundation beneath an 
embankment loading is unfeasible, both technically and 
e conorai cal ly . 



87R 



4. Field vane shear testing is a more practical method of 
measuring the undrained shear strength. This method allows 
for more rapid determination of shear strength, and elim- 
inates the need for assumptions regarding the extent of con- 
solidation beneath and adjacent to the embankment by making 
direct measurements. The limitations of vane shear testing, 
as discussed in Chapter 3, should be considered when inter- 
preting test results. 

5. In order to construct embankments over deposits of amor- 
phous peat or muck, stage loading will be required in most 
instances, especially when a surcharge is to be applied. 

The strength gain from consolidation will allow the place- 
ment of subsequent loads without inducing failure in the 
foundation. 

6. To reduce the amount of settlement experienced during 
the service life of the embankment, a surcharge should be 
placed to accelerate compression of the foundation. 

7. A procedure for the design and construction of low 
embankments (± 10 ft) over amorphous peats and mucks has 
been developed, and is presented in Chapter 4. 

RECOMMENDATIONS 



Based on the findings of this report, a number of 
recommendations have been made: 



88R 



1. The construction procedure outlined in this report 
should be utilized for construction of low embankments (± 10 
ft) over amorphous peat and muck. 

2. For deposits of these materials that are extremely soft, 
geotextiles may be required for successful construction. To 
supplement the information provided in this report, the 
reference by Christopher & Holtz (1985) should be consulted. 

3. A test embankment should be constructed over a deposit 
of amorphous peat or muck, including installation of pres- 
sure transducers, inclinometers, and settlement plates. The 
test embankment should be used to prove the usefulness of 
the recommended procedure for design and construction. 



4. As this procedure is implemented, the results of field 
performance should be collected for development of figures 
correlating laboratory and field behaviors similar to Fig- 
ures 2.1 through 2.4. 






LIST OF REFERENCES 



89 



LIST OF REFERENCES 



1. ASTM Standard Specification D854-83, "Standard Test 
Method for Specific Gravity of Soils," 1986 Annual Book 
of ASTM Standards, Vol. 04.08. 

2. ASTM Standard Specification D2573-72, "Standard Test 
Method for Field Vane Shear Test in Cohesive Soil", 
1986 Annual Book of ASTM Standards, Vol. 04.08. 

3. ASTM Standard Specification D2974-84, "Standard Test 
Methods for Moisture, Ash, and Organic Matter of Peat 
Materials," 1986 Annual Book of ASTM Standards, Vol. 
04.08. 

4. ASTM Standard Specification D4318-84, "Standard Test 
Methods for Liquid Limit, Plastic Limit, and Plasticity 
Index of Soils," 1986 Annual Book of ASTM Standards, 
Vol. 04.08. 

5. ASTM Standard Specification D 4544-86, "Standard Prac- 
tice for Estimating Peat Deposit Thickness," 1986 
Annual Book of ASTM Standards, Vol. 04.08. 



6. 



7. 



8. 



9. 



Attewell, P.B. and Taylor, R.K. (1984) "Ground Move- 
ments and their Effects on Structures," Surrey Univer- 
sity Press, London, 441 pp. 

Barsvary, A.K., MacLean, M.D. and Cragg, C.B.H. (1982) 
"Instrumented Case Histories of Fabric Reinforced 
Embankments over Peat Deposits," Proceedings Second 
International Conference on Geotextiles, Las Vegas, 
Vol. 3, pp. 647-652. 

Bjerrum, Laurits (1972) "Embankments on Soft Ground," 
Proceedings of ASCE Specialty Conference on Performance 
of Earth and Earth Supported Structures, Purdue Univer- 
sity, Vol. 2, pp. 1-54. 

Bjerrum, L., Clausen, C.J.F. and Duncan, J.M. (1972) 
"Earth Pressures on Flexible Structures: A State of the 
Art Report," Proceedings Fifth European Conference on 
Soil Mechanics and Foundation Engineering, Madrid, Vol. 
2, pp. 169-196. 



90 



10. Boutrup, E. and Holtz, R.D. (1983) "Fabric Reinforced 
Embankments Constructed on Weak Foundations," Final 
Report, Joint Highway Research Project, Project No. 
C-36-3 M, File 6-14-13. 

11. Campanela, R.G. and Vaid, Y.P. (1972) " A Simple K 
Triaxial Cell," Canadian Geotechnical Journal, Vol? 9, 
pp. 249-260. 

12. Carpenter, James R. (1986) "STABL5/PC STABL5 USER 
MANUAL," JHRP-86/14, Joint Highway Research Project, 
Purdue University, West Lafayette, Indiana. 

13. Christopher, B.R. and Holtz, R.D. (1985) "Geotextile 
Engineering Manual," Prepared for Federal Highway 
Administration, National Highway Institute, Washington, 
D.C. 



14. Dhowian, A.W. and Edil, T.B. (1980) "Consolidation 
Behavior of Peats," Geotechnical Testing Journal, Amer- 
ican Society for Testing and Materials, Vol. 3, No. 3, 
pp. 105-114. 

15. Edil, Tuncer B. and Mochtar, Noor E. (1984) "Prediction 
of Peat Settlement," Proceedings, 

Sedimentation/Consolidation Models, American Society of 
Civil Engineers, San Francisco California, pp. 411-424. 

16. Edil, Tuncer B. and Simon-Gilles , Dixie A. (1986) "Set- 
tlement of Embankments on Peat: Two Case Histories," 
Advances in Peatlands Engineering, National Research 
Council Canada, Ottawa, Canada, pp. 1-8. 

17. Fowler, J. (1981) "Design, Construction, and Analysis 
of Fabric-Reinforced Embankment Test Section at Pinto- 
Pass, Mobile, Alabama," Technical Report EL-81-8, USAE 
Waterways Experiment Station, Vicksburg, Mississippi, 
238 pp. 

18. Gruen, H.A. Jr. (1983) "Use of Peats as Embankment 
Foundations," MSCE Thesis, School of Civil Engineering, 
Purdue University, West Lafayette, Indiana, 149 pp. 

19. Gruen, H.A. Jr. and Lovell, C.W. (1983) "Controlling 
Movements of Embankments Over Peats and Marls," 
IN/JHRP-83/6 , Joint Highway Research Project, Purdue 
University, West Lafayette, Indiana, 180 pp. 

20. Hannon, J. (1982) "Fabrics Support Embankment Construc- 
tion over Bay Mud," Proceedings Second International 
Conference on Geotextiles, Las Vegas, Vol. 3, pp. 653- 
658. 



91 



21. Hollingshead, Garry W. and Raymond, Gerald P. (1971) 
"Prediction of Undrained Movements Caused by Embank- 
ments on Muskeg," Canadian Geotechnical Journal, Vol. 
8, pp. 23-35. 

22. Humphrey, D.N. (1986) "Design of Reinforced Embank- 
ments," Ph.D. Thesis, School of Civil Engineering, Pur- 
due University, West Lafayette, Indiana, December 1986. 

23. Hutchins, R.D. (1982) "Behaviour of Geotextiles in 
Embankment Reinforcement," Proceedings Second Interna- 
tional Conference on Geotextiles, Las Vegas, Vol. 3, 
pp. 617-619. 

24. Joseph, P.G. (1986) "Behavior of Mucks and Amorphous 
Peats as Embankment Foundations," MSCE Thesis, School 
of Civil Engineering, Purdue University, West Lafay- 
ette, IN. 

25. Jurgenson, Leo (1934) "The Shearing Resistance of 
Soils," Contributions to Soil Mechanics 1925-1940, Bos- 
ton Society of Civil Engineers, pp. 184-217. 

26. Krizek, Raymond J. and Sheeran, Donald E. (1970) 
"Slurry Preparation and Characteristics of Samples Con- 
solidated in the Slurry Cons olidome ter , " Northwestern 
University, for U.S. Army Corps of Engineers Waterways 
Experiment Station, Vicksburg, Mississippi. 

27. Lambe, T.W. and Whitman, R.V. (1979) "Soil Mechanics," 
John Wiley & Sons, New York, 553 pp. 

28. Landva, A. (1986) "In-Situ Testing of Peat," Use of In 
Situ Tests in Geotechnical Engineering, Proceedings of 
In Situ '86, Geotechnical Special Pub. No. 6, pp. 191- 
220. 



29. Lefebvre, G. et al. (1984) "Laboratory Testing and In 
Situ Behavior of Peat as Embankment Foundations," Cana- 
dian Geotechnical Journal, Vol. 21, No. 2, pp. 322-337. 

30. Lo, K.Y., Bozozuk, M., and Law, K.T. (1976) "Settlement 
Analysis of the Gloucester Test Fill," Canadian 
Geotechnical Journal, Vol. 13, No. 4, pp. 339-354. 

31. Lovell, C.W., Sharma, S.S. and Carpenter, J.R. (1984) 
"Slope Stability Analysis with STABL4," JHRP-84/19, 
Joint Highway Research Project, Purdue University, W. 
Lafayette, Indiana . 



92R 



32. MacFarlane, I.C. (1969), Muskeg Engineering Handbook, 
Muskeg Subcommittee of NRC Associate Committee on 
Geotechnical Research, University of Toronto Press, 
1969. 

33. Mitchell, W.K. and Villet, W.C.B. (1987) "Reinforcement 
of Earth Slopes and Embankments," National Cooperative 
Highway Research Program Report 290, Transportation 
Research Board, Washington, D.C., June, 323 pp. 

34. Petrik, P.M., Baslik, R. and Leitner, F. (1982) "The 
Behavior of Reinforced Embankment," Proceedings Second 
International Conference on Geotextiles, Las Vegas, 
Vol. 3, pp. 631-634. 

35. Raymond, Gerald P., (1969) "Construction Method and 
Stability of Embankments on Muskeg," Canadian Geotechn- 
ical Journal, Vol. 6, No. 1, pp. 81-96. 

36. Slope Indicator Company "Geotechnical, Geophysical, 
Groundwater and Structural Instrumentation," Seattle, 
Washington. 

37. Stamatopoulos , A.C. and Kotzias, P.C. (1985) "Soil 
Improvement by Preloading," John Wiley & Sons, New 
York, 256 pp. 

38. Tavenas, F. and Leroueil, S. (1977) "Effects of 
Stresses and Time on the Yielding of Clays," Proceed- 
ings, Ninth International Conference on Soil Mechanics 
and Foundation Engineering, Tokyo, Vol. 1, pp. 319-326. 

39. Terzaghi, K. and Peck, R.B. (1967) "Soil Mechanics in 
Engineering Practice," Second Edition, John Wiley & 
Sons, Inc. , New York. 

40. Vesic, A. A. (1973) "Analysis of Ultimate Loads of Shal- 
low Foundations," Journal of the Soil Mechanics and 
Foundation Division, Vol. 99, No. SMI, pp. 45-73. 



41. Watson, G.H., Crooks, J.H.A., Williams, R.S. and Yam, 
C.C. (1984) "Performance of Preloaded and Stage-Loaded 
Structures on Soft Soils in Trinidad," Geo te chni q ue , 
Vol. 34, No. 2, pp. 239-257. 

42. Weber, W.G., Jr. (1969) "Performance of Embankments 
Constructed over Peat," ASCE Proceedings, Vol. 95, No. 
SMI, pp. 53-76. 

43. Winterkorn, H.F. and Fang, H-Y (1975) "Foundation 
Engineering Handbook," Van Nostrand Reinhold Company, 






APPENDICES 



93 



APPENDIX A: CREEP TEST RESULTS 



94 




I 

CO 
I 

o 



(0 

H 

<u 
u 
u 



e 

•i-l 

H 

B 
u 

•H 

CO 
00 

o 

1-1 

CO 

3 
(0 

l-l 

> 



CO 
^i 

4-1 
CO 



cu 

3 

•H 



95 




o 



, 




i 

i 






l 

►J 






o 






4-1 
HO 


ft 




(U 


. o 




H, 


#4 




0. 

<u 

01 
M 
U 




r\ 


(U 


01 

. o 


Of 


E 

•r-l 

H 




i> 






3 


s 




C 


J3 




• *4 


u 




E 


•H 


■ 


\/ 




«4 
«4 


E 




1-1 




-i 






h- 


U] 


' 




3 


i 




w 








m 




c 


. O 




•H 


v4 




(0 






h 



w 



< 

01 

3 
00 



96 




«4 




I 

hJ 

o 
4J 

CO 
<U 

H 


o 




n 

o 

cu 






r\ 




o 


VI 


H 


«4 


a* 






•»* 


B 




3 


.C 




C 


4-1 
■r-l 






u 




fc 


cO 




\^ 


00 

o 


«4 


(b 


kJ 


e 


E 






»4 


m 




H 


D 
(0 
M 

cu 
> 

a 

•H 


a 
O 




CO 


«4 




u 



o 


o 


o 


o 


o 


o 


o 


10 


a 


o 


<\l 


in 


o 


T* 


aj 


' 


O 





o 


1 


1 



o 

o 

K 

o 



o 

o 
o 


a 

3 


o 
o 
o 


o 


cu 


ID 


If) 


vfl 


N 


O 


a 






r 

o 

(0 





o 
o 
o 
o 
o 
-I 



< 
cu 

M 

D 
00 

•H 

m 



uiejig 



97 




o 


o 


o 


o 


O 


O 


o 


o 


o 


o 


10 


o 


ID 


O 


in 


o 


r> 


o 


o 


^ 


in 


N 


o 


r>- 


in 


(\j 


o 


o 


m 


N 


•4 


io 


CD 


(U 


<& 


o 


a 


T 


CD 


O 


N 


»* 


<a 


o 


in 




O 


O 


•4 


r* 


01 


01 


P) 


0) 



I 

I 

►J 
O 

4J 

og 

0) 
H 

(X 
U 

01 
M 
O 



a 



e 

•H 

u 

CO 

M) 
O 

w 
3 
00 

kl 
> 






< 
U 

3 
00 



uteris 



98 




i 

I 

Pi 
.J 



H 

»•■ 

01 

u 
u 



B 
.C 

4-1 
•H 

M 

(0 

oo 
o 
►J 

(0 

D 
(0 

u 

<u 

> 



CO 



< 

01 
M 

D 
60 



ute.ng 



99 




» 






o 

94 




1 

vD 
1 

(*S 

u 

v; 


«» 




OJ 


o 




H 


«4 




• 

a 

0) 

n 

c_> 




•N 


0) 


n 


(A 


a 


W€ 


Of 


•H 




*> 


H 




3 


B 




C 


£ 




•M 


W 




E 


•rH 




\/ 


P 
fl 


V* 


Si 


60 


O 


E 

•«4 


o 




1- 


M 

> 


• 




C 


O 




•i-t 


*« 




n 



w 



< 

OJ 
M 

3 
00 
■H 

(14 



100 




to 
6> 

t» 
3 
C 

E 
\^ 

lb 

E 



i 

<3\ 

I 
OS 

►J 



(0 

<u 

H 
i 

(X 

01 
01 
M 
U 



ID 

e 



a 



u 

CO 

o 

►J 

ia 

3 
ca 
t-i 

(U 

> 



CO 

u 



M 

D 
60 



uipjis 



101 







0/ 




E 




■iH 




H 




e 




JZ 




i-l 




•H 




1-1 




(0 


• 


CO CN 


O 


1 


-J 


o> 


gg 


OS 


3 


hJ 


w 




p 


u 


0) 


Bl 


> 


<D 




H 


c 




•H 


R. 


(0 


01 


M 


o 


u 


u 


co 


u 


00 




< 




0/ 




M 




3 




oo 





o 


o 


o 


o 


o 




o 


e 


o 


e 


o 


in 


o 


» 


o 




bl 


O 


ii 


o 


o 


M 


in 


N 


o 




01 


in 


N 


o 


o 


«4 


(U 


P) 


m 




U) 


N 


CD 


8 


o 


O 


in 


Ok 


. w 




in 


CD 


v4 




o 


o 


o 


*4 




•4 


w* 


M 


(U 


o 


'" 


1 


1 


1 

uiej 


is 


1 


1 


I 


1 



102 



APPENDIX B: 



K TRIAXIAL TEST CHECK LIST 






103 



Date Test No. 

K Triaxial Test Check List 
o 

Specimen Installation 

1) Place sample on the bottom pedestal with the 

plastic former on 

2) Fasten rings on bottom pedestal 

3) Install top platen 

4) Remove the plastic former .. 

5) Connect the drainage lines to top platen.... 

6) Check position of sample 

7) Place cell wall 

8) Place top lid with piston locked at position 

leaving room for six inch sample 

9) Tighten the top rod nut 

10) Place rods and fasten top lid securely 

11) Place piston top and strain gage 

12) Place the load cell, check level and 

position of triaxial cell 

13) Lock the load cell frame 

14) Attach counterweights 



104 



Date Test No. 

K Triaxial Test Check List 
o 

Specimen Saturation 

1) Open the cell top drainage valve , 

2) Place transducers, connect drainage lines..., 

3) Pump water into the cell , 

4) Drain the cell transducer port , 

5) Close the top cell valve , 

6) Shut all valves at cell bottom , 

7) Apply cell pressure of 6 psi, pore 

pressure of 5 psi , 

8) Open cell valve 

9) Open drainage valves 

10) Flush top and bottom platens 

11) Flush cell top valve 

12) Lock Bellofram piston 

13) Unlock piston , 

14) Switch drainage lines to PWP 

15) Record burette, DCDT, and load cell readings 

16) Increase cell, pore, and Bellofram pressure 



105 



Date Test No, 

K Triaxial Test Check List 
o 

Preliminary Consolidation 

1) Record DCDT, pore and cell pressures .... 

2) Lock drainage lines 

3) Increase cell and Bellofram pressures ... 

4) Wait for pore pressure to stabilize .... 

5) Record initial readings 

6) Lock the cell 

7) Open drainage lines and start timer .... 

Axial Loading 

1) Lock Bellofram piston 

2) Adjust cell pressure 

3) Open cell line 

4) Lock drainage lines 

5) Adjust axial pressure and hand crank 

the load frame 

6) Wait for pore pressure to stabilize .... 

7) Check time 

8) Start the motor 



106 



APPENDIX C: COMPUTER PROGRAMS 



107 



c *•»»»* •**>» 

c »**»* GIBSON. F *«•»« 

c *«»»* <><«> 

C »■»»# *#***#»**-»*-»*H«-»*»*-l»-»'-« *•*»*•»»»••* »♦»**#♦•• » •»«*##***♦»*»*■»■« • «<••«*«• 

c This program was developed to calculate the parameters required 

c for the Gibson-Lo model using the m=thod developed by Lo, 

c Bozozuk, and Law (1976). 

q «♦»««•»■*•**«•»»♦•** «#«***«*«#*4* ♦ »■**#*■■** *»•»**#* »**-ih»#«ih>«»*;»#*««*-*«'<i-» 

C #-»«»-ini»-»*»#»#»###*»i»ini)i*-«*»-inn« #**#-<--nnnnnt«tnnnnnnnn« ###*«##* #»«##» 

c DEFINITION 0^ VARIABLES 

c 

c a = coefficient of primary compression -from Lo, Pozozuk and 

c Law 

c , ambda = lambda from method o* Lo. Bozozuk and Law 

c. b = coefficient of secondary corpression from Lo, Bozozuk 

c deltasig = stress increase 4 or creep test 

c esubf = last straio reading in secondary compression 

c ratef = rate factor from method of Lo, Bozozuk and Law 

c eouals 1 ^mbda/b 

c slope = line slope 'rom method of Lo. Bozozuk and Law 

c trcppt = y-intercept -from method of Lo, Bozozuk and Law 

c tsubf = time of last strain reading corresponding to esubf 

C *■**#* »»* -nnnnnni -nnnnni. » + •**•* -urn »» + >♦#♦»*■»*»«*•»■•#*■««»#»*«#»*■»#*•»##» 

c 

c t* **##♦****♦■*«■**■*»**•» *#•»****#« *■*•»■»»+»« »#*♦*♦*«#<»*«»«*♦**»*«#»♦» 

c Read in y-intercept, line slope, last strain reading, time of 

C last strain reading, and stress increase 

write<6,100> 

format </2x , 'Enter y-intercept, slope, last strain reading', 
?/2x,'time of last strain reading, and stress increase:') 
read(6,«) trcept, slope, esubf, tsubf, deltasig 

Calculate parameters for Gibson-Lo model using method developed 

by Lo. Bozozuk and Law 
»**»*•«#«##»*♦**** »«»«»»< *»♦*■»*#♦»■»#♦«»«» #*»*»■»#«#*«#*#•»»*« »•»»••** 

ratef =sl ope/0. 434 

ambda= <10.0**trcept)/deltasig 

b=ambda/ratef 

a= (esubf /del tasig) -b+ (b»e;:p (-r at ef * tsubf ) ) 

write(6,«) 'GIBSON. F OUTPUT' 

write<6,») 

wri te (6,*) 'The calculated value of a equals', a 

wri te(6,*) 'The calculated value of b equals' ,b 

wri te(6,«) 'The calculated rate factor (lambda/b) eouaJs", ratef 

stop 

end 
c *++ »***«*«**#**»■»#**•»■** «#»*•*<•»♦•»«■***»♦■*■*»#■*«*##•**** »#•»*«#■»**»«* 



100 



c 
c 
c 
c 



108 



c **«•« **•*• 

c #«*** '.^ SETTLE. F «***• 

c *•**• *«**• 

c •••*•#••*••*•*«*«••**••*•«•••*•••«***««**••*•««••*•«««•«*•*«*«•** 

c This program will make settlement predictions using the 

c Gibson-Lo model. Plots o-f the prediction analysis will also 

c be created with this program. During early portion o-f the 

c settlement prediction, short time increments are required to 

c provide a smooth curve. However, as time increases, larger 

c increments are acceptable. This program will automatical 1 / 

c increase time increments to reduce computational effort. 

C #*##«*#**•«*«**■*»**#*•»••#»•»*»*#•«•***«*•*** *«#-*«#««#**-»ii*#i«***-»-inf.»»*»* 

c DEFINITION OF VARIABLES 

c 

c a = coefficient of primary compression from Lo, Eozczuk and 

c Law 

c ambda = lambda from method of Lo, Bozozuk and Law 

c b = coefficient of secondary compression from Lo, Bozozuk 

c and Law 

c deltasig = stress increase for settlement prediction 

c dtime = initial time increment 

c ratef = rote factor from method of Lo, Bozozuk and Law 

c equals lambda/b 

c time = ti(ni of settlement prediction 

c strain = strain of deposit calculated from Gibson-Lo 

C #»*#■»*«#*«*«»**»»**-• »«■*■»♦** •**■***■»■*****»*#•««»•»•#■»****»««**•«*■»- t*r 

c 

c Dimension time and strain 

dimension time(0 : 37000), strain (37000) 

c Read in a, b, lambda/b, stress increase, ^r.d initial time 
c increment 

write(6, 100) 
100 format (b2x , ' Enter a, b, rate factor, ' 

?//2x,' stress increase for settlement prediction, ' 
?//2x f ' and initial time increment: ') 

read (6,*) a,b,rcrtef.deltasig,dtime 
c •*«*#*•**■»*«*■»»#♦-**« *«*«*« *♦«*«**»«******■»•»•■»«■«■*■»■•»■**•»■»* ••**»* *•»*#»« 
c Create plot files 
c #*«*««*««*»*«*«»*•«**»»»»»»«*■»«#»»«****###**»*«»*«»*«»«***■»*»■»■»•« 

open (7 ,f i le= ' ti me ' ) 

open (8,fi le= 'strain ' ) 
c **»*»*»*#■»«*»*»«*•»•»«*♦#«»*#*■»*#*«»«*«»«#»»#*##«**«#*»**««****«** 

c Calculate strain using Gibson-Lo model 

c #*#*»*»**«*«*»»•**#-*#»***«***««###****##♦#«**»*«*#■»*#+•*»*•»*#**•* 

time(0) =0.0 

do 10 i= 1,1 0000 

ti me (i ) =dti me+ti me (i-1 ) 

strain (i ) =-del tasi g* (a+b* < 1 . 0-exp (-ratef «ti me d ) ) ) ) 
c ««******#«#«»««**»*«#**»*♦***»♦***•+*«#***«***«****«•***«******** 
c Write output file, and store data in plot file 

wri te(« .*) ' At time eaual to ' .time (i ),' strain ecual s ' .strain (i > 



109 



write<7,#) time(i) 

write(B,*> strain(i) 

i-f (timed ) .gt. 100000. ) dtime= 10000.0 

if (timed). gt. 3000000.) dtime= 100000. 

if (timed) .gt. 10000000.) goto 20 
10 continue 
20 continue 

stoo 

end 

c «■»#**#■»*■*«•#■« *#*■*-***■*»*-»*** *»*•»■»■» »»♦♦♦♦#* »■» «■•»*•!>#»■*»#»*#■*###« *»•»»# 



1 10 



APPENDIX D: DESIGN EXAMPLES 



1 1 ] 



APPENDIX D: DESIGN EXAMPLES 

UNREINFORCED EMBANKMENT 

This example illustrates the design of the 
unreinforced embankment shown in Figure Dl . 



10' 



50'- 



y = 130 pcf, <J> = 30° ^^H] 




15* 



amor phou s peat 
Y = 68.3 pcf, s u = 330 psf 



30' 



soft clay 
Y = 1 10 pcf, s u = 330 psf 



Figure Dl Embankment Configuration for Design Example. 

Design Procedure 

1. Overall Bearing Capacity: 

q . =cN 
ult c 

N c =5.14 (pg.112, Das 1984) 

q ult =(330 psf)(5.14) 
=1696.2 psf 



1 12 



1696.2 
all 1.3 

=1304.8 psf 



Find allowable height: 



M all 

Y 
1304. 8psf 

130pcf 



10.04 ft 



2. Lateral Squeeze; 

P=icBL 2 
a 



P=wt. of unit length of embankment 
-j(9 0+50)(10ft)(130pcf)(lft) 



=91000 lb 






Find required shear strength to prevent lateral squeeze 

9 1000= T ^ T c(l ft) (45ft) 2 

c =337.0 psf 
req 



Calculate factor of safety 
c 



F.S.=- 



' a vail 

* req 
330 



337 

=0.98 < 1.3 NG 



At this point, there are two options. Either reduce 
the height of the first load, or decrease the embankment 
slope to widen the base. Try changing slope to 1:4 as 
shown in Figure D2 . 



I 13 



10' 



-50' 



Y = 130 pcf, <}) = 30° 




15' 



amorphous peat 
y = 68.3 pcf, s u = 330 psf 



30' 



soft clay 
Y = 1 10 pcf, s u = 330 psf 



Figure D2 Revised Embankment Configuration 
for Design Example. 



P=y(50+130) ( 10ft) (130pcf)( lft) 



=117,000 lb 



1 17,000 = (y^-)c(lf t)(65ft) 2 

c =207.7 psf 
req 



F.S.=- 



aval 1 

c 
req 

330 



207.7 
=1.59 > 1.3 OK 

3 . Embankment Spreading 



The most crucial location for lateral spreading is 
at the crest of the embankment, as long as the slope is 
less than 1:1. At the crest, the lateral earth pressure 
is equal to the maximum value, yet the resistance to 
sliding is at a minimum. 



1 14 



Calculate lateral earth forces 

1 u 2_ 2 ,, c 3CK 
P„=rTfH tan (45-tt- ) 
a I I 

1 2 2 30 

=Y(130)(10) tan (45-jp) 

=2167 lb 



Calculate resistive forces 



P =cL 

r 



'33O(4O)=13200 lb 



Calculate factor of safety: 
P 

a 

13200 



2167 
=6.1 > 2.0 OK 

4. Stability Analysis 

Perform stability analysis using STABL4 or STABL5 . 
The input used is presented below, and the 
resulting output is illustrated in Figure D3 . 
The calculated minimum factor of safety against 
rotational failure using the modified Bishop method 
is equal to 1.64 > 1.3 OK. 



1 15 



prof i 1 

embankment stability 

3 3 

0. 40. 100. O 40. 2 

100. 40. 140. 50. 1 

140. 50. 165. 50. 1 

100. 40. 165. 40. 2 

0. 25. 165. 25. O 3 

soil 

3 

130. 130. 0. 30. 0. 0. 1 

68. 3 68. 3 330. 0. O 0. 0. 1 

110.0 110.0 330.0 0.00.00.0 1 

water 

1 62. 4 

2 

0. 40. 

165. 40. 

circl2 

5 30 60. 95. 110. 155. 0. 2. 5 0. 0. 



5. Settlement Prediction 

From the results of Creep Test OL-9-2, a plot 
of log strain rate versus time is constructed, 
as shown in Figure D4 . The y-intercept and 
line slope are found as indicated on the Figure 
These values are then used in GIBSON. F for 
calculation of the parameters required for the 
Gibson-Lo model. 



GIBSON. F Input: 

-4.95 0.000194 0.333557 2900.0 8.3 



1 16 





T> 




(b 




t-> 




<t> 




i. 


<-> 


c in 


D 


Si <^ 


a 


cr ^ 


<-> 


W r^ 


J 


ID 

U II 





ro 




«♦- j 


_i 




PQ 




<r 


V (/) 


h- 


<4- 


Cfl 


- 




T> 




U L. 


L. 


•H 


O 


4-> •-> 

■-< u 




L <U 


h- 


U L. 


o 


*> E 


_l 


Ul D 
£ 


Q_ 


£ n 




c 




o ••* 




1-1 E 




VI 

■ri 

X 

1] 



c 

a 

c 

XI 

a 
u 



o 
o 



o 

U-l 



3 
Q. 
tJ 
3 
O 

PQ 
< 
H 





o 



rl 

Q 

<D 
u 
3 
00 
•H 



sixe-h 



1 17 



o 
o 
o 
m 



o 
o 
o 



o 
o 
o 
n 



c 

•H 

S 



S 

•H 

H 



O 
O 

o 



o 
o 
o 




0) 

a 



3 

0] 

<U 
> 



PS 

n 
(d 

M 
4J 

CO 

• 
CM 
sz I 

4-> CT\ 

•H I 

U ,J 

(0 o 

60 

O 4J 

►J to 

<D 

u-i H 

O 

U <U 

o <u 



Q 
<U 

u 

3 
60 
• i-t 



ajBy uiBjqs 3oq 



1 18 



GIBSON. F OUTPUT 

The calculated value of a equals 3.79906e-02 
The calculated value of b equals 3.02420e-03 
The calculated rate factor (larabda/b) equals 4.47005e-04 



These parameters must now be corrected for field 
conditions: 

For this embankment, Ao=9.0 psi. 
From Fi gure 2.3, 



field 
'lab 



= 8.8 



b CJ . =0.0266 
field 



Using (f) i i_ » find the average 
b lab 

laboratory strain rate from Figure 2.4 



Laboratory strain rate=2.8x!0 



-5 



Edil & Mochtar (1984) recommend assuming a field 
strain rate two to three orders of magnitude smaller 
than the laboratory strain rate If field values are 
not known from previous experience. 



Try using field strain rate= 



2.8x10 



-5 



= 1x10 



280 
-7 



From Figure 2 .4 , 



( F>field =2xl ° 



-6 



1 19 



Now, using SETTLE. F make settlement predictions, 

SETTLE. F Input: 

0.03799 0.0266 0.000002 9.0 1000. 

From SETTLE. F Output, ultimate strain is equal 
to 0.58. A plot of strain versus time is shown 
in Figure D5 . 



6 . Surcharge 



Find maximum height of the second load. In order 
to accomplish this, a new round of field vane 
shear tests should be performed to find the 
strength gain beneath, and adjacent to the 
embankment . 

For purposes of this example, assume 

s beneath embankment = 500 psf 
u r 

s adjacent to embankment = 400 psf 
u J r 

Stability analyses as illustrated in Steps 1 through 4 
must now be performed to calculate the safe height 
of the surcharge . 

Overall Bearing Capacity: 



q . =cN , where c=average s 
u 1 1 c u 



120 




o 

o 
o 
o 

o 



o 
o 
r> 

fv 
m 
o 



o 
o 
o 
in 



o 

o 


o 
o 


o 

o 


in 


o 


in 


(U 


o 


N 


SD 


m 


o 


w 


m 


3- 




uiejis 





o 
o 
o. 
r> 
m 
10 



o 
o 

4 
I 



o 

o 
o 
o 
o 

^ 



c 
<u 

B 

a: 

c 

01 

B 
W 



o 

o 



o 

c 
o 



T3 
0L, 



c 
<u 

B 

01 






Q 

u 

a 



121 



400+500 .__ , 
c = : = 4 5 p s f 



q , =450(5.14) 
u It 



=2313.0 psf 



2313.0 
'all" 1.3 

=1779.2 psf 



H= 1 130 * 2 =13 - 7 ft (t °P width=20.4 ft) 



Lateral Squeeze: 



P=y( 130+20.4 )(13.7f t)(130pcf ) 
=133,931 lb 



13 3,9 31 = (rp-r)c( lft) (65ft) 2 

c =237.7 psf 
req 



FS = 



avail 



c 
req 

450 



237.7 



=1.89 > 1.3 OK 



Embankment Spreading: 



P 4yH 2 tan 2 (45-|) 
a 2 2 



=j(130)(13.7ft) 2 tan 2 (45-|^) 
=4067 lb 



P =cL 

r 



=450(54.8) 



122 



=24660 lb 



FS =|A fi 60 =6 . 1 > 2.0 OK 
4 067 



Stability Analysis: 

The input used is shown below, and the resulting output 
is shown in Figure D6 . 
FS= 1 .9 1 > 1.3 OK 

prof i 1 

embankment stability 

S 3 

0. 40. 100. 40. O 4 

100. 40. 154. 8 53. 7 1 

154. 8 53. 7 165 53. 7 1 

100. 40. 127. 4 40. O 4 

127. 4 40. 165. 40. 2 

0. 25. 126. 4 25. O 3 

126. 4 25. O 127. 4 40. 2 

126. 4 25. 165. 25. 3 

soil 

4 

130.0 130.0 0.0 30.0 0.00.01 

68. 3 68. 3 500. 0. 0. 1 

110.0 110.0 330.0 0.00.00.0 1 

68. 3 68. 3 400. 0. 0. O. 1 

mater 

1 62. 4 

2 

O. 40. 

165. 40. 

c ire 12 

5 30 60. 95 110. 155. 0. 2. 5 0. 0. 



Prediction Analysis 



According to Gruen & Lovell (1983), the parameters 
of the Gibson-Lo model are valid for stress levels 
less than twice that used during testing for 



123 



3 

, a 
^> 
3 


_J 

<E 

co 

O 

h- 

O 

_l 
CL 









"c wV 






Q) \ 1 






♦* \ 






m \ 






i. \ 






<b \ 






C CO \ 






Q) ^i \ 






CT O \ 






in ^-i \ 




1 


Q) \ 






U II \ 






10 \ 






t- J ' 






i. t^ 






D Oj 
1/1 <*- 

1] < 


f)l 


' 


v- en 


/ 







' I 




<4- 


/ 1 




r-. 


/ 




<T> 


V 




U k. 


/ 




•^ 


/ 




4-> fc) 






•-< u 






L t) 






U L. 






♦> e 






in 3 






€ 






£ M 






c 






o •*< 






*-• "Z. 




■ 


,,.,.. r . 1 




1 ' — f 





B 




0) 




B 


o 


.» 


3 


C 


•4 


CO 




ja 




e 




u 


O 

3- 


T3 



o 
■(u 



X 

I 

x 



01 

00 

M 

(0 
JC 

O 

M 

3 
C/> 

u 
o 



3 
P. 
•u 

3 
O 

cq 
<! 
H 

CO 

<4-l 

o 
o 

l-l 
a, 



vo 
o 

<u 

M 

3 
60 



o 
o 



o 
CO 



o 
•■0 



o 

T 



o 
CO 



stxe-h 



124 



initial determination of these parameters. 

Ao=12.37 psi < 2x8.3 psi OK 

A settlement prediction is now performed using SETTLE. F. 

SETTLE. F Input: 

0.03799 0.0266 0.000002 12.37 1000. 

The resulting settlement prediction is illustrated 
in Figure D7 . From this Figure, it is observed that 
the strain occurring during the service life of the 
10 foot high embankment will occur in approximately 
200,000 minutes, or 4.6 months. Therefore, after 
the surcharge is applied for approximately 5 months, 
it can be removed, and settlements will be minimal. 

GEOTEXTILE REINFORCED EMBANKMENT 

This example will illustrate the design of the embankment in 
the first example when geotextiles are to be used at the 
base . 

Design Procedure 

1. Overall Bearing Capacity: 



As a result of the geotextile, the pressure resulting 
from the embankment can be calculated as the total 
load, P, over the length of the embankment, 2L. Check 






125 




in 
a> 

■>> 

3 
C 

E 
\^ 

Hi 

E 



c 

0) 

e 
x 
c 
m 

.O 

B 

U3 
•V 

m 

ct) 

x: 
u 

M 

3 
w 

M 

O 



C 

o 



u 

•H 

u 

Pu 



C 
CU 

e 

01 



HI 
CO 



Q 

cu 
u 

3 

ao 



uipjis 



126 



to see if the entire load, including the preload, can 
be applied in one stage. 



q , =1696.2 psf 
M ult 



app (L)(l) 



P=133,931 lb 



133,931 
app " 130 



=1030 psf 



FS> 



1696.2 
1030.2 



=1.65 > 1.3 OK 



2. Lateral Squeeze: 



c =237.7 psf 
req 



c , =330.0 psf 
avail 



FS = 



330.0 
237.7 



1.39 > 1.30 OK 



3. Embankment Spreading: 



P =4067 lb 
a 



P r = I YLHtan<D sf 



■y (130) ( 54.8) ( 13.7 )tan<f> gf 



1 4P 
<t = tan ( — -LH) 
s f y 



= tan [ 



(4)(4067) 



( 130) (54.8) ( 13.7 ) 



] 



= 9.46° < -f, 



Specify <(> ..=20 

' sf 



P r =y(130) (54.8 )( 13.7 )tan20° 



127 



FS 



=17762 lb 

17762 
'4067 



■4.4 > 2.0 OK 



4. Stability Analysis 



Perform the stability analysis using STABL6 . The 
input used is presented below, as well as a list of 
points defining the most critical failure surface. 
The calculated minimum factor of safety against 
rotational failure without a geotextile is 1.31. 
Therefore, the geotextile will not be necessary to 
resist rotational failure. 



128 



PROFIl. 

REINFORCED EMBANKMENT STABILITY 

wJ ■-• 

. 40 . 1 00 . 4 . 2 
1 00 . 40 .0 1 54 . 8 53.7 1 
154. B 53.7 176.0 53.7 1 
1 00 -. 40 . 1 76 . 40 . 2 
0.0 25.0 17.r>.0 25.0 3 
SOIL 

1 "~0 . i 30 . , C 30 . O . . 1 
68.3 68.3 330.0 0.0 0.0 0.0 1 
1 1 . O i 1 . 330 . . . . 1 
WATER 
1 62. 4 

0.0 40.0 
176.0 40.0 
RE INF 

1 

100.0 40.0 0.0 0.0 

156.0 40 . 00 . . 

176.0 40 . = . 

CIRCl.2 

5 30 60.0 95.0 110.0 165.0 0.0 2.5 0.0 0.0 

EXECUT 



5. Find Required Fabric Strength: 

Since the geotextile is not required to resist 
rotational failure, the required fabric strength 
is controlled by the forces developed in the fabric 
as a result of embankment spreading. 



T =1.5P 
f a 



129 



Following Are Displayed The Ten Most Critical Of The Trial 
Failure Surfaces Examined. They Are Ordered - Most Critical 
First. 



♦ * Safety Factors Are Calculated By The Modified Bishop Method * * 
Failure Surface Specified By 35 Coordinate Paints 



Paint 


X-Surf 


Y-Surf 


No. 


(ft) 


(ft) 


1 


95.00 


40.00 


2 


96.78 


38.24 


7, 


9B.66 


36.59 


4 


100.63 


35.06 


5 


102.69 


33.64 


6 


1 04 . B3 


•rr> 35 


7 


107.04 


31. 19 


B 


1 09 . 32 


30. 16 


9 


1 11 . 66 


29.26 


10 


114.04 


2B.51 


11 


116.46 


27.90 


12 


118.92 


27.43 


13 


121.40 


27. 10 


14 


123.89 


26 . 93 


15 


126.39 


26.90 


16 


128.89 


27.02 


17 


131. 38 


27.29 


IB 


133.84 


27.70 


19 


136.28 


28.26 


20 


138.68 


28.96 


O 1 


141.03 


29.80 


1*? 


143.33 


30.78 


97, 


145.57 


31.90 


24 


147.74 


33. 14 


25 


149. B3 


34.51 


26 


151.84 


36 . 00 


27 


1 c-y 7c 


37.61 


2B 


1 JJl uif 


39.33 


29 


157.28 


41, 15 


30 


I 58. 88 


43.07 


31 


160.37 


45.08 


"T'y 


161.73 


47. 17 


y.y, 


162.97 


49.34 


34 


164.08 


51.59 


7C 


164.97 


53.70 



Circle Center At X - 125.6 ; Y = 69.2 and Radius, 42. 



1.310 



130 



1.5(4067) 



=6100 lb 



6. Find Required Geotextile Tensile Modulus: 



E =(6100)(10) 



=61,000 psf 



7. Settlement Prediction: 



The use of a geotextile will not affect the total 
settlements experienced beneath the embankment. 
Therefore, the prediction made in the previous example 
for the surcharged height of embankment is still 
valid . 



131 



APPENDIX E. 



NEGATIVE NUMBERS FOR CONTACT PRINTS 



132 



APPENDIX E: NEGATIVE NUMBERS FOR CONTACT PRINTS 



Figure 


Ne 


g 


ative 


Number 


3.1 






1 


* 




3.2 






25 


* 




3.6 






9 


* 




3.7 






14 


* 




3.8 






26 







Location of Negative 
Stewart Center, Room 65 
Stewart Center, Room 65 
Stewart Center, Room 65 
Stewart Center, Room 65 
Grissom Hall, Room 140 



* Information required for retrieval: 

3/30/87 Civil Engineering-Tim Crowl-Equipment 35 



z I 
DC I 

ol