UNDRAINED CREEP BEHAVIOR

OF ATCHAFALAYA CLAY

FROM CU DIRECT-SIMPLE SHEAR TESTS

by

CHARLES E. WILLIAMS

S.B., Massachusetts Institute of Technology

(1971)

Submitted in partial fulfillment

of the requirements for the degrees of

Master of Science and Civil Engineer

at the

Massachusetts Institute of Technology

May 1973K : R--

Signature of Author.. ...-...Department of Ciril Engineering,

May 11, 1973

Certified by........... ... '.. . .. . .. .-.-................Thesis Supervisor

Accepted by.Chairman, DepartmEAital Committee on GraduateStudents of the Department of Civil Engineering

Archives

(JUN 29 197)

ABSTRACT

TITLE: UNDRAINED CREEP BEHAVIOR OF ATCHAFALAYACLAY FROM CU DIRECT-SIMPLE SHEAR TESTS

by

CHARLES E. WILLIAMS

Submitted to the Department of Civil Engineering onMay 11, 1973 in partial fulfillment of the requirementsfor the degrees of Master of Science and Civil Engineer.

The creep behavior of normally consolidated and overconsolidated samples of a highly plastic deltaic clayfrom the Atchafalaya Basin in Louisiana is investigatedin consolidated-undrained direct-simple shear tests.From 49 to 95 percent of the undrained shear strengthwas applied over a typical period of one week in thestress controlled tests. The results of "single incre--mental" tests are compared to those in which the samplecrept under several successive increments.

The rate process theory and other rheologic models forcreep are reviewed and both undrained and drained creepdata are summarized in light of the various theories,including those for creep rupture.

For initial shear stress increments below 85 percent offailure, the creep behavior of the clay follows theSingh-Mitchell rate process model after an initialtransient pore pressure condition. For additional in-crements, the creep strain rate is initially very slow,but it changes with log time to eventually converge withthe Singh-Mitchell model.

The creep rupture models of Saito and Singh-Mitchell werecompared to laboratory creep rupture behavior and foundreasonably acceptable for this clay. Creep strengthsin terms of applied shear stress for a given consolidationstress were found to be 85 percent of that obtained from

-2-

conventional controlled strain CK UDSS tests. This valueowould probably be lower if the creep process was allowed

to continue for longer periods of time (greater than oneweek) to develop sufficient pore pressures for failure.

Thesis Supervisor:

Title:

Charles C. Ladd

Professor of Civil Engineering

-3-

ACKNOWLEDGEMENTS

The author expresses his sincere gratitude to the

following:

Professor Charles C. Ladd, his thesis supervisor, whose

guidance and thought provoking comments made this investi-

gation possible;

Dr. Lewis Edgers, who suggested the topic and aided the

author in all phases of the investigation;

Ms. Patricia Potter, who typed the manuscript;

The U.S. Army Corps of Engineers, Waterways Experiment

Station, Vicksburg, Mississippi, whose financial support

made this investigation possible.

-4-

TABLE OF CONTENTS

Page

Title Page 1

Abstract 2

Acknowledgements 4

Table of Contents 5

List of Tables 8

List of Figures 9

I INTRODUCTION 13

1.1 Background 13

1.2 Descriptionof EABPL Foundation Clays 15

1.3 Scope of Research 16

II REVIEW OF CURRENT LITERATURE 22

2.1 Rate Process Creep Theory Development 22

2.2 Creep Rupture 35

2.3 Creep Strength 38

2.4 Conclusions 41

III DESCRIPTION OF EQUIPMENT AND PROCEDURES 55

3.1 Introduction 55

3.2 Shear Apparatus 56

3.3 Sample Preparation 57

3.4 Testing Procedure 57

-5-

TABLE OF CONTENTS (Cont.)

Page

IV EXPERIMENTAL INVESTIGATION 60

4.1 Soil Tested 60

4.2 Test Program 61

4.3 Test Results 62

4.3.1 Controlled Strain Tests 62

4.3.2 Controlled Stress Creep Tests 63

4.3.2.1 Initial stress levels 63

4.3.2.2 Additional Shear in 64Increments

4.3.2.3 Creep Failure 64

V DISCUSSION OF TEST RESULTS 77

5.1 First Load Increments 77

5.1.1 Normally Consolidated Samples 77

5.1.2 Over Consolidated Samples 81

5.2 Additional Load Increments 85

5.3 Creep Rupture 86

5.3.1 Singh-Mitchell Creep Rupture 86Relation

5.3.2 Saito's Creep Rupture 88Relation

5.3.3 Strength Decrease Due to Creep 90

5.4 Comparison of Simple Shear Stress 91Levels with Triaxial Stress Levels

5.5 Use of Multi-Increment Testing 93

VI CONCLUSIONS AND RECOMMENDATIONS 123

6.1 Apparatus Performance 123

6.2 Test Results 124

6.3 Recommendations 128

-6-

TABLE OF CONTENTS (Cont.)

Page

VII REFERENCES 130

Appendix A

Appendix B

Appendix C

List of Symbols

Tabulated Test Results

Plots of CK UDSS Creep TestsO

-7-

133

138

164

LIST OF TABLES

TableNo. Title Page

4-1 Summary of Samples for Test Program 65

4-2 Summary of CK U Direct-Simple Shear 66Tests on EABPE Clay

4-3 CK UDSS (Creep) Test Results - EABPL 67Clay

4-4 Description of Measured Test Parameters 69

5-1 Application of Singh-Mitchell Creep 97Rupture Theory of EABPL Test Data

5-2 Comparison of Laboratory CK UDSS Creep- 98Rupture Behavior to Saito's Creep-Rupture Relationship

5-3 Strength Decrease Due to Creep 100

5-4 Relationships Between D, D , D and 101Rotation of Principal Planes in CK DSSTests

5-5 Computations for Theoretical Strain-Log 102Time Curves

5-6 Duplication of D = 65% Stress Level 103T

-8-

LIST OF FIGURES

FigureNo. Title Page

1-1 Pre-Construction Effective Vertical 18Stress Profiles, Test Sections II & III

1-2 Estimated Maximum Past Pressure, Sections 19II and III, 180 ft. Offsetts to LS and FS

1-3 Average Unconfined and Field Vane Strength 20Data From Test Sections II & III

1-4 Average Natural Water Content Profiles 21for 105 ft. and 180 ft. Offsets forSections II & III

2-1 Derivation of Berkley Three Parameter 42Relationship Based on Rate Process Theory

2-2 Variation of Strain Rate with Deviator 43Stress from CIUC Creep Tests on RemoldedIllite

2-3 Influence of Time on Strain Rate During 44Creep of Remolded Illite

2-4 Rheologic Model for Christensen and Wu 45(1964) Rate Process Development

2-5 Rheologic Model for Murayama and Shibata 46(1961) Rate Process Theory

2-6 Stress Controlled CIUC Tests on N.C. 47Remolded Clay

2-7 Barden's Rheologic Model 48

2-8 Drained Triaxial Creep Results on London 49Clay

2-9 Drained Triaxial Creep Results on 50Pancone Clay

LIST OF FIGURES (Cont.)

FigureNo. Title Page

2-10 Demonstration of the Saito Creep 51Rupture Relationship

2-11 Singh-Mitchell Creep Rupture Concept 52

2-12 Influence of Consolidation Conditions 53on the Pore Pressure Generation ofTwo Clays

2-13 Typical Stress Path to Failure for an 54Undrained Creep Test on NC Clay

3-1 General Principle of Direct-Simple- 59Shear Apparatus - Model 4

4-1 Range in Plasticity Characteristics of 70Atchafalaya Backswamp Materials

4-2 CK UDSS Tests on Normally Consolidated 71EA9PL Clay

4-3 Stress vs. Strain from CK UDSS Tests on 72EABPL Clay o

4-4 Stress Paths from CK UDSS Tests on Normally 73Consolidated EABPL Cday

4-5 Normalized Stress Paths from CK UDSS Tests 74on EABPL Clay o

4-6 Compression Curves, CK UDSS Tests, EABPL 75Clay

4-7 S /5 vs. OCR from CK U Direct-Simple 76SRearCTests on EABPL Cday

5-1 Shear Strain Vs. Log Time from CK UDSS 104Creep Tests on N.C. EABPL Clay o

5-2 Log Strain Rate vs. Log Time from CK UDSS 105Creep Tests on N.C. EABPL Clay 0

-10-

LIST OF FIGURES (Cont.)

FigureNo. Title Page

5-3 Log (Strain Rate X Time) vs. Log 106Time from CK UDSS Creep Tests onN.C. EABPL Cday

5-4 Log Shear Strain Rate vs. Applied 107Stress Level from CK UDSS Tests onN.C. EABPL Clay

5-5 Shear Strain vs. Log Time from 108CK UDSS Creep Tests on O.C.EAPL Clay

5-6 Log Shear Strain Rate vs. Log Time 109from CK UDSS Creep Tests on O.C.EABPL Clay

5-7 Log (Strain Rate X Time) vs. Log Time 110from CK UDSS Creep Tests on O.C.EABPL Cday

5-8 Log Strain Rate vs. Applied Stress 111Level from CK UDSS Creep Tests onO.C. EABPL ClRy

5-9 Comparison of Induced Pore Pressures 112for Various Creep Stress Levels

5-10 Comparison of Plots of Log Strain 113Rate vs. Log Time for N.C. and O.C.Samples of EABPL Clay

5-11 Comparison of Stress Paths from 114CK UDSS Creep Tests on N.C. andO.e. EABPL Clay

5-12 Comparison of Pore Pressure 115Generation During Creep of N.C.and O.C. Samples of EABPL Clayat Various Stress Levels

-11-

LIST OF FIGURES (Cont.)

FigureNo. Title Page

5-13 Variation of Creep Parameter m 116with Stress Level from CK UDSSCreep Tests 0

5-14 Log Shear Strain Rate vs. Log Time 117During Later Increments from CK UDSSCreep Tests on N.C. EABPL Clay o

5-15 Log it vs. Log Time from CK UDSS 118Creep Tests on EABPL Clay

5-16 Plot of Log Strain Rate vs. Log Time 119for Steady State Creep at Various StressLevels of EABPL Clay

5-17 Initial Shear Strain (at t = 1 minute) 120vs. Applied Stress Level

5-18 Stress Conditions in the Direct Simple 121Shear Device - Assumed Pure Shear

5-19 Comparison of Laboratory and 122Theoretical Creep Curves at VariousStress Levels

-12-

I INTRODUCTION

1.1 BACKGROUND

For the past fifty years a steadily increasing volume of

water has been diverted from the Mississippi River into the

Atchafalaya distributary. This is accomplished by means of

a flood levee system within the Atchafalaya basin. Progressive

siltation of the southern half of the basin along with the

increased capacity requirements for the distributary have made

it necessary to raise the flowline within the Atchafalaya

floodway by means of levees.

Levee construction within the southern portion of the

Atchafalaya Basin Floodway in south-central Louisiana has been

in progress for more than thirty years (Kaufman and Weaver,

1967). Construction has proceeded in stages so that the thick

deposits of soft alluvial and deltaic foundation clays can

consolidate and achieve sufficient strength to support subse-

quent lifts of fill. This procedure has proved to be very

time consuming and unsatisfactory. Moreover, settlements in

excess of 10 to 15 feet have made it impossible to attain

design heights (20 to 25 feet above the original grade) along

-13-

large portions of the floodway. Large lateral movements,

thought to be caused by undrained shear deformations, yielding

and creep of the soft foundation clays, have been a major

contributor to the excessive vertical settlements of the levee

crown.

In 1964-65 three extensively instrumented test sections

were constructed along the East Atchafalaya Basin Protection

Levee (EABPL) (Kaufman and Weaver, 1967; USCE, 1968). Test

Section I, with a four foot crown and a height of six feet,

had a design factor of safety of 1.1 based on a 4 = 0 analysis

using UU triaxial compression test data. Test Sections II

and III, with a wider crown, a height of ten feet, and wide

stabilizing berms, had factors of safety of 1.1 and 1.3 based

on similar total stress analyses.

Despite very large berms, both Test Sections II and III

experienced very large lateral deformations with a resulting

drop in the crest elevation, and for all practical purposes

failed. These lateral shear deformations, which continued

long after the end of construction, were particularly large

between elevations -20 to -40 feet (original ground surface

El. = 3-4 ft MSL). These large deformations are considered

to be caused by undrained shear, which is the subject of sep-

arate investigations, and undrained creep.

-14-

Considering the fact that a substantial portion of the

zone of large deformations (El. -20 to -40 ft) is approximated

by a simple shear stress system, any detailed undrained creep

investigation should attempt to duplicate this stress system.

The subject of this report is undrained creep behavior of

EABPL foundation clays with the direct-simple shear stress

system.

1.2 DESCRIPTION OF EABPL FOUNDATION CLAYS

The site of the EABPL construction is composed of sand

and gravels overlain by approximately 120 feet of clay which

can be divided into three categories (Krinitzsky and Smith,

1969).

(1) Poorly drained backswamp deposit (El. -25 to +2 ft)

Presence of considerable organic matter with

several inches to several feet of peat; dark gray

to black in color; concretionary matter includes

carbonates, sulphides and vivianite; soil is typically

CH clay with WL = 110 + 30.

(2) Well drained backswamp deposit (El. -120 to -25 ft)

Light to dark brown in color due to oxidation;

stratification is absent due to reworking by plant

roots; fissures due to dessication are common; soil

is typically a CH clay with WL = 90 + 30.

-15-

(3) Lake deposits

Deposited in shallow fresh water; gray in color

mostly highly plastic CH clays with some silt and

sand (WL = 95 + 25); well defined layering with occa-

sional silt and fine sand layers; presence of shells

and carbonate concretions; well developed fractures

and slickensides due to shear failures during deposi-

tion.

The area, which was originally level at El. +2 ft prior

to levee construction, has the net overburden stress profile

shown in Fig. 1-1 which was obtained from total weight and

pore pressure measurements. Maximum past pressure profiles

for offsets from the levee centerline indicate overconsolidated

layers at elevations -45 to -65 ft and -70 to -100 ft (Fig. 1-2).

These findings are also supported by field vane strengths and

water content variations at these depths as shown in Figs.

1-3 and 1-4. These layers are probably overconsolidated due

to drying of exposed backswamp deposits.

1.3 SCOPE OF RESEARCH

The experimental program involved obtaining representative

samples of the clay in the zone of greatest levee foundation

-16-

deformations (El. -20 to -40 ft). A series of standard con-

trolled strain Consolidated-Undrained Direct Simple Shear

(CKfoU DSS) tests were conducted in order to establish a

normalized shear strength versus overconsolidation ratio

relationship. Subsequently, a program of constant load un-

drained creep tests were carried out with the Direct Simple

Shear device in order to establish strain and strength rela-

tionships with stress, time and OCR.

The Singh-Mitchell Theory (see 2.1a and 2.2b) for creep

strains and creep failure and Saito's creep rupture theory

(see 2.2a) will be applied to the test results in order to

evaluate their validity with the EABPL clay.

-17-

HHH

HH

LU.V)

lbU)U)wI--

h0

L&J/C.)wUwLLLLw

l-

a:w

U)

z

(7S1W) 1I' NOILVA313

Figure 1-1-18-

0S -I-)

o-plL-- U)

0

P4

o

--I4-

0L U)

IUUP4 -rzl

-I)

0vo and n, TSF+Ic

C

-IC

-3C(

2.0

S

* 0s

O

S(

C----Estimated om

IJ

--/

0FS

0

S0

* Corps Data0 MIT Data from 300 ft offset

ESTIMATED MAXIMUM PASTAND III , 180 FT OFFSETS

PRESSURE,SECTIONS IITO LS AND FS

Figure 1-2

0O AAA

I .-50

-70

Ave105ft

-90

-110

to LS

v uI v

-19-

---- f

1 r

r 30

AA

I

• IIII

-

LL.C)

I-

I,-U)

w

C_1

WIJ,04d

w,4~

LA.U)

ILl

C0a.wLL.z0z

w4

(7SIN) 4; 'NOI1IVA313

Figure 1-3-20-

WN ,%

I-

ww

AVERAGEPROFILES

NATURAL WATER CONTENTFOR 105 FT 8 180 FT OFFSETS

FOR SECTIONS II 8 III

Figure 1-4-21-

II REVIEW OF CURRENT LITERATURE

The rate process theory and various rheologic models have

been proposed to represent the creep behavior of soils. Creep

behavior of particular concern are: strain and strain rate

as functions of time, creep rupture (failure) phenomenon, and

the effects of strain rate on the shear strength of soils.

Several of the more pertinent publications dealing with this

subject are reviewed in this section.

2.1 RATE PROCESS CREEP THEORY DEVELOPMENT

(a) Mitchell, Singh, et al

Singh, Mitchell and Campanella (1968) use Fig. 2-1

to describe soil deformations in terms of rate process

theory. The theory considers the displacements of flow

units as requiring a free energy of activation, AF, to

surmount the energy barriers as shown in Fig. 2-1. Shear

forces in the flow units, f, distort the energy barriers

as indicated by curve B and cause preferential flow unit

displacements in the direction of shear force. X repre-

sents the distance between successive equilibrium positions

and 6, the elastic distortion of the material structure.

-22-

The resulting theoretical expression for strain rate is:

= 2X kT -AF FX= 2X -- exp (- )sinh(2-) (1)

where k = Boltzmann's constant, T = absolute temperature,

h = Planck's constant and R = universal gas constant. If

(fX/2kT) is greater than one, which is generally the case for

stresses sufficient to cause creep in soils, then:

flsinh (2kT) Z 1/2 exp (fX/2kT) (2)

and

kt -AF fX= - exp (R- )exp(Tkf) (3)

"X" is a time and structure dependent parameter which is also

dependent on the number of flow units in the direction of

deformation and the average component of displacement in the

same direction due to a single surmounting of the barrier.

To simplify analysis, Eq. (1) may be written as:

S = K(t)sinh(aD) (4)

in which

-23-

Is

kT -AF(t)K(t) = 2X(t) -- exp RT (4a)

and

aD = XF/2kT (4b)

The authors present consolidated undrained creep data for

remolded illite as shown in Fig. 2-2 which conforms to the

form of the theoretical log C/k versus aD relationships. For

the range of stresses where the exponential approximation of

strain rate is valid, Eq. (4) can be written as:

K(t)2 exp(aD) (5)

and "a" can be evaluated from the slope of the straight line

sections of the log & versus D curves where D is the applied

deviator stress:

d log (6)dD a (6)dD

Figure 2-3 presents the typical relationship observed by

the authors between log strain rate and log time and indicates

that the following expression is possible.

-24-

t mK(t) 12 = A(t-) (7)

d log &where m =d log (7a)d log t

and "A" is found by extending the log E versus D line for tl

to the value of D = 0 in Fig. 2-2. The resulting modifications

yield the following equation for strain rate:

a t1 M (8)tA( exp (aD) (8)

A and a appear to be dependent on the particular test

conditions, such as the stress system, while the authors claim

that "m" is a material property which is only slightly influ-

enced by test conditions. They further point out that m can

be used as a "creep indicator" since low values of m (m<l)

are found for clays that exhibit high creep strains, lower

strength after creep and creep rupture under sustained loads.

Values of m greater than unity are indicative of rapidly

decreasing strain rates with time and decreasing de/d log t

under subfailure conditions. The condition where m = 1 implies

a constant value of ds/d log t and long term strength equal

to short term strength.

Strains are computed by integrating Eq. (8) over time,

with knowledge of the strain at a known time; a necessary

-25-

condition. The equations for strain are as follows (Singh &

Mitchell, 1968):

A aD 1-mE=a+ 1 e (t) (m 1) (9a)

where

A aDa = e---- e

and

S= + Aea D In t (m = 1) (9b)

The authors present laboratory data on London Clay, Osalsa

Clay, Illite and San Francisco Bay Mud to support their theory

and good agreement appears to be present for stress levels

which are 30 to 90 percent of ultimate strength. The theory

is rather flexible and can handle cases where strain and log

time are either linearly or non-linearly related, which is

an improvement over most rheologic models. However, the

greatest advantage of this method is the comparative ease with

which the creep parameters can be obtained in the laboratory.

This model will be called the Singh-Mitchell rate process

model.

-26-

(b) Christensen and Wu (1964)

Another form of the rate process theory is presented

by Christensen and Wu (1964). Clay deformations are de-

scribed mathematically by the Kelvin-Maxwell model in Fig.

2-4. The spring K, represents the initial elastic response

and the dashpot, the flow process, the spring K2 models

the bond strength at particle contacts which resists soil

creep. The combination of springs K1 and K2 represent

the initial elastic response of the soil.

From this rheologic model, the authors obtain the

following expression:

u 1 + A In tanh [Z(t) + tanh exp(-A)]

where

K1A = aD

3 Kl+K2

1 KIK2Z(t) = a2 t2 K +K2

U* = 1 -(l ) o (dimensionless strain)la loE

-27-

and a, 8, K1 and K2 are soil parameters which can be de-

termined from experimental creep results. (el) is the

instantaneous axial strain, l1 is the axial strain at

time, t, and (El), is the axial strain at large times.

The authors assume that creep stops at some large

time after load application and a unique (el)0 is reached.

Although this assumption may be true for soil with little

creep tendency (m > 1 from Singh-Mitchell theory) it is

not generally acceptable.

Two major drawbacks of this treatment of the creep

process are:

(1) it only applies to soils with low creep potential

(m > 1), of which the EABPL clay clearly is not one;

and

(2) there are no established relationships between

strain rate and time or strain rate and stress level.

(c) Murayama and Shibata (1961)

The authors use the Eyering rate process relationship

to model soil creep. The theory is modified to account

for a lower restraining resistance at particle bonds, below

which movements are elastic and time independent. They

arrive at the following:

-28-

dab babdt - Ab(a-a o ) sinh(b) (1)O

where

-EAb = 2aakT exp ) (2)

Bb 2bkT

where K = Boltzmann's Constant

T = absolute Temperature

h = Planck's Constant

A = average distance between unbalanced molecules

n = number of molecules wiht activation per unit

length in the direction of stress

N = number of molecules with activation per unit

area perpendicular to the direction of stress

a = restraining resistance0

E = free energy of activation

Strain equations are developed with the aid of a

rheologic model as shown in Fig. 2-5 and result in:

+ (a-a o ) (a-a) A2 t (4)E + E +B B2E log B2E2t (4)

1 2 2 _2

-29-

(where Ab + Bb from Eq. (1) are changed to A2 and B2 ) and:

a (a- o)E - + (5)t-+OO E E1 2

These equations predict a linear variation of strain with

log time, followed by a decrease in slope and the attain-

ment of a constant ultimate value. They also point out

that the above relations do not apply when an ultimate

stress is reached which brings on failure. From laboratory

results from long term anisotropic consolidation tests

the authors report that ds/d log t varies linearly with

applied stress, which is in agreement with their equations,

but in disagreement with Singh and Mitchell who find that

log (de/dt) varies linearly with - = D/Dmax . The authors

also report that long term undrained strength decreases

linearly with the time to failure.

Although their theory development is straight forward

and agreement between laboratory data and theory is

obtained,the Murayama-Shibata relations do not appear to

be as advanced as the work of Singh and Mitchell and there-

fore are not as useful.

(d) Shibata and Karube (1969)

Shibata and Karube mainly extend the work of Murayama &

-30-

Shibata and arrive at the same questionable conclusion

that 6d = de/d log t varies linearly with deviator stress

(where *d is deviatoric strain). While never stating

whether they still believe that strain varies linearly

with log time, they present the following empirical

equation, which is in disagreement with the Singh-Mitchell

model:

Sd = Constant • exp (Bad)

where B is a stress factor and ad is the deviator stress.

They maintain that "B" is a function of overconsolidation

ratio and that overconsolidated samples yield a steeper

slope on a d versus deviator stress plot.

Their most credible work has to do with creep strengths

of clays. The authors state that there exists a yield

value past which sufficient pore pressures are generated

to produce undrained failure, or so-called "creep rupture".

This relationship, which was developed completely empiri-

cally through laboratory observation, can best be shown

by Fig. 2-6.

Although this effort is an improvement over the work

of Murayama and Shibata, the same drawbacks still exist

-31-

and the equations concerning strain rate are questionable.

Their section on creep strength is nevertheless definitely

useful.

(e) Barden (1969)

Barden maintains that soil deformation is a function

of stress ratio (U1/U3 or Singh-Mitchell's D/Dmax ) . He

separates the volumetric and deviatoric components of creep

and by use of rate process and the rheologic models in

Fig. 2-7, arrives at the following:

2.3 a(SKtV = v + log (FINAL Ka+ 2

2.3 aBGtY = Y + -- log (FINAL Go- 2

The volumetric strain and its corresponding equilibrium

value are represented by v and vFINAL and the shear dis-

tortion and its equilibrium value by y and YFINAL The

parameters a and 8 are constants from the rheologic model

and can be obtained in the laboratory. The strain rates

follow directly from the above equations:

dv/d(log t) = 2.3/Kc

dy/d(log t) = 2.3/Ga

-32-

where Ka and Ga are functions of the effective stress ratio.

Although Barden's laboratory data are completely from

drained tests, he presents the undrained laboratory results

of various investigations (including Singh and Mitchell)

which in general support his equations.

Although Barden concludes that his theory is accept-

able, he points out that during undrained creep, possible

build up of pore pressures will change the effective

stress ratio and thus, the strain rate. This problem can

be bypassed by using the deviator stress. The main draw-

back of Barden's theory is that K and G are probably as

difficult to obtain in the laboratory as E, the undrained

Young's modulus.

(f) Bishop and Lovenbury (1969)

Bishop and Lovenbury conducted long term triaxial

compression drained creep tests on London and Pancone clay.

Using D = (a1-a3)/(al-a 3)f of 17-98 percent, they conducted

the tests until either failure was reached or problems

with equipment developed (some of the tests were still

in progress after three years).

Axial strain versus log time was approximately linear

-33-

for samples that did not fail and plots of log strain rate

versus log time were approximately linear (London Clay

data was especially good in this respect) as shown in Figs.

2-8 and 2-9. Sudden increases in strain rate at large

times or "instabilities" which were accompanied by substan-

tial volume decreases were unexplained although the authors

maintain that these decreases were due to fundamental

modifications in soil structure and not experimental error.

These excellent laboratory data lend great support

to the Singh-Mitchell theory. Log strain rates vary

approximately linearly with log time over large time peri-

ods and it is demonstrated that creep rupture can occur

at stress levels less than full strength. The only real

drawback in the paper are the instabilities which are yet

to be explained.

-34-

2.2 CREEP RUPTURE

(a) Saito, et al (1961, 1965, 1969)

Saito and Vezawa (1961) conducted a series of uncon-

fined and triaxial (CU) compression tests on a number

of Japanese soils in an attempt to determine the range

of stresses and elapsed times for which slow creep move-

ments generate into failure. On the basis of these tests,

the authors claim to have found a unique relationship

between creep rupture life (time after loading) and strain

rate. This relationship which is linear on a log-log

plot is shown in Fig. 2-10 and appears below as:

log tr(min.) = 2.33 - 0.916 log e(l/min.) + 0.59 (1)

They also add that the following is approximately true:

• * tr = 2.14 (2)

Saito (1965) later defines ý as the constant strain

rate which develops prior to failure and that tr is the

elapsed time from this point to failure. Saito claims

that the above relationships apply to all soils (this

is obviously the reason for the rather large range for

-35-

tr in Eq. (1)). He supports his work with several field

cases which are of questionable worth due to the rather

arbitrary way in which strain is measured in the field.

In a later paper Saito (1969) concerns himself

with the transient creep range which occurs at a time

closer to failure than the steady state creep which he

had studied previously. From his basic steady state equa-

tion, Saito develops a three parameter equation to model

transient creep behavior and presents a graphical solution

for analyzing field problems.

Saito's work is completely empirical and his claim

as to the generality of his equations with regard to

soil type is questionable but it is a step in the right

direction. The major drawback of his work is that is

appears to be of little practical significance. Failure

is not indicated until it is actually occuring and by

then remedial action is out of the question.

(b) Singh and Mitchell (1969)

The authors deal with the problems of creep rupture

and time to failure by modifying their basic equation

(Eq. (8) in section 1.21a).

-36-

-37-

= At1m exp (aD)t1 - m

Strains at time "t" (for m # 1) are proportional

to it at that time as shown below:

E = C + l-1 1-m

where

aDAe tD1C = E1 l-m

For m > 1, it is constant or decreases linearly on

a plot of log 6t versus log t. However, for m < 1, t

increases linearly with a slope of 1 - m until a given

value of ýt is reached at which time there is a rapid

change to a much greater constant slope as shown in Fig.

2-11. Singh and Mitchell (1969) define the 6t at which

this event takes place as indicative of failure conditions.

Thus (it)f is independent of stress level and appears

to be unique for each soil type. Analytically tf can be

found by:

1loge (tf)= 1 -m (C2 - D)

where

C2 = loge (Et)f - loge (Atl m)

In comparing this method with that of Saito and

Vexawa (1961), we find it to be much more attractive.

Although Saito and Uezawa also arrive at a unique "ft ,

they weaken its development by making it too general while

Singh and Mitchell claim a unique (ýt)f for each soil.

The main advantage is that the Saito relationship is re-

ferring to a time just prior to creep rupture (long after

(ýt)f has been reached) and Singh and Mitchell are con-

cerned with a modest change in creep rate long before

shear failure which makes it a more valuable indicator

and eliminates one of the major drawbacks of the Saito

relationship. But most important, the Singh-Mitchell

relationship is an extension of a much larger development

of a theory for creep in soils and is not merely an

empirical tool.

2.3 CREEP STRENGTH

(a) Arulanandan, et al (1971)

The authors conducted a series of undrained triaxial

compression creep tests on isotropically consolidated

San Francisco Bay Mud in order to prove the Roscoe (1963)

-38-

relation:

M oq = logM(1-K/A) loge P

where M = slope of failure envelope on p,q plot

p0 = isotropic stress

p = mean normal stress

q = deviator.stress

A = C c/2.3 = Compression Index/2.3

K = Cs/2.3 = Swell Index/2.3

Stress levels ranged from 30 to 90% of failure stress

and loads were maintained for two weeks after which time

the unfailed samples were taken to failure.

Results showed that the yield stress was only 60 -

75% of the ultimate strength for all consolidation stresses.

Creep stress paths crossed both experimental and Roscoe's

(1963) theoretical stress paths. The large pore pressures

generated during the test were responsible for this phe-

nomenon. Possible reasons for these pore pressures are:

(1) leaks through membranes and seals;

(2) osmotic pressure due to the salt concentration

of San Francisco Bay Mud;

(3) structural rearrangement of particles or a

-39-

drainage into micropores from compression of the

microstructure, which is comparable to secondary

compression in their viewpoint.

The last reason is most likely due to the fact that

mercury jackets and a saline cell fluid were used.

Reason three above was further investigated by con-

solidating samples of San Francisco Bay Mud and Kaolinite

for 30 hours and then closing the drainage valve. The

results, as shown in Fig. 2-12, indicate that the Bay

Mud, with a high secondary compression tendency, generates

pore pressures due to the arresting of secondary compres-

sion. The Kaolinite, with lower secondary compression

tendency, begins to generate pore pressures but "thixo-

tropy" is thought by the authors to eventually dominate

the process and pore pressures decrease.

The authors therefore conclude that the arresting

of secondary compression during undrained creep generates

large pore pressures which can fail samples at deviator

stresses much lower than ultimate values as shown in

Fig. 2-13.

These results, which tie in well with the creep

failure data of Shibata and Karube (1969), Bishop and

-40-

Lovenbury (1969), Singh and Mitchell (1969) and Walker

(1969) are very useful and their proposed mechanisms may

be helpful in explaining such phenomena.

2.4 CONCLUSIONS

Of the various rate process and rheologic models, the

Singh-Mitchell treatment appears to be the most attractive.

Although this method is of a more empirical nature than some

of the others, it is by far the most advanced and unrestrictive.

The model parameters can be obtained in the laboratory with

relative ease.

Although the Singh-Mitchell treatment of creep rupture

appears to be somewhat superior to that of Saito, it must be

kept in mind that these two theories apply to two different

events in the creep rupture process and both can be evaluated

in the laboratory.

-41-

/ 1 Shear force

force acting

no force acting

DISPLACEMENT

(a) REPRESENTATION

IUUU

500

200

100

50

20

10

1.0

0.5

0.2

0.1

OF ENERGY BARRIERSPOSITIONS

SEPARATING EQUILIBRIUM

0 'f

VERSUS c< 'f ACCORDING TO HYPERBOLICEXPONENTIAL FUNCTIONS

SINE AND

DERIVATION OF BERKELEY THREE PARAMETER RELATIONSHIPBASED ON RATE PROCESS THEORY (After Mitchell et al, 1968).

Figure 2-1

(b) 6/K

I AJ•. A

-42-

Ix 10 4

-54x1O6

Ix 105

-64 x1O

Ix -6

-7

Ix 10 1.0 1.4 1.8STRESS, 1 - 3- D- kg/cm2

VARIATIONCIU C CRE

OF STRAINEEP

RATE WITH DEVIATORTESTS ON REMOLDED ILLITE

Mitchell,-43-

1968)

STRESS FROM(After Singh and

Figure 2-2

-2Ix 10

-44x104

0.6DEVIATOR

Q2

10 100 1000

TIME ,MINUTES

INFLUENCE OF TIME ON STRAIN RATE DURING CREEP

OF REMOLDED ILLITE (Mitchell et, al. 1968)

-44-

0.1

.01

'001

.0001

.00001

1;·~i -"----YII a~~~~i~~~o~a·nnai

/

RHEOLOGIC MODEL FOR CHRISTENSEN AND WU(1964)

RATE PROCESS DEVELOPMENT

-45-

RHEOLOGIC MODEL FOR MURAYAMA AND SHIBATA(1961)

RATE PROCESS THEORY

Figure 2-5

-46-

O 00(5

Figure 2-6

-47-

~~:;a~i~ :~b-`~,''r`·""-br .~~l~*Z b~C ~ JII ~ .~; ~I

ro

·._ o

cHl-)

4-)

o

Uo

0

N

0a. nU

ToQ)

r-i

U

El)4-

ri)

v

~

VbAvUHE7rA/e-

I PWA PR--- F -11- 1 I

I .,r.

COdfPAE~SS/OA'

5 HA. P /S ,O R/7-cIO

FIGURE 2-7 Barden's Rheologic Model (Barden,1969)

Figure 2-7

-48-

=,off l

/ /0 /00 /o000T/ffME, PR YS

DRAINED TRIAXIAL CREEP RESULTS ON PANCONE CLAY

(Bishop and Lovenbury, 1969)

Figure 2-9

-50-

_i~o!--- _

I O

t.

01O ,o0

.0001

0.o

H IV,

(fW)..71 /7 /7Jd.f7 d-7-33V

Figure 2-10

-51-

r%

04-)

HU)

4

HOH

12

4)

77 i 77-.~N1*?J;~F ~ I

I i·.rýV-w ;r.·-n;:;·n

oa

LO

01

C)

Ci)

oo0r HOOo4 a

a)

a'

ClW,Q)

,O~ 4J

Figure 2-11

Figure 2-11

-52-

0

Cd

Q)

a)

a)

mý-::IU)

0

O •

oO' r-

n0 0

S(C

0O m0 >)

o ~i0 H

u <

40 >

HOC)0) 0N 7

e, J

L1O y

7 Ah'/)/ .y/d5-5$ 3Y/ N(.7 H•/•> .".•n -s3 • d W

Figure 2-12

I

r%.

-53-

vQJ

00 eIc

(5c· d~z Y'IOz/4ap) g

0

4-

U)

(2)

0

UrO

12,Enra

U)la)CdCd

Cd

Sa)

Cd7

Ug

Li

7igure 2-13

-54-

0

0

2'

III DESCRIPTION OF EQUIPMENT AND PROCEDURES

3.1 INTRODUCTION

The Direct Simple Shear device used for this investigation

was developed by the Norwegian Geotechnical Institute and

manufactured by Geonov. It is very similar to the machine

used for tests described by Bjerrum and Landva (1966). The

primary requirements that the machine satisfies are:

(1) Ability to induce fairly uniform horizontal shear

deformations in a cylindrical sample.

(2) Capability of performing drained or constant volume

tests.

(3) Ability to keep the upper and lower sample surfaces

parallel during testing.

(4) Controlled stress on strain.

(5) Maximum vertical load of 800 kg, horizontal load

of 400 kg.

(6) Designed for 50 cm2 by two cm sample confined in

a wire reinforced membrane.

-55-

(7) Capability of testing soft clays as well as stiff

clays, silt and sand. A method of sample prepara-

tion was devised that minimizes sample disturbance.

3.2 SHEAR APPARATUS

The shear apparatus (see Fig. 3-1) is composed of the

sample assembly, vertical and horizontal load units and a hori-

zontal load hanger assembly. The sample assembly consists

of lower and upper filter caps (16), pedestal (17),

and a plastic reservoir for inundating samples during testing.

The vertical load unit consists of the base (18), tower

(19), lever arm (12), load gauge (4), piston (20), dial gauge

for vertical deflection measurements (6), and lever-arm

adjustment (21). The lever-arm has a load magnification ratio

of 10 to 1 and is built into a U-formed hanger. A counter

weight (22) balances the weight of the load-gauge, piston and

sliding box.

The horizontal unit consists of a gearbox (11), proving

ring load gauge (10), precision ball-bushing (9) and connection

fork (23), and the sliding box (7). The present machine,

Model 4, strains the sample by moving the top cap while holding

the bottom cap and pedestal stationary. The horizontal load

-56-

is transmitted from the horizontal piston through the fork

to the sliding box and top cap. The limit of travel is 10 mm

in either direction from zero strain.

3.3 SAMPLE PREPARATION

By a system of vertically aligned cutting cylinders, top

and bottom caps, pedestal and membrane streching device, the

cylindrical sample is mounted in position and confined by the

wire reinforced rubber membrane. The procedure is designed

to support the sample at all times while minimizing distrubance

and is much easier than sample preparation for the triaxial

test (see Landva (1964) and Edgers (1967) for more detail).

3.4 TESTING PROCEDURE

For all tests, the sample was installed in the machine

and consolidated by dead load increments until the desired

consolidation stress was reached. All samples were consoli-

dated to a stress sufficiently greater than the insitu vertical

stress so as to minimize the effects of sample disturbance.

The reservoir was not filled until v = 0.4 kg/cm was applied

in order to avoid swelling. The final increment for normally

-57-

consolidated tests and the maximum and final loads for over-

consolidated samples were allowed to consolidate for approximate-

ly two days in order to develop a constant degree of secondary

compression.

For constant volume controlled strain tests, the shear

procedure was that described by Edgers (1967).

For controlled stress constant volume creep tests, the

horizontal proving ring was removed and the horizontal load

was applied by means of the hanger assembly. After consoli-

ation was complete, the lever-arm was connected to the vertical

load adjustment screw and the appropriate horizontal load

applied to the hanger assembly. The test was started by un-

clamping the horizontal piston and readings of horizontal

strain with time commenced. As in the standard constant

volume controlled strain test, the vertical load was adjusted

in order to maintain constant sample volume throughout the

test. Most of the adjustments occurred in the first two hours

and usually only one adjustment per hour was needed thereafter

on the first day and only one or two adjustments per day were

needed on subsequent days. The total number of adjustments

needed for an average test duration of one week (\ 10,000

minutes) was on the order of 25. Room temperature during the

test was approximately 690 F. Fluctuations of + 20F were common

during a 24 hour period.-58-

1 Sample 2 Reinforced rubber membrane 3 Wheels forapplying dead load 4 Load gauge for vertical load5 Ball bushing 6 Dial gauges for measurement of verticaldeformation 7 Sliding box 8 Dial gauge for measurementof horizontal deformation 9 Ball bushing 10 Load gaugefor horizontal force 11 Gear box 12 Lever arm13 Tieights 14-15 Clamping and adjusting mechanism usedfor constant volume tests 16 Lower and upper filterholders 17 Pedestal 18 Base 19 Tower 20 Verticalpiston 21 Adjusting mechanism 22 Counterweight23 Connection fork 24 Horizontal piston

Fig. 3-1 General Principle of Direct Simple-Shear Apparatus,Model 4 (Edgers, 1967)

Figure 3-1-59-

i

IV EXPERIMENTAL INVESTIGATION

4.1 SOIL TESTED

The soil for this investigation was obtained from five

inch diameter undisturbed fixed piston tube samples obtained

from three borings offset approximately 300 feet to the land-

side of Test Section III.

Boring StationNo. No. Offset (ft) Samples El. (ft)

94ues 1395+50 310 LS of BL 10A to 13C -30.5 to -45.8

95ues 1395+45 305 LS of BL 10A to 13C -30.6 to -45.6

96ues 1395+55 305 LS of BL 10A to 13C -30.5 to -45.3

Samples used in the investigation ranged in elevations

from -35.9 ft to -44.15 ft. Values of vo for this layer,

which appears to be either normally consolidated or slightly

over consolidated (see Fig. 1-2), range from 0.55 to 0.70

kg/cm2 according to Foott and Ladd (1972). Consolidation

curves for the Direct Simple Shear Tests conducted in the

program yielded maximum past pressures in the range of 0.70

to 0.85 kg/cm2

-60-

Atterberg Limits were conducted on a number of representa-

tive samples and the results are shown, along with those ob-

tained by the Army Corps of Engineers, in Fig. 4-1.

A series of constant volume controlled strain direct-

simple shear tests were conducted on a number of samples at

various overconsolidation ratios.

4.2 TEST PROGRAM

As previously stated, the testing program was conducted

to achieve the following:

(1) A relationship between s vc and OCR for CKoU DSS

controlled strain tests.

(2) Investigation of the applicability of the Singh-

Mitchell Rate Process Theory for EABPL Clay with

regards to strain and strain rate as functions of

time, stress level and overconsolidation ratio for

first load increments.

(3) Investigation of the applicability of the Singh-

Mitchell and Saito creep rupture theories for EABPL

Clay.

(4) The effect of time on the strength of EABPL Clay.

-61-

In addition, the testing program was designed to investigate

the behavior of additional load increments with respect to

strain and strain rate as functions of time, stress level

(applied horizontal shear stress/horizontal shear stress at

failure, TH/Tmax) and overconsolidation ratio. It was hoped

that a method could be developed for translating second and

third load increment data into the form of initial load per-

formance.

The program designed to achieve these objectives is

outlined in Table 4-1 and is basically of two parts: (1)

the CK U DSS controlled strain shear tests, and (2) the con-o

trolled stress creep tests. Tests Nos. 2 through 7 were used

to develop the sU/-vc versus OCR relationship from which

various stress levels for the creep tests can be calculated.

The remainder of the program consists of undrained creep tests

on normally consolidated and overconsolidated samples in which

the number of increments, the stress level and the load dura-

tion time serve as variables.

4.3 TEST RESULTS

4.3.1 Controlled Strain Tests

CK U DSS tests were conducted on seven samples of EABPL

Clay. The results are shown in Table 4-2 and Figs. 4-2 to 4-7.-62-

CK U DSS-l was lost due to equipment failure and was there-

fore not tabulated. The normalized shear strength (Su/6 vc)

for normally consolidated samples was very consistent at

0.23. Failure occurred at very high shear strains of 20%

and more for normally consolidated samples but this value

was somewhat reduced for highly overconsolidated specimens.

It was interesting to note that the maximum value of TH/'v

occurred simultaneously with the maximum value of TH for most

of the overconsolidated samples though this was not the case

with normally consolidated tests. A plot of normalized shear

strength versus log OCR yielded the smooth graph in Fig. 4-7.

4.3.2 Controlled Stress Creep Tests

A summary of the 12 CK U DSS Creep tests appears in

Table 4-3. Appendices B and C contain tabulated test data;

strain, log strain rate and log tt versus log time curves;

stress paths and stress-strain curves for the 12 tests. Summary

plots of the data are presented in Chapter 5.

4.3.2.1 Initial Stress Levels

Various shear stress levels were applied to the samples

after consolidation was completed. These levels consisted of-63-

I

-64-

50, 65, 73, 80, 85, 90 and 94 percent of failure for normally

consolidated samples and 50, 72 and 94 percent for samples

with OCR = 2. Samples were allowed to creep for one week

(10,000 minutes), except CKoU DSS-12 where D = TH/Tmax = 73%

was left on for 40,000 minutes.

4.3.2.2 Additional Shear Increments

Additional higher shear stress levels were applied to

samples that did not fail under their initial loadings. These

second and third increments were also allowed to creep until

either creep failure occurred or a week passed. These loadings

are listed in Table 4-3.

4.3.2.3 Creep Failure

Creep failure occurred in 11 of the 12 tests. The first

test (CK U DSS-8) was failed by controlled strain after three

shear stress increments had been applied over a three week

period.. For the remaining tests, creep failure eventually

occurred at all applied shear stress levels of 85% or more.

Test No.

CKoUDSS-1

CK UDSS-2oCK UDSS-3

oCK UDSS-4oCK UDSS-5

CK UDSS-6o

CK UDSS-7

CK UDSS-8

CK UDSS-10o

CK UDSS-11o

CK UDSS-12

CK UDSS-16o

CK UDSS-17

CK UDSS-18o

CKoUDSS-19

CK UDSS-20

CKoUDSS-21

CK UDSS-22oCK UDSS-23

Type

Controlled c

Controlled s

Controlled

Controlled

Controlled

Controlled

Controlled

Creep

Creep

Creep

Creep

Creep

Creep

Creep

Creep

Creep

Creep

Creep

Creep

SUMMARY OF SAMPLES FOR

TEST PROGRAM

Boring Sample W,%

96UES

96UES

96UES

96UES

96UES

96UES

95UES

94UES

94UES

94UES

94UES

94UES

95UES

95UES

95UES

95UES

95UES

95UES

94UES

11B

liB

11B

11B

11C

11C

10B

1IC

13C

13A

13A

12C

13A

13A

12D

12D

12D

12D

10B

72.2

72.6

76.3

91.7

86.3

78.1

73.8

84.9

27.2

86.7

73.4

79.1

79.2

67.7

80.1

75.1

77.3

79.5

Elevation, ft

-35.9

-36.1

-36.2

-36.5

-36.9

-37.1

-32.1

-36.7

-45.0-42.7

-42.9

-40.5

-44.0

-44.1

-41.2

-41.3

-41.5

-41.6

-32.0

-65-

TABLE 4-1:

U) 0

<O .c 000-

o

1c E 0 E

L

3 8 omo, o o o o,8,80 o o o 0 ( O OO 1o 0 0 ao

o oo + U n re p U - r o o 1 ro I)

So 0 0 C 0 0 0 00 0 +1 (D = OD

'- (0 = r cK) o 0 (e C6 N w

-0 N 0XI e I - -to --z 0" 0 D 00 000 0 0 0 0 0L0 0 •D C -• 0 -

• N O r) U a ON) N N +c0 ) c.0 ' d o0 d o 0

O- 00 0 -- 0 0 0 0 00 O _o

0 o (D0 0 O

0 0 0 0o 0 0 00 0 O Ln 0 LO roa Ob' ca N It

Ur) U)) U) U) U~ U)

_ D 0 nN NO

I __ _ - -N N N d _ _ ___ N -

U)00 ow50 N< 0 o (i N 0

Table 4-2

-j

O

-Ja-

z0

U)

U)

>--U)rU0

o

0

U)

cm Eo

cC

(,

cnU)(,L.

u,

0

4-c

1-

C

U)-ISc

, .,0V

-66-

0000 0 0

4-4 Hco "-I kD

r-I H-

0

o N

o0

ir-t

CH I

4-- 0CI

-I -o cI1 c

I L

Ln0

Sco4IJ 0

N-

..-.%oP 'q

oo

ci*e 000

44 re)

4- ý

LO0coo H m

o o

0 Ln

CC)I O

U)0;

Oo

CO· o

10

No

0o

N

C)

or

H

I NO r.im 0

I 0CY)O

L(

O•

0

%D

LnO(N

O

0I I I I0I I I 10

(NC)

CO

I c

co 0)C)

O

L)

0

0

N,N0 cl rHcoo

) Ln.

Cý UýNr0O

*O

coco

0

O H

CN r

00o r- N %D,-' r-I -• 'I I I I

C:I CI] V1 VU) U) U) U(

00

o : O 0U bU IU

00 LOLONrrn 00l_;~ Cý

N, co

I I I0 00U)l U) UU) U) U)

a a a

U U

-67-

0 0

r-! r-t

0 ot-f 00Ln r-Ir-I CN

000r

H rH

0 0o N

HH0r-I ,-r-t ,-I

0 0ko H

r-t r-I

o 0CN Cm

CN

0 0

o o

,- riH NC

u, i-U)I -

E-4O I

OD L

(Nj

U)

vl

oP4

ýD

E-4

z

i-IH

E-U)U)-I

1F::FZ4

0

z0HE

HP4uU)

0 0 00 000

"N •r Ck rN rHN Hr

0 0 0 0 000

0 0 . .0 . .0

Oml Lf YoIN OMLOH- I I I I I0•J I I I I IL Imf

H Co N-e~- 00 O

N N

II IIH E

-Iv

IIC)

44 00 r--. LC)

.- I

H H OH

Nco

0 N m Lo Ou H NL CN 0N 0H

0 rlf 00 6ONIV

* 0 0 0

0 O C C0_ c o oo

o r- OO

0 0 0

o U

LnN

NOHO

<m i

co00 ('.

HHrO CO

Hl r iL) r-tcoco

moHLO C

N O0)0 0

OO

0N

.- I U)rz3

E-U

U)

0 0

Z0HE-1

m pI Ha4

Urr U- i

H7v

0

0

-,4-)-cr

0U

-68-

TABLE 4-4: DESCRIPTION OF MEASURED TEST PARAMETERS

Stress Level (D T ) = (TH/TH ma x )

Yo (Initial Shear Strain, %) = y @ t = 1 Minute

% (Initial Strain Rate, %/Min.) = ' @ t = 1 Minute

yf (Final Shear Strain, %) = y @ tf

(rt)f (Final Strain Rate X Elapsed Time, %) = ýf X tf

tf (Final Elapsed Time) = Length of Time the Shear Stress

Increment (DT) Was Left on Sample

Y1 (Steady State Strain) = Strain at Which a Second or Third

Load Increment Began to Follow the

Singh-Mitchell Theory (See Section

5.2)

nl (Steady State Strain Rate) = Strain Rate at Which a Second

or Third Load Increment Began

to Follow the Singh-Mitchell

Theory

tl (Elapsed Time) = Elapsed Time at Which y1 and ~l Occur

-1(tan- TH/ v ) = Y at t = 1 Minute for a Given DT

ý -f (tan- TH/ v ) = v at tf~f(tan TH/a ) = $ at tf

-69-

J )-j

oj o oU o

>-u

I wjI-.JI ,UuZnoo

z z0

31'*ow

0 a w i < GCL 1LA OgJ):OoW0a ~

0 0 0 0X30NI A)lI1S.•1d

Figure 4-1

Cdr4

PH

Cl)

C+

*H

C)

0

Q)

4-,

C)

C

4-)Cl

P4O:O;

-70-

------ oSOO

SHEAR STRAIN , 7,%

SHEAR STRAIN Y, %,

CKoUDSS TESTS ON

NOR M A LLY CONSOLIDATED

EABPL CLAY

Figure 4-2

Ib

164Ib

I

I

I1a

C

C

C

Symbol No.---- 2 72t 41 3.1 1.0 -36- 3 72.- 63.7. 1.02 L5 -36

- -- 4 763 5&91 205 1.7 -36

-71-

0.8

0.6

'-C0.4

0.2

0

0.4

IIb9 0.0

< -0.4

-n) A0'' 5 10. 15 20 30

SHEAR STRAIN r, %

STRESS VS. STRAIN FROM CKoUDSS TESTS ON EABPL CLAY

Figure 4-3-72-

0tr

z.I-

I-OZ

V-

I-

U.

-r

Ua,W

wLLwF-

a:

w-i

oE

_-o4 / 'SS3I.LS iV3HS 03ZI-VI•VON

Figure 4-4-73-

U)/ 'sssWA5 U1 'SS3d.LS

0

ba.m

LUIf

w Z0

U)w I> U)

w

HL

0 H,al--

crw

0a crw LU

-J-

Z0

N-J Cl

' 08V3HS 033ZIflVlIVON Z

Ficure 4-5-74-

0

z

-J

0i-cQui

vc , TSF

COMPRESSIONCKoUDSSEABPL

CURVES

TESTSCLAY

Figure 4-6-75-

Symbol No. Wn,% El.,ft.

-- 2 72 t -36-- o-- 3 726 -36

_- - 4 76.3 -36- 5 91.7 -37- - 6 86.3 -37

- 7 78.1 -32

U

I16>

q)

OCR = m/ "vc

SU /Fi& VS OCR FROM CoU DIRECT- SIMPLESHEAR TESTS ON EABPL CLAY

Figure 4-7-76-

V DISCUSSION OF TEST RESULTS

5.1 FIRST LOAD INCREMENTS

The term first load increment refers to the first

increment of shear stress which is applied to the sample

after consolidation is complete. Prior to this process the

shear strain is zero. The Singh-Mitchell Rate Process Theory

is concerned with the undrained creep that occurs due to this

loading.

5.1.1 Normally Consolidated Samples

Results from first loadings on normally consolidated

samples are presented in Figures 5-1 through 5-4. Within

the first minute of loading there occurred an instantaneous

shear strain (y ) which can be determined from Figure 5-1.

From that point the creep process for stress levels (5 ) of

80 percent or less appeared to follow the linear relationship

between log strain rate (f) and log time which is the basis

for the Singh-Mitchell theory (see Fig. 5-2 and 5-3). For

stress levels greater than 80 percent, the plots of log strain

rate versus log time were initially linear, but the strain

-77-

-78-

rate then became constant in each case and then increased

as the sample creep ruptured. This phenomenon will be covered

in detail in a later section.

Unlike the Singh-Mitchell Theory where log-log plots

of strain rate versus time yielded parallel lines for the

various stress levels (D), EABPL clay yielded straight lines

which were divergent with time or in other words, "m" was

not constant but a function of stress level, D , as shown

in Fig. 5-13.

(Note: Singh-Mitchell's D refers to the deviator

stress ratio in the triaxial test:

al- 03(a1 - a3)f

Since the deviator stress cannot be determined in the simple

shear stress system, DT refers to the applied horizontal

stress ratio, TH/T H max). Therefore, "m" which is supposedly

a constant parameter for a given soil and the index of that

soil's creep tendency, is in fact, for EABPL clay, a variable

which demonstrates that this clay has greater creep tendencies

(decreasing "m") at higher shear stress levels.

Another point which is of interest concerning Fig. 5-2

is the presence of discontinuities in the linear plots which

N

I

-79-

take place at approximately t = 400 minutes. For all stress

levels above 50 percent (except for those that caused failure)

the strain rate begins to decrease a few hours after loading

at a greater rate than that of the initial log-log plot of strain

rate versus time. At approximately t = 400 minutes, a new

log-log linear relationship is established which results in

a straight line which is "below" the initial line and has

a higher value of "m". The sudden application of a shear

stress induces positive pore pressures within the sample. The

vertical stress is reduced to compensate for this but since

Cv is not equal to infinity, time is required before equili-

brium returns to the sample. No doubt the average pore

pressure in the sample is close to zero (because the sample

volume remains constant) but pore pressures within the sample

interior are probably greater than zero and those near the

edges are negative. The time required for equalization is

on the order of 400 minutes, which is approximately the time

required for primary consolidation in normally consolidated

EABPL clay based on Cv = 0.01 ft2/day. This value of Cv is

within the range recommended by Ladd et al (1972). The

hypothesis is further supported by the fact that this

phenomenon is only barely noticeable for stress levels (iT)

of 50 percent. As can be seen with the hypothetical stress

path in Fig. 5-9, the induced pore pressures for higher stress

levels are much greater.

-80-

41

Assuming this hypothesis to be correct, the pore

pressures within the sample interior (i.e. on the potential

failure plane) decrease the effective vertical stress. Since

the shear stress is unaffected by pore pressures, the resul-

tant obliquity (TH/av) is greater than that applied and

is equal to the obliquity that one would normally obtain from

a higher application of DT. Only when the induced pore

pressures are dissipated will the obliquity within the sample

and the resultant creep behavior correspond to that level

normally associated with the applied DT .

This situation makes determination of the remaining

two parameters in the Singh-Mitchell equation particularly

difficult. The two constants (A, U) are obtained from a

plot such as that presented in Fig. 5-4. Since the obliquity

at small times (t < 400 minutes) is larger than that normally

associated with the applied 5T, the strain rates at small

times for a given DT are higher than they should be. Since

this effect is more marked for larger stress levels, the

resulting slope (1) is too large. A more reasonable result

can be obtained by extending the steady state portions of

the plots (t > 400 minutes) in Fig. 5-2 back to small times.

The resulting data points yield slopes (M) which are probably

more representative of the true soil behavior. As can be

seen in Fig. 5-4, all laboratory data points lie above the

I

modified data for all times except t = 1000 minutes where

there is uniform scatter, which is consistent with the

hypothesis.

5.1.2 Over Consolidated Samples

Results from first loadings on three over consolidated

speciments (OCR = 2) are presented in Figs. 5-5 through 5-8.

The creep behavior is essentially that of the Singh-Mitchell

Theory, with "m" decreasing with increasing stress level.

In contrast with the creep behavior of the normally consoli-

dated specimens, there occurred no discontinuity at t = 400

minutes on the log-log plot of strain rate versus time. The

author believes this to be due to the fact that the stress

path for OC samples is such that initial pore pressure

development is not significant. As a result, "m" remained

essentially constant throughout a given test.

It was also found that "m" for any given stress level

is significantly less than that for normally consolidated

EABPL clay. Inspection of Fig. 5-10 shows that initially

the normally consolidated samples exhibit a higher strain

rate but that the curves converge with time due to the lower

"m" value for the OC samples. At large times the OC samples

exhibit greater strain rates.

-81-

Initially, for a given value of D-, the OC sample is

closer to the failure envelope (see Fig. 5-11). Nevertheless,

the strains obtained at t = 1 minute are essentially equal

for the two stress histories (see Fig. 5-17). This is to

be expected since Fig. 4-3 shows that equal values of 5 T

yield nearly equal values of strain during standard controlled

strain CK UDSS tests.o

The plot of normalized pore pressure versus shear strain

for standard CK UDSS tests in Fig. 4-3 shows that initiallyo

two entirely different phenomena occur during the shearing

of the NC and OC (OCR = 2) samples. It is only after

approximately 6 percent shear strain has occurred that the

OC sample begins to develop positive pore pressures (i.e.

develop a positive slope in Fig. 4-3) and behave in a manner

similar to that of the NC sample. Strangely enough, the

plots of the two samples become parallel at this point and

failures occur at nearly equal strains. Figure 5-12 is a

plot of Au/o versus y for standard CK UDSS tests and CK UDSSvc o o

creep tests at D = 0.50, 0.72 and 0.94 for NC and OC samples.T

The author believes that pore pressure generation during

undrained shear and creep behavior (i.e. strain rate) are

closely linked. Since the NC and OC standard CK UDSS testso

behaved quite similarly with regards to stress versus strain,

it would appear that one could compare the creep behavior

-82-

of NC and OC samples by looking at the amount of pore pressure

generation with strain (slope of the curve) relative to

that of the standard CK UDSS test for each group (NC and OC).O

For instance, in the case of D = 0.50, the two tests (NC

and OC) begin with nearly equal strains at t = 1 minute

though the OC sample is at a stress state in terms of effec-

tive stress that is closer to failure. Inspection of Fig. 5-12

shows that pore pressure generation for a given strain incre-

ment for the NC sample has increased over that of the standard

CKRDDSS test. At the same time the OC sample shows a signifi-

cant decrease from that of the standard test. At a later

time, the rate of pore pressure increase with strain for

the NC sample decreases while that of the OC sample radically

increases. As a result, the rate of pore pressure develop-

ment with strain relative to that of the corresponding

standard test is greater for the OC samples. Therefore,

the rate of strain rate development is greater for the OC

sample though the magnitude of the strain rate may not be

greater. Inspection of creep curves for b = 0.72 and theirT

positions relative to the standard CK UDSS curves shows ao

similar phenomenon. On the other hand, the curves for T =

0.94 are very similar to their parent curves, the standard

CK UDSS test results, but are only displaced to the right.o

As a result, their creep behavior are very similar. Figure

5-13 shows that the two plots for NC and OC samples tend to

-83-

converge at high stress levels, which is what Fig. 5-12

implies. The author realizes that the "m" value for the

NC samples was obtained at small times and is therefore

not the true value which is representative of the stress

level. But it must be kept in mind that the initial unequal-

ized pore pressures developed in the NC sample effected

strain rate level to a far greater extent and that "m" was

not greatly affected by the phenomenon.

Due to the small amount of data for the OC samples and

the difficulty with which good pore pressure data can be

obtained in the direct simple shear device, this proposed

explanation is mainly conjecture. A similar test program

with triaxial equipment would definitely be a great help in

evaluating this hypothesis.

Calculation of T and A presented little problem though

their reliability is questionable due to so little data

(three tests). The value for T was greater than that obtained

from the normally consolidated tests and "A" was somewhat

lower (see Fig. 5-8).

-84-

5.2 ADDITIONAL LOAD INCREMENTS

For all tests where failure did not occur as a result

of the initial shear stress increment, the stress level was

increased and creep was allowed to continue.

Examples of this second (or third) load creep behavior

are shown in Fig. 5-14 along with results from first load

tests for comparison. Other examples can be found in Tests

8, 16, 19 and 23 in Appendix C. In all cases, creep is

initially almost nonexistant. The strain rate at small

times for a given stress level is much smaller than that of

a comparable first load test (same D ). Apparently, creep

from the previous increment has caused a modification in soil

structure that results in a more creep resistant sample.

After a period of time, the creep process begins to accelerate

on a log time scale and the plot of log strain rate versus

log time begins to approach that of the first load test for

the same f (see Fig. 5-14). At some elapsed time, tl, this

plot intersects the "firstload plot" for that stress level

and takes on the log linear behavior of the rate process

theory. This is true for both normally consolidated and

over consolidated samples.

-85-

There does not appear to be any way of expressing this

phenomenon in quantitative terms. The values of the initial

strain and strain rate as well as the strain, strain rate

and elapsed time at which the creep process begins to follow

Singh and Mitchell's theory appear to be rather random. No

doubt they are dependent on ADT as well as the magnitude of

the previous increment and its length of duration.

Analysis of this phenomenon was hindered by the lack

of data as there were only four tests in which either first

or second increments did not cause creep rupture (D < 0.85).

5.3 CREEP RUPTURE

Creep rupture is the condition when strain rate increases

abruptly and strains become excessive. As can be seen in

Figs. 5-1, 5-2, 5-5 and 5-6, this phenomenon occurred for

stress levles (DT) of 85 percent or more.

5.3.1 Singh-Mitchell Creep Rupture Relation

As mentioned in Section 2.2 on creep rupture, Singh and

Mitchell define creep rupture as a sudden increase in the slope

-86-

-I..

of the log 't versus log time plot as shown in Fig. 2-11.

They claim that the value of it (or it) at which the change

in slope occurs is unique for a given soil.

In the EABPL test program eleven samples were brought

to failure through creep rupture. For each case a value of

(it)f (i.e. the point at which there is a sudden increase

in slope) was obtained from the appropriate plot by laying

off tangents to the two line segements of significantly

different slope. For the four tests in which the samples

were brought to failure under the initial load increment,

a consistent (tt)f of 5.2+ % was obtained independent of

stress level and OCR (see Table 5-1). There occurred only

one test (CK-~bSS-19) in which a steady state condition was0

reached by an additional increment just prior to failure

(see Fig. 5-15). In all other cases the additional incre-

ment which led to failure never attained a strain rate

consistent with the Singh-Mitchell Theory before it reached

the failure condition. Because of this, the resultant curve

of log it versus log time was much too smooth and yielded

a value of (it)f which is much too low as shown by C-K~ DSS-10

in Table 5-1 and Fig. 5-15. By contrast, the log it versus

log time curve for CK UDSS-19 quickly changed slope at

t = 500 minutes when steady state was reached and assumed

-87-

a "first increment-like" appearance as shown in Fig. 5-15.

The resultant (it)f of 6.0 percent was also consistent with

first load increments. All other tests which were failed

by additional increment were rejected due to the fact that

steady state was not reached prior to failure.

It was interesting to note that the average total strain

at t = tf (see Table 5-1) for the five tests (all those in

Table 5-1 excluding CK UDSS-10) was approximately 20 percent

which was the average strain at Tmax obtained for the con-

trolled strain CK UDSS tests at OCR = 1 and 2. Apparentlyo

the occurrence of a unique (it)f which has units of strain

is coincident with a unique undrained yf which are 5.2

percent and 20 percent, respectively, for EABPL soil.

5.3.2 Saito's Creep Rupture Relation

As presented in Section 2.2, Saito (1961) has developed

an empirical relation for creep rupture. The equation is

log tr = 0.50 - 0.916 log E + 0.59. The value of e is the

strain rate at which the slope on the log-log plot of

strain rate versus time becomes zero (i.e. constant E with

time). This phenomenon can be seen in Fig. 5-2 for values

of 5D greater than 80 percent. The variable "t "is theT r

-88-

elapsed time that will occur from the attainment of a

constant strain rate to creep rupture (E + c).

This relation was applied to the eleven EABPL tests

and the predicted times to failure compared to those actually

obtained in the laboratory (see Table 5-2). In all cases

the time to failure fell within the range presented by the

Saito relationship, but that range is so large, due to the

gererality of the equation, that it is of little use. The

+ 0.59 in Saito's relation represents a time range of more

than one log cycle. For short times to failure such as

that in CK UDSS-17 this range only constitutes two hours,o

but at large times to failure of perhaps one to four

weeks, a log cycle of 90,000 minutes, or more than two months

is the range.

Another major drawback with the Saito relation is that

failure is not indicated until it is obvious. By inspection

of Fig. 5-1, it can be seen that failure was imminent for

all stress levels of 85 percent or more as early as 10 min-

utes after their application to the sample.

-89-

5.3.3 Strength Decrease Due to Creep

Inspection of Table 5-3 shows that stress levels of

85 percent or more caused creep rupture. Since the values

of 4 at failure were approximately those obtained in

conventional undrained direct-simple shear tests, these

failures were due to the generation of pore pressure of

much greater magnitude than are normally obtained for a given

stress level. This process can be seen quite easily by

inspection of the creep stress paths in Appendix C. For

all stress levels, generation of large pore pressure during

creep produced stress conditions which are much closer to

failure than are obtained in the conventional CK UDSS test.o

This phenomenon is particularly marked with the first

increment, such as that (D = 0.729) in CK UDSS-10.

These results agree quite well with those obtained

by Shibata and Karube (1969), Bishop and Lovenbury (1969),

and others. It is interesting to note that the average

value of f obtained from the creep tests is consistent

with that obtained from standard CK UDSS tests. Apparently

the failure envelope for undrained shear of a given soil is

unique and independent of the mode of failure (controlled

stress or controlled strain). The mechanism of failure

must be the same for both cases.

-90-

It is difficult to believe the concept of a yield value

as presented by Shibata and Karube (1969) (see Fig. 2-6).

If one accepts the Singh-Mitchell model, which the author

does, then given enough time on a log scale, sufficient

strains and pore pressures will be generated to bring about

a failure condition for any significant stress level

(DT> 30%). The threshold value of 85 percent obtained in

this investigation is the minimum stress level at which

sufficient pore pressures could be generated to cause

creep rupture in a convenient time period (less than a

week).

5.4 COMPARISON OF SIMPLE SHEAR STRESS LEVELS WITH TRIAXIAL

STRESS LEVELS

The original Singh-Mitchell laboratory investigation

which was undertaken as part of their rate process develop-

ment was conducted with triaxial equipment. Therefore, the

Singh-Mitchell definition of U(= Aq/AqF) differs from that

used in this investigation with the Direct-Simple Shear Device.

For CK UDirect-Simple Shear Device:o

(Pore Shear Assumption)

From Ladd and Edgers (1971),

-91-

= N (I-Ko (THT 4vc vc

HvcTan = H VC

p (1-K )/2 + q/U

S= 900 - a, Fig. 5-18 indicates rotation of principal

planes.

For CK U Triaxial Compression:0

T (1-K)vc 0qi= 2

Aqf = qf-qi = c Su

vc

S q qi + Aq

qf q + Aqf

S

qf = vcvc

(1-K )2 , Aq = * Aqf

S 1-K+ [-Ku o2 o vc

D (S /=U vcCalculations are shown in Table 5-4.

Though the exact state of stress in the direct-simple

shear sample is not known at any time during shear, the

-92-

=

,1, .2

direct shear assumption probably gives one the closest

approximation to the actual stress state.

As one can see in Fig. 5-18, the difference in stress

states for the triaxial and direct simple shear samples is

not extremely large for the ranges of stress level with which

the test program was concerned (D- > 50%). This difference

reduces considerably with increasing stress level and OCR.

5.5 USE OF MULTI-INCREMENT TESTING

By utilizing the Singh-Mitchell equations,a series

of "first increment" strain-log time curves can be developed

from one test. This can be accomplished by applying a series

of three or four shear stress increments to one sample and

allowing creep to develop for each loading. If each

increment (after the first) is allowed to pass through its

transient state and begin to behave in the manner predicted

by Singh and Mitchell, an "m" value for each stress level

(D ) can be found and plotted as in Fig. 5-13. The various

log-log plots of strain rate versus time for the steady state

at large times can then be extended back to small times

(t < 400 minutes), and a plot such as Fig. 5-16 can be

-93-

obtained. From these data the parameters "A" and "a" are

calculated as in Fig. 5-4. The final data to be obtained

are the initial shear strains at t = 1 minute which are

shown in Fig. 5-17.

These various parameters can then be used in conjunction

with the Singh- Mitchell equation for strains:

Ae [ t1Yt = Yo l-m [I - (:--) t]t o 1-rnt

Since the creep parameter "m" is not constant but varies

with stress level, DT, the values of strain rate at a given

time become more divergent with time in Fig. 5-4. Although

the curved plots at larger times in Fig. 5-4 are no doubt

correct, the resulting slopes, 3, do not represent the

curves for stress levels greater than 50 percent. In order

to obtain representative slopes, a "best-fit" line must be

drawn through the data points for t = 10,000 minutes (see

dashed line in Fig. 5-4). Such a procedure yields an average

value for """ of 4.6 which is approximately the value ob-

tained for t = 1 minute.

This resultant parameter modification along with the

other creep parameters previously mentioned yielded the

theoretical creep strains tabulated in Table 5-5 for

-94-

D = 0.50, 0.65, 0.72 and 0.80. The resultant curves are

compared with the laboratory strain-log time curves in

Fig. 5-19.

The results compare quite well at large times which is

the result of the cancellation of two errors.

(1) The initial shear strains (yo) are too large due

to the fact that they are obtained from second and

third increments and not first loadings. The

discrepancy can be seen in Fig. 5-17 where the two

curves are compared.

(2) The initial creep rate (t < 400 minutes) is too

slow. This is due to the fact that for a given

stress level, the parameters used to calculate

strains are for that increment in the steady state.

In the laboratory, the internal obliquity (Th/v)

is initially higher than that applied, and there-

fore, the creep rate is greater.

As a result, the theoretical curves start out too high

but lag behind in the first 400 minutes. After that period

the two curves are parallel and in most cases, very similar

in magnitude.

-95-

A further example of this point concerning the initial

development of pore pressures can be made by attempting to

duplicate a first loading curve with modified data. If one

takes the initial shear strain from the first loading condi-

tion (y = 1.71%) (see Fig. 5-17) for DT = 0.65 and uses

the "m" obtained for D = 0.72 for the first 400 minutesT

and "m" for D= 0.65 for the remainder of the time period,T

one duplicates the first loading curve. These results

are shown in Table 5-6.

Since field problems are concerned with strain-log time

curves at times much larger than 400 minutes, the multi-

increment approach can prove to be quite useful. Strains

at much greater times can be calculated from the Singh-

Mitchell equation for strain and the initial portion of the

curve can be completely ignored by setting tl > 1000 minutes,

which would be a common procedure for field problems.

-96-

z m4-)

ZrdH (

0 0u z

E-i

I

U

C)0

OzZ0H

UHH

i-

mI:I

o oo H

o o

.. H

-Ic

I I

Cl) Cl)

U Url r

U)

0X )

Od

E

N- .C- mcvl

[-, -,-- IH cv cda 3rd * r'I

H- -H A- U

El Cl) 0

-97-

Z

O0U

0\0

00 mn m C0to f H o

'I o L

rD r.D rD ru u u

vH

oo-C**4-)

o0

ý4

C)pqu]u]

ZH

4-I

L O

4 0H H

E-'UE-

0D

I4 4I

0 MU ZN MH

C3 E E " 0U)4 -

0 -

O

H

n o mooiI H

OO

ZE-40O

*O Z

E - -1

I 0

>)n

LE--t

ClrN

0 0 O 0-L -. LO m 4

v .Lo LO U-..mcto r-i rl - toLn (Y N r - mo00 I i 0 I

0o C % 00 CNC r-mo%0 00 c r-

c0 q:: r CY) CN

r-i

CC 0 L 0 0C-oM - m m

o C

Co o o0CY) 0 LA

e ri rN m 0 ONo o 0 C 00 0

o o o 0o o

C Cc C) 0 Cc H

Cr O O HO rlo H I N c

I I i I Q

) r-) OD U U

U U U) U ) U)

-98-

a,

0*C.

-I 004

0 0-P10 +O

-HU) 0IHO O

0P-P

0 0

x 0 n ")4 LA

-P wC•)S00

m) a) 4JE 0 4 +

04 O Um 04HH -H CI0 U) O -iN +u

QE~ aHO

H

0H Lf. U)

I-

a U)0 -H

p 0

N " H

u) Hw P4OUEI H

H

0 I04 O

00u zH

U E-iO I

O U)I * E- 4

[-1

•C 0O O-- -C

0 om m CrN CLO CD Y),-I NI LOo I

Ho-~ HI 1IV HI --

0 r- 0 0 00 H 0i 0 N

o C D C) C)

;.o r1 1r-I 0O H H C)

q CO N HC LO

m O m ONC) C\ H N

+ +I o U .I I I

o1 o- co 01 01

C) C) C) 0 Qr• • r r. r

-HO4-)

*H 0H

0rX4

a-

*00 Lo-'-30 +*WA4U) 0

0 O

C)

) U 4.10 0 C)H P-P

4- d -H

( 0 ) . 4 .,

H- 0 O- - HrX4 E-1 Ia )

p 0) l ro p r

SH -O- U p

H ro 0 H l 0 H

CHO

W- - 0 4-4 -z uH 0

H (Nm tC4'WO.-, +

-99-

TABLE 5-3: STRENGTH DECREASE DUE TO CREEP

TEST NO.

(OCR=1)

a aFvc vm

(kg/cm2 )

LOAD HISTORY

(Values of D )

f =tan (T /h )f h v

CK UDSS-8*OCK UDSS-10

oCK UDSS-11

CK UDSS-12o

CK UDSS-16o

CK UDSS-20o

CK UDSS-21o

CKoUDSS-22

CK UDSS-230

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0

0.49,0.73,0.82

0.73,0.94

0.94

0.73,0.94

0.50,0.65,0.80,0.95

0.90

0.850.80,0.90

0.65,0.80,0.95

20.0

19.75

21.7

21.8

24.4

22.5

22.5

22.25

23.3

Average 22.0* vs. 21.60+

(OCR=2)

CK UDSS-17o

CK UDSS-18oCK UDSS-19

o

1.0 2.0

1.0 2.0

1.0 2.0

0.94

0.72,0.94

0.50,0.72,0.94

Average 25.10 vs. 25.10+

* Failed at Controlled Strain

+ Obtained for Constolled Strain CK UDSS Tests on EABPL Clay0at (T h)max.

-100-

24.75

23.3

27.4

N LO O . O O .O O0

H-

--

ON C0) N

r4 r~ 0

C; C; Cý N

0 00 r--l

00 H o

N. OO

> O N NID 0 * N *mD O • 6

****

O

ZOH

E-4

z

0

o

SZ:4

H

0HH4

o N

II II 0

o o zI

LC) 0 LO 00O L( 0NV Ln 'Q 40 6 0i C

Hl

M~ Lfe4o "o• uo

C) ; C CO

O c O

H NýO m 00Hn N N CN m H

OH Oo

000000tn NU MM d

od0000

O Lm O Ln O OCN Ln N- O C O

H-

O0

SIIOzt

C.

0I I

C-

u~iii ll>

-101-

O O o O OLf N" O', OH-

o\O

II 0

IIo

iko\oIIioo

U

OIQ

0O

IIto

I11

O

II

II

--

,o0

U 0O

--- 0-

ý4

I

TABLE 5-5: COMPUTATONS FOR THEORETICAL

STRAIN-LOG TIME CURVES

AeYt = o 1-m

1 m1- ( ) t ; c = 4.6, A = .014%/MIN.

D = 0.05

D = 0.65T

D = 0.72T

D = 0.80T

t (MIN)

10100

1000

10000

t (MIN)

10100

1000

10000

t (MIN)

10

100

1000

10000

t(MIN)

10

100

1000

10000

Y = 0.8%

Yo = 2.1%

y = 3.0%0

y = 4.85%

m = 0.937

Y(%)1.152

1.547

2.004

2.542

y(%)2.796

3.639

4.628

5.811

y(%)4.003

5.165

6.608

8.120

y(%)6.318

8.246

10.777

14.083

m = 0.924

m = 0.915

m = 0.882

-102-

TABLE 5-6: DUPLICATION OF D = 65%T

STRESS LEVEL

Y = 1.71%O-d = 4.6 A = 0.014 D = 0.65

T

Aet = Yo 1-m

t1 m

1- (C-) tt

m

.915

.915

(.915+.924)/2

.924

y Theory (%)

1.71

2.70

3.86

5.06

6.23

y Lab (%)

1.71

2.45

3.75

5.05

6.10

-103-

t

1

10

1001000

10000

I" N c

0

-JIo.

2z0

z W

a-w w

w

" )0

w

C,

co wD~· c z(%) 'NIV.LS 8V3HS H

C/)

U)

-104-Figure 5-1

W

w

4

Co

ELAPSED TIME (MINUTES)

Log Strain Rate vs. Log Time from CK UDSS Creep Tests on

NC EABPL ClayFigure 5-2-105-

w

wI-4

z

I-(1,

10 100 1000 1000 100,000

ELAPSED TIME (MINUTES)

Log (Strain Rate x Time) vs. Log Time from CK UDSS Creep

Tests on NC EABPL Clay

-106- Figure 5-3w- -

0 20 40 60D= l'h/Thf,9%

80 100

LOG SHEAR STRAIN RATE VS. APPLIED STRESS LEVEL FROMCKoUDSS CREEP TESTS ON N.C. EABPL CLAY

Figure 5-4

1.0

1IO

1-2

10IO

Z

.o

w

n-

10

-10

O IINEL

-107-

or IU9v--1%0

0Rho mo <

w-- O

dz0Cl)o HC')

z wZ W

UC)woE w- a

Sok

z-o2j

w " 0 I-

(%) NIV8iS V3HS

W

I-rC/)

Figure 5-5

v n Um

-108-

z

0o

Z

cr-

c)it

wI,(I)

ELAPSED TIME (MIN)

LOG SHEAR STRAIN RATE VS. LOG TIME FROM CKoUDSSCREEP TESTS ON O.C. EABPL CLAY

-109- Figure 5-6---

w

w

zI-

(I,

K)10 100I 1000 o10pOO 100000

ELAPSED TIME (MINUTES)

Log (Strain Rate x Time) vs. Log Time from CK UDSS Creep

Tests on OC EABPL ClayFigure 5-7-110-

1.0

-I

10

163

I-.

w

LT6

0 zo 40 60 80 100

Log Strain Rate vs. Applind Stress Level From

CK UDSS Creep Tests on OC EABPL ClayFigure 5-8-111-

U)

~~00 0

41

S0U

Os-Ir)

0

O

a)

0

0

ar

Figure 5-9

-112-

uJwI-z

w

z;aIw•9

ELAPSED TIME (MINUTES)

Comparison of Plots of Log Strain Rate Vs. Log Time

for NC and OC Samples of EABPL Clay

-113- Figure 5-10

Eo lb

d7

o9o)

a.0 m

L.L w

a<)I

C,)C)I-lLL)O9C')

LL0

z0OCV)

[1-

>:r

C.

Figure 5-11

ýýJJE:ý

-114-

U)

aC

ro

0rdU

0a

$-4-o HH 000

-4-i ,4rr aO 4

0

U0

UO03 a

*

It)

a Is

Figure 5-12-115-

wwW

OU.C)Cl)0

0I--Jww

=E

O-jU)Cl)

SW.U)wE

ww

0~crwwcr0

LL0

z0

I-44r

Figure 5-13

-116-

I \S I I I I I II I I I I -I ILines from first loadingsof other tests at

corresponding stress levels

(95

K _

Stead state3 ) condition

CKoUDSS -16;c = 2.0 kg./cm W, 73.4%0

4 ( ) =values of I,%

INCREMENT 6,,,% SYM

I 50 02 65 A \3 80 -E

5 4 95 0I

I

22 100ELAPSED TIME

1000 10000(MIN)

LOG SHEAR STRAIN RATE VS. LOG TIME DURING LATERINCREMENTS FROM CKoUDSS CREEP TEST ON N.C. EABPL

CLAY

-117-Figure 5-14

z

0o

w IC

n-

z

I

<)

It"

rn

10d

-210TEST

CKdJDSS- IOII -II

II -16

i, -17

1" -197, -20

i, -21

-7

II-

S('9)

(10)

INCREMENTNUMBER

700

94

94

50

94

9490

85

(r,

-7

OCR

1.01.01.0

2.0

2.01.0

1.0

(20) (10), kI,,.

(19

_L

(17)7(19)

/' T 1Steady state

r Typical final increment/no steady state

(16)

Typical first incrementnot leading to failure (•r = 50%•

STEADYSTATE

NO

YESI,

100ELAPSED TIME (MIN)

SYML

1000

LOG ' t VS. LOG TIME FROM CKoUDSS CREEP TESTSEABPL CLAY

ON

Figure 5-15

c

4-

0<l10

tlO

w

10000

(20)

A•iiialoop

(-ill,. 21)

1/ I

110

IIE\

I _I_

P

w

I~------1

I I I

' 4

I II I I i I I

u

\IPI

loe

.00ý

'e

D5

I I

-118-

91

o000 I1ooo

ELAPSED TIME (MINUTES)

Plot of Log Strain Rate vs. Log Time for Steady State

Creep at Various Stress Levels of EABPL ClayFigure 5-16-119-

WI-

zi

I-

4

z

U)

Q)

S --H

Lo

(1)U)C)

IIIU)

C,0)

H

x

4-

ý4N

4-)

CHr--I I

·ro.rq

4-)k0)U)3

H

H

to k"N/@'I~a12 Wb7y5~ 7cY/.1/A'

Figure 5-17-120-

CONDITIONS IN THE DIRECT SIMPLEASSUMMED PURE SHEAR

0

SHEAR DEVICE-

Figure 5-18

100

80

60

o= Aq (%)&qf

40

20

0

80

60

0C

40

20

C

STRESSc~r h h/hfl

-121-

I 1 i1

Q

\-o\8\

\\

rI

0

U)

0

,-Cd

U)

o3U

a

w 4

.- U

4-)ýr0

E-w

00

-,-0 )

OHr)

Q4 i0c 4)

C< C(%) NIV.IS

Figure 5-19

-122-

VI CONCLUSIONS AND RECOMMENDATIONS

6.1 APPARATUS PERFORMANCE

The Geonor Direct Simple Shear Device has proven to be

a convenient tool for undrained creep testing. No modifica-

tions were necessary as the device is equipped with a "fric-

tionless" load hanger for controlled stress testing. This

is contrasted by triaxial creep testing where constant

temperature precautions are necessary in order to stabilize

pore pressures, and a number of sample area corrections must

be made throughout the duration of an increment load to

insure the condition of constant stress. On the other hand,

all dimensions of the simple shear sample are kept constant

during shear and there are no great quantities of water in

the system to be affected by temperature. In addition,

there are no problems with membrane and equipment leakage

as there is with the triaxial device.

While advantageous in many areas, the simplicity of the

device also presents a number of problems. The three major

drawbacks are:

-123-

(1) Lack of knowledge concerning the sample state of

stress;

(2) Lack of knowledge concerning sample pore pressures;

and

(3) The unsteady state which exists when the initial

shear stress increment is applied and the first

major adjustment is made on the vertical stress.

6.2 TEST RESULTS

The Singh-Mitchell Rate Process Theory with modifica-

tions can be used to model creep behavior for EABPL clay.

Results show this clay to have great creep potential (m < 1.0).

It was found that, contrary to the theory, lines of log 9

vs. log time at various values of DT were not parallel

(m = constant) but were divergent with time (m = f(D ) (see

Fig. 5-16). Normally consolidated and over consolidated

samples behaved similarly with OC samples more creep

susceptible at large times.

Additional load increments (values of D ) behaved

much differently from initial increment behavior. Instead

of yielding a linear relation between log i and log time;

the strain rate initially was much lower than the linear

-124-

relationship would yield. With time this new curve would

converge with the linear relation and adopt its behavior.

Apparently, the previous increment(s) had installed within

the sample a greater structural stability than was initially

present. The new increment had to overcome this new stability

before first increment-like behavior could develop.

Unfortunately, no empirical relationships could be developed

concerning this initial transient behavior.

It was found that by applying several load increments

to one sample, one could reconstruct the equivalent first

stress level behavior for each increment applied. This

can be accomplished by allowing each increment to pass through

the initial transient phase and establish the steady state

first increment-like behavior. This established linear re-

lationship between log j and log time can then be extended

to all times to obtain equivalent first load performance.

It was found that loads of 85 percent or more cause

creep rupture. This phenomenon was due to high pore pressure

generation rather than a decrease in friction angle, •.

-1This angle (defined as ( = tan (Th/v )max) was essentially

the same as that obtained from controlled strain tests.

-125-

The creep theories of Saito and Singh-Mitchell were

found to be somewhat useful for predicting failure in the

laboratory. The Saito relationship was able to yield a time

range for failure which included the actual time to failure

in all cases. Unfortunately, this time range which covers

more than one log cycle of time can be very large and on the

order of a week for large times to failure. The Singh-

Mitchell Theory was somewhat of an improvement but only applied

to first load failures and to those additional increments

which achieved steady state creep prior to failure. The

average value of (it)f (the value of 't at which there is

a sudden increase in slope) was found to be approximately

5.0 percent. For the five cases in which steady state

creep existed prior to creep rupture as defined by Singh-

Mitchell, failure occurred from 30 to 1700 minutes after

attainment of (it)f = 5.0%. The test with the upper bound

time to failure was the lone additional increment test

which achieved steady state prior to failure. The Saito

time range for this test was 2800 minutes. It was found

that the average strain attained prior to failure* in the

five tests for which the Singh-Mitchell creep rupture

* This strain is defined as that attained when the log-

log plot of -t vs. time has passed (ft)f and becomes

nearly vertical.

-126-

_1

theory applied, was 20 percent, which is approximately the

average strain at failure for the controlled strain CK UDSSo

tests. Apparently, the mechanisms which bring about failure

under both controlled strain and controlled stress creep are

somewhat similar for the strains as well as the obliquity

(Th/Uv ) at failure were the same for the two failure modes.

Major drawbacks in the program were:

(1) The inability to ascertain accurate values of

D(=Aq/Aqf) for the various stress levels; and

(2) A great change in creep behavior after approximately

400 minutes of load duration.

The first problem is due to the stress system and made the

selection of - for use in the theory's equations a diffi-

cult task of questionable success. The second problem,

the author believes, has to do with pore pressure equaliza-

tion in the sample. The rapid application of shear stress

increases pore pressures within the sample. Reduction of

vertical stress maintains the sample volume and the average

sample pore pressure is in equilibrium with the sample

stresses. Unfortunately, pore pressures at the sample edges,

which can dissipate more quickly than those on the potential

-127-

failure plane, are now too low and those near the center are

too high. Sample deformations which are more strongly governed

by the stress state within the interior of the sample (i.e.

on the potential failure plane) are proceeding at an accel-

erated rate. Therefore, the linear log strain rate versus

log time relation for an applied stress level, -T, is too

high as the obliquity (Th/Iv) within the sample is greater

than that which would normally result from the applied

stress level, DT. Only after these excess pore pressures

dissipate does the linear creep rate relation drop down to

that level which is representative of the applied stress,

D . All data obtained prior to this event is misleading

and presently of little value.

6.3 RECOMMENDATIONS

The two major problems encountered during this investi-

gation were concerned with a lack of knowledge of the stress

state within the sample and measurement and dissipation of

initial pore pressures within the sample.

The first problem has been the object of considerable

effort with little resulting success and the author has

-128-

no suggestions for improving the situation. The second

consideration can either be handled by measuring this tran-

sient excess pore pressure or eliminating it. Measurement

would indeed be difficult and at this time not practical.

Elimination appears to be the only reasonable recourse. This

can be accomplished by either applying a desired stress

level in smaller increments over a time period within which

a major portion of these excess pore pressures could dissipate

or else loading the sample by means of controlled strain

until the desired stress level is reached. After transferring

the proving ring load to the hanger assembly, the creep test

could be commenced. Unfortunately, the data obtained during

the execution of these two techniques is also of little

use so one accomplishes nothing. Apparently, the best

procedure is to ignore all data obtained prior to t = 400

minutes.

Other aspects of the program which need further investi-

gation are the effects of OCR on creep behavior, long term

testing (t > 1 month) and creep rupture behavior. Any

program designed to investigate these problems could be

very useful.

-129-

VII REFERENCES

1. Arulanandan, K.; Shen, C.K.; and Young, R.B. (1971),"Undrained Creep Behavior of a Coastal OrganicSilty Clay". Geotechnique, Vol. 21, N8.4,pp. 359-395.

2. Bishop, A. and Lovenbury, A., (1969), "Creep Character-istics of Two Undisturbed Clays". The SeventhIntl. Conf. of SMFE, Vol. , pp. 29-37.

3. Bjerrum, L. and Landva, A. (1966), "Direct Simple ShearTests on a Norwegian Quick Clay". GeotechniqueVol. XVI, No. 1, pp. 1-20.

4. Brooker, E.W. and Ireland, H.). (1965), "Earth Pressureat Rest Related to Stress History". CanadianGeotechnical Journal, Vol. 2, No. 1, pp. 1-15.

5. Christensen, R.W. and Wu, T.H. (1964), "Analysis ofClay Deformation as a Rate Process". ASCE JSMFD,Vol. 90, No. SM6, pp. 125-156.

6. Edgers, L. (1967), "The Effect of Simple Shear StressSystem on the Strength of Saturated Clay". M.S.Thesis, Dept. of Civil Engineering, M.I.T.

7. Edgers, L. (1973), "Undrained Creep of a Soft FoundationClay". Ph.D. Thesis, Dept. of Civil Engineering,M.I.T.

8. Fisk, H.N.; Kolb, C.R.; and Wilbert, L.J. (1952),"Geological Investigation of the AtchafalayaBasin and the Problem of Mississippi River Diversion".U.S. Army Engineer Waterways Experiment Station,Vicksburg, Mississippi.

9. Foott, R. and Ladd, C.C. (1972), "Prediction of Endof Construction Undrained Deformations ofAtchafalaya Levee Foundation Clays". M.I.T. Dept.of Civil Engineering, Research Report R72-27.

10. Kaufman, R.I. and Weaver, F.J. (1967), "Stability ofAtchafalaya Levees". ASCE, JSMFD, Vol. 93, SM4,pp. 157-176.

-130-

11. Kolb, C.R. and Shockley, W.G. (1959), "Engineering Geologyof the Mississippi Valley", Trans. of the ASCE, Vol.124, pp. 633-645.

12. Krinitzsky, E.L. and Smith, F.L. (1969), "Geology ofBackswamp Deposits in the Atchafalaya Basin,Louisiana", U.S. Army Engineer Waterways ExperimentStation, Tech. Rept. S-69-8, Vicksburg, MS.

13. Ladd, C.C.; Williams, C.E.; Connell, D.H. and Edgers, L.(1972), "Engineering Properties of Soft FoundationClays at Two South Louisiana Levee Sites".M.I.T. Dept. of Civil Eng. Res. Rept. R72-26.

14. Laing Barden, "Time Dependent Deformations of NormallyConsolidated Clays and Peats". ASCE, JSMFD,Vol. 95, No. SMI, Jan. 1969, pp. 1-31.

15. Landva, A. (1964), "Equipment for Cutting and MountingUndisturbed Specimens of Clay in Testing Devices".NGI Publication No. 56, pp. 1-5.

16. Mitchell, J.K.; Campanella, R.G.; and Singh, A. (1968),"Soil Creep as a Rate Process". ASCE, JSMFD, Vol.94, No. SM1, Jan. 1968, pp. 231-253.

17. Mitchell, J.K., Singh, A. and Campanella, R.G. (1969),"Bonding, Effective Stresses, and Strength ofSoils". ASCE, JSMFD, Vol. 95, No. SM5, pp. 1219-1246.

18. Murayama, S. and Shibata, T. (1961), "RheologicalProperties of Clays", Proceedings of the 5thInterntl. Conf. on Soil Mech. and Fdn. Engr.,Paris, Vol. 6, pp. 269-273.

19. Shibata, T. and Karube, D. (1969), "Creep Rate andCreep Strength of Clays". The 7th Intl. Conf.on SMFE, Vol. 1, pp. 361-367.

20. Singh, A. and Mitchell, J.K. "General Stress-StrainTime Function for Soils". ASCE, JSMFD, Vol. 94,No. SM1, Jan. 1968, pp. 21-46.

21. Singh, A. and Mitchell, J.K. (1969), "Creep Potentialand Creep Rupture of Soils". The 7th ICSMFE,Vol. 1, pp. 379-384.

-131-

22. Saito, M. and Vezawa, H. (1961), "Failure of Soil Dueto Creep", The 5th ICSMFE, Vol. 1i, pp. 315-318.

23. Saito, M. (1965), "Forecasting the Time of Occurrenceof a Slope Failure". The 6th ICSMFE, Vol. II,pp. 537-541.

24. Saito, M. (1969), "Forecasting Time of Slope Failureby Tertiary Creep". The 7th ICSMFE, Vol. II,pp. 677-683.

25. USCE (1968), "Interim Report on Field Tests of LeveeConstruction, Test Sections I, II and III, EABPL,Atchafalaya Basin Floodway, La.". Dept. of theArmy, New Orleans District, Corps of Engineers,New Orleans, Louisiana.

26. Walker, L.K. (1969), "Undrained Creep in a SensitiveClay", Geotechnique, Vol. 19, No. 4, pp. 515-529.

27. Watt, B.J. (1969), "Analysis of Viscous Behavior inUndrained Soils", Sc.D. Thesis, Department ofCivil Engineering, M.I.T.

-132-

APPENDIX A

LIST OF SYMBOLS

Note: Suffix f indicates a failure condition.

Prefix A indicates a change.

A bar over a stress indicates an effective stress.

Stresses

q-9

h

h

v

01

a3

hc vc

VO

vm

T

max

(a - ah)/ 2 or (a1 - a3)/2

(v + Uh)/ 2 = (a1 + 73) / 2

Pore Water Pressure

Horizontal Stress

Vertical Stress

Major Principal Stress

Minor Principal Stress

a hF v at Consolidation

Initial In Situ Vertical Effective Stress

Maximum Past Pressure

Shear Stress

Maximum Shear Stress

-133-

LIST OF SYMBOLS (Cont.)

Stress Ratios

q/qf

DO Aq/Aqf

Th/Thf in the Direct-Simple Shear Test

--h/cv or --hc/vc

OCR Overconsolidation Ratio =

/vo or a /vmvm vo vm vc

Strains

Linear Strain

Shear Strain

Strength Parameters

PU,' 4' I f

max

and Stress Strain "Constants"

Young's Secant Modulus for Undrained

Loading in Terms of Total Stress from

an Undrained Shear Test

Undrained Shear Strength

tan-1 v ) at f in the Direct-Simple

Shear Test

at (T/ )mv max

T at Creep Rupture

-134-

sU

LIST OF SYMBOLS (Cont.)

Strain Rate and Creep Parameters

E Strain Rate = de/dt

Y Shear Strain Rate

tr Time to Rupture

tf Time to Failure

t Time at Which Strain Rate Becomes

Constant Prior to the Development of

Creep Rupture

A

D

D (or -D, D )

Dmax

m

El' Y1

t

tl

Creep Parameters in

Singh-Mitchell Relationship

See Section 2.1

-135-

LIST OF SYMBOLS (Cont.)

The following symbols are used in the derivation of the

Singh-Mitchell relationship from rate process theory.

Shear Force Acting on a Flow Unit

Planck's Constant = 6.624 x 1027-1

erg - sec

kT AF2 x exp (- )h RT

Boltzmann's Constant = 1.38 x 1016

erg K

Universal Gas Constant = 1.98 cal °K-1-imole-1

Number of Flow Units

Absolute Temperature, OK

Reference Time

Parameter Relating Activation Frequency

to Strain Rate

4SkT

2kT

Free Energy of Activation, Calories

per mole

Distance Between Equilibrium Positions

of Flow Units

-136-

S

T

tl

AF

LIST OF SYMBOLS (Cont.)

Miscellaneous

C Virgin Compression Indexc

Cr Recompression Index

CR Compression Ratio = Cc/(l + e o )

c Coefficient of Consolidation =v

k v/(mv w )

e Void Ratio or Napierian Base 2.7183

e Initial Void Ratioo

k Coefficient of Permeability

log Log to Base 10

N.C. Normally Consolidated

O.C. Overconsolidated

t Time

w Water Content

w Natural Water Contentn

P.I. Plasticity Index

L.I. Liquidity Index

-137-

APPENDIX B

TABULATED TEST RESULTS

-138-

CKoUDSS - CREEP TEST

SOIL DESCRIPTION EAB PL CL• YBORING No. 9 UEgsSAMPLE No. le/ELEVATION -36.7 "rINITIAL CN 73. a /o

TEST Na CK/UDS- 8DATE /0 -,Z/ -7/

CONDUCTED BY CEWvc( KG/CM2 ) Z. 0

OCR /-0PRETEST Ht.(MM) .2z379

DATE VERTICAL VERT. " (0/6) Y(%/MIN) ht(%) ELAPSED APPLIEDME PROVING RING STRESS TIMEN PER CENT)

(MINUTES, (PER CENT)/0-2-7/

/.'oo 3057.6 2.0 0 0 0 o . 9

3o05 /89 1.,'/5 . 538 ,/1~8 (.2- ,)

,99 1.9 Z3 o220o ./3,• 6

-2986 , /07 ,o o9/5 ,.o0/3 2 z

296t /.80o /.29 ,00o39 .,oks /7

2 9 138/ ,0oo0229 .19,3 8s

29YS 1 , ,7oo/85 .o8/, i2.5

Z 9 15 /1,54 .000oo93 ,/735 /18

2928 /1,73 1,78/ , 000o60 ,992 18q3

29 1/8 /,882 , oo.Z5o ,/ S/ 660

289 93 Z, 69/ o000/3,9 -. 331 /6go

_ 28.93 .2.17/ ,00009/0 ,2225 .2 7/5

.2 87-4 /.6bZ ?2.;20 ,o0,•bl .1099 32752 8 76 0, 26/ .O00 0o•0 . 3327 q /&S

284&0 4 -.325 .0ot09.29 Y-ZsI/ y-5 7529_59 1s_9 z, 399 .O0'tS8,6 .••95(, Sov/c

Zs,6o 2. 387 C000s77 .32Y9 r-.25

z29 8 ..-399 , 0oc338 .A.o31 o 0-5

_2_ / 1.5-6 , 34 -• , - 1 7080

1/-7-7/

&3n /,-56 ~. .3. 3 ---- o 72.9

28. a.688 .098_ 3 ,o98Y3 K (35,o Kc)

8 1,, ;2,731 . o/8i1 ,osS. 3

_8__ Z 2780 ,0/5y ,ofgY 4

28 y 2.8.29 .o/o93 ./1093 o

aS _ a3.020 .00o67/ .2/0223 /SB-ze 3,o0a .20067 ,Z/53 38

Dr = T (APPLIED) / T MAX.

-139-

PAGE Z

VERTICALPROVING RIN(

2 792.

-P769

DATE/

1300

Z 83 7

.8 26

• 9I

za Ii

z 7o80 s

27790*2785S

27 814

a 7 82-2Z 777

277/

27656

a 7 65a2765

.2765

:yss

VERT.STRESS

I.S2

/, /93

I, 3"93

1.393

/ ,0

Y(%))

3. 099~

$.s4 /3. 78•

3. 752,

3.8/49

,,g/.o5.o23

9. 269

9·72

-. 71

(.98

s. 56

s-. 8, S"

. 333

s.'1255.523

5.523

o603

61063l

(53

67/

733

6,oa7

6.071

6.102.

b. I (44. 2•

*(/(oMIN

00 2,360

,& 29 1

.00 197

,oo00 19

.0012o 5

,0oo009,000 6,31

.000519

,000 3SS,

,0O06785

S0ooo000 7

, o•123 ),oo219'00219

,002-2-

,oo loG

, oeo 4 71,ooo 6.

, o0 332,

,000 2,63

.000. 21.000241

, OOo .2.

,oo/ 7y, 0o00 73

ELAPSEDTIME

(MINUTES:

53

t(%)

.1/909

.'t353

.5,308

.s1 I

, '980

,72"/o

,609(,qgag

q400

580-5J 2,0.5

7156

IS

so

260

//25

1330

/6.5

.2.545

APPLIED 1:(PER CENT)

-140-

/59

.270

.3/o

.3.S

4b6

7o 9729

/395

Lf6s(

.0/3/6

.#3399

,03979

O.302-7.,3o2•7

.38•36

3552

.3423

,3/o .'310Yr C

.so 2

(.39, . /., )

I

PAGE 3

VERTICALPROVING RING

S760

DATE/ METME

17r(o 02761

1761

.2- 76-17 '77

ý2 7s k _A 7L9 I

.179A

VERT.STRESS

/,38

I. 3L

; Si9:A/3t 9SR99

Y(%/)

6. 366

6(S38

ý, 58 1(c.64.7b. 670

6.785

61872-

(, 877

Q =1

)ý%/MINJS000 /- Z

,oool2.-7, oo /27,OmoI 2-

,0000 90qYoCc gog,ooo o&S

I-

ELAPSEDTIME

(MINUTES)

ý897

Yt(%)

,S'1/33

,/398

,5334

, 5.a23

,'777

-o - ---

APPLIED D(PER CENT)

-141-

L-1275

•9cso

6900

7Bas

8 310

8 _L~Zio

CKoUDSS - CREEP TEST

SOIL DESCRIPTION EABBPL CI 4YBORING No. 99 uVsSAMPLE No. / 3ELEVATION -15 FrINITIAL WN ,p . 9,

TEST Na _ua ss -/clDATE i- 3-7/

CONDUCTED BY C

vc(KG/CM2 ) 2.0OCR /.o

PRETEST Ht.(MM) .2 -0o/

DATMVERTICAL ERT. 'Y (%)z (MIN)() ELAPSED APPLIED NMD MEPRING RING STRESS TIME R CENT)

/•-2 Y- 7/CENT)75- .3 5os2 0 2 .o 0 72.9

3osz7 _ __ bo /_5/ 7 /2 7 / 's. 5 I

302o. _ 3.o33 I 2285 , ISo 2,

29 75 / 83 3.2 f9 ,u/o. ,70 2 5

.90oo , ,7 0/7 ,o 7860 ,7o7lY 9

A-883 /11/ ;,39 ,057oss ,7725 /52883 , be 7 , 0-250 9,Z96 222 S83 -,3,7 . 2 -08 8 Y J/ 30

288 3 o83 O;Z3 ,78 .838z 3

z2_.3 £7.2 ,02079 ,.938 93S 883 537 / ,O/?y2 ,6723 so

._ _I. G .42 .~ 7 , o/s-I/ ,9go/ o6

_8_-&_ s2472, .o0/2z7 ,2733 70

s_-_ $ 8e8 , 0099/ .a23 go0

-.••-0 157 6 IA,/2 ,oo777 ,o90aS 17

__ ,288 ,00oo53 l,9/Y /9o

Y y7 _ ,Y/_ .I/o , I/8 ,9 77/ 1,58

Sp•4 / /,5 ..Y56 ,ooo/89 ,639/ 172-

a'89' b __596 9 , po a/'7 , 87/1 /9•

c 279 6 / Yb ~-998 ,o63 e9 ,Zs-y Z99

,..Z297 7./68 0oo007•4 .3997 5-22

,772/ 9.10' , 00 67i ,936 /395

2195 8,/97 , o0sS- ,5239 /6/5

2.•9-/ &.73 ,ooViy .%,C28 /870

S 7289 P.1 P/b ,oI 9 000/6 "5a/?5-0 /1,36 Pi 6/, 14,000 /1(c -5v 3 6,I; .- 's, I AzoOO .L ..,,7 t-/1./-,,,

Dr= T (APPLIED) / r MAX.

-142-

DATE/nTME

7t;

PAGE 2,

VERT.STRESS

/. 3(.

VERTICALPROVING RING

2 75/

2,76,

27(,3a72V

.2 72 S2-7 (C

276327 &63

27 62

-222327552 ).0

. 700

/.37

/.39

/.3 2

X :t•-8

R. 77?

9. IC6

9.157

9,.87

2. 67f

9. 7989,W-9,926

/o. 25;2

/0. 6/9

// 98/7

12,929'/

13S7

.ooo i37

'000 12S

,ooooS'72

,0 23 2- P

,o /7'9.?/2,70

,00 b17

,o060-3,oo60/7

, o 582

,oo 629

, oo 699

,oo 82, o lo69

0 /0oI

, 0 /2a 7

Uz E

.X324

,25-2B

, 3043

,Y523

.52935

,6// z

2,/32.

3.34L

--- --- --

ELAPSEDTIME

(MINUTES)

ys30S&. 2.28

.6295

,'I097

.(,68o

, 329

.oqs, S5,o67253

, "3.lo5

33

Y/5-0

1,o0

75

38

I03

/3 /

34

53/

APPLIED ]•(PER CENT)

9"/, //-(.'52 z

-143-

so Y•-

/o 06

I

2

9/6

,• •

CKoUDSS - CREEP TEST

SOIL DESCRIPTION •ABPl. c ,9 YBORING No. 99/uESSAMPLE No. /314ELEVATION - 42, 7 FrINITIAL WN 72.2 %

TEST Na C K0s55-/IDATE /2-/o-7/

CONDUCTED BY WcWvc(KG/CM 2 )

OCR .oPRETEST Ht.(MM) 2-• 3/61

VERTICALPROVING RIN

30526

DAT% ME

/2-1/o-7/

/;--/2--7!.x,3o

VERT.STRESS

2.6

/ 774,/. 71

1,33

/,23

/. /3 t

r (%)

. 21/6. 78/

2.5/7

/. o, 7'

1o. 663

/7, 987

/3. 3/o7

/3 8 7

/l. 270

/7.97

/8867

(%/MIN

,,. 9b P

,325o, "197

, /0230

.0b 75

.03 &(72

.0366/

,02.70, c'293

,0 2J1

,O IA6 o

* t/ u

/,30]

b S83i, 8 (..

I, V/5/1"02

.07 C,

2. Sa

-7L 39.Alt

3.3 oc

,3. 803,535-

/.3 /o0

ELAPSEDTIME

MIN UTES

0.s-

/

q/127/7

28Z23

3o

//9

APPLIED NC(PER CENT)

y4/

Dr= t (APPLIED) / T MAX.

-144-

2950

19/8

2-800

e771

7 7/4

2 7.2-7Y7

~z. 737

2730

27036b87

.2- /

.2.6 0

a•&o

CKoUDSS - CREEP TEST

SOIL DESCRIPTIONBORING No.SAMPLE No.ELEVATIONINITIAL WN

EI FQBPz- CL•)Y9q u='3/)

- 'L2,9 P7' ," ,

TEST Na C•KIDSS -/2DATE I2.-2--71

CONDUCTED BY e Wrc( KG/CM 2 ) .o

OCR 1oPRETEST Ht.(MM) 2"; ./Y9

VERTICALPROVING RIN(

o30z57

3 osr 2

os7,

30//5

1909

49 00

28~oo

S8220

2 z20

28/3

2 797

277X

2 72//9

; 73 7

.. 22Z9

VERT.STRESS

2.0

/9/

Z.o

4-5-7

//,33

/J19 ?

Vr (%]

/ ?Jo

2.233

2. 88/

3.o0'/3- 220

.3. 39o

3.68933- L89

3~866

(, /7I

(9'32

( ,/8

,77 69

4,77/

DATM

I -7-"

X(V/MIN

/ 71? 31

0. 30S7

0,17//

&, /03 7

, o.3/• •

S0/ //70

,oo 1 53

,000608

,00030•/

,OOO/39

P0O0 683

,0000s56

ELAPSED

(MINU ESYo3%S2

o.3y$7

,5327

,5,33

.bBOc.6238c

59 82.

,6900

7333

6P03

,7/88

,ss5oT

gbooo

. Mo?

APPLIED 1(PER CENT)

72, 9%( f*h J/,lcv!4

Dr= T (APPLIED) / r MAX.

-145-

0.5

7

16

30

Yo

o30J90

2 70

i-5-79

2 82 7

13/o

573 o

7/92,• LE

&-o2. .

- 7 /,

PAGE 2

DATE/M

/-3/-- 2..

...

;270 /

22;0/

2-7O/

27 o/

270/

Dzo /

220/

-22 77

2997

b2413

Z6 -34

263 :

as 5 j

VERT.STRES!

1,2/0,zo

/,26

VERTICALPROVING RIN

-A 7/9

270 /

2 -o /

,&cv o ~;r-

,06a90 /92

7, 632

7R1.27.7 1

78 7/

7. 28/

8.o3 /

. .9o

. /165

-/, 7/8

9.122

2. 216

3.609

·,.lsif /

9.223

, 12- 2 1y _

, 0 /o3-SY

, loos-9,Do7"32.

,0o . 22

, 190282.

.oo £2681 00o289

,003/7

,0030,0

,oolb

1 00326, Co 28"00329

, o 9'/8,00 39i.oot

,g23

,92-3

,1-29,0572ý

aL- 1

,-701(l

,/563

,>.,Pyoy

,75/31

012-3

/,/9'6 /

.33e

.357

2307 713

2/6

ELAPSEITIME

(MINUTEs

/o o63

o30 A/0•

/

la 7. -?r

2yo

725

goo

/o3o

650h/2o

APPLIED 'U(PER CENT)

s vf r ./)

qz/. /c7.

-146-

, 0ooo0/00

.r.

,Sd 6~

72,5

c

~----

I ,-,ý6 3-5- -Pl

"'

CKoUDSS - CREEP TEST

SOIL DESCRIPTION EA-13PL CL-A YBORING No. 9,Vu7-rsSAMPLE No. //2ELEVATION - /-o,5 FrINITIAL WN 73 Y- %

TEST No.& C,4aS -//DATE 3 -I5-7_

CONDUCTED BY cErw"c( KG/CM 2 ) .~.0OCR /o

PRETEST Ht.(MM) 2• /73

VERTICALPOVNG RN

3c> , 7 K

DAT ME

S-z;-72

73-5

VERT.STRESS

.2.o

SIB

/ 88 ,

1,71

/,67

/,S7

/, gy7

/ s7 1

7 (0%

o 976

/.2~

I, s-31, 5f•3I. 77

7.2

Y

3, 073

(%/MIN

0. 1?b,

r, o27

0,01-3, 0//3

/ 9023

'Z23 7

,000 067

.pooo0 -/,O Oo 4 9 1.ontThl ?, oo /"^ 00 0 /

5o3s 2 &

J3c- 7, Z

7:? 0o

2 ?B/

2-- s-

2 1/I2

2933

.2,/o 2.

-2 I F28 71

28 y7728 47

a, 7B8

., lSY

£12. /

,332 1

,AY7, X3'03,S•71.•.V31

,),qo3

ELAPSECTIME

(MINUTES

o, •

2

7

.2.0

3. o

x/72,

3385

r63 -3•zSo

s-PS3

APPLIED A(PER CENT)

-5- ,,

Dr = r (APPLIED) / t MAX.

-147-

7 277 ,oo,

yy1I

VERTICALVERTICAL

PROVING RIN

?r239

- 8 /762_8?

DATE/TIM

.-- Z-3

2-

2

7

8

28 og

A60)

a 7 Gs5"

d/s/

a -) 765zeoc

PAGE -2-

VERT.STRES.

A 153"

/, '13 /

/. _5/,~s'-

//. B' ..

(a/c

3. 3'?

3. 3 ?

3, •5"

3,.3;Z

-, 702

;. 72.

,/72_

1333Y,

Y, sa•

/ 768

'Y €3 7

•.0/t'/

,0000o3 2

o, /31o

, oo 233

, Oo3 7

,00i,7o 006 8,000 596

, 000700

Oo5Y-'2

-po0o6/

,doo ,gg?

, oo0/3

gooo//Y/

, cooo o3g, oooo8 •,o00o b5?

,000o632-

0~/o?

,238/

,3/3/

'om,0(73

/0-00 j

ELAPSEITIME

(MINUTES

7-1139

1O on "-

0

2

7

/ 1

-3 l

0!A i c

.23 3

-283

-2 83 3

,572

J/L%

71o1130/7/-5o

7790

SZo

2.

I

APPLIED t-(PER CENT'

So s/

-148-

s2&04f7

/352

3768

S75•y

0/07

/,foI

_ /,

.

I I -

-, 00,-298

oog&g7 ,0/7.3o009o .o0,3ly

65 ~o

~3"~9~m~

I

PAGE J

VERTICALPROVING RIN(

z . >eo

DATE/TI ME

Y-11-72- 2_.

; 2' /

2760

Z? 7> )

.2 2 3

.27~ 29

27 7,z

A 75•

2 7 Yo4

27 22/

2 72/o

1716

a.72 /

;t7 /

/ Z 9

VERT.STRESS

J, /2-

S35,

l, 3o

-Jo 2

Z,,30

Y(%)

./a) C

\L.oyo

-.0 2L5o

•.334

, 21 23

R. 70

.5, /,o6.357

,0& B79. 053

I0/,571

.o/.-I.N

,0 0• •"£

, oo efI,00 / ?

n9o ./.

_ 0o77Z

, Oo13Z

S00o /3z3.ooo 13

,0, 1.20

,0o,- 7

,o/o77,n /Z /1

,· o sy~/

,oob:,?

20 I23o &3 63

ELAPSECTIME

(MINUTES

,o043

,.25-Z

.5238

,5-Z37

1 .3/i3

1,2027

/1377A-/,o'•7o.yoy

o08"

. YS-2f,3/6,fJA/t

APPLIED :(PER CENT)

Io/0o

/320loo/723

09388633

/o 0o3

o

o.,r/

712,

30

27•

ty/3

-149-

YS,f ign~

PAGE -

VERTICALPROVING R1NW

,Z7 Ic,

DATE/ METME

tf--I 21--72.Jo 2r

2Q0 3

Z6 5

MO 3,:2 68 cl5

25 7 7

VERT.STRESS

/I 2/1,25

1,0 oc

(%/MIN.

oo 0 "

m~o)

11. 71-3

12. ?Zo

_12. 3)0

IL- B2

19i. 32VQ1'a/u.i go_L 8R3

)rt(%)

3.Y12

~.ý 72t,

•,/

ELAPSEDTIME

(MINUTES

0T

39/

72-

/ Y0

APPLIED "r(PER CENT)

-150-

* oc 7&

o0p 71, oo 73 4,Oo 2ý3

0 o 7/ 0

DL 7•b00 00 &-y.0 i) 3 1

CKoUDSS - CREEP TEST

SOIL DESCRIPTION E/•SPL C/-A"YBORING No.SAMPLE No.ELEVATIONINITIAL W"N

- '¼.O FTr

79. / 0•

TEST No Cooss- /7DATE 5- /i - 72-

CONDUCTED BY _E W'vc(KG/CM2 ) /. o

OCR 2.PRETEST Ht.(MM) ,J-R-g9

DAT ME

5-10-72-7 LS

ELAPSEDGTIMF E0NUTF.

VERTICALPROVING

APPLIED a(PER CENT)

(. i/

D-T= t (APPLIED) / T MAX.

-151-

VERT.STRESS

2.S7. " -

abo 2b

IG I

6t, 1 r

.t 2S

as L

&793

7 (%)

o

7, So3

//,~ s

/2.21/7

IY.17-. /a7

/. O

/.0•

1 /o3

/ 03 5

0, 969

Y(VMIN)

',3 2B/.,1/ 7 6

, , /0,S

C, /6o3-t', 2-7/

-),d

3.127

3, 092

2 222-

lg. "/

8

12-

aa(oS

CKoUDSS - CREEP TEST

SOIL DESCRIPTION F4ABPI- (~.LA YBORING No. Q u_ - sSAMPLE No. /3 4ELEVATION -_ /__" FTINITIAL WON 7

TEST No C~D~SVS-/8DATE 1- /--72,

CONDUCTED BY =s Wv

avvc( KG/CM 2 ) /.0OCR 2.0

PRETEST Ht.(MM) 23. I//

VERTICALPROVING RING

-. 575

262. Z

2Z•33

26 439

2633

.217

•30

2 6 22-

2t.6 -

24 £/

22 b /3

.1J 7/

VERT.STRESS

/1

1. 13

/, oL

7' (%)

1.932.,

7,5-/0/

2.3/6

0.22-/33

2.3 33/

3.S2,-

•, 71/

5.6Ae"

6. 727

7.o08 7

S7//

& 02

DATME Y(9/MINM

,2./o/

o. z 9/36, /9f

,0 g3 7

6 9$26

, o/83

, 0096,o/J8, 00 75, Oo 7s-

,0o "-/5

, oo 4/3 7,6c9379,003 2, ooa.•9

,00149

St0 / i7'3

,oo/16, oo1•/7• o b17 /

,ooo6a/ O

,Ooo /#o,0o0/So !

,6'00o. /0

,ot0 A50

ELAPSEDTIME

(MINUTE;/t(%)

-3Y~, 21/77&LS3(, 7-3-5

.590

8 60/6

, 720J

,7227

.58/8

, •o73

,783•

,6735

/.0/ /

• 7370

,9970

/. · 1##1.-- ;o

APPLIED 5(PER CENT)

72~/o

(3/7//7') )

Dr= T (APPLIED) / T MAX.

-152-

o,3-0

60

/50/2-

33 L

7/75

10 300

.5-2Z-72-74'o

_jI -

PAGE

DATE/TIME

5-2 9- 72

72.0J7 -,30 -72,

VERTICALPROVING RING

25b802zs'co

.26-eo'

2•9 7

6zS,

,3,5 72.;

.2• 7

.255-/

VERT.STRESS

/. o3

0199

.9503

j=I IL

Y"("/o)

,o7/

8.173

A. /-36B.s"-/

8. 855

9.//13

_, 4/1

/I, / 7/

/I. 07

Is. 984

)(/o/M INJ

,36o0o

, 0/8010207

,0/8

,O/87

,0/899· o/Z;I

ELAPSEDTIME

(MINUTES)

,/goo

.323 7

,/ •60

,839

//22

&2 ~

83.3

6. ?99

APPLIED ":(PER CENT)

( / 4/ )

-153-

.2.

.22

30

3/o

360,///0

CKoUDSS - CREEP TEST

SOIL DESCRIPTION EA.P/& C•A YBORING No. 92'- IlesSAMPLE No. /42ELEVATION -,//. -FINITIAL WN g77%

TEST No. CX_0,DSs-/9D ATE _-__- 72

CONDUCTED BY te••;W

vc( KG/CM 2 ) Z oOCR .2.0

PRETEST Ht.(MM) 2 ,Z 920

VERTICALPRONG RIN

.25 75

DATM E

6- 7- 72,

-lq2Z.I2-

VERT.STRESS

/ /2-

/, /9

/ /1/S///I

6 oZ

I.o6,-

7 (0/]

A 307

1597

6 76-/45/_

2.03 y

2 73

-237:

,7. / 2

2.682

8~3 L

?Y(VMIN

/ 3S8,0720

,03 79

, 0/3 2

, ooqzs',ooG397

,003•, 7

,00,2683

,oo2/13,001962

,0006/

,0)0 77,040 9

,1'4mt 7L, Odoo.003/7

0,/ 193

It(%o

o.67ý,D720"

,0759

, MI•

,///92

,/ 3Y/

,O '/2,

./72

,3071

,2 1/6

O 09s"

ELAPSECTIMEN

(MIN UTES

0.56'

2

7

/0

72

loo

/-.o

X36

/,20

o

o ,-

APPLIED "(PER CENT)

So n

7 /o

.:7 ,i•--mDr= r (APPLIED) / r MAX.

-154-

.263 8

.,2 25"

,24 2 6

,~2 /B

12-

,a /2-Z'/7

'Z( o7

2 & bo-a//or

PAGE 2-

VERTICALPROVING RING

2&o6

DATIME

.5-24 -72-7a-,s"

;2• 93

VERT.STRESS

/ tl<

Z592

.240,7-/

,-.4o 7

2_6o 72607

Z6 0

a6062-.,Z03

2S8 -2,

258 28

,sr 78

2•88

/.7

1 035

40/2/. o //, e /

Y(/o)

3. 10/

, /10

3,/77

3.233

3, 30/1J.932

3.55/

i2/7

S933

1, /040

6. M2&,so/,5.o/

6.2/o

.03 o00

,o 2q7

,0/3 6

,005- 28

,101/33

,000•779

, 100s-8

,0~ /23

Ah0 9X 0.000077

.0000o5;

,00008 63

,os0 7

- 0.36-.

, o/5-

,oz/3

'09

,'15-777

.5o2

.0,.9 ?,0680,o5R73

,/3881,-5-/ 2-,s)jo

,/50oo

.os27

, /~'/S, /,-1

ELAPSEDTIME

(MINUTES)

,o3•O

,0.S77

D06 72

,9/3 g

,.2.20.20'9/2L/M~

1690

.30 /

J/7606255

73 Yo

I/ bo,

N'- 7S3"

Z7 ;ao60

()

7/0

/5ý2,

APPLIED 1r(PER CENT)

-155-

Z

YL,f7

/O

14

70

/o0e

,agr2fzffidz92-

-- Y

.9 4%(m A'/cr

PAGE 3

DAT E/ME

5-2e-72/72L

VERT.STRESS

/ 0.2

VERTICALPROVING RING

25987

2YSL

2 y72y b9

a273

0 f)5

~

R, A t/

C? ~9

h x./o IN I7 Olo)'.535

7. los!

7,450

9. 331

o10. 7/2

/2. 263

A5/79

12 /16

PA

ELAPSEDTIME

(MINUTES)

5-O

t,5 o,.7472,o oso1.353

1, o 3

9.762-

3,o2L9

,30,-96

7l12,1

.610LSor oi

.010o

.ooft510 2ok %

,00 778

t oo ro I

co v93 0

"0063o'0060 3

.00593·o s/Ll/RE 9

APPLIED D(PER CENT)

-156-

9,702Z

75

/oy/J L

/95

/A43 /

/ b3 3

/ 786

e 6 & 5/sA£

/ eo_.D

i7•g

CKoUDSS - CREEP TEST

SOIL DESCRIPTIONBORING No.SAMPLE No.ELEVATIONINITIAL (ON

EARPL CL4 VY

- VI,3 FrSo, /%

TEST Na Co.sS -o20DATE 7-/7 - 7 -

CONDUCTED BY tEI

yvc(KG/CM2 ) ,oOCR / o

PRETEST Ht.(MM) YV, 6,3

DAT% VERTICAL VERT. 7 (% ,I) (MIN) t() ELAPSED APPLIED r1ME PROVING RING STRESS (MNIFES, (PER CENT)

o7- -72 g/ 90%,800 009 9 i 517 .0 o o

303o /9 /7 lf (,//7 ;of og (32/m)

.2990 / 6 6,'/17 I. 9- /,2fo I

£9'<5 /, 7R p 2-1•3• ,7/63 / Z 33 2

lZR/ s 7 17 ' 3 '91 1/ 1,259/, o 3,6/ .½ A "B 2so 7

7 o //' /1,/2 /0.33•" ,2/33 .2,33 /0

27/o / .Z, 33 139 , /338 2.q3 22-. 5 72. /3,5-9 ,/0/9 3,2/ . -3z

z 6e9 / 1"2 y 5o xý2 39 ,9o5 8 v-.77/ 2 o2b60 /1851 ,o?/o0• 107! 70

-r•b7 /1 /7 ly/457 ,Y/,7 1,739 1oSZ, /, o z ,o.yo"06 , Y• 33 & 1c6 1,z

z9 /7 /as59 ' 0o9'5 /3,/ o3 l /7

6 Z02 /1c 0 25•98/ l/85'

_ _ /L 'ZE6 /s

Dr= "r (APPLIED) / r MAX.

-157-

CKoUDSS - CREEP TEST

SOIL DESCRIPTIONBORING No.SAMPLE No.ELEVATIONINITIAL CON

E-,ABPL CLAgY95 UES

-7-. 9"Fr7.5./%

TEST Na ctu'u•ss-2/DATE 7-z -72-

CONDUCTED BY c.LA/

Kc(KG/CM 2 ) 2.0OCR I. o

PRETEST Ht.(MM) 22, 700

DATMVERTICAL ERT. (/o )'(MIN) tt(%) ELAPSED APPLIEDDTMEPROVING RG STRESS TIME CENT)(MIN US) (PER CENT)

7-3o -72-9o4 0s7-4 2.0 0 _ ____

.3_ ol / C 18 8e9 3 5l / (2,'751 ' Ic?)

29,5 A, 79 ...s5B/ / 0/7 /, 0/7

.Z9oo 4 , "7 ,/9 ,s8/3 1 /./ 228 7o 7 32,68 .353, /3 ,V/1

2830 , l86 ..,a 1-7 / 7/3 7

se00o /, y:5 8,796 , I/,e / /gq /o2 790 9.533 , /7/8 . zg~ /3~

2 78 7 ./.V38 /o, ~68 O 7,,208 .22-

2z 787 / 9. 3 ,0&75 2, o.2L/ 3 e

S7(. / 2 ,i , o5/o 2. 5-7)

;?7,1o - Z3/ /' ./,l1f/ ,Zs59 7,51Y 7019 73 23 /3.293 ,68?Z .82/ /o7

.2 7/ /•95s ,02/9 3.071 /Yo

S700 / 2 , 9'/7 , o/'/' Z2.23 7 /g o

____ _/_ !/2 .%/(.z V~2/0 .3(

S2670 /9,33/ ,0o917 •f3& qzT.K

.6z 1 /, 7 zo, /Y3 , Do275-2 323 57516 oZ A2/ /73 .o./3 J,9/L 7.20

.z6 0/ ,,A 22B ,2 20'/ 8 220 /5-•0

.59" .2(t.7 7 , 007/7 /3/. '3 1760

,.oos o. .o 6: 6 9 /y .7,.

90 3 ! L4Z~L ___

Dr = T (APPLIED) / r MAX.

-158-

CKoUDSS - CREEP TEST

SOIL DESCRIPTION f-•BLP CL, c YBORING No. 9,5 •vcsSAMPLE No. /2 DELEVATION -_ I,_ r-INITIAL (N 77 %

TEST No CAfDss--2ZDA TE 8-1/t'-7_

CONDUCTED BY CEhV

"c(KG/CM ) .oOCR _o

PRETEST Ht.(MM) 22.073

DAT VERTICVERTICAL ERT. r (%)/ Y(VMIN) ELAPSED APPLIED rMEPROVING RING STRESS TIME(PER CENT)

.2930 436I9 j3,( N7/ 0.5 4• O/cmE

._700 /_ _7 _3.f9'• ,•.13.3 ,523/ /

_ .28- /,t-35 3-r57V ,33/1 ,b6l2' z

_____ .2 8 p.' />89 A1I ,,9/29 /

.ý s • 007 ,/y-/1 /ro/o5 7

8 7/ 5. SY ,0 9 5•2- , 6 ,92 • o0

,Z 6 2- S.76 ,o7/172 , A070 /y

-oz -•15 . 7 4. 6.1 , 535O- /,177 1.z2_-

_2- .30 6,s71 ,oa76 7 7,/196 30

.2 8.2 7.083 ,0/935 ,7 o7 Y/

S(" 2435 .'O/129 /200 7028oo 8.0 9 01/25" /.2-St, /00

,7 87 J /1 ,, .63 ,00,77 /3,2ý/ /S/

-97,6 ./1/ ,oo6l/3 / 3f YA l&

276/ 9,546 , o0016 p5 , 139 oo

S2,57 9,932 ,o2, 352 /, 371 90

,Z T75- 10 252, , oo .z 1, zy3 q P2 737 / Ay 1/7i ,H/00159 Z, 22 0 76;•

700 ;z g, Z•o7 ,oo /00 /3 / Io/ /227/1o /2. 07 .,00~ &13

,6 9/ /2,1,77 p, a '27 /,21ff 23$9D

S63 I-2/3 , 0oooZo 30~/ / 1 oo

.676 /3.o2-. ,t 90, 20 ,/71/8 1 73S7 -25-72 1,

_ _ _ _ /,205 /3.78/ ______ ____ o 90I<

-DT= T (APPLIED) / r MAX.

-159-

(72 1</4m

PAGE 2-

DAT MEVERTICAL

PROVING RINC

£47o2670

2670

2670

.24b 71

2622

;..• 72

~f 73a, 73

2670

-24 ?D..2_ 570

•2-570

.26352do. a.

VERT.STRESS

/,,A o

I, L9S

09 71

Y(o)

/3. f8f

/3.8&3/3. 88,

/3.956

/1/77

IY /27

/16 072,

/57 782

/ý,/83

/6, 201

•4-997

FA IL

ELAPSEDTIME

(MINUTES)

0.5

ho%/,0337

,0233

,o837

, 2/50

, /990o,7330

/, / /f

1 0/2 6

, 008/, oo j/

, 00 3 9

,O0zo.

, 0077

,0028/

IR f

hr '(Yb

.2. 35)

3.£52

sgo

7/6

7 2D

13& 1

12, f

APPLIED 1r(PER CENT)

-160-

,_5"l~ -/

Y/2/2

3•"

..Z00

3•57oo

.•,s

Lfso

CKoUDSS - CREEP TEST

SOIL DESCRIPTIONBORING No.SAMPLE No.ELEVATIONINITIAL WN

••BP PL C L,49 uESioB-32.079. 5/

TEST Na CkVD4s-23DATE 9-3-72

CONDUCTED BY cewvc( KG/CM 2 ) 2.0

OCR /.oPRETEST Ht.(MM) 23.3 Y3

DATIME

4'-o-;

VERT.STRESS

2,o

VERTICALPROVING RK

3 o 35

2o /s

.2999

-z-,980

.29 392925

29/2-,

290/'

2902-

z89.2-

2785

2- 79/

80ol7•790

.Z783

2783

/ 9-3

y (M%)

2.06632.306 S.2.o/39

.2. " 7

3,039

3,26 /

3.'127

3. ••-

_.. 9oj/

.087

€99/3c 307

j. 78257035776/3

5 8-8335"e257A

Y(%/MIN

/ 6/3 6

, 0 9759-

Zb6 3 9

,o03280

, I6/87

,0/,19,0/293,6•oB79

,o00,35

· 9 '32#

,e'a232

,ool71-.oo963O

, 00 /0D

,0000 95

ooo 72 9

3o3/

,g902

.58/9

,52 79'/

.sB I*5837

,8 X5

, 5859.9s/8

.S/32

, yjSc.5999

,9/S94

, Y.97

ELAPSED

anTI~uE$

O.s

22

I0I7

1o

30

95

(0

/ B8

3 YO

G, 2.,

-3230

-71748,s y,_•

577£,-

APPLIED N(PER CENT)

33/12 /ý/07

Dr= T (APPLIED) / r MAX.

-161-

.,79

/ 7/

l,7 S.

/I5-/7

7•2.,5"9-8~/;zzy

2 28S- IV 11 000 b62I

PAGE 2.

VERTICALPROVING RING

.227/

DATE/ EnIME

2 780

277,5

2 775

2 76S

2 762,

2 263276e 3

2260

2772-

277/

2 7 272772

2772.772. -

22773

.2 2; 2-

2 75'2• 25*-21/7,1777

VERT.STRESS

4 Y!

/ 3-5

/3•

Y(%O)

S,/777

k 099&

-o99/6.2/0•1

7,2/of

1.27/56. 2/77'k. 271

1,37/3

A. YJZS

6.5:972. 06.7

6, 7113

6~./7 7

&.09;

7A2 ~

7. 77db

•,o903

6',2622.6,32B7

8 539'6 7/13

.00053

.00/, 70, ooo 24(7_

~-

o. /l .

,000g

,ooI/ /

,tooo f/

,000 165

,000 •Y•o 6/ c

, Oo3657

.P0 0 '57

,oco.o I

,o000o15 Y

,000o/5

. 000/ 212

ELAPSEDTIME

(MINUTES)

7/70

,3750

,025

651,O/y/tS

,/570

,isZo

, 7ygy,.

, 617

, 9y/7/

APPLIED l•(PER CENT)

-162-

77/5

B6551o o8o

0

2

/o

7

/_3/

.30

Y/0.ob

g yo

I/93

.2370

go35

38 75

132S-

63 75

7/8o

Doo //),9 1

'o60

p

p loop

PAGE 3

VERTICALPROVING RING

-Z ? '/3

DATE ME~ETIME

9- 2 2 -7z

730

7 '0

2? 1/6

2-) /4

2 7 y'

.,, 2A18

.2-2 YB

27 y27. f

-zz24 ~

x ;,3

VERT.STRESS

/ 3s'?/3y'7

/,3/

/,06/3.;

8.866•

9,0/9~

9,of Y2

Y. -9Ym7

Woo28

2L7o293

//, 783

/3.26 '

E9L

£, ' IN8

o o /2,r

,oo 32

, o 5b 93

,0go •32, oo ffy, oo Soy, oo ps7

, 00 6V•, 00 2/o

, oo 79q,00 -29

,01/69

ELAPSEDTIME

(MINUTES)

/Io o 6

,ouYt

106 9(;b

,/2 1b

/b9 7

bo

/33

170

325

775

APPLIED 5r(PER CENT)

95%s/Of"A

-163-

0o0.5"

I.z.

7

1o

3/

32

I 225Z6

/,730

3 77

LJ 627

9, ~

,,

APPENDIX C

PLOTS OF CK UDSS CREEP TESTSo

1. STRAIN VS. LOG TIME Cl to C12

2. LOG STRAIN RATE VS. LOG TIME C13 to C24

3. LOG(STRAIN RATE X TIME) VS. LOG TIME C25 to C36

4. SHEAR STRESS VS. SHEAR STRAIN C37 to C48

Note:Lines with symbols represent the

test being plotted. The remaining

curve is the stress-strain curve

for the corresponding (NC or OC)

controlled strain CK UDSS test.o

5. CREEP TEST STRESS PATH C49 to C60

Note:Curved stress path is that of

the corresponding CK UDSS controlled

strain test. The remaining stress

path is that of the creep test.

-1max = tan (T/-v') from controlled

strain CK UDSS testo

-1T = tan (T/- v ) at Tmax from controlledU v max

strain CK UDSS testo

= t-1S= tan 1 (T/l ) at creep rupturec v

-164-

-.4

U'

F, I I

tY _ _ _ _ _ _ _ _ _

L 1

(A

ov4L2

I. 4- 4 4 .1I __________

-165-

c)wI-

o0

HE-1

0

E

U

aw

-J

i~ i r ~

c (%) NIVII S

0N IV,,' (%) NIVdII.S

-166-

00-I

ulzi-

LfLIA

l-

(I\

zrN

1

9-=i

rH

a

0

00

0

(Iw

-Jw

1 \

r-i

I

rw E

0zHE-1

oo0n.-Jw

H

U

co Nr S

-167-

-R (M) NIVNIS

OO

"4

(A

(4 zrý

I I -I

. '(%) NIVaIS

-168-

rCl

E-4

H

H

Uso

a

w

w-

1%eN

rl

Cf) Cf.w 0

w

E-

0

0(I)a._

-J

H

LOU

-169-

. c N oZ,R 1 %) NIVHISS

0

(I)

wr

(0

-170-

HU)

0

0

E--

U

I r

-r

t

[(%) NIISco

18,

1

F- L_

L

' (%) NIVUlS

-171-

00

v,Owz

U

HE-A

0o O

i-iw

V)a.

U

0 %x . IW

, (%q) NIV8lS

-172-

r-

I0

z

0SO0w

U0-eu

o

I I

I-D :3- N

-V'(%) NIVb.LS

-173-

0-J

ULw..J

cIn(A

I-U)

cu

E--rU0H

0ul

)

I

,"

i-- i r--

-1_I

I i

L

~

1

I

--i1 1 u

C 4

. (%) NIV•lUS

-174-

r-II

U)

S COw

z

w 0

0o OwC,*

-Jw

H

E-4

rCl

U

'.0 cNLCO (% NI" I• (%/) NIV•IIS

0O

08

NI

0o0

8z 0

w

Oo

EJwO a. [

HEH

**

0

-175-

OD 1

o O U

F2 `

- c

C

+ I ~ I- t -I

N (% )NIVIIS

wI-

z

w

w

w

- n(

-176-

CN

a

U

H

r--

Ul

)OO

ELAPSED TIME (MINUTES)

C13: LOG STRAIN RATE VS. LOG TIME, CK UDSS-8o

-177-

w

zi

w

(I,

ELAPSED TIME (MINUTES)

C14: LOG STRAIN RATE VS. LOG TIME, CK UDSS-100

-178-

x>

ELAPSED TIME (MINUTES)

C15: LOG STRAIN RATE VS. LOG TIME, CK UDSS-11o

-179-

3'

w

zi

w1-4r

ELAPSED TIME (MINUTES)

C16: LOG STRAIN RATE VS. LOG TIME, CK UDSS-12

-180-

)00

w

z

w

z

C,,

ELAPSED TIME (MINUTES)

C17: LOG STRAIN RATE VS. LOG TIME, CK UDSS-16O

-181-

10 100 1000 10000

ELAPSED TIME (MINUTES)

C18: LOG STRAIN RATE VS. LOG TIME, CK UDSS-17

-182-

w()

4

1oQ000

1O 100 1000 14000

ELAPSED TIME (MINUTES)

C19: LOG STRAIN RATE VS. LOG TIME, CK UDSS-18O

-183-

w

zi

iocooo

ELAPSED TIME (MINUTES)

C20: LOG STRAIN RATE VS. LOG TIME, CK UDSS-19O

-184-

e•

I--

IiJwz

I--4

ZnF"

I--CO

I-

)0

10 100 1000 10,000

ELAPSED TIME (MINUTES)

C21: LOG STRAIN RATE VS. LOG TIME, CK UDSS-20O

-185-

w

zi

lo0ooo"'

'>0

w

zi

wI-

I0 IUU IUU 1u000u

ELAPSED TIME (MINUTES)

C22: LOG STRAIN RATE VS. LOG TIME, CK UDSS-21

-186-

I100UO·--

10 100 1000 10,0ooo

ELAPSED TIME (MINUTES)

C23: LOG STRAIN RATE VS. LOG TIME, CK UDSS-22o

-187-

-)ow

z,I'-zi

100,000

IV 1WU 1000 10,000

ELAPSED TIME (MINUTES)

C24: LOG STRAIN RATE VS. LOG TIME, CK UDSS-23

-188-

o00ooo

10 100 1000 10000

ELAPSED TIME (MINUTES)

C25: LOG(STRAIN RATE X TIME)VS. LOG TIME, CK UDSS-8

-189-

10,000

I' loo 1ooo Iooo ioo,oooELAPSED TIME (MINUTES)

C26: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-10-190-

101

I0

9w

wiJ

zI-(I)12Z,

10'

10

Iv o00 1000 1Q00ELAPSED TIME (MINUTES)

C27: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-11-191-

100,000

10 100 1000 I0,p00 100,000ELAPSED TIME (MINUTES)

C28: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-120

-192-

lIuu 1000 10000 100,000ELAPSED TIME (MINUTES)

C29: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-16o

-193-

Iu 100 1000 10=000 100,000oooELAPSED TIME (MINUTES)

C30: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-17-194-

w

I-

ZwIvzI.-

u 100 o000 IooooELAPSED TIME (MINUTES)

C31: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-18

-195-

oo00,00ooo

lU 100 1000 1P000 1I0000,ELAPSED TIME (MINUTES)

C32: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-19

-196-

w

w

z

U,

IV 100 1000 1•000

ELAPSED TIME (MINUTES)

C33: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-20-197-

ioo0ooo

I0 100 100) IU0J00 I00,000

ELAPSED TIME (MINUTES)

C34: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-210

-198-

---

'u 100 1000 10,000 oo100,00ELAPSED TIME (MINUTES)

C35: LOG(STRAIN RATE X TIME) VS. LOG TIME, CK UDSS-22

-199-

w

z

w

w:E

zI

Un

luu 1000 10000=ELAPSED TIME (MINUTES)

C36: LOG(STRAIN RATE X TIME) VS. LOG TIME, CKoUDSS-23-200-

100,000

-201-

a

IdI

')wTV)

oIUl

O

U)J

UC

U)

U)crj

o-o;

C) Nd cS

2U)

0

CD

UD

o

00Cci3cor..

-202-

d .

I.-

cat

4.

U)

U)

U)

U)

U)

C)

-203-

-204-

U)k'

to

NrH

U

H

c-I:

ti--

0r

U

dV

IJU'to)

U)

U3

zrHI

U)O

U)

U)l

U)

U);

U)

U)

UI

-205-

-206-

I.-

4k

,)

V)

'-N

0U•

Fi

H

E--

U.

CN

Uvl

cO

5O

u]

', In mo V da.'

OdSci dKJ6 d2

-207-

Li

U,

COoo

r-1

0rl

eoU)U)

U)

U)l

U)

U

o d oY o'%P A6

-208-

A

I-vt

V)

o0

l"

H

LO3

U:

**:

p:

(NI

U3

o

U)

u

U)

Un

od d

-209-

49

U)

/oH

U)

U)

U)

U)

U)

Ur

210-

uj:tVj)

-211-

S

t4:

P11

NN

Ul)Lo

0

z

H

U)U)E

U)

UwzC3

NC)U3o9Pot•l

P:

6

-212-

0

()cc

in

M

o

zH

ry

Ul<00,

-213-

SU

03

I/

o3O

01UE-ýPL4

a)4ý1u

Y

4

-214-

Nd

o

U

u)

4

C>

UE-

**!

Or

E-0

0C4l v

-215-

u6

rU)

U)0

U

a4

*4

L)

Y~.

-216-

lb~

CN

rl

U)

ul

U,

E-i

ua

0pq

Lnu

Io

Cll

o

L)U

N2\0

-217-

Cj 0

* U

>

16S10

116

--

PJ)0

to

-218-

C)

U)C)

r-4

Kh

.)

oo

in

'-I'

-219-

-220-

.1

Nb

r-I

U

uUeH-

H:

~ Y

E-

o

U)

LtrU<,E-1[0

CmI

-221-

urb'

\1

-222-

u>b

r--CN

0

U)

LU

Uu;

-223-

ub

ljý

NN

a4

o

01u

LU

UU•

bt

I

Ci)

U)

Ue

~ X cV0

-224-