3n-

...4..7..( W ,
.2. . v- f

. s.

:2. :3...:.

3a....

a

4L .3
:1

.71:

£1.53: a
4 2:. $14»...
, 3a..
{anus

i .

I... it}?!
1.11.:th 3 in: .119

avfkat t) 5.. at!
‘ .1123;

1 as. . Mme!

.9. 33?.3?: $43...-
421.27., Lung?!

. .Y .
514 ’3
. an" «um.

in
1 :3... fixarbvflru:

|I .X.§Is 5.4.1.; a

51.3,! 12.1...
I. 53.13! 3“...
II

 

‘ .I.‘ x.
v) 018:!
V 4.90.1...
0:? .. wit it.
cit-$1.... \qi‘. 3.
. 13”: xv 4:! I.
1117...!

 

 

 

 

 

.. sanuxnr.... ,,. 1.: 15.65! . ‘ a. L
.. _ . . . ‘ uh.
. ‘ . a ‘ . guvxfiwunx .. .

“£5,918 7

 

 

V“)
a
2008
This is to certify that the
thesis entitled
LABORATORY SUBGRADE RESILIENT MODULUS DESIGN
VALUES FOR THE STATE OF MICHIGAN
l a)
4—» presented by
>— 59 ....>;:
0: (D
<12 5: x:
Q: (13 > . . .
lm 2') E: Colin Patrick Sessnons
:1 ‘5, :3;
2
._._.______' has been accepted towards fulfillment

of the requirements for the

MS degree in Civil and Environmental
Engineerim

 

 

MJ/i/A 74

Major P’ofessor’ 5 Signature
5/8/0 8

Date

 

MSU is an affirmative-action, equal-opportunity employer

— —‘-°--l-n-I—c-I-vpi-o-o-p-n-o-o-o-c—u--o-09-I-O-.-.-n—n---.—-—._

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K IProi/Acc8Pres/ClRC/DateDue indd

 

 

LABORATORY SUBGRADE RESILIENT MODULUS DESIGN VALUES FOR THE
STATE OF MICHIGAN

By

Colin Patrick Sessions

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree
MASTER OF SCIENCE

Department of Civil and Environmental Engineering

2008

ABSTRACT

LABORATORY SUBGRADE RESILIENT MODULUS DESIGN VALUES FOR THE
STATE OF MICHIGAN

By
Colin Patrick Sessions

The Michigan Department of Transportation (MDOT) currently uses several
different procedures for determining the resilient modulus (MR) of the roadbed soil.
Therefore, a consistent procedure is needed to unify the state and to meet the
requirements of the Mechanistic-Empirical Pavement Design Guide (M-E PDG). To do
this, the State of Michigan was divided, by soil type, into 15 clusters and 99 areas.
Disturbed soil samples were collected from 75 areas along with 10 undisturbed Shelby
tube samples. The samples were then tested in the laboratory to determine the natural
moisture content, grain size distribution, and Atterberg limits. Selected samples were
chosen to undergo cyclic load triaxial tests to determine the MR value. Empirical
correlations were then developed for estimating the MR value for the different soil types
by comparing them to the results of the moisture content, grain size distribution, dry unit

weight, and Atterberg limits.

TO MY PARENTS

iii

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Gilbert Baladi, for his
effort, guidance, and patience during this study. I would also like to extend my thanks to
the other members of my advisory committee, Dr. Neeraj Buch, Dr. Karim Chatti, and
Dr. Syed Waqar Haider.

I would also like to extend my thanks to the Michigan Department of
Transportation (MDOT) for their financial support. Thanks also to the MDOT technical
advisory group headed by Mr. Dave Weber for all their valuable comments. Many
thanks are due to my officemate and co—researcher Tyler Dawson for all his help and
support. As well as the undergraduate students who assisted in the laboratory.

Finally, I would like to thank my parents for their encouragement and support

throughout my college career. Without them this thesis would not be possible.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................... viii
LIST OF FIGURES ................................................................................. x
CHAPTER 1
INTRODUCTION AND RESEARCH PLAN
1.1 Background ................................................................................... 1
1.2 Problem Statement ......................................................................... 2
1 .3 Objectives .................................................................................... 3
1.4 Research Plan .............................................................................. 3
1.5 Thesis Layout .............................................................................. 10
CHAPTER 2
REVIEW
2.1 Review of MDOT Practices ............................................................. 11
2.2 Role of Roadbed Resilient Modulus in the M-E PDG ............................... 14
2.2.1 Procedures for Determining the Resilient Modulus of the Roadbed 15
Soil for Design Level One — Laboratory Testing
2.2.2 Procedures for Determining the Resilient Modulus of the Roadbed 17
Soil for Design Level Two
2.2.3 Procedures for Determining the Resilient Modulus of the Roadbed 18
Soil for Design Level Three
2.3 Literature Review ......................................................................... 19
2.3.1 Resilient Modulus and the Soil Classification Systems ..................... 19
2.4 Laboratory Testing ........................................................................ 20
2.4.1 Resilient Modulus Test Procedure ............................................. 21
2.4.2 Issues with Current Test Standards ............................................ 26
2.5 Field Testing .............................................................................. 27
2.5.1 Destructive Testing .............................................................. 27
2.5.1.1 California Bearing Ratio .................................................... 27
2.5.1.2 Dynamic Cone Penetrometer .............................................. 28
2.5.1.3 Plate Load Test .............................................................. 29
2.5.1.4 Pocket Penetrometer ........................................................ 30
2.5.1.5 Pocket Vane Shear Tester .................................................. 30
2.5.2 Nondestructive Testing .......................................................... 31
2.5.2.1 Backcalculation of Resilient Modulus from Deflection Data 32
2.5.2.2 Backcalculation of K Value Using Pavement Surface Deflections .. 34
2.6 Correlations Between Backcalculated Modulus, Laboratory-Based .............. 4O
Modulus, DCP, and Soil Physical Properties
2.6.1 Correlations Between Laboratory and Backcalculated ..................... 40

Resilient Modulus

2.6.2 Relationship Between Laboratory and Backcalculated .................... 42
Resilient Modulus and Physical Soil Properties

CHAPTER 3
FIELD AND LABORATORY INVESTIGATIONS
3. 1 Introduction ............................................................................... 44
3 .2 Soil Delineation .......................................................................... 44
3.3 Roadbed Soil Sampling ................................................................. 45
3.4 Field Testing .............................................................................. 60
3.4.1 Penetration Resistance Using Pocket Penetrometer ........................ 60
3.4.2 Pocket Vane Shear Tester ...................................................... 61
3.4.3 Falling Weight Deflectometer (FWD) ....................................... 61
3.5 Laboratory Testing ....................................................................... 62
3.5.1 Moisture Content ................................................................ 62
3.5.2 Grain Size Distribution ......................................................... 63
3.5.2.1 Sieve Analysis .............................................................. 63
3.5.2.2 Hydrometer Analysis ....................................................... 64
3.5.3 Atterberg Limits ................................................................. 64
3.5.4 Cyclic Load Triaxial Test ...................................................... 68
CHAPTER 4
DATA ANALYSES AND DISCUSSION
4. 1 Introduction ............................................................................... 80
4.2 MDOT Practice ........................................................................... 82
4.3 Field Data Analyses and Discussion ................................................... 83
4.4 Soil Classification ........................................................................ 86
4.5 Cyclic Load Triaxial Test Results ....................................................... 87
4.5.1 Poorly Graded Sand (SP) ....................................................... 90
4.5.1.1 Univariate Analyses ......................................................... 91
4.5.1.2 Multivariate Analyses ....................................................... 98
4.5.1.3 Validation ................................................................... 102
4.5.2 Silty Sand (SM) ........ , ................. , ........................................ 103
4.5.2.1 Univariate Analyses ........................................................ 104
4.5.2.2 Multivariate Analyses ..................................................... 110
4.5.2.3 Validation ................................................................... 115
4.5.3 Clayey Sand (SC), Low Plasticity Clay (CL), and ........................ 117
Low Plasticity Silt (ML)
4.5.3.1 Univariate Analyses ........................................................ 120
4.5.3.2 Multivariate Analyses ..................................................... 124
4.5.3.3 Validation ................................................................... 126
4.5.4 Poorly Graded Sand — Silty Sand (SP-SM) ................................. 128
4.5.4.1 Univariate Analyses ....................................................... 128
4.5.4.2 Multivariate Analyses ..................................................... 132
4.5.5 Clayey Sand —— Silty Sand (SC-SM) .......................................... 135
4.5.5.1 Univariate Analyses ........................................................ 137
4.5.5.2 Multivariate Analyses ..................................................... 142

vi

4.5.6 Gravelly Sand (SG) ............................................................. 144

4.6 Damage Model ........................................................................... 145
CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

5.1 Summary .................................................................................. 150
5.2 Conclusions .............................................................................. l 50
5.3 Recommendations ....................................................................... 1 52
Appendix A ........................................................................................ 159
Appendix B ....................................................................................... 166
References ......................................................................................... 1 81

vii

2.1

2.2

2.3

2.4

2.5

2.6

3.1

3.2

3.3

3.4

3.5

4.1

4.2

4.3

4.4

4.5

4.6

4.7

4.8

4.9

4.10

LIST OF TABLES

MDOT procedures for determining resilient modulus ............................ 11, 82
Models relating material index and strength properties to MR ..................... 18
Comparison between three soil classification systems ............................... 21

Typical resilient modulus values for unbound granular and subgrade materials . 23

Regression coefficients for 6: ...................................................... 39
Range of k value for soil type, density, and CBR .................................... 43
Soil percentages for each area within the 15 clusters ................................ 49
Locations of pocket penetrometer and vane shear tests .............................. 56
Locations of Shelby tube samples ...................................................... 60
Laboratory test results .................................................................... 65
Laboratory MR test results ............................................................... 75
Sample backcalculated k and MR values .............................................. 84
Number of samples per soil type ....................................................... 87
Location of SP subgrade soils ........................................................... 93
Locations of SM subgrade soils ....................................................... 105
Location of SC, CL, and ML subgrade soils ......................................... 118
Locations of SP-SM subgrade soils ................................................... 128
Locations of SC-SM subgrade soils .................................................. 136
Locations of SP—SM subgrade soils ................................................... 144
Damage factor calculation ............................................................. 149
Design resilient modulus values for M-E PDG design level 3 .................... 149

viii

5.1

Al

A2

A3

B.l

Summary of predictive equations for each soil type ................................ 154

AASHTO soil classification system .................................................. 160
Possible AASHTO soil classifications per USCS group ........................... 165
Possible USCS classification per AASHTO group ................................. 165
Laboratory resilient modulus results .................................................. 167

ix

2.1

2.2

2.3

2.4

2.5

2.6

2.7

3.1

3.2

3.3

3.4

3.5

3.6

3.7

3.8

3.9

3.10

3.11

4.1

4.2

4.3

LIST OF FIGURES

MDOT regions ............................................................................ 12

Range of soil support values, structural coefficients, and resilient ................. 13
modulus for various materials

Soil classification related to strength parameters ..................................... 24
Resilient modulus testing apparatus for soils .......................................... 25
Resilient modulus concept ............................................................... 25
Schematic of a dynamic cone penetrometer ........................................... 29
Photo of plate load testing apparatus ................................................... 30
Western portion of the Upper Peninsula ............................................... 46
Eastern portion of the Upper Peninsula ................................................ 47
Northern portion of the Lower Peninsula .............................................. 48
Southern portion of the Lower Peninsula .............................................. 48
Wet sieve test .............................................................................. 63
Dry sieve test .............................................................................. 64
Liquid and plastic limit apparatus ...................................................... 69
Vibrating table setup ..................................................................... 70
Stress influence with depth .............................................................. 71
Cyclic load test setup ..................................................................... 73
Typical cyclic load test results .......................................................... 74
Pocket penetrometer versus vane shear tester ......................................... 85
Typical particle size distribution curves ............................................... 88
Resilient moduli at 10 and 15 psi cyclic axial stresses .............................. 89

4.4

4.5

4.6

4.7

4.8

4.9

4.10

4.11

4.12

4.13

4.14

4.15

4.16

4.17

4.18

4.19

4.20

4.21

4.22

4.23

4.24

4.25

4.26

Grain size distribution curves for SP soils ............................................. 95
Resilient modulus versus the percent passing sieve number 10 for SP soils 96
Resilient modulus versus the percent passing sieve number 10 for SP1 & SP2 .. 96
Resilient modulus versus the dry unit weights for one SP1 soil sample ............ 97
Resilient modulus versus the dry unit weight for 20 SP soil samples ............. 97

Resilient modulus versus the dry unit weight for SP1 and SP2 soil samples ...... 98

Resilient modulus versus the moisture content of one SP soil sample ............. 99
Resilient modulus versus SVSPl and SVSP2 ........................................ 102
Predicted resilient modulus values for the validation points ....................... 103
Resilient modulus versus water contents of the SM soil samples ................ 106
Resilient modulus versus dry unit weight of 13 SM soil samples ................ 107
Resilient modulus versus the degree of saturation of SM soil samples .......... 109
Resilient modulus versus the liquid limits of SM soils ............................. 111

Resilient modulus versus the average particle size at thirty percent passing 112

Resilient modulus versus the sample variable for SM (SVSM) subgrade ...... 1 l3

Resilient modulus versus the moisture index of the test sample .................. 114
Measured and calculated resilient modulus values, Equation 4.5 ................ 116
Measured and calculated resilient modulus values, Equation 4.6 ................ 116
Resilient modulus versus the moisture contents of the samples .................. 121
Resilient modulus versus degree of saturation for SC, CL, and ML soils ....... 122
Resilient modulus versus dry unit weight for SC, CL, and ML soils ............. 123
Resilient modulus versus degree of saturation ....................................... 126

Laboratory measured and calculated MR values for SC, CL, and ML soils ..... 127

xi

4.27

4.28

4.29

4.30

4.31

4.32

4.33

4.34

4.35

4.36

4.37

4.38

5.1

5.2

5.3

5.4

A.1

A2

A3

A4

Resilient modulus versus moisture content for SP-SM soils ...................... 129
Resilient modulus versus the dry unit weight of the test samples ................ 130
Eight gradation curves of the SP—SM subgrade samples ........................... 131

Resilient modulus versus percent passing sieve number 200 for SP-SM soils .. 132

Resilient modulus versus the sample variable model, SP-SM soils .............. 134
Resilient modulus versus SVSP-SM soils ........................................... 136
Resilient modulus versus moisture content for SC-SM soils ..................... 137

Resilient modulus versus saturation for SC-SM soils .............................. 140
Resilient modulus versus the dry unit weight for SC—SM soils ................... 140
Resilient modulus versus liquid limit for SC-SM soils ............................. 141

Resilient modulus versus SVSC-SM soils ........................................... 143
Locations of SP-SM subgrade soils ................................................... 144
Recommended M-E PDG level 3 design modulus values for the western ...... 155

portion of the Upper Peninsula

Recommended M-E PDG level 3 design modulus values for the eastern 156
portion of the Upper Peninsula

Recommended M-E PDG level 3 design modulus values for the northern ...... 157
portion of the Lower Peninsula

Recommended M-E PDG level 3 design modulus values for the southern 158
portion of the Lower Peninsula

AASHTO Atterberg limit ranges ...................................................... 161
Casagrande’s plasticity chart .......................................................... 162
USCS coarse grained soil classification ............................................. 163
USCS fine grained soil classification ................................................. 164

xii

CHAPTER 1

INTRODUCTION AND RESEARCH PLAN

1.1 Background

The state of Michigan is geographically located within the glaciated section of
North America and most of its soil has developed from glacial deposits. The ice sheet
advanced over the state in three lobes, one along Lake Michigan, one along Lake Huron
and the third along Lake Eric. A branch from Lake Huron lobe advanced southwesterly
and connected to the other two lobes. During the advance of ice a large amount of soil
and bedrock along the path of each ice lobe were pulverized and incorporated into the ice
sheet to be later redeposited. When the Wisconsin ice sheet retreated to the north, these
materials (known as glacial drift) were superimposed on sedimentary rock of the
Michigan Basin in the Lower Peninsula and the Eastern part of the Upper Peninsula and
on igneous and metamorphic rocks in the Western part of the Upper Peninsula. The
thickness and composition of the drift varies from one location to another. For example,
the thickness of the drift in the Alpena area is only few inches whereas it is more than
1200 ft thick in the Cadillac area. The glacial drift also varies from clay to gravel; the
granular texture may be segregated or mixed heterogeneously with boulders and clays.
Because of these complex arrangements, about one hundred and sixty-five different soil
types were formed and are being used for engineering purposes by the Michigan
Department of Transportation (MDOT) (MDOT 1970). The engineering and physical
characteristics of these soils vary significantly from those of gravel and sand in the
Western side of the Lower Peninsula, to clay in the Eastern side and to varved clay in the

Western part of the Upper Peninsula.

For a given type of roadbed soil, its mechanical (engineering) properties (the
resilient modulus (MR) and the plastic properties) are a firnction of the physical
parameters (moisture content, grain size, grain angularity, Atterberg limits, etc.) of the
soil and have a major impact on the performance of pavement structures. In the past,
MDOT funded several research projects to study the engineering properties of certain
types of roadbed soils (Goitom 1981, Lentz 1979). The results of those studies will be
incorporated into the results of this research.

In this study, the MR of various roadbed soil types will be determined in the
laboratory using cyclic load triaxial tests and in the field using the Falling Weight
Deflectometer (FWD) deflection data. Statistical correlations between the laboratory and
field MR will be obtained. The major objective of this study is to determine whether or
not the MR of a given roadbed soil type can be estimated using the results of simple tests.

In the laboratory, the MR of a given soil type is calculated as the ratio of the

applied deviatoric stress, ed, (the difference between the axial and lateral stresses) to the
recoverable axial strain (8,) of the soil (Goitom 1981, Lentz 1979, Young and Baladi

1977, and Yau and Von Quintus 2002). Mathematically, the MR is expressed as follows:

MR = — Equation 1.1

1.2 Problem Statement

The roadbed soils in the state of Michigan consist of glacial soils with distinct
seasonal stiffness changes due to temperature (possible frozen condition) and moisture
levels. MDOT’S current pavement design process follows the procedure outlined in the
1993 American Association of State Highway and Transportation Officials (AASHTO)

Design Guide. One of the inputs of said procedure is the effective value of the resilient

modulus of the roadbed soil, which is a function of seasonal changes. The pending new

AASHTO Mechanistic-Empirical Pavement Design Guide (M-E PDG) procedure is even

more stringent for defining MR. in terms of seasonal effects. Currently, MDOT’s various

regions provide the “adjusted” MR value used for pavement design. The MR value is
derived from either backcalculated deflection data or a correlation with known Soil

Support Values (SSV).

1.3 Objectives

The main objectives of this study are to:

0 Evaluate the existing processes used by all regions of MDOT for determining the MR
value of the roadbed soil for flexible pavement design and the modulus of subgrade
reaction (k) for rigid pavement design.

0 Determine the needed modifications to make the process compatible with the M-E
PDG.

1.4 Research Plan

To accomplish the objectives, a research plan consisting of five tasks was
developed and is presented below.

Task 1— Review and Information Gathering

In this task, the research team will become familiar with MDOT’s current and
historical processes/procedures for selecting MR and k values for the design of flexible
and rigid pavements. The information could be obtained, by calling the various regions
and talking to the soil engineers. The research team will also obtain information from

MDOT that is needed for the other tasks in this study. These include:

. The locations of FWD tests that were conducted in the past and the availability of the

measured deflection data and the pavement cross—section data that existed at the time

of testing.

. The depth of frost penetration especially in the northern part of the Lower Peninsula

and in the Upper Peninsula.
. The repeated load triaxial test data that were obtained as part of research projects that
were sponsored by MDOT from 1975 to 1979. The data will be digitized and

tabulated along with the roadbed soil type and will be used in later tasks.

. Traffic data in terms of average daily traffic (ADT) and percent commercial.

The effort of this task should produce:
. Tabulation of the procedures used by the various Regions for selecting MR and k
values and the basis of such selection. Based on the information, differences and

similarity in these procedures will also be tabulated.

. Tabulation of the range and typical MR and k values used by the regional soil

engineers for the various soil types.

. A brief summary of the background and the development of the SSV-resilient
modulus chart provided in Figure 2.

. Assessment of the adequacy and sufficiency of the existing process for estimating
MR and k values to be used in the new M-E PDG.

For all available deflection data, tabulation of the locations of all FWD tests that were
conducted in the past and the pavement cross-section that existed at the time of
testing.

. A map or a chart showing the depth of frost penetration where data are available.

7. Tabulation of the cyclic stress, confining pressure, vertical and horizontal
deformations and strains, and the resilient modulus of the various roadbed soils
included in the MDOT sponsored research projects during the period of 1975 to 1979.

Task 2— Partitioned State Map

Based on the MDOT Field Manual of Soil Engineering, the information obtained
from the various regions in Task 1, the trunkline locations, and the soil maps of the US

Soil Conservation Services (USCS), the state will be partitioned into geological zones

for the purpose of field testing and soil sampling. The state will be divided into a

maximum of 15 coarse clusters where the soil within any given cluster would have

similar range of engineering and physical characteristics. Each coarse cluster will then be
divided into areas to narrow the range of the soil characteristics. A maximum of 99 areas
will be produced. The results will be presented to members of the technical advisory
group (TAG) for review and possible modification. The main use of the partitioned soil
map is to determine the locations of field testing and soil sampling.

Task 3— Field and Laboratory Testing and Soil Sampling

In this task, the research team will finalize the field sampling and the laboratory
testing plans based upon the information obtained in Tasks 1 and 2. The total number of
tests to be conducted is based purely on cost and available budget. The field sampling
and the laboratory testing plans are presented in three subtasks below.

Subtask 3.1 - Soil Sampling Plan

From each area on the State Partitioned map, soil samples will be obtained. In
areas where the roadbed soil is predominantly sand, only disturbed bag samples will be

collected. In clay areas, both disturbed and undisturbed thin Shelby tube sample will be

obtained. In total, 75 disturbed roadbed soil samples and 11 undisturbed (Shelby tube)

samples will be collected. All samples will be transported to the laboratory for testing as

presented in Subtask 3.2 below.

Subtask 3.2 — Laboratory Testing Plan

The laboratory plan consists of moisture content, sieve analysis, Atterberg limits,
and cyclic load triaxial tests. All tests will be conducted according to MDOT, AASHTO
or ASTM standard test procedures. Results of the laboratory testing will be analyzed (see

Task 4) to determine:

1. Soil classification - For each of the 75 disturbed samples (bag samples), the soil will
be subjected to sieve analyses to determine the breakdown fractions between sand and
clay/silt particles. Any sample where the fine fraction (passing sieve number 200) is
more than seven percent, plastic and liquid limit tests will also be conducted. Results
of the sieve analyses and Atterberg limit tests will be used to:

0 Classify the soil according to the AASHTO and the USCS soil classification
systems.

0 Develop, if possible, statistical correlations between the resilient modulus of the
roadbed soils and the gradation and Atterberg limits of the material.

2. Cyclic load triaxial tests - For each location where Shelby tubes were collected,
repeated load triaxial tests will be conducted. The samples will be tested at three
moisture contents to simulate the effects of seasonal changes on the resilient modulus
of the soils. The water content of the samples will be changed to the desired level by
either drying or by using back pressure technique in the triaxial cell. For sand

subgrade soils, the test specimens will be compacted at three moisture contents and

subjected to cyclic load triaxial tests. Since, the resilient moduli of sand roadbed soils
are heavily dependent upon the deviatoric stress; the laboratory tests will be
conducted at three stress states which will be estimated through mechanistic analyses
to simulate the probable in-situ field conditions.

Subtask 3.3 -Field Test Plan

This plan consists of Falling Weight Deflectometer (FWD) tests. The FWD tests
will be conducted at the network- and project-levels. At the network level, one FWD tests
will be conducted at 500 feet interval along the state trunkline. At the project level, 20
FWD tests will be conducted within : 50 ft from all locations where Shelby tubes
(undisturbed soil samples) will be extracted.

All FWD tests will be conducted in the spring and in the late summer — early fall
seasons). For those areas where FWD tests were conducted in the past and the deflection
and pavement cross-section data are available at MDOT, the data will be used and the
number of FWD tests (to be conducted in those areas in this study) will be reduced
depending on the availability of spring and fall deflection data. .

It should be noted that analyses of various damage models including AASHTO
indicate that the two point FWD testing (spring and fall seasons) is adequate to assess the
relative pavement damage caused by the roadbed soil due to different degrees of
saturation.

Task 4 — Data Analyses

The data analysis, in this study, will be accomplished according to the three

subtasks presented below. First, it should be noted that for all soil types, the relationship

between the MR and k found in the M—E PDG was used. Since the relationship applies to

all MR and k values, the analyses stated in the subtasks below will be conducted on the
MR values and the results will be converted to k values.
Subtask 4.1 — Backcalculation of Layer Moduli

All deflection data, whether collected during this study or other studies, will be
used (depending on the availability of the pavement cross-section data) to backcalculate
the layer moduli using the MICHBACK computer program. Although the moduli of all
pavement layers will be backcaleulated, only the resilient modulus of the roadbed soils
will be subjected to further analyses. The moduli of the pavement layers will be reported
without further analyses. For each test area on the partition map, two sets of moduli will
be backcalculated; one set is based on the spring deflection data and the other on the late
summer—early fall data. The two sets will be further analyzed to estimate the seasonal
damage factor as presented in task five below.
Subtask 4.2 — Laboratory Test Data

Results of the cyclic load tests conducted on Shelby tube and reconstituted bag
samples at three moisture contents were analyzed to determine the laboratory values of
the resilient modulus of the roadbed soil. Results of the analyses were used to assess the
impact of moisture (season) on pavement damage and to compare the values to those
obtained from backcaleulation.

In addition, the digitized cyclic stress-strain data of those research studies that
were sponsored by MDOT from 1975 to 1979 were analyzed. This pool of information
will be used as supplement to verify the relationships or to increase the pool of data to

develop more accurate relationships.

The Atterberg limits (liquid limit, plastic limit, and plasticity index) and sieve
analysis data will be used to classify the soil and to develop correlations to MR whenever
possible. Correlations will also be developed between the laboratory and the
backealculated MR values.

Task 5— Damage Assessment Analyses

The damage assessment analyses (noted in subtask 4.1) was conducted based on
the seasonal MR and k values obtained from the backealculation of the FWD deflection
data. The purpose of the analyses was to determine the effective MR and k values to be
used in the design and rehabilitation of flexible and rigid pavements. The effective
roadbed resilient modulus is an equivalent modulus that would result in the same damage
as if the various seasonal resilient modulus values were used (Huang 2004). In the
analyses, three methods will be used as follows:

1. The existing AASHTO 1993 damage model shown below.

=1.18><108x(MR)_2'32

u f Equation 1.2
Where u]: relative seasonal damage

2. The existing damage model in the M—E Design Guide, which is based on Miner’s

hypothesis of cumulative damage represented by the following equation.

D _ nijklmn
r — 2 E ' 13
N quatron .

ijklmn
Where, D, = damage ratio

nijklmn = actual number of load repetitions

Nijklmn = allowable number of load repetitions, for the ith age, the jth season,
the k'h,axle combination, the l‘h,load level, the mth temperature and the nth
traffic path.

3. Mechanistic analyses of stresses and strains induced in the roadbed soil due to traffic
load. The magnitudes of the induced stresses and strains for various roadbed moduli
were compared as to arrive at a damage model or to verify the above models.

1.5 Thesis Layout

This thesis is composed of five chapters and three appendices as follows:

Chapter 1 — Introduction

Chapter 2 - Review

Chapter 3 -— Field and Laboratory Investigation

Chapter 4 — Data Analysis

Chapter 5 — Conclusions and Recommendations

Appendix A —Soil Classification Systems

Appendix B — Laboratory Resilient Modulus Results

10

CHAPTER 2

REVIEW

This review chapter consists of two parts; review of the existing MDOT practices
regarding the determination of MR and literature review.
2.1 Review of MDOT Practices

As shown in Figure 2.1, MDOT divides the State of Michigan into seven self
administered regions; Superior, North, Bay, Grand, University, Southwest and Metro.
The practices of each region in determining the MR of roadbed soils differs slightly from
one region to another. Table 2.1 summarizes the procedures used by each region. In
general, the soil engineers use Figure 2.2 to estimate the MR of roadbed soils. The chart
in Figure 2.2 is based on the soil support values (SSV) and the USDA soil classification
system. It provides correlations between the SSV, the AASHTO layer coefficient for
subbase and base materials and the resilient modulus.

Table 2.1 MDOT procedures for determining resilient modulus

 

 

 

 

 

 

 

 

 

 

 

 

Region Procedure 3:111:21 (11:: 3
Bay Soil boring & visual identification 3600

Grand FWD data (if available) or soil boring & 2700 _ 8600
Visual identification

Metro Soil boring & visual identification 3000 - 4500

North FWD data (if available) or soil boring & 2500 _ 6000
Visual identification

Southwest California Bearing Ratio correlations
Superior Soil boring & visual identification 4500 - 7000
University Soil boring & visual identification 3000 - 4000

 

 

ll

 

 

Figure 2.1 MDOT regions (MDOT)

limestone

Gravel 22A

Sand Gravel 20A

Loamy Sand

SOIL SUPPORT VALUES

Salvaged
Gravel plus
bit surface
seal

Sandy Loam

0.15

Used as subbase

m
a)
(1:
D
O
0
w
(I)
<
m
m
<

 

85,000

69,000

'Jr
DJ

,

STRUCTURAL COEFFICIENT

O
O
O

J

 

39,000
26,000

15,000 —
7,500 -

3,000

RESILIENT MODULUS (PSI)

 

Figure 2.2 Range of soil support values, structural coefficients, and resilient modulus for

various materials (MDOT)

2.2 Role of Roadbed Resilient Modulus in the M-E PDG

The required inputs to the 2002 Design Guide can be broadly classified under four
main categories - general, traffic, climatic and structural inputs. The Design Guide uses a
three level hierarchical approach for selecting traffic and structural inputs. This gives
designers the flexibility of using very specific or very general input data for the design
process, depending on the agency’s resources and the requirements of each specific
design project (Coree et. al 2005).

The procedure for level one design can be thought of as "first class" and requires
high accuracy inputs. The level one procedure will typically be used for obtaining inputs
for the design of pavement sections subjected to heavy traffic or wherever there is safety
and/or economic consequences of early failure. The procedure requires laboratory or
field testing, such as the dynamic modulus testing of hot-mixed AC or site-specific axle
load spectra data collections, or FWD deflection testing. Hence, the inputs for level one
design require more resources and time than other two levels.

Level two is an intermediate design level whose required inputs are similar to
those used for many years in the earlier editions of the AASHTO design guides. This
level is used when resources and/or testing equipment are not available to obtain level
one input. The required design data inputs for level two could be selected from an agency
database, derived from a limited testing program, or estimated through correlations.
Examples would be dynamic modulus estimated from binder, aggregate, and mix
properties, or PCC elastic moduli estimated from unconfined compressive strength tests.

The required pavement design inputs for level three have the lowest level of

accuracy. This design level might be used for pavement sections where there are

14

minimal consequences of early failure (low volume roads). The data inputs consist of
typical default or average values used by the agency.

Further, the input data requirements can vary from one input parameter to another,
which makes the procedure more flexible. For example, on a given project, the pavement
designer could use level one for the subgrade resilient modulus input and level three for
the traffic distribution data. Regardless of the selected input level, the 2002 design
process is the same (Prozzi and Hong 2006).

Finally, regardless of the design level used, the resilient modulus of the roadbed
soil is a required input to the pavement structural response model. It has a significant
effect on the computed pavement responses and on the dynamic modulus of subgrade
reaction, k-value, which is computed internally by the Design Guide software. The
resilient modulus of the roadbed soil can be measured directly from the laboratory or
obtained through correlations with other material parameters such as California Bearing
Ratio (CBR) and Dynamic Corie Penetrometer (DCP). The procedures for obtaining the
resilient modulus for the various design levels are described in the following subsections
(NCHRP 2004).

2.2.1 Procedures for Determining the Resilient Modulus of the Roadbed Soil for

Design Level One

For level one design, the resilient modulus of the roadbed soil is determined using
cyclic load triaxial tests in accordance with one of the following standard test methods:

0 NCHRP 1-28A, “Harmonized Test Methods for Laboratory Determination of

Resilient Modulus for Flexible Pavement Design.”

15

o AASHTO T307, “Determining the Resilient Modulus of Soil and Aggregate
Materials.”

The stress conditions used in laboratory testing must represent the range of stress
states likely to be developed beneath the pavements. Stress states used for modulus
testing are based upon the depth at which the material will be located within the
pavement system (i.e., the stress states for specimens to be used as base or subbase or
subgrade may differ considerably).

The M-E PDG recommends Equation 2.1 for calculating MR. The nonlinear
elastic coefficients and exponents of the model are determined by using linear or

nonlinear regression analyses to fit the model to laboratory generated MR test data

 

(NCHRP 2004):
k k
MR—kP i 2 To“ +1 3
_ 1 a P Equation 2.1
a 0
Where, MR = resilient Modulus (psi);

6 : bulk stress = 0'l + 0'2 + 0'3 ;

 

0.1. : major principal stress;

0.2 : intermediate principal stress;

0—3 = minor principal stress/confining pressure;

Tact 2% (0'1"0'2)2 +(01 03):; +(0'2 “03)2

16

Please note that the above described procedure for level one design applies
equally to reconstruction and rehabilitation where destructive material samples can be
obtained. Alternatively, for pavement rehabilitation, nondestructive deflection tests can
be performed and the MR values can be determined using back calculation routine.

2.2.2 Procedures for Determining the Resilient Modulus of the Roadbed Soil for
Design Level Two

The MR values for pavement design level two can be estimated using existing
direct or indirect correlation equations between MR and other material parameters.
Indirect correlation implies, for a given soil, correlating parameter A to B and then B to
the soil MR. Table 2.2 provides a list of correlation equations included in the M-E PDG.
Further, the Design Guide software allows the use of the following two Options:

0 Input a representative value of MR and use Enhanced Integrated Climatic Model
(EICM) to adjust the MR value for the effect of seasonal climate (i.e., the effect of
freezing, thawing, and so on).

0 Input an MR value for each month (season) of the year (total of 12 months).

The primary use of EICM in this design procedure is to estimate the temperature
and moisture profiles within the pavement system throughout its design life. The
estimated temperature and moisture profiles within the subgrade layers can also be used
to modify the representative MR to account for the effects of climate. The procedure for

pavement design level two is applicable to new, reconstruction, and rehabilitation design.

17

Table 2.2 Models relating material index and strength properties to MR (N CHRP 2004)

 

 

 

 

 

 

 

 

Strength/index Test
property Model Comments Standard
_ 0.64 CBR = California AASHTO
CBR MR - 2555 (CBR ) bearing ratio T193
AASHTO
R—value MR :1155+555R R=R—value T190
AASHTO
AASHTO ai . 2 guide for
layer MR 2 30,000 1:;erfiifirfgt the design
coefficient 0. 14 of pavement
structures
wPI = P200*PI
P200 = percent AASHTO
PI and R = 75 passing sieve No. T27,
gradation 1+ 0.728(WPI) 200 AASHTO
P1 = plasticity T90
index
DCP CBR : 292 DCP = DCP ASTM
DCPI . 12 index, min/blow D6951

 

 

 

 

 

2.2.3 Procedures for Determining the Resilient Modulus of the Roadbed soil for

Design Level Three

For level—three design, the MR value is determined based on the classification of

the soil. Table 2.4 provides a list of MR values that are recommended in the M—E PDG.

For this design level one typical representative MR value at the optimum moisture

content is required. Users have the option to use the EICM to modify the MR value for

the effect of climate.

This design level can be used for new, reconstruction, and rehabilitation projects.

The material type can be obtained from historical boring record, material reports or

county soil maps. The bedrock depth is important and it should be investigated.

The MR values presented in Table 2.4 are approximate and should be cautiously
used. The reason is that these values are based on the assumption of a semi-infinite
media. For a finite roadbed soil thickness (less than 5ft), the MR of the lower and weaker
material should be used to obtain a composite MR value.

2.3 Literature Review

Early in this study, an extensive literature review was conducted to study and
summarize the results reported by previous investigators regarding:

- The advantages and shortcomings of the laboratory and field test procedures used to
determine the MR of roadbed soils.

0 The relationships between the laboratory determined and the backealculated resilient
modulus using deflection data.

0 The resilient characteristics of various types of roadbed soils.

0 The factors affecting the MR of roadbed soils including moisture contents (seasonal
effects), particle size, Atterberg limits, and grain size distribution.

0 The reported correlations between MR and the modulus of subgrade reaction (k) of
roadbed soils.

0 The reported correlations, if any, between the results of simple tests such as Atterberg
Limits, grain size distribution, pocket penetrometer, and hand held shear vane and the
MR of the roadbed soils.

Results of the literature review are summarized below.

2.3.1 Resilient Modulus and the Soil Classification Systems

There are currently several common soil classification systems. The most popular

of these are the United States Department of Agriculture (USDA), the Unified Soil

19

Classification System (USCS), and the AASHTO soil classification system (Holtz and
Kovacs 1981). Table 2.3 provides comparison between the three classification systems.
Such comparison chart is important because it allows the users of one highway authority
to compare their roadbed soils to another agency that uses different classification system.

Nevertheless, several correlations between the soil classification systems and the
resilient modulus of the roadbed soils can be found. These include:

o The data in Figure 2.2, which is used mainly by MDOT.

o The data in Table 2.4 of the AASHTO mechanistic-empirical pavement design
procedure (M-E PDG), which provide estimate of a typical range of values of the
resilient modulus of roadbed soils based on their AASHTO classification system and
USCS.

o The data in Figure 2.3, which provide estimates of various roadbed soil parameters
based on their AASHTO and USCS classification systems (NHI 1998).

Although the data in Table 2.4 and Figure 2.3 provide, for each soil classification,

a range of values, the exact value to be used in the pavement design process is a decision

that must be made by the engineer on the job.

2.4 Laboratory Testing

The resilient modulus (MR) of a soil is an index that describes its stress-strain
relation under cyclic loads (Maher et. al 2000). Mechanistic-based pavement and overlay
design procedures require as input the subgrade MR to determine layer thicknesses and
the overall system response to traffic loads. MR can be obtained in the laboratory and

from the backealculation of nondestructive deflection test (N DT) data. The laboratory

20

Table 2.3 Comparison between three soil classification systems (USDA 1992)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

USDA Classification Percent Passing Sieve Number Liquid Plastic
texture USCS AASHTO 4 10 40 200 Limit Limit
Muck PT A-8 100 100 90-100 404% 0-14 NP
- SP'SM’ SM’ [33:11, 5i;- 40- 25
Sand SP, GP, GP- ’ ’ ' 15—90 0—35 <25 NP
A—2, A-3, 100 100
GM, GM
A-2
SM, SC- A-2, A-4,
Loamy SM, ML, A-l-b, A— 85- 60-
Sand CL-ML,SP- 1,A-2-4, 100 100 ”‘90 3’55 <30 NP
SM, SP A-3
. ML, CL,
3”” CL-ML, SC, A'4’A‘6’ 95' 85' 60—100 3095 <45 NP/P
Loam A-7, A-2 100 100
SM, CH
SM, 80 A—2-4, A-
Sandy SM, ML, 4, A-2, A- 70- 60-
Loam CL—ML, SC, l,A-1-b, 100 100 ”‘90 ”'75 <35 NP
CL A-6
CL CL-
Clay ’ A-6, A4, 95- 75-
Loam ML, SC, A”, A_2 100 100 70-100 3590 25-45 NP/P
SC-SM
CL, CL- A-4, A-6, 90- 75-
Loam ML’ML A_7 100 100 704% 50-90 15-45 NP/P
Mucky SM, SP, SP- A-l-b, A- 95- 75-
Sand SM 2-4,A-3 100 100 30'70 0'15 0'14 NP
A-6, A-7- 90- 85-
Clay CH, CL 6 100 100 65-95 45-95 30-65 P
Silty CL, SC, CL— A-4, A-6, 85- 60- _ _
Clay ML A_7 100 100 50-100 30 90 2550 NP/P
NP = non-plastic, plastic limit<10
P = plastic soil, plastic limit>10

 

 

determination of MR of roadbed soils is reviewed below, along with factors that affect

the MR values.

2.4.1 Resilient Modulus Test Procedure

In general, laboratory test procedures for the determination of MR value are

essentially based on the existing cyclic triaxial test methods used for the determination of

 

soil properties under repeated loads. A schematic of the test apparatus for conducting
MR tests is Shown in Figure 2.4. Figure 2.5 shows a typical hysteresis loop output
(stress-strain of one load unload cycle) used in the calculation of MR (Hall et al. 2001).

Figure 2.5 also Shows the recoverable (CR) and plastic (81)) portions of the axial
strain of the sample and the equation for calculating the MR of the soil.
Guidelines for conducting laboratory MR testing are given in the 2001 AASHTO T 307
standard test procedure. The procedure calls for placing a compacted soil sample in a
triaxial test apparatus, applying confining pressure and a sustained load to the sample,
and then applying a repeated axial load and measuring the resulting vertical
deformations (AASHTO 2001).

The history of the development of the AASHTO T 307 standard test procedure
traces back to the Strategic Highway Research Program (SHRP) protocol P 46-94
“Resilient Modulus of Unbound Base/Subbase Materials and Subgrade Soil.” The
protocol is based on determining MR in a repeatable, practical, and productive way.
After eight years of implementation, the protocol was adopted in 1992 by AASHTO as
test method T 294 replacing its predecessor T 274-82. The procedure has further evolved
to incorporate additional technical requirements and currently is labeled AASHTO T 307
standard test procedure (Groeger et al. 2003). According to Maher et a1, 2000, some of
the most recognized changes from the AASHTO T 274-82 and T 294-92 procedures to

the most recent AASHTO T 307 are:

22

Table 2.4 Typical resilient modulus values for unbound granular and subgrade materials

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(NCHRP 2004)
Classification Material Jounds/square mCh .

System Classification MR Range T121131
A-l -a 38,500 - 42,000 40,000

A-l -b 35,500 - 40,000 38,000

A-2-4 28,000 - 37,500 32,000

A-2-5 24,000 - 33,000 28,000

A-2-6 21 ,500 - 31,000 26,000

A-2-7 21,500 - 28,000 24,000

AASHTO A-3 24,500 - 35,500 29,000
A-4 21,500 - 29,000 24,000

A-5 17,000 - 25,500 20,000

A-6 13,500 - 24,000 17,000

A-7-5 8,000 - 17,500 12,000

A-7-6 5,000 - 13,500 8,000

CH 5,000 - 13,500 8,000

MH 8,000 - 17,500 11,500

CL 13,500 - 24,000 17,000

ML 17,000 - 25,500 20,000

SW 28,000 - 37,500 32,000

SP 24,000 - 33,000 28,000

SW - SC 21,500 - 31,000 25,500

SW - SM 24,000 - 33,000 28,000

SP - SC 21,500 - 31,000 25,500

USCS SP - SM 24,000 - 33,000 28,000
SC 21,500 - 28,000 24,000

SM 28,000 - 37,500 32,000

GW 39,500 - 42,000 41,000

GP 35,500 - 40,000 38,000

GW - GC 28,000 - 40,000 34,500

GW - GM 35,500 - 40,500 38,500

GP - GC 28,000 - 39,000 34,000

GP - GM 31,000 - 40,000 36,000

GC 24,000 - 37,500 31,000

GM 33,000 - 42,000 38,500

 

 

 

 

 

23

 

1 1523456810

ASTM Soil Classification System
Unified Soil Classification System (USCS)

MH

AASHTO Snil Classification
A-l-b
A-2-4, A-2-5

A-2-6, A-2-7

A-6
A-7-5, A-7-6

Resistance Value — R
5 10 20 30 4O 50 60 7O

Modulus of Subgrade Reaction — k (psi per inch)
100 150 200 250 300 400 500 700

Bearing Value (psi)
20 30 40 50 60

 

Figure 2.3 Soil classification related to strength parameters (NHI 1998)

24

  

Cell pressure inlet

  
 
 

 

Loading piston
.. F Load cell leads
. pasta

 

 

 
   

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DiSplacement ( “3;. i'
transducers fl Sample cap
Sample membrane
Clamps for
displacement
transducers
Sample ,
base
Vacuum Vacuum Displacement
inlet saturation transducer inlet
inlet

 

 

 

 

 

 

Figure 2.4 Resilient modulus testing apparatus for soils (NI-H 1998)

/\ I Total | _|

 

 

‘4 I MR = —
Resilient 5R
l
1:. Plastic
m +————_——..|
L:
O
E
>
Q.)
Q
\
I

 

 

 

I Deformation of soil sample (strain)

 

Figure 2.5 Resilient modulus concept (NHI 1998)

25

1. The number of loading sequences has been decreased from 27 to 15 and the number.
of loading cycle per loading sequence has been decreased from 200 to 100 cycles.
The decrease in testing sequences and load cycles led to a reduction in the sample
deformation and testing time from approximately five hours to two.

2. The maximum axial stress range was changed from 1.0 to 20.0 pounds per square
inch (psi) to 3.0 to 40.0 psi for base and subbase materials, and from 1.0 to 10.0 psi to
2.0 to 10.0 psi for subgrade materials.

3. There was a change of confining stress in the subbase testing sequence from 0 to 2.0
psi.

4. The implementation of a constant sustained stress of 10% of the applied deviatoric
stress (the difference between the axial stress and the confining pressure) to ensure
full contact between the loading piston and the sample.

5. Granular soils (Type I) are tested using a sample size of 6.0 inches in diameter and,
12.0 inches high. Cohesive soils (Type II) are tested using a sample size of 2.8 inches
in diameter and 5.6 inches high.

2.4.2 Issues with Current Test Standards

The AASHTO requirement to apply and remove the deviator load in 0.1 seconds
is difficult and costly. It requires high performance servo-valves and fast electronics. It
is believed that dynamic effects may become significant for 12 inch high specimens with
stiffness less than 20,000 psi and for 6 inch high Specimens with stiffness less than

10,000 psi (Marr et a1. 2003). While the rapid loading rate is used to model moving

vehicles on a pavement system, it is not clear that this fast loading rate is necessary. The

26

test would be a lot Simpler to run and the equipment less expensive if the loading period
is increased to 0.5 seconds.

Accurate measurements of the axial deformations are essential in obtaining
reliable MR results. The AASHTO requirements for such measurements have been
changed from two linear variable differential transducers (LVDT) mounted internally at
180° along the specimen’s axis, to two LVDTS clamped to the loading rod inside the
triaxial chamber to one LVDTS externally mounted to the loading piston and resting on a
rigid surface. These changes have made the test procedure manageable. It should be
noted that using the average of two LVDTS mounted at 1800 would decrease
measurements error if the sample is subjected to a slight rocking motion.

2.5 Field Testing

Field testing for determining or estimating MR is divided into two categories;
destructive and nondestructive.
2.5.1 Destructive Testing

The following five destructive tests are often used to estimate MR or k values.
2.5.1.1 California Bearing Ratio

The California Bearing Ratio (CBR) is the ratio between the soil’s resistance to
0.1 inch penetration of a standard piston to the resistance of a well graded and crushed
stone to the same penetration level. The test can be conducted in the field and the
laboratory as described in the AASHTO T193 standard test procedure (AASHTO 2001).

The 1993 AASHTO Pavement Design Guide uses Equation 2.2 to estimate the
MR from the CBR (AASHTO 1993). It Should be noted that the 1500 constant in

Equation 2.2 can vary fi'om 750 to 3000 (NHI 1998).

27

MR -—— ISOOCBR ' Equation 2.2

Whereas the new M-E PDG recommends the use of Equation 2.3 for estimating

the MR value (NCHRP 2004).

MR : 2555(CBR >064 Equation 2.3

2.5.1.2 Dynamic Cone Penetrometer

The Dynamic Cone Penetrometer (DCP) Shown in Figure 2.6 is a graduated rod
with a metal cone on one end and a mass which is repeatedly lifted and dropped to drive
the cone into the soil. The DCP is an efficient and inexpensive way to estimate the in-
place CBR. The cone’s penetration rate (PR) is measured after every drop and is labeled
the DCP index. The DCP index (DCP = mm/blow) correlates well to CBR for fine
grained soils up to CBR value of about 15 percent. Equation 2.4 from the M-E PDG

provides a correlation between the CBR and the DCP index (NCHRP 2004).

 

CB R _ 292
_ Equation 2.4
D C Pl . 12
Equations 2.3 and 2.4 are combined in Equation 2.5.
1 0.72
— Equation 2.5

Equation 2.6 provides another correlation between the CBR and the penetration

rate (PR) ofa 60° cone (NHI, 1998).

405.3

CBR =
PRI.259

Equation 2.6

28

Where, PR = penetration rate, mm/blow

. , *—'Hammer
3 (17.6 lb)

Cone angle 60° 1‘

 

  
 

 

= a.
—-> 0.79 in It— I
39.4 in
(variable) “Steel rod
I (0.64 in)
l in = 25.4m
1 lb = 0.454 kg '
«- Cone

Figure 2.6 Schematic of a dynamic cone penetrometer (NHI 1998)

2.5.1.3 Plate Load Test

Plate load testing of subgrade soils is not commonly used because it is laborious
and Slow destructive test that requires the removal of segments from the pavement
surface and base layers (Yoder 1959). It is, nonetheless, a direct method for determining
the modulus of subgrade reaction which is a required input in the current AASHTO
concrete pavement design procedure. Figure 2.7 depicts a photo of the plate load testing
apparatus. Guidelines for repetitive static plate load testing are given in ASTM D1195
“Standard Test Method for Repetitive Static Plate Load Tests of Soils and Flexible
Pavement Components” for use in evaluation and design of airport and highway
pavements, and in AASHTO T 221 standard test procedure (AASHTO 2001). The static
elastic k value is calculated as the ratio of the applied pressure to the elastic deformation,

which is the recoverable portion of the total measured deformation (Yoder 1959).

29

 

Figure 2.7 Photo of plate load testing apparatus (NHI 1998)

2.5.1.4 Pocket Penetrometer

The pocket penetrometer is a small hand held device with a Spring loaded probe at
one end. The probe is pushed to penetrate the soil 0.25 inches; the Spring measures the
resistance of the soil to penetration. The pocket penetrometer is used to estimate the
unconfined compressive strength of the soil. Since the MR is a nonlinear elastic soil
property it is logical to assume that there is a relationship between soil strength and MR
(Han et a1. 2006). Thompson and Robnett (1979) proposed estimating the MR value

from the unconfined compressive strength using Equation 2.7.

MR = 0.86 + .307q,

Equation 2.7

Where, MR = resilient modulus, ksi

qu = unconfined compressive strength, psi

2.5.1.5 Pocket Vane Shear Tester

The pocket shear tester is used to estimate the undrained shear strength of the

30

soil. The shear tester is inserted 0.25 inches into a flat soil surface and rotated until

failure. The maximum pressure required to cause failure is the 5,, value. Sukermaran et

al (2002) suggested a relationship between the undrained shear strength and MR.

MR = 100 — 5003, m,

Equation 2.8

MR 2 500 — 15005,, “<30

Equation 2.9
Where, MR = resilient modulus, psi

Su = undrained shear strength, psi

P1 = plasticity index

2.5.2 Nondestructive Testing

The nondestructive deflection test (NDT) is the most popular test in this category.
In recent years, the use of NDT has become an integral part of the structural evaluation
and rehabilitation of pavement structures. Various types of equipment are available
including the Road Rater, Kuab, Dynaflect and the falling weight deflectometer (FWD).
NDT devices are used by state highway agencies to apply patterns of loading and record
deflection data along the pavement surface. The deflection data measured along the
pavement surface at different distances from the center of the load are typically used to
backealculate the modulus values of the various pavement layers and the roadbed soil.
Numerous backealculation software packages are available either in the public domain or
can be purchased. Most of these use more or less the common procedures presented in

the next sections.

31

2.5.2.1 Backcalculation of Resilient Modulus from Deflection Data

The subgrade modulus can be determined by using the pavement surface
deflection measured at distances of 48-inches or more from the center of the load.
Because of arching effects, at these distances, the pavement surface deflection is
influenced mainly by the roadbed soils. Hence, the subgrade MR can be backealculated
from a Single deflection measurement. The most widely used routine to backealculate the

subgrade MR from a Single deflection measurement is the Boussinesq equation (George

 

 

2003)
2 2
CP(1—v ) CP(1—v )
dr= 0rMR= E t_ 210
quaion .
IrrMR 727d,
Where, CIr = the surface deflection in inch at a distance r in inch from the the load

P =applicd load in pounds
C = correlation/adj ustment factor accounts for the difference between the
backealculated and the laboratory obtained MR value.
MR = resilient modulus in psi
By assuming a Poisson’s ratio of 0.5, equation 2.10 can be reduced to the

following equation (AASHTO 1993).

0.24CP

d ,- Equation 2.1 l
r

MR:

 

AASHTO recommends the use of C value not greater than 0.33
The minimum distance (r) in Equations 2.10 and 2.1 l is given by the following

relationship.

32

 

   

r=0.7 (12+ D3

Equation 2.12

Where, 212 = radius of load plate
D = total thickness of pavement layers above the subgrade
E p = effective modulus of all layers above the sub grade

Ep in equation 2.12 can be calculated by using the following equation:

 

 

 

 

 

r 1 N
1 ‘ —”—2
D
1 + —
M Rdo _ a >
qa E p Equation 2.13
MR

Where, do = deflection measured at the center of the load plate after adjustment to

a temperature of 68 0F
q = pressure on load plate
The Washington State Department of Transportation (WSDOT) developed, for
asphalt pavements, Equations 2.14 through 2.16 and, for concrete pavements, Equation
2.17 to estimate the subgrade modulus from deflection sensors located at various

distances from the center of the load (Pierce 1999).

33

(12892

d 24 Equation 2.14

1000

 

MR(p‘si) = 9000 ><

 

MR( psi) = —466 + 9000 XW .
( d36 ) Equation 2.15
1000
. __ 0.00567
MR(psz) — —l98+9000X——( d48 ) Equation 2.16
1000

MR = 9000 0‘00577 — 1 11

48 Equation 2.17

1000

 

 

Where, d24, d36 and d48 are the pavement surface deflections in inches measured
at 24, 36, and 48 inches from the center of the load.

2.5.2.2 Backcalculation of K Value Using Pavement Surface Deflections

Deflection testing at the center of a slab can be used to determine the value of the
dynamic modulus of subgrade reaction (k) and the effective thickness of the concrete
slab, based on an assumed modulus of elasticity of concrete of 5,000,000 psi (Hall and
Crovetti 2000).

Two different analyses were presented with and without the use of the deflections
recorded at the center of the loading plate. Initially, deflections are used to compute the

deflection basin AREA using the following equations:

34

6

ARE/14 = 7(‘10 + 20112 + 2d24 + d36) Equation 2.18
0

AREAS 2 “ES—(1 + 2dis + 3“'24 + 661136 + 461,60)

12

Equation 2.19

Where, AREA i = deflection basin AREA, inches

di = surface deflection measured at 1 inches from the load center

The calculated AREA values are then used to backealculate initial estimates of the

radius of relative stiffness (dense-liquid foundation model), using the following equations

 

 

 

 

 

 

 

 

(Hall and Crovetti 2000).
— 1 4.387
In 36 — AREA4
l _ 1812.279
k4—est _
—2.559 Equation 2.20
- - 2.22
48 — AREAS
In
I _ 158.40
k5—est _
_0476 Equation 2.21
Where, 1km, = estimated dense-liquid radius of relative stiffness, inches

Hall and Crovetti initially estimated the dynamic foundation k value, based on

infinite slab Size assumptions, are backealculated using the following equations (2000).

35

—0.14707e(‘°'°7565’k4-85')]

_ 0.12456 P

est 2 ,
d ( l ) Equation 2.22
0 k4—est

 

k4

[—0.79432e(‘°'°7074’k 5-61)]

 

__ 0.12188e P

est 2 .
d ( l ) Equation 2.23
12 kS—est

k5

Where, kiest = estimated dynamic interior foundation k value, psi/in

P = applied load, lb

di = utilized maximum deflection at i inches from the load, inches

[kl--95, = estimated radius of relative stiffiiess, inches

Based on Slab size, correction factors for the estimated radius of relative stiffiiess

are computed using the following equations.

1.04831

CF,,_,S, =l—0.89434exp —O.61662 —‘?’-’— 5.111.110.1224

k—est

0.80151

 

CFD, = 1 —1.15085 exp —0.71878 5’7

k—est

Equation 2.25

Where, CF 1”,, = correction factor for 1km

CF 0,- = correction factor for utilized maximum interior deflection

L47: effective Slab length, inches

36

The effective Slab length is computed based on the length and width of the test

Slab, using Equation 2.26 (Hall and Crovetti 2000).

Lefl : V145 X Lw Equation 2.26

Where, L, = Slab length, inches
1... = slab width, inches
After computation of Slab Size correction factors, the adjusted radius of relative

stiffness and dynamic foundation k value are computed using the following equations.

 

lk—adj : CPIk—est X lk—est Equation 2.27
k _ _ K est ,
adj — 2 Equation 2.28
(CFlk—est) XCFDi

Other backealculation procedures exist for pavement surface deflections including
one presented by Frabizzo (1998). From FWD data a backealculation procedure can be
used to calculate the deflection basin area (AREA), radius of relative stiffiiess (I), elastic
modulus of the concrete (EC), and the modulus of subgrade reaction for rigid pavements.

AREA is the cross-sectional area of the deflection basin between the center of the
FWD load plate and the outer most deflection sensor. It is calculated using deflection

‘6 99
I'

data at various distances from the center of the load plate. AREA calculation is

normalized with respect to the sensor beneath the load; resulting in units of length

(Frabizzio 1998).

AREA: 4+6é8— +5$ + E +9.52; +153: +1 1%
a, a, a, a, a, 50

Equation 2.29

37

Where, AREA 2 deflection basin area, inches

: . th .
6;- deflection of the 1‘ sensor, inches

The radius of relative stiffness, which characterizes the stiffness of the slab-
foundation system, can now be calculated. Equation 2.30 is used to calculate I if the load

radius is 5.91 in (Frabizzio 1998).

 

6O AREA 2.566
I = LN 289708 /(— 0698) Equation 2.30
Where, I = radius of relative stiffness, in

Now the elastic modulus of the concrete can be calculated at each sensor location.

:12(1—v2)P126:

C 3 E at'on 2.31
6’}? qu 1

 

Where, E c : elastic modulus of the concrete, psi

v = Poisson’s ratio for concrete = .15
P = FWD load, pounds

2 radius of relative stiffness, inches

(C ’9

5*
r — non-dimensional deflection coefficient at distance r

h = concrete slab thickness, inches

: . th .
6r deflection of the 1' sensor, inches

38

5* : ae[_be(_d)]

r

Where,

I = radius of relative stiffness, inches

Table 2.5 Regression coefficients for

a, b and c = regression coefficients (see Table 2.5)

5*

r

 

 

 

 

 

 

 

 

 

 

 

 

Radial Distance, r
. a b c
(inches)
0 0.12450 0.14707 0.07565
8 0.12323 0.46911 0.07209
12 0.12188 0.79432 0.07074
18 0.11933 1.38363 0.06909
24 0.11634 2.06115 0.06775
36 0.10960 3.62187 0.06568
60 0.09521 7.41241 0.06255

 

Equation 2.32

 

The seven elastic moduli of the concrete are averaged to obtain a representative

EC. Now, the modulus of subgrade reaction, which characterizes the stiffness of the

foundation, can be calculated.

Ech3
12(1—v2114

k = modulus of subgrade reaction, pci

 

k:

Equation 2.33

Where,

E:

c concrete modulus of elasticity, psi

h = concrete slab thickness, inches

1 = radius of relative stiffness, inches

39

2.6 . Correlations Between Backcalculated Modulus, Laboratory-Based Modulus,

DCP, and Soil Physical Properties

Many correlations exist to convert laboratory modulus to backealculated modulus.
There are also correlations between soil properties and their MR. An introduction to
these correlations can be found below.

2.6.1 Correlations between Laboratory and Backcalculated Resilient Modulus

The primary purpose of establishing relationships between backealculated FWD
modulus and laboratory modulus is for the design of pavement overlays. The laboratory
MR values are stress dependent. Therefore, in order to compare the different modulus
values, the stress state in which the FWD test was performed must be known (George
2003)

Whether the laboratory modulus or field modulus of the subgrade soil is used in
the pavement design and analysis depends on the input required for the model being used.
For example, the original AASHO road test was calibrated to the laboratory MR of the
soil. Therefore, when using the 1993 AASHTO pavement design or overlay procedures
the appropriate input for the subgrade soil is the laboratory MR (AASHTO 1993).

MR values obtained from laboratory tests may be considerably lower than the
backealculated MR values due to differences in the magnitudes of the deviatoric stress,
confining pressure, and loading rate (George 2003). Similarly, field MR values for fine
grained soils, obtained by backealculation from FWD deflections, have been reported in a
number of studies to exceed the laboratory resilient modulus values by factors between 3

and 5 (AASHTO 1993).

40

Layer theory was employed for the analysis of the stress state under a 9000 pound
FWD load. It was found that a reasonable correlation exists between FWD
backealculated moduli and the laboratory moduli based on the in-Situ conditions with

identical stress states (Ping et al. 2002).

MRFWD = 1‘6539MRlab Equation 2.34

From Equation 2.34 the FWD backealculated moduli were about 1.65 times
higher than the laboratory MR. The ratio is in agreement with the suggestion by the
AASHTO design guide (AASHTO 1993) that the FWD backealculated moduli are
approximately two to three times higher than the laboratory determined moduli,
considering that the AASHTO relationships were based primarily on clay soils. In
addition, for this comparison the FWD tests were performed under in-situ soil conditions
and the laboratory determined MR were obtained from the reconstituted soil samples;
simulating the in-situ moisture and density conditions under identical states of stress.
The possible causes for the difference between the lab MR and backealculated values as
reported in this study (Ping et al. 2002) were:

0 The FWD backealculation program is based on the linear elastic theory of multiple
layer pavement structures while the pavement materials are not elastic.

0 The FWD backealculation method is not a unique solution method; therefore,
different layer moduli could be obtained from the same FWD data.

0 The lab Specimens were tested almost immediately after they were compacted, and
the confining pressure for the triaxial test was applied by air; the in-Situ soil had been
there for a long time, and the confining pressure was caused by vertical load and soil

weight.

41

Von Quintus and Killingsworth believed the reasons for the differences in the
laboratory and field moduli were related to the inability of the laboratory tests to simulate
the actual in-situ confinement and effect of the surrounding materials in both the lateral
and vertical direction (1998).

For rigid pavements the dynamic k value obtained from backealculation is about
two times greater than the static elastic k value that would be obtained from plate load
testing of the same soil. This is due to the difference in the 5011’s response to dynamic
and static loads. Correlations have been developed to estimate the k value as a function
of CBR, density, and soil class. Additional correlations between soil properties
(gradation, density, moisture content), soil classification, CBR, DCP penetration rate, and
MR are available in the literature.

2.6.2 Relationship between Laboratory and Backcalculated Resilient Modulus and

Physical Soil Properties

Many research studies (George 2003; George 2004; George et al. 2004; Janoo et
a1. 2003; Janoo et al. 1999; Maher et al. 2000; Rahim and George 2003; Yau and Von
Quintus 2002) are available in the literature which investigated correlating the laboratory
or backealculated MR to soil index or physical properties.

Correlations have been developed to estimate soil k value as a function of CBR,
density, and soil class. Several of these correlations are summarized in Table 2.6.
Additional correlations between soil properties such as gradation, density, moisture
content, soil classification, CBR, DCP penetration rate, and MR are given in the Illinois
Department of Transportation’s Guidelines on Subgrade Inputs and Subgrade Stability

Requirements for Local Road Pavement Design (Hall et al. 2001).

42

Table 2.6 Range of k value for soil type, density, and CBR (Hall et al. 2001)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dry Static k
AASHTO Class Soil Description Claslsjisfiiition Density (ES/31; value
(lb/1’13) 0 (psi/inch)
Coarse grained soils
“fax“ 125 - 140 60 - 80 300 - 450
A lgr 1 Gravel GW, GP
’ 'a’ poor y 120 - 130 35 - 60 300 - 400
graded
A-l-b Coarse sand SW 110 - 130 20 - 40 200 - 400
A-3 Fine sand SP 105 - 120 15 - 25 150 - 300
A-2 soils (granular materials with high fines)
A-2-4, 11 S1 1
gave y , ”y grave GM 130 -145 40 - 80 300 - 500
A-2-5, gravelly Silty sandy gravel
A-2-4, sandy Silty sand
, SM 120 - 135 20 - 40 300 - 400
A-2-5, sandy Silty gravelly sand
A-2-6 11 C1 1
’ng€ y my grave GC 120 - 140 20 — 40 200 - 450
A—2-7, gravelly Clayey sandy gravel
A-2-6 d l (1
’5“ y C aer 3““ SC 105 - 130 10 - 20 150 — 350
A-2-7, sandy Clayey gravelly sand
Fine grained soils
S'lt 90 -105 4 - 8 25 -165
A-4 _ ' , ML, OL
Silt/sand/gravel mix 100 - 125 5 - 15 40 - 220
A-5 Poorly graded Silt MB 80 - 100 4 - 8 25 - 190
A-6 Plastic clay CL 100 - 125 5 - 15 25 - 225
A-7-5 MOderati’ly 1’13“” CL, OL 90 - 125 4 - 15 25 — 215
elastic clay
A-7-6 H‘gh‘y 9323‘: 6133‘” CH, OH 80 - 1 10 3 — 5 40 - 220

 

 

 

 

 

 

It should be noted that the k value of fine grained soil is highly dependent on the degree

of saturation. Adjustments to the k value are required for embankments less than 10 feet

thick over a softer subgrade, and/or for bedrock at a depth within 10 feet.

43

 

3.]

CHAPTER 3

FIELD AND LABORATORY INVESTIGATIONS

Introduction

At the outset, field and laboratory investigation plans were designed to

accomplish the objectives of this study. The plan consisted of the following activities

and tests:

0 Soil delineation in the State of Michigan

0 Soil sampling

0 Field tests which consist of:

O

O

O

Penetration resistance using pocket Size penetrometer
Shear strength using pocket vane shear tester

Deflection using falling weight deflectometer (FWD)

0 Laboratory tests which consist of:

O

O

3.2

Moisture content

Sieve analysis

Atterberg limits (liquid and plastic limits and plasticity index)
Hydrometer analysis

Cyclic load triaxial test

Soil Delineation

As stated in Chapter 1, the State of Michigan is geographically located within the

glaciated section of North America and most of its soil has developed from glacial

deposits. The thickness and composition of the drift varies from clay to gravel depending

on the location. Because of the complexity of the glacial drift, about one hundred and

44

sixty-five different soil types were formed and are being used for engineering purposes
by MDOT. To characterize the resilient modulus of the glacial drifts in an economical
and practical manner, the State of Michigan was divided into 15 clusters where the soil in
each cluster has Similar engineering and physical characteristics. The boundaries of the
15 clusters were established based on the 1982 Quaternary Geology map of Michigan
(DEQ 1982), inputs from members of the research Technical Advisory Group (TAG) of
MDOT, and inputs from the soil engineers in the various MDOT Regions. After
establishing the cluster boundaries, each cluster was divided into areas based on the
percentages of each soil type found in the Natural Resources Conservation Service
(N RCS) Web Soil Survey (Web Soil Survey). Once again, the boundaries of each area
were slightly modified based on inputs from the TAG members and from the soil
engineers in the various MDOT Regions. The final state divisions consisted of 99 areas
within the 15 clusters. Figures 3.1 through 3.4 depict the boundaries of the clusters
shown by the dashed lines and the boundaries of the 99 areas (lightly shaded) shown by
the green lines. Once again it Should be noted that the division between the clusters was
based on Similar (not the same) soil types whereas the boundaries between the areas were
based on narrowing the range of the soil parameters within each cluster.
3.3 Roadbed Soil Sampling

After dividing the State of Michigan into 15 clusters and 99 areas, the percent of
each soil type (sand, clay, Silt, etc) in each area was obtained from the Natural Resources
Conservation Service (NRCS) Web Soil Survey (Web Soil Survey). Table 3.1 shows the
percentages of each soil type in each of the 99 areas. Because of budget constraints and

based on similar soil makeup, some areas within some clusters were grouped together

45

 

 

 

Figure 3.1 Western portion of the Upper Peninsula

46

‘_ “'9‘K ”in, it} _,

' sew-N

     
  

  

 

 

 

 

Figure 3.2 Eastern portion of the Upper Peninsula

47

 

 

 

 

 

 

 

 

 

 

 

His:

{’3' 06

 

 

Figure 3.4 Southern portion of the Lower Peninsula

48

.Hfiom 2% an :30qu E Eng 05 .«o 8528305 2E. $20 ecu Em 6:8 2:328 Smog ...

69$ :8 85 mo Savanna o 8865 £60 beau ”Bo Z

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

x 2: mom was a
x is on 5.8 8
x we 2 3. mm Em 8 3
x 3. m2 is 8
N2 is a: R S
x S Em to Na: 8 8
2 n: as ea New 8
x 2 Rm 2 mi 8
x 8 2 w as 3 8
x a: 2 Ni ea RN om S
n: so 2 Na 2 a a
x 5.3 oz 8
n2 SN em a: ad 2 8
x 3: 3a 3 ea 2 New 3 S
3 4.2 m 2 98 Ba 8
x a :2 mm 3: ”.2 S
x 53 oz 5 S
eoaoaea Ea Eu 382 .885 sass seam new .2585 Saw 0332

 

 

 

 

 

 

 

 

 

 

 

 

 

 

£833 2 05 :32? «one some new mowfieoaom mom mm 2an

49

.Emom $5 an 565?: E 88— 2: mo 850335 2: $30 Cam :6 Beam 292:8 884 ...

.25 :8 BE 68 €083 o 2865 £00 .395 ”302

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fin m.» 06m mém 2 co
X I: 3 odm mam fimfi S
X 52 men Qmm N6 8 no
X we as 3: :3 N 3
X _.mm m fine no
X 03 v.: We 3: we
m: N? v.3 98 no
X om w mdm 5mm vo
X :0 Qm meme mm mo co
X Nw ms mi. mm mo
X md Vdm m.mm m.m S
X Now. 2 «.mm «.3 :0 co
X >4 QR mo
v.2 .13 Wow vo mo
X N M: 5K 3: mo
mdm 9:0 <2 No
X Yaw m mm 5
as a. a... a... e... a. a.
eomomoE bum ED 3032 *Smoq BED Beam bmm .380.“ ceam x032

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6.83 E ozee

50

.EBQ mg “a 9505?: mm 88— 2: no 550385 BC. $20 was Em .eemm mESCOQ Smog ._.

.099. :8 35 mo 33qu o 8865 £00 395 ”802

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.20 3% 9w N.N co
X QNm m.wN mNN NN~ vo
Nm NAN N. 2 m wé no
v.0 vNN mic m3: m6 mo
X VNH m6 2: Wow wNN No mo
X vAN fiwfi v.3 imm mo
X mm v.0 ow mo
X wNo ed fig _.N_ we
X ms m.m véo 50 m: S
X Wmm ed N.m m.mv N.w S
X w 2 N.NN Yam Ndfl co
X _. N 0.3 de No
5.3 fie NdN m _ mo
X w.wm m.: 0.2 v.2 3 mo
X Wow 0N 9w N we
X _.N\. méN N .8
eases S S 8 S S S S 8 S 8
. >30 “Seem Emoq Smog Ewe; Comm «ca 5520
womomoi bzm .35 30:2 .4884 >935 Exam bzm .8804 2mm x032

 

 

 

 

 

 

8.83 E 2.5

4:6: is: E 8505?: mm :82 one. no :Boexmoen BE. $20 ES :6 £53. £580 :50: a.

09$ :8 :2: .«o E080: o 8885 360 meo ”802

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2: 92 e 3
x 2; Se 8
x 2: v.8 e.: 4.: mo :
x mam wee we as S
x on e: 3m 2 S
is ode E 8
x Ge 0% w 2 8
new is me 8
x 2% 3a a: 2: 8
x 2:. n; we: 3
x _.: mas. o: 3m 8 S
x an new 4.: 2 8
ace :2 w: 2: 8
x was a: a: :
x we. 2: 2 S
was 3... we
weasem g g g g axe g g g g g
. >20 gem Smog Smog Smog gem meg 8330
womomobw baa >20 493:2 Smog xoxflU hfigm \sz caged Guam £032

 

 

8.83 E osee

 

:50: m5 us 8592:: v: :82 05 mo 550385 2F >20 :5 am :53, 85:8 :83 ...

69>: =8 35 mo :5on o 23.65 mzoo >380 ”802

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

X M; vv v.3. v.2 S
X mém om 9mm mé :2 mo
X mém fl?” 9: 2 No
X o0 NE 0.2 v No vo
. . . m:
X m mm m nm e 0 mo
:6 Q: 3% mém NH co
X w.m w.mN Dom m.wm fim no
X WE mdm fiwm 5m :6 mo
3 ON: w.m No 3
X 5.: Wow 2: 8
X vw > No no
X 08m Dom 03 Q in we 2
X o.: mic. v.2 ENN co
v.0 Nmm v.2 9m Na 06 No
X new vdm NM: Qm No S
mEEEmm MMW Afiv “MW 33 :meNH :meva :mwoovq “WWW ax: £9 32 5320
vomomoi >H=m >20 3032 8:04 >o>20 >::mm bmm >884 93m x032

 

 

 

 

 

 

 

 

 

 

 

9.603 3 29¢

53

:Eom £5 “a 8505?: a :82 2: no 850333 one >30 98 :6 6:3 25:8 Smog *

69>: :8 3:. mo E080: o 886:: E8 39:0 ”802

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cm 9mm :6 mo

X hm 2h v.3 ms mo

X m.w Qmm mdc wé 8

X Hm m.mm mm 2 S 2
fiom NS cm no

X “.mm 02 3: Wm v.0 mdm No

X w.mo n w.m Wm: fim co

X 5.: 5mm 0;: mdm #0 vs :0

X <H<Q OZ co

X <H<Q OZ mo 4

X 02 n Em a 0.: Wm B 5
3% ~23 m.m N: mo

X mdm fiom a mo 3

X 0:2 OS 0.3 $2 ms v.8 no

X om N42 iwm 3

X m.m N.mm 0mm 2 NW mo

X m.mm v.3 Ywm m Wm No

X M3; :2 mdm om Wm S

wEEEwm MMW Q5 “WWW OXVV :mWooW :mWqu :monvA “WNW AXV go 83 5530
womomoi 32m >20 30:2 :80: >o>£0 >v:am >26 >Emoq 98m 3032

 

9,380 2 22¢

which reduced the number of areas from 99 to 75. The combined areas are collectively
marked by the letter “X” in Table 3.1. For each of the 75 areas, disturbed roadbed soil
samples were obtained. Table 3.2 provides a list of the designation number and the
location of each of the disturbed roadbed soil samples. The designation number consists
of 9 characters A-BCD-E-(FG-HI) where A designates the road type (I=interstate, U=US
road, M=Michigan road), BCD represents the route number, B shows the traveling
direction (N=North, E=East, S=South, W=West), FG is the cluster number (01, 02,

15), and H1 is the area number (01, 02, 10). For example, the sample designation
number M-059-W-(1 3-02) implies that the sample was obtained from Westbound M—59
in cluster 13 and area 02. Table 3.2 also shows the results, when available, of the pocket
penetrometer and the pocket vane shear tester. It should be noted that empty cells
indicate no tests were performed at those locations.

In addition, 10 undisturbed Shelby tube samples were collected at locations where
the disturbed soil samples indicated the presence of clay subgrade soil. The locations of
the Shelby tube samples can be found in Table 3.3. Both the disturbed and undisturbed
soil samples were properly transported to the Geotechnical Laboratory at Michigan State
University where each soil sample was subjected to various laboratory tests. These tests

and the test results are presented in section 3.5 below.

55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E 83 a: 8:3 3: 883. 8 82 808 82 28.88 282.2
8.8 8 8-2 8 883 888 888 8888 388-2
38.88 :88 :8 8.88 3: as»: .8 888 8.8 on» 8888 888.2
to 83 88 8: oo 8 882 88.: em: 9888 2.5.:
H 88.8 888 88$ 8888 58822 8 £88 88.: 88 2888 2.5.:
:8 8 B: 85: .8 :8: 8.8 888 8888 388-2
3885: :88 88 8.88 83: am 8 £88 80.: 82 98-80 888.2
8.8 no 83: 8830 28: 8 83 80.: o: 2888 388-2
8 no 888: .8 83 888 88: 28-80 388.2
8.8 8.8 3: oo 8 :8 £82 888 88 28-80 888-:
3828: :88 88 8.88 888 8888 0:8 88.80 288-:
8.8 88.0 888 8888 0:8 98.80 288-:
8.: 2 t 72 8 88m 888 88 2.8.80 888-:
8 8.8 3: 288m 8 83 :88 88 28-80 388.2
3 8.8 R2 8 882 888 82 A888 3882
8.8 88.0 2 83m 8 88m :88 88 2888 888-:
28888 8.88 =8 8.88 3: 888:8 .8 88m :8: 88 @888 888-:
E 88.0 88-2 .8 s8: 8...: 8882 28-80 388.2
8.8 8 :2-2 8 83 8.8 822 28-80 388.2
8.8 8 88-2 8 88m :88 888 2088 :89:
388:8 8.88 8: :88 838 882880 8 £88 808 83 :38 888-2
8808885: 888
Hoxoom 80:8 05> :ocaoo-H 85:5: 038%

 

 

 

 

 

$88 828 38> 0:8 886885: 8880: 00 32884 Nam 038:.

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.: 88.: 3: 828m 8 882 88 88 28-80 888-2

88 8838838 888 88.: 888 28-80 8.88.:

2+5 88888 28-80 8-38-2

>833 8 88888. E :8 8:5 2 So 3 8 E

Z 88 :88 2mg 88 88888. ooswomaw 888% 8.3.8 288-80 8; 8 2-2

2 8.8 838 82 2 8 888 88.: 82. 28-80 8.82.:

8E 883 882 8 883 88.: 2: 28-80 3.88.:

8: 88382 83 8 888 88 82 28-88 8-: 82-:

2.8 8.2 82 88 88 8888 8 882 888 O8 8088 2-88-:

38888 :88 :8 8.88 82 2,28 8 888 888 8: 8088 8-88-:
3888: 8.88 :8 :88 8: 388882 8 883 8...: 82 @888 3.08.:
8.8 :8 888 82 888 8888 8828 8 883 8.8 88 2888 388.2

8.8 8.0 888802 8 883 888 82 2888 828.2
28828: 8.80 8-2 8 as 888 828 88: .88 08 88-88 838.2

8.8 8 82 85> 8 882 8.8.: 88 280-80 2-882-:
28888 888 :8 :88 82 888888 8 882 8.8 22 88-80 2-882-:
85. 8838980 8 88m 888 882 28.88 8288.2
82 8:2 8_ 8 88m 88 8.2 28-80 3-08-2

8.8 2 82 88888 8 882 888 S 2 280.88 2-88:

88-2 8 882 88 88 28-88 2.28.:

8.2 88.8 888 82 oo 8 882 :88 82 28-88 2.88.:

88:80:08.2“ :88
woo—com 880:8 28> 55804 89.82: 228%

 

 

 

 

 

8.803 8.8 8888

57

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.8 88 888 88 88828 885 88 888 88.8: 3.18.:
88:2 8888 88-8: 3888-2
8 8.8 88 8% 8 882 88 888 :88: 8.88.:
8.8 8.8 8a: 8888 8 885 88 888 :88: 88-85-:
8.8 8 88 2888 82 8 883 88 88 88-8: 3-8888
8 88 888 88 2888 8 88m 88 882 88.8; 8.88.:
8.8 2 888 88 888 888888 88s 88 88 88.8: 2188.8
8.8 8 88 882 88 8 88s 88 88 28-8: 88.8888
88 88 8% 888cm 8 888 88 88 88.8 8 8888-8
8.8 8 88888 88-8 888888 8 883 88 888 88.8 3 3888-2
8.8 8.8 888 8 8: 88 8 88,8 88 882 288-8 3 3888-8
88 8888 888 8 882 88 888 :8; 3 2.88.8
8.8 8 88 88888 8 888 88 888 2 8.8; 888.2
88888 888 88 888 2 888m 8 888 88 882 88-8: 8888-8
8.8 8.8 88 885 8 885 88 8: 888-8: 3888-8
8.8 8.8 8a: 8888 838 8 882 88 888 888-8: 2888-8
8 8: as: 88888 882 88 888 @883 2888-8
8._ 8 88 88 8888 88888 883 88 888 288-8: 3888-8
8 8 8am 8: 888 8 882 88 88 38-8: 2-8888
2 8.8 88 838 8880 8 888 88 8: 88-888 8888-8
8288088qu 888 88583 598:: macaw
886088 880:8 28> .

 

 

 

 

 

8.888 8.8 888

58

 

N872 mo £28m 0:8 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.8 8.2 32 8888228 8 8288 88 888 288-828 8828-22

8.8 8.2 82 28882 8828 882 88 8: 288-828 2.8.2

8.8 8.2 32 8882 8 882 88 888 288-8 28 8-888-:

2288 828: 88.8225 8 888 88 88

88888 888 88 8.288 88 88.882 8 882 88 888 228.828 82888-22
8.8 8.8 32 88822 8 888 88 828 288-8 28 82888.2

8.8 8 88-22 8 888 88 888 288-828 8888-22

8 228 8.288 32 88282 8 888 88 888 282-828 3888-2

8 8.8 282 8882883 8 888 88 888 288-828 3888-2

32 28280 8 8888 88 8882 288-828 8-888-:

8.2. 228 888 32 88885 8 88m 88 8882 288-828 82882-22

8.8 8.8 2:282 88822 8 885 88 888 288-828 3888-2
8888 8288 288-828 8-88-2

8.8 88.8 82882 8 882 88 882 228-828 8.88.:

8 8.8 32 82832 8 8888 88 88822 288-828 8.88.:

8.8 8.8 32 882828 888 88 88 288-828 8-88-2

8.8 8 32 88880 8 8288 88 88 228-828 8-88-2
8.8 88 888 8+8 82888 288-8 28 82828.2

8.2 88.8 2888.28 8 882 88 88 288-828 2.88.:
88888 8.288 228 8.88 838 2888282 828 32 828822 82,882 888-828 82828-22
88+88 282888 228-828 3-8882
8888220822028 888

8828A 8888 28> 22058004 832282282 uEEmm

 

 

 

 

8.888 8.8 8288.2.

59

Table 3.3 Locations of Shelby tube samples

 

 

 

 

 

 

 

Region Sample number Control Section Number of Shelby tubes
M-153-E (14-06) 82081 1
Metro M-010-E (13-08) 82111 1
I-094-W (14-09) 77111 2
University I-075—S Q 4-01) 58151 2
Bay U-010-W (08-04) 9101 2
U-127-N (07-05) 37014 2

 

 

 

 

 

 

3.4 Field Testing

The field testing consists of measuring the soils penetration resistance using
pocket penetrometer, the shear strength resistance using pocket size vane shear tester, and
pavement surface deflection using FWD. The pocket penetrometer and pocket vane
shear tests were conducted at the same time and at the same location where some of the
disturbed roadbed soil samples were obtained. The FWD tests, on the other hand, were
conducted during the fall and spring seasons to assess the impact of the two seasons on
the measured deflection and consequently on the backealculated roadbed soil modulus.
The three tests are presented below.
3.4.1 Penetration Resistance Using Pocket Penetrometer

The pocket penetrometer is a small hand held device that consists of a spring
loaded probe that slides into a cylinder. The maximum pressure required to push the
probe 0.25 inches into the soil is recorded. Pocket penetrometer is typically used to
estimate the bearing capacity of the soil surface (Liu and Evett 2008). However, for this
project the penetration resistance will be recorded and correlations between MR and

penetration resistance will be developed (if possible). A total of 67 pocket penetrometer

60

tests were conducted and the locations where pocket penetrometer tests were performed
and the results can be found in Table 3.2.
3.4.2 Pocket Vane Shear Test

The field vane test is used to estimate the undrained shear strength of the soil. To
perform the test the full depth of the vane is inserted into the soil. The vane is then
rotated by applying torque at the top of the rod until the soil fails (Das 2004). The
maximum torque required to fail the sample will be used in an attempt to develop
correlations to the MR. The locations of the 67 tests are listed in Table 3.2 along with the
results of the tests.

3.4.3 Falling Weight Deflectometer (FWD)

FWD tests were/are being performed as a supplement to the laboratory tests as a
second technique to calculate the MR. MSU requested and received 502 FWD files from
MDOT from FWD tests that were conducted prior to the start of this research project.
After the locations of all FWD files were established, MSU obtained thickness data
during the time of the test for every FWD file that could be found at MDOT. The total
thickness data for 103 files were obtained. Therefore, only 103 of the existing FWD files
were backealculated for this project because the cross section data is required to
accurately backealculate the MR values. To supplement the existing FWD data, test
locations were requested for this project. FWD tests were conducted at every Shelby
tube location to provide laboratory and backealculated values at the same location so the
values can be compared. The Shelby tube locations were chosen based on the results of
the disturbed soil sampling. Additional FWD tests were/are being performed along

pavements in intervals of 528 feet (0.1 miles) in order to determine the variability in the

61

roadbed soil. Initially, 208 locations were requested for FWD testing. Of those, 124
were ranked as high priority for the fall season and 84 for the spring season. However,
due to delays in starting the FWD testing the list had to be reduced to 86 high priority fall
testing locations and 58 for the spring season. Then, due to difficulties in obtaining
traffic control for the FWD tests the list was further reduced to contain locations only on
interstates. This created 52 high priority spring and fall test locations. Due to FWD
testing delays at MDOT, FWD data is not discussed here. Discussion on the FWD
testing can be found in Tyler Dawson’s thesis (Dawson 2008).
3.5 Laboratory Testing

Once the disturbed and undisturbed Shelby tube samples were received in the
laboratory, they were subjected to a battery of tests that include: the natural moisture
content, grain size test that included dry and wet sieving and hydrometer tests, Atterberg
limits (liquid and plastic limits and plasticity index), and cyclic load triaxial tests. These
tests and the test results are presented in the next subsections.
3.5.1 Moisture Content

All 81 soil samples collected underwent natural moisture content tests according
to ASTM C 29 standard test procedure. The results of the moisture content tests can be
found in Table 3.4. It should be noted that samples with. an “X” under the Shelby tube
column were taken from undisturbed Shelby tubes and those with nothing in the cell are
from disturbed samples. A detailed review of the affect of moisture content on the MR
values was conducted and. it is included in Chapter 2. The effects of moisture content on

the MR values in this study are discussed in Chapter 4.

62

3.5.2 Grain Size Distribution

The grain size distribution for soils with more than 10 percent passing sieve
number 200 was determined using sieve and hydrometer analyses. For soils with less
than 10 percent passing sieve number-200 the grain size distribution was determined by
sieve analysis only.
3.5.2.1 Sieve Analysis

All bag samples were subjected to either wet (see Figure 3.5) or dry sieve (see
Figure 3.6) analysis according to ASTM C 117 and ASTM C136 standard test
procedures, respectively. First all soils were subjected to dry sieving. When the test
results showed more than 10 percent passing sieve number 200, the soil was subjected to
wet sieve and hydrometer analyses. The objective of the test is to determine the particle
size distribution and the classification of the roadbed soil. In all analyses, the sieve sizes
and the sieve arrangement were chosen based on the MDOT Uniform Field Soil
Classification (see Appendix A). A total of 81 dry sieves and 56 wet sieve tests were run.

Results of the dry and wet sieve analyses can be found in Table 3.4.

 

Figure 3.5 Wet sieve test

63

 

Figure 3.6 Dry sieve test

3.5.2.2 Hydrometer Analysis

Soil samples which have more than 10 percent passing sieve number 200 were
subjected to hydrometer analysis according to the AASHTO T 88 standard test
procedure. A total of 56 hydrometer analyses were conducted.
3.5.3 Atterberg Limits

Soil samples with more than eight percent passing the #200 sieve were subjected
to Atterberg limit tests. The Atterberg limits consist of liquid limit, plastic limit, and
plasticity index. The liquid limit is the water content at which soils change behavior from
plastic to liquid. Whereas the plastic limit is the water content at which soils possess
plastic behavior (Liu and Evett 2008). Both the liquid and the plastic limit tests were

conducted according to the AASHTO T 89 standard test procedure. Figure 3.7 shows the

64

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3.4 Laboratory test results
- Natural Sample Percent passing sieve # Atterberg limits Cu : CC 7: Classification
Sample number Shelby water weight .3/8 4 10 20 40 100 200 D10 D30 D60 1360/ th/
tube content (0. 1nch L1. P1- P1 D10 (D60) AASHTO USCS
We) 3 9.500 4.750 2.000 0.850 0.425 0.150 0075 (D10)
M—045-S (01-01) 11.5 298.8 99.5 99.3 98.9 96.8 96.7 77.2 66.7 26 16 10 0.0030 0.006 0.040 13.33 0.30 A-6 CL
U—002—E (02—01) 16.8 303.3 99.1 97.8 96.6 92.3 68.1 46.4 39.2 18 — NP 0.008 0.040 0.300 37.50 0.67 A-4 SM
M-028—W (02—02) 21.0 200.0 100.0 99.4 98.0 93.4 83.2 64.5 56.1 23 - NP 0.0080 0.024 0.110 13.75 0.65 A—4 ML
M-028—W (02-03) 6.6 535.8 100.0 99.3 97.2 92.1 81.8 23.4 6.1 16 - NP 0.091 0.175 0.285 3.13 1.18 A-l—b SP-SM
U-002—E (02—04) 10.8 200.0 100.0 99.4 98.0 93.4 83.2 64.5 54.1 19 — NP 0.0100 0.050 0.110 11.00 2.27 A—4 ML
U—002-E (03—01) 5.0 525.3 100.0 99.8 99.6 98.5 92.6 15.8 6.5 13 — NP 0.130 0.190 0.275 2.12 1.01 A-3 SP—SM
M—028-W (03-02) 3.1 519.1 99.9 99.6 99.3 97.9 89.7 14.0 3.0 NA NA NP 0.150 0.190 0.280 1.87 0.86 A-3 SP
U—002-E (03—03) 13.1 222.9 100.0 96.8 93.7 88.7 77.8 3 i- .7 25.1 15 — NP 0.002 0.120 0.300 150.00 24.00 A-2—4 SM
M—028—W (03—03) 4.8 520.2 94.1 87.5 82.6 71.2 45.5 11.1 6.4 21 - NP 0.140 0.285 0.600 4.29 0.97 A-3 SP—SM
l-O75—N (03-04) 9.4 549.2 99.9 99.8 99.5 98.4 91.3 10.0 1.5 NA NA NP 0.160 0.200 0.280 1.75 0.89 A-3 SP
1—075—N (03-05) 21.2 197.8 100.0 99.9 94.1 92.4 80.9 60.3 48.2 55 22 33 0.001 0.002 0.150 150.00 0.03 A-7—6 SC
U—023—S (04-01) 22.0 547.2 98.8 98.8 98.5 96.4 90.3 10.3 4.3 NA NA NP 0.170 0.200 0.280 1.65 0.84 A-3 SP
M-068-W (04—02) 4.0 205.0 99.9 98.6 91.0 51.3 25.2 16.0 14.1 18 12 6 0.040 0.500 1.000 25.00 6.25 A—2-4 SC-SM
M-068-W (04—03) 33.3 515.6 100.0 100.0 99.7 98.7 89.8 14.3 3.7 NA NA NP 0.160 0.190 0.280 1.75 0.81 A-3 SP
M—065—S (04-04) 8.1 201.5 99.3 95.4 91.3 87.5 72.7 30.4 21.5 30 — NP 0001 0.150 0.300 300.00 75.00 A-2—4 SM
M—032—W (04-05) 9.6 203.4 100.0 99.8 99.6 99.0 95.0 64.6 48.7 19 12 7 0.001 0.006 0.130 130.00 0.28 A-4 SC-SM
U—131—N (05-01) 13.1 199.4 99.8 99.2 96.4 95.0 78.7 43.5 29.2 14 - NP 0.016 0.140 0.280 17.50 4.38 A-2-4 SM
U-127—N (05-04) 8.9 527.6 91.8 84.4 79.1 73.3 53.6 6.4 3.7 NA NA NP 0.180 0.260 0.500 2.78 0.75 A—3 SP
M—033—S (05-05) 3.5 525.7 63.1 57.5 45.4 35.7 26.7 7.8 4.6 NA NA NP 0.185 0.510 6.000 32.43 0.23 A-l-a SG
M-072—W (05—06) 14.3 201.0 100.0 99.6 98.8 97.3 91.4 56.1 39.9 22 11 11 0.0070 0.035 0.160 22.86 1.09 A—6 SC
M-132-N (06-01) 15.0 521.7 99.5 99.0 98.5 96.8 78.7 8.8 4.2 NA NA NP 0.160 0.220 0.320 2.00 0.95 A-3 SP
1-075-N (06-02) 3.4 518.0 95.1 93.7 92.8 90.4 63.4 5.8 4.1 NA NA NP 0.170 0.260 0.400 2.35 0.99 A—3 SP
U—031—N(06—03) 5.8 1060.3 99.5 99.1 98.4 97.4 87.2 7.9 0.5 NA NA NP 0.170 0.210 0.300 1.76 0.86 A—3 SP
I—196-N (06—05) 10.5 1085.6 99.6 98.4 96.2 91.2 84.4 26 5 5.9 15 — NP 0.089 0.160 0.275 3.09 1.05 A—2—4 SP-SM
M—020-W (07—02) 4.2 1003.7 99.6 99.3 98.7 97.9 88.0 2; 0.8 NA NA NP 0.180 0.220 0.300 1.67 0.90 A-3 SP
M-020-E (07-03) 4.5 513.3 99.2 96.8 94.5 89.6 21._2_ 3.3 NA NA NP 0.110 0.190 0.280 2.55 1.17 A-3 SP
U-127-N (07-04) 10.9 200.8 100.0 98.8 96.6 95.4 90.3 ii 26.9 22 12 10 0.001 0.100 0.230 230.00 43.48 A—2—6 SC
U—127—N (07—05) X 11.2 203.9 W" 92.6 87.3 79.9 53.7 40.5 23 14 9 0.0011 0.006 0.190 172.73 0.17 A-6 SC
J-127-N (07—05) 14.4 213.7 99.8 98.2 85.2 81.0 74.8 L.5_2_'1__ 43.7 24 14 10 0.0010 0.008 0.210 210.00 0.30 A-6 SC

 

 

65

 

 

 

 

-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3.4 (cont’d)

‘ Natural Sample Percent passing sieve 14 Atterberg limits Cu : Cc : Classification
Sample number Shelb water weight .3/8 4 10 20 40 100 200 D16 D46 D60 1360/ D3“

ytube content (g3 1nch 11. PL P1 “ D10 (D60) AASHTO USCS

(“/01 9.500 4.750 2.000 0.850 0.425 0.150 0.075 (D10)
M-061—E(07—06) 22.1 198.5 100.0 98.8 93.3 84.7 59.3 23.7 17.9 19 — NP 0.040 0.190 0.430 10.75 2.10 A-2—4 SM
M—061-E(08-02) 20.3 223.1 100.0 99.7 93.9 77.8 51.9 26.1 23.2 11 — NP 0050 1.000 0.520 10.40 38.46 A-2-4 SM
U—010-W(08-03) 21.4 200.2 100.0 100.0 99.8 99.7 97.6 61.0 55.2 32 14 18 0.001 0.002 0.140 140.00 0.02 A-6 CL
U-OlO—W(08—04) 8.2 200.1 99.9 99.9 98.8 96.6 84.5 48.8 36.7 29 13 16 0.001 0.011 0.200 200.00 0.61 A—6 SC
U—010—W(08-04) x 15.0 205.1 98.0 98.9 96.5 95.8 80.3 42.5 33.3 27 13 14 0.0009 0.018 0.200 222.22 1.80 A—6 SC
1-075—S(08-05) 8.9 201.0 100.0 99.9 97.7 94.5 69.4 40.3 33.5 25 12 13 0.001 0.011 0.300 300.00 0.40 A—2-6 SC
l—075—N(08-06) 11.8 201.5 100.0 99.2 96.8 93.7 85.4 36.6 26.2 17 10 7 0.001 0.011 0.270 270.00 0.45 A-2—4 SC—SM
U—131-S(09—OI) 4.6 1056.3 99.0 98.0 97.4 97.0 83.7 2.5 0.5 NA NA NP 0.180 0.220 0.300 1.67 0.90 A—3 SP
1—096—W(09—02) 9.9 206.2 100.0 99.0 97.3 93.8 82.7 40.9 30.5 17 13 4 0.001 0.075 0.240 240.00 23.44 A-2-4 SC-SM
U-131—S(09-03) 1.9 530.4 100.0 100.0 99.9 99.8 97.2 6.0 0.4 NA NA NP 0.180 0.200 0.290 1.61 0.77 A-3 SP
U—131-S(09-05) 3.6 1025.6 97.5 90.2 80.8 69.5 45.8 3.1 1.3 NA NA NP 0.185 0.295 0.605 3.27 0.78 A-3 SP
M—044—E(09—07) 8.7 206.5 100.0 99.5 97.7 94.1 85.5 37.7 26.7 14 — NP 0.020 0.110 0.250 12.50 2.42 A—2—4 SM
1—075-S(09-08) 20.2 216.1 99.1 96.1 91.8 89.7 85.5 62.3 45.8 31 14 17 0.001 0.004 0.140 140.00 0.11 A4 SC
M—024-S(09—09) 13.3 198.6 100.0 99.6 97.6 95.4 93.2 45.0 24.1 20 — NP 0.012 0.090 0.200 16.67 3.38 A-2-4 SM
1—069—E(09-10) 7.1 527.8 98.3 93.4 83.0 66.3 36.8 5.2 3.1 NA NA NP 0.190 0.340 0.700 3.68 0.87 A-3 SP
1—069-N(10-01) 10.1 534.1 94.9 88.7 81.1 67.6 49.2 16.7 8.0 16 11 5 0.093 0.230 0.600 6.45 0.95 A-3 SP-SM
1-096-W(10-03) 14.7 199.74 100.0 98.4 93.9 90.1 82.0 29.5 17.5 29 14 15 0.0600 0.150 0.280 4.67 1.34 A—2-6 SC
1-069-N(10—04) 11.1 198.5 100.0 99.3 94.1 86.4 74.9 30.1 17.6 16 — NP 0.010 0.150 0.200 20.00 11.25 A—2-4 SM
1-069-N(10—05) 24.0 204.0 100.0 100.0 97.8 87.6 54.9 43.2 37.3 19 — NP 0.010 0.070 0.500 50.00 0.98 A-2-4 SM
1—096—W(10-09) 15.1 200.9 100.0 99.6 93.7 91.0 61.1 38.0 30.4 19 — NP 0.006 0.075 0.410 68.33 2.29 A-2-4 SM
1—069-E(10-10) 12.8 204.9 98.0 96.1 92.4 90.5 84.7 57.2 37.7 26 15 11 0.001 0.009 0.170 170.00 0.48 A6 SC
M—021-E(10—11) 15.0 230.2 99.4 92.1 85.9 79.5 72.2 46.3 33.8 23 14 9 0.001 0.030 0.270 270.00 3.33 A—2—4 SC
1-069-N(11-01) 9.1 1032.9 90.3 87.1 83.0 77.8 63.9 15.9 6.9 14 — NP 0.120 0.210 0.390 3.25 0.94 A—3 SP—SM
1-094—W(11-02) 7.1 1022.7 95.0 91.7 87.1 77.5 51.2 6.2 2.7 NA NA NP 0.170 0.270 0.510 3.00 0.84 A-3 SP
M-060-W(11-03) 10.5 199.3 99.7 99.0 97.4 90.6 67.0 37.6 31.1 22 15 7 0.004 0.025 0.330 82.50 0.47 A-2-4 SC—SM
I-069-S(11-05) 6.6 201.1 100.0 99.1 93.9 86.9 77.3 49.3 38.6 15 11 4 0.002 0.034 0.210 105.00 2.75 A-4 SC-SMJ
1-094-W(12—01) 8.6 199.8 100.0 95.2 81.8 73.9 51.8 26.5 20.0 16 12 4 0.038 0.180 0.560 14.74 1.52 A—2—4 JSC-SM
I—094-W(12~03) 13.2 527.4 97.4 95.4 91.6 83.0 68.3 18.7 7.4 16 — NP 0.095 0.195 0.345] 3.63 1.16 A—3 [SP—SM
U-012-E(12-04) 4.9 200.4 99.9 98.9 94.2 89.4 73.7 36.6 23.0 16 — NP 0.003 0.110 0.300110000113441 A-2—4 ] SMJ
66

 

 

Table 3.4 (cont‘d)

 

Atterberg

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Natural Sampl Percent passing sieve # limits C : CC ,: Classification
Sample number Shellby ‘water e 1 3/8 4 7 D10 D30 D60 D160/ D307
t” C “8223’“ W25 “ inch 10 ‘0 40 100 200 LL PL P1 D10 gig; AASHTO USCS
9.500 4.750 2.000 0.850 0.425 0.150 0.075
1-094-W (12—06) 12.1 213.7 100.0 99.8 92.2 90.5 86.0 35.2 23.8 15 - NP 0005 0.130 0.250 50.00 13.52 A—2-4 SM
J—012—E(12-07) 7.0 513.8 67.5 57.0 42.2 25.8 16.0 10.0 8.1 18 — NP 0160 1.000 6.000 37.50 1.04 A-l-a SG
M—024-S (13—01) 10.6 196.0 100.2 98.4 93.4 90.2 85.2 59.4 45.1 18 15 3 0.001 0.013 0.150 150.00 1.13 A-4 SM
M-059—W (13—02) 11.6 1033.3 99.4 97.9 95.1 91.2 65.7 8.9 1.7 NA NA NP 0.160 0.220 0.380 2.38 0.80 A—3 SP
M—014—W (13—03) 9.3 198.1 100.0 99.1 94.0 90.0 85.7 62.7 49.2 22 13 9 0.001 0.006 0.130 130.00 0.28 A—4 SC
1—094—W (13—04) 8.0 1005.6 98.1 95.8 90.5 82.8 65.9 13.1 3.5 NA NA NP 0.140 0.210 0.390 2.79 0.81 A—3 SP
U—012-E (13—05) 14.9 205.0 100.0 99.9 99.0 97.8 95.5 65.6 56.7 33 17 16 0.001 0.002 0.100 111.11 0.04 A—6 CL
U-023—N (13-07) 9.8 529.5 94.1 83.4 66.2 53.5 43.3 12.0 5.7 13 — NP 0.130 0.280 1.350 10.38 0.45 A3 SP-SM
M-010—E (13—08) 14.0 201.0 100.0 99.7 98.1 95.0 90.8 74.3 59.9 24 14 10 0.0010 0.003 0.075 75.00 0.12 A—6 CL
M—010—E (13—08) X 12.3 207.0 100.0 98.0 95.6 93.5 88.3 72.6 54.8 23 14 9 0.0009 0.015 0.090 100.00 2.78 A—6 CL
1—075—S (14—01) X 18.4 204.5 100.0 99.9 89.4 87.9 67.6 54.2 48.2 42 21 21 0.0090 0.015 0.250 27.78 0.10 A-7-6 SC
1—075—S (14-01) 25.4 200.6 100.0 96.9 78.9 76.2 68.4 47.8 41.2 45 19 26 0.0007 0.003 0.270 385.71 0.05 A—7-6 SC
I—O75—S (14—02) 18.7 201.0 100.0 98.3 97.6 92.6 85.5 64.1 46.1 41 19 22 0.001 0.003 0.190 211.11 0.06 A-7-6 SC
U—024-S (14—03) 19.2 202.3 100.0 99.4 98.8 91.8 79.7 55.3 41.4 40 13 27 0.001 0.003 0.190 271.43 0.07 A—6 SC
1—075—S (14-04) 15.8 200.8 100.0 99.9 99.8 99.7 96.4 59.4 46.9 34 17 17 0.001 0.003 0.260 288.89 0.04 A-6 SC
U—024-S (14-04) 22.2 543.7 100.0 100.0 99.8 99.6 96.3 23.3 2.5 NA NA NP 0.100 0.170 0.255 2.55 1.13 A-3 SP
1—094-W(14-05) 21.6 199.0 99.7 97.6 97.5 89.7 78.0 56.7 46.7 34 21 13 0.001 0.013 0.160 160.00 1.06 A—6 SC
M—153—E (14—06) X 26.0 209.4 100.0 99.8 99.0 98.3 92.7 70.1 51.1 51 19 32 0.0090 0.018 0.100 11.11 0.36 A-7—6 SC
M—153—E (14—06) 21.6 202.9 100.0 100.0 98.4 98.1 94.1 64.4 49.9 52 20 32 0.0007 0.001 0.140 200.00 0.02 A-7—6 SC
M-053—S (14—07) 5.9 529.1 93.1 87.5 81.5 70.3 55.0 9.3 4.7 NA NA NP 0.170 0.240 0.500 2.94 0.68 A—3 SP
1—094-W (14-09) X 26.3 205.1 100.0 100.0 98.5 97.9 85.2 59.8 55.8 42 23 19 0.0010 0.010 0.150 150.00 0.67 A-7-6 CL
I—094—W (14—09) 21.9 197.3 99.7 99.2 97.7 96.6 90.8 66.8 60.9 44 21 23 0.0010 0.002 0.075 75.00 0.05 A—7—6 CL
1—094-W (14—10) 21.5 198.9 100.0 99.5 93.3 91.6 80.3 65.2 56.3 42 19 23 0.001 0.002 0.100 166.67 0.07 A—7—6 CL
M—053—S (15-02) 17.2 200.4 100.0 99.5 96.8 94.4 87.5 42.8 26.2 14 — NP 0.008 0.100 0.210 26.25 5.95 A-2—4 SM
M—090—E (15—03) 38.0 204.1 100.0 99.9 98.8 96.1 90.7 73.1 55.8 35 20 15 0.001 0.005 0.088 88.00 0.28 A—6 CL
M—090—E (15-04) 12.4 199.6 100.0 99.7 97.4 95.0 90.6 67.4 52.8 24 15 9 0.0010 0.006 0.100 100.00 0.36 A4 CL
M—025—S (15—05) 4.4 532.8 99.3 98.7 98.2 97.3 84.4 1.9 1.1 NA NA NP 0.180 0.210 0.300 1.67 0.82 A—3 SP
M—25—N (15—06) 16.4 206.4 100.0 98.9 94.0 90.8 85.1 54.2 42.3 24 13 11 0.001 0.007 0.190 190.00 0.26 A—4 SC
M—019—S (15—07) 11.4 199.4 99.9 95.1 83.9 76.4 61.5 29.0 17.2 14 — NP 0.065 0.160 0.400 6.15 0.98 A-2—4 SM

 

67

 

 

 

 

devices for both tests. After obtaining the liquid and plastic limits, the plasticity index
was calculated as the difference between the two limits. A total of 60 Atterberg limit
tests were performed. Results of the Atterberg limit tests are listed in Table 3.4.
3.5.4 Cyclic Load Triaxial Test

Cyclic load triaxial tests were conducted. to determine the resilient modulus of
laboratory compacted sand and clay samples and Shelby tube samples. The sample
preparation procedures are presented below.
Preparation of Sand Samples - All sand samples were compacted in 2.125 inches
diameter and 4.8 inches high split mold using a 10 pounds static load and a vibrating
table. The split mold has two outlets connected on the outside of the mold by small
diameter drainage tubes and protected on the inside by two small porous stones tightly '
fitted to the holes. First, the split mold was assembled and a rubber membrane was
stretched along the interior walls of the mold. The membrane was then flipped over the
edges of the split mold and a vacuum was applied between the rubber membrane and. the
interior walls of the mold. The vacuum forced the membrane to stick to the wall and
hence eliminates wrinkling. The split mold and the rubber membrane were then placed
around the bottom pedestal on the base of the triaxial cell and a paper filter was then
placed on top of the pedestal. The entire assembly was then placed on the vibrating table.
The total weight of the sand was then measured and recorded and the sand was then
compacted in five lifts. Each lifi was subjected to a 10 pound static load while being
vibrated at the maximum amplitude for a period of three minutes. Figure 3.8 shows the
vibrating table, static load, vacuum pump, split mold, and triaxial cell base. When the

last sand lift was compacted, a paper filter and the top pedestal were respectively placed

68

 

Figure 3.7 Liquid and plastic limit apparatus

in orderly fashion on top of the sand. The rubber membrane was then secured around the
top and bottom pedestals using rubber bands. The vacuum lines were removed from the
split mold and connected to the drainage lines at the bottom of the triaxial cell to apply
vacuum to the sand. The split mold was then removed. After assembling the triaxial cell,
a confining pressure of 7.5 psi was then applied while the vacuum was decreased to zero
by disconnecting the vacuum line. The weight of the lefi over sand was measured and
recorded. The difference between the initial and final weights represents the weight of
the sand sample. Afier determining the water content in the sample, its dry unit weight
was calculated.

Preparation of Clay Samples - Disturbed clay samples were compacted according to the
AASHTO standard proctor test procedure T99. Afier compaction, the samples were then
trimmed to the desired length of 4.5 inches and diameter of 2.25 inches. The samples

were then placed in a rubber membrane (which was stretched tightly around the interior

69

 

Figure 3.8 Vibrating table setup

of a split mold). The sample was then sealed from the atmosphere by securing the rubber
membrane to the bottom and top pedestals. The entire mold assembly was then
transferred to the triaxial cell where the cell was assembled and a confining pressure of
7.5 psi was applied to the sample.
Preparation of the Undisturbed Shelby Tube Samples — First, the Shelby tubes were
cut to several segments 6 inches long. The clay soil was then extracted from the tube
segment and was trimmed to a height of 5.6 inches. The diameter of the soil was kept the
same as the interior diameter of the Shelby tube (2.8 inches). A rubber membrane was
placed around the soil (see clay samples above) and the sample was placed in the triaxial
cell and was subjected to 7.5 psi confining pressure.

Afier placing the sample in the triaxial cell and subjecting it to 7.5 psi confining

pressure, the cyclic load test commenced. The detailed steps are presented below.

70

Cyclic Load Test Procedure — All cyclic load triaxial tests were mainly conducted
according to the AASHTO T307 standard test procedure. Because of the type of tests
and equipment available, the following three modifications to the AASHTO standard test
procedure were made:

0 The load cell of the available MTS system is located below the sample instead of
above it as stated in the standard test procedure.

0 The loading and unloading time of 0.1 seconds and the relaxation time of 0.9 second
were changed to 0.5 and 0.9 seconds, respectively to more accurately simulate the
load pulse experienced by a roadbed soil located about 30 inches below the pavement
surface (see Figure 3.9). The figure Shows the distances from the load that the stress
is felt in the pavement based on the shown thicknesses.

0 The sample was conditioned by applying 498 load cycles instead of the conditioning

sequences outlines in the standard test procedure.

 

 

 

 

 

 

 

P1 P1 P1

 

 

 

Dl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

D3 J
‘
I A
AC = 5”
I 17
I A
|// Base = 12”
l v
I
I Subbase = 12”
l

Roadbed soil

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.9 Stress influence with depth

71

Cyclic load triaxial tests are difficult to conduct and require extreme care and

patience. The resulting MR values obtained from the test are typically affected by

several test and sample variables including: confining pressure, deviatoric stress, loading

frequency, soil type, moisture content, and specimen conditioning. In this study, all

cyclic load triaxial tests were performed using:

A sustained load of 10 pounds was applied to maintain contact between the MTS
actuator and the piston of the triaxial cell.

A confining pressure of 7.5 psi.

A frequency of 0.7 1 hertz.

Cyclic axial stress of 10 psi followed immediately by cyclic axial stress of 15 psi.
For each axial stress level, the samples were conditioned to 498 load cycles.

The axial sample deformations were measured using two linear variable differential
transducers (LVDTS) located at 180 degrees from the longitudinal axis of the sample.
The resilient moduli of the samples were calculated as the average resilient modulus
obtained at cycles 499, 500, 501, 799, 800, 801, 999, 1000 and 1001. The resilient
modulus at each cycle was calculated using the average deformation of the two
LVDTS.

The test setup used for the cyclic load tests is shown in Figure 3.10. Figure 3.11

shows two Hysteresis loops from load cycles 800 and 1000. The figure shows that as the

load increases the sample deformation increases and vise versa. The shift in the two

loops represents the cumulative plastic deformation that took place in the sample between

cycles 800 and 1000.

72

Results of the cyclic load tests for both the 10 and the 15 psi cyclic stresses are
listed in Table 3.5. The highlighted rows in the table indicate tests which were run to
verify the developed models. The verification is discussed in Chapter 4. The table also
includes the moisture content and degree of saturation, the dry unit weight of the test
sample, the USCS and the AASHTO soil classifications, and the sample type (disturbed
or undisturbed). It should be noted that the data in the table are sorted based on the
USCS. As can be seen from the table, 81 samples were subjected to cyclic load tests.
These include 25 SP tests, 16 SM tests, 9 CL tests, 14 SC tests, 2 ML tests, 7 SC—SM

tests, and 8 SP—SM tests.

m- m
on.”
a

(1 'V‘
66660.:

1
. 1 z
t :1,
. .
V 1
.1 |
1
4 .
1..
7% ‘
1
:5- . 1
l
, 1

 

Figure 3.10 Cyclic load test setup

73

 

-0.063 -0.0635 -0.064 -0.0645 —0.065 -0.0655 -0.066 —0.0665 -0.067

Cyclic stress (psi)

._i____l__.-.fil___l..___i

 

Deformation (in)

 

 

Figure 3.1 1 Typical cyclic load test results

74

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

was 36.2 98 am 3.: mm 3. x El: mane-2
05mm 5.: ma 3. 3: am 3. x A 8-68 2-3-2
20.3 68% a: 3 . 3: mm 34 x 8708 mag: .
Nam-E 33m 4.8 mm 62: mm m-< x Ame-8V 3-02-2
48.x wow-8 £4 3: 32 mm m-< x 2.3.: 38-:
max: 9%.: 3m 3 3: mm m-< x Ave-m: 3463
6312 84.3. 62 Ne 68 H mm m-< on so; : B463
$38 2 13 6.3 3 E : am .3. x Ave-me 2.33
56.3 33m 02 0.6 3: am 34 x Ave-me 2.8 7:
35.2 9.38 3. ed :2 mm m-< x Ame-m: 3.68.:
86% w :48 6% n: 3: mm 3. x Ame-8V 3.80.:
$0.2 666.6 93 0.8 6.2: mm 3. x A848 348.:
$68 was 2: me am: ET; m-< on El: z-mmo-a
66.2 23: was a: e. 5 zm-mm m-< x Ame-N: 3-38-H
81% 53m «.3 3 3w: 2mg; m-< x E: a 2-62:
$6.2 wmmf NS 6.6 E: 2mg; 3. x 20.9: 2-62:
36.: 68.8 62 Z 2 : 22m .33. x A888 2-82
28.2 3&2 ed 2 on: zm-mm ,3. x Ame-8v 348.:
None SEN SN 3 ES zm-am 3.. x :38 ”+89:
9%.: 3:2 wt. 3 am: 2mg; fa. x Ame-Ne 388.2
0 3
o. m fl ca: 33 0:23 flab: womb Ohmm<< moccasin Swohm
ammo mmobm coca-83 com “nausea Emmi» 898:: 295m
0198 E .52 “855 “E: .05 mouse—mamas 08$ 295m

 

 

 

 

 

 

 

 

 

 

 

 

£38 58 ME b88093 m.m 2an

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

68.2 23: 6.8 we 32 5 4-3. x A888 mic-2
32m 83m mam 3 3: SE 4.3.. x Ame-we mace-2
85.2 34.: woe q: cam 2w 4.3. x Go-SV mace-2
‘ «8.3- 338 6.8 3 3: 2m 2-... , x :38 2-5-:
mam: NS; _ new as 3.6 2m «.3. x 9.3.8 mace-2
3.2 $6.2 2:. E 2 : 2w 4.3. x Ame-me m-Nooé
2%: N32 at. no 02: gm 3. x :o-Ne wage-D
$3 a _ m...“ 3 No 6.2: mm m-< x Ame-SC 358-2
SSS 86.2 6o 2 2.2 mm 3.. x Ame-8v 388-2
858 09.. _ m 3 do 32 mm 3. x A848 3.08.:
80.2 9.2; E 2 £2 am 3.. x Ame-me m; m 7:
84.8 63% 3 2 3.2 mm 3. x Ame-8v sews-2
23m 23: 68 3 3: am 3. x Ame-68 m-E-D
3 3m 54% S: i 2: mm 3.. x A868 2-3.3
85m 8?: 6.2 2 3: mm 3.. x 20.68 mac-D
3.6m 2 2:. 4.2 3 2: mm a... x Ame-m: mas-z
83.8. 89% 6.: mm 3: am 3.. x Ame-8v memo-E
68.8 Ea: 4.: m8. 2 z _ mm 3. x A888 238-:
E. a 68.3 EN 2 3: mm 3. x 23.8 mama-D
o
o. 2 QB 58 0:93 AEBC m UmD 0.5-Ea}. 32535 .EWMm
£va wmobm coma-53m How “Gouda-co Emmog sauna vac-5m
0:93 3 ME 883 :5. CO 838E320 25 293m

 

 

95:03 Wm 033-

76

 

 

 

 

 

 

 

3.8 20.8 mi 3 £2 Emém #3. x :83 333
30.2 02.8 0.8 to 32 Emém 3.. x Go; 8 Waco;
$82 3.2 3m .8 32 85.8 v-3. x E: 8 3.80.:
Ewe «3.8 3 2 32 Emom I-.. x @898 3-8%-.
«8.2 85.2 E No 22 35.8 «.3. x 638 2-3:
33: 2S: «.2 5 32 35.8 I... x $3.8 3&3:
8.3m mgdm . 3.. x $1.8 388-2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

RS: :5 3-... x ASA: 33-2
02.8 $6.: 3:. Wm md: 2m v5... N Co-m$ mémoéz
3de 03.3 5.3. mg: m m2 2m v-m-< X Go-NC 34.34
Rmdfi «2.3 WE Wm 3:: 2m Wmé. X AXES m-NS-D
God oomfi qma 5mm «.03 Em v-m-< N 890$ 2-02:
3%? $13 04% 1w «#2 2m v-~-< N 9563 2&8;
3&2 «3.2 m. :. wd ode 2m v-m-< N Amodov m-vmo-2
33
. . 0 “Ba
o m H o 3 38 0:93 AMMBD mUmD OHEmE w a .Q xflonm a
38 $05 grandam 5m 3380 E “M3 9 5985: 2 Sum
0298 3 m2 53>? ES .5 cosmomwmmfio «.93 2 Sam

 

 

 

8.388 nm 2%;

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

£3 2? «.3 a: 2: 0m c-< x $38 2-27:
33 23 a? m. : nos 8 o-< x 338 2-5-:
SEN 818 NE 2: S: Um o-< x 638 3&5:
89$ wté 3m 3: 32 go v-< x 9.0.3 3%.:
08¢ mad 0.3 08 v.8 5 3% x 83: >303
325 2&8 03 m. : 32 d 0-3V x lace: >303
33K 3%? a? 2: SS 5 3% x 83: 2153
and an: 0.3 as 3.2 ,8 3‘ x 30.3 38.2
Sq? 33% n3 2 32 go o-< x $3: $5-2
N32 02.: a? E: 32 go ©-< x $3 s 38.2
2% 33 mg oi 3w: 18 o-< x 93.3 35.2
8%: gem mac 2: M32 go o-< x :38 @292

0.2 0.2 58 0228 Cars: m UmD Ohmm<< @3565 zmwmm
A58 £05 53838 How 68:00 £303 528:: 038mm
0:98 3 52 63>? ES NCO coumomumflo 25 038mm

 

 

 

 

 

 

 

 

9:53 2 05¢

78

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

52m 23m 3% 2: 9m: 1:2 w% x $38 @805
8n 3 Exam 9% q: 32 a: v% x $28 3-392
$32“ «3,? 2% 2: w. :: Um 3% x 63: mi 22
:3 cm? 0.8 tom 98 cm 5% x 63: mi 72
m 8a SE 3m 0.3 0.8 Um 3% x 63: mam 72

4 Eu: 52.2 was 3: ‘ ma Um 5% x, w :3: WE:

83 $3 N8 0.8 32 mm 5% x :3: £22
£3 53 25 X: 32 mm 3% x :3: mic;
3de $3M Em A‘ 3 . 2: um 5% 1 x :3: was;
$0.? $5.? 1% o. : 32 8 3% x 5:5 3-83
3% 3.? w? 5.8 3: Um 0% x :5-on 3-05-:
81m 2? mi 0.2 2 : 0m 0% x $38 3-05-:
SEN Scam 0% 2: E. : 0m , 0% x ‘ 338 2.487:
$3 $3 NS 0.2 mm: Um 0% x 518 2-5-:
89mm 23% 3% E SE Um 0% x 938 2-5-:

0 s
0.3 QB 58 0:98 Cm}: mUmD 0:5: 383me Ewohm
A53 mmobm comuwbdmm how EQEOQ “5&5? 59925 038mm
0198 3 m2 H885 ES 50 cosmoEmmEU 33 2mem

 

 

520$ 2 23

79

CHAPTER 4

DATA ANALYSIS AND DISCUSSION

4.1 Introduction

Recall that (see Chapter 1) the objectives of this study are:

0 Evaluate the existing processes used by all regions of MDOT for determining the MR
value of the roadbed soil for flexible pavement design and the modulus of subgrade
reaction (k) for rigid pavement design.

0 Determine the needed modifications to make the process compatible with the M-E
PDG.

To accomplish these objectives, a research plan was carefully designed and
executed. The plan consists of the following approaches:

1. Information Gathering - Various conversations were held with the Soil Engineers
from each MDOT Region. During the conversation, the method used by the region to
estimate the resilient modulus of the roadbed soil was documented. The results are
listed in Table 2.1 of Chapter 2.

2. Soil Delineation —- The State of Michigan was divided into 15 clusters and 99 areas
where the roadbed soils are similar. The 99 areas were later reduced to 75 areas for
soil sampling.

3. Soil Sampling — From each of the 75 areas, disturbed bag roadbed soils samples were
obtained. In addition, 10 Shelby tube samples were collected where the disturbed soil
samples showed that there was a substantial amount of clay.

4. Field Tests — The field tests consist of:

o Penetration resistance using pocket size penetrometer

80

o Shear strength using pocket vane shear tester

o Deflection using falling weight deflectometer (FWD)
The field test results are presented in Chapter 3.

5. Laboratory Tests — The laboratory tests consist of:

0 Moisture content

0 Dry and wet sieve, hydrometer, and Atterberg limits (liquid and plastic limits and
plasticity index) to analyze the grain size distribution and to classify the soils
according to the USCS and AASHTO soil classification systems

0 Cyclic load triaxial test
The laboratory test results are presented in Chapter 3 and in Appendix B.

6. Data Analysis — The data analysis and discussion are presented in this chapter in the

following order:

a) Existing MDOT practices regarding the estimation of the resilient modulus of
roadbed soils.

b) Field data analyses and discussion (including a sample backealculated roadbed
resilient modulus).

c) Laboratory data analyses to classify the soil according to USCS and AASHTO
soil classification systems.

d) Laboratory data analyses and discussion regarding the effects of test and sample
variables on the MR values.

e) The analyses used to develop correlation equations between the resilient modulus
of the roadbed soils and simple test results that would satisfy the M-E PDG

requirements for design levels 2 and 3.

81

t) The seasonal effects on the resilient modulus of the roadbed soils and the

determination of the effective design resilient modulus using the damage model

found in the 1993 AASHTO Design Guide.

It should be noted that the full analysis and discussion of the FWD deflection data

and the correlations between the backcalculated and the laboratory obtained resilient

modulus values are presented elsewhere (Dawson 2008).

4.2 MDOT Practice

Existing MDOT practices regarding the estimation of the design resilient modulus
of roadbed soils were discussed during interviews, telephone conversations, and e—mail

communications with Soil Engineers from the seven MDOT Regions. The findings are

listed in Table 2.1, which is repeated herein for convenience.

Table 2.1 MDOT procedures for determining resilient modulus

 

 

 

 

 

 

 

 

 

 

 

 

Region Procedure {IEEZI (1;): i)
Bay Soil boring & visual identification 3600

Grand FWD data (if available) or soil boring & 2700 _ 8600
Visual Identlficatlon

Metro Soil boring & visual identification 3000 - 4500

North FWD data (if available) or soil boring & 2500 _ 6000
Vlsual 1dent1ficatlon

Southwest California Bearing Ratio correlations
Superior Soil boring & visual identification 4500 — 7000
University Soil boring & visual identification 3000 - 4000

 

82

 

Examination of the information provided in Table 2.1 indicates that the existing
practice for estimating the resilient modulus of the roadbed soil is mainly based on soil
boring, visual identification of the soil type and estimating the resilient modulus of the
roadbed soils using Figure 2.2. One region uses correlations between the California
Bearing Ratio (CBR) and the resilient modulus of the roadbed soils. On the other hand,
two regions use the FWD deflection data, when available, to estimate the resilient
modulus of the roadbed soils. However, the exact procedure used by either region to
estimate the resilient modulus of the roadbed soils from the deflection data is not clear.
The above scenario implies that, the practice in most regions may satisfy the requirement
of design level 3 of the M-E PDG. The practice of the Southwest Region may satisfy
level 2. Finally, the use of FWD data (when available) in Grand and North Regions may
satisfy the pavement rehabilitation requirements in design level 1 of M-E PDG.
4.3 Field Data Analyses and Discussion

Recall that the following field tests were performed during this study:
0 Falling weight deflectometer (FWD)
0 Pocket penetrometer
0 Pocket vane shear tester
The data analyses and discussion for each one of these tests are presented below.
Falling Weight Deflectometer — Unfortunately, due to unforeseen constraints during this
study, FWD testing was delayed by MDOT for a period of about one year. During the
first year of the study, only 19 FWD tests were conducted. The deflection data from
these tests was analyzed and the resilient moduli of the roadbed soils were

backealculated. Due to a limited number of tests, only a sample of the backealculation is

83

presented in this section. Analyses and discussion of the entire FWD tests and the
correlations between the backealculated and the laboratory obtained resilient modulus
values are included in (Dawson 2008).

In the backealculation, the area method presented in Chapter 2 was used to
backealculate the modulus of subgrade reaction (k) of the roadbed soils under rigid
pavements. Whereas the resilient moduli of the roadbed soils under flexible pavements
were backealculated using the MICHBACK computer program (Harichandran 1994).
Nevertheless; sample results of the backealculated k and MR values for one rigid and. one
flexible pavement sections are included in Table 4.1. As can be seen, both the k and the
MR values presented in the table are reasonable. Unfortunately, no further analyses can
be conducted at this time because of the delays in the FWD testing.

Table 4.1 Sample backealculated k and MR values

 

 

 

 

 

 

 

 

 

 

 

 

Pavement Load Deflection (mils) at distances behind the load (in)
type (lbs) 0 8 12 18 24 36 60
Rigid 8957 9.06 7.7 6.63 6.07 5.01 3.71 1.93
Flexible 9729 12.38 9.66 7.70 5.49 4.07 2.48 1.30
Cross section data (in) Backcalculated Values
Surface Base Subbase Stggg er k value (pci) MR (psi)
Rigid 9 4 8 NA 246 NA
Flexible 3 8 '18 700 NA 27,898

 

 

 

 

 

 

 

Pocket Penetrometer and Pocket Vane Shear Tester - The locations where pocket
vane shear and pocket penetrometer tests were conducted are listed in Table 3.2 of
Chapter 3. The pocket penetrometer and vane shear tester results were initially compared
to each other. It was found that the pocket penetrometer and vane shear tester correlate
with each other. Figure 4.1 shows the pocket penetrometer plotted against the vane shear

resistances, the trendline, and the correlation equation relating the two sets of data. As

84

can be seen, the pocket penetrometer resistance (PPR) is related to the vane shear

resistance (VSR) by Equation 4.1 with R2 = 0.59.

0.4685

PPR = 0.9888 ln(VSR) Equation 4.1

This correlation was expected because both tests measure indirectly (PPR) or directly

(VSR) the shear strength of the soils at the same locations.

Although the results of the two tests correlate to each other, as it was expected,

neither sets of data were found to correlate to the MR of the soils. This is because the

MR values are obtained under low strains (primarily elastic). Whereas, the pocket

penetrometer and vane shear tests were conducted at the critical state of stresses and

strains (elastic, viscoelastic, and plastic).

 

Pocket penetrometer (psi)

 

2.5 4

1.5

0.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l 1
0.4685 ' ‘
MR = 0.9888(VSR) I l
R2 = 0.5867 ; ° 1
* l
L v £ //
l ‘k 0
o XV/l/
4» ° . .
A o l *
° 13 I c
. 1
o ‘ , .
<. . l . l
0 1 2 3 4 5 6

Vane shear (psi)

 

Figure 4.1 Pocket penetrometer versus vane shear tester

85

 

4.4 Soil Classification

For each disturbed soil sample the dry and wet sieve test data, the hydrometer test
data, and the Atterberg Limits data were analyzed to determine:

0 The grain size distribution curve and the coefficients of uniformity and curvature.
o The AASHTO and the USCS soil classifications.

The results of the analyses are listed in Table 3.4. Table 4.2, provides a list of the
number of disturbed soil samples for each soil classification type according to the USCS
and the AASHTO soil classification systems. As can be seen, the majority of the roadbed
soils in the State of Michigan can be divided, in general, into 8 soil classification types
for both the USCS and the AASHTO soil classification systems. It is very important to
note that the soil in each classification has a wide range of grain size distribution
parameters. To illustrate, consider the grain size distribution curves for SC, SM, ML,

CL, and SP-SM soils shown in Figure 4.2. It can be seen that the six soils have different

grain size distributions and different coefficients of curvature (CC) and uniformity (Cu).
The Cc values for the six soils ranged from 0.02 to 1.18 and the Cu values ranged from
1.67 to 200. Whereas the CC range for all the collect soil samples is 0.02 to 75 and the

Cu range is from 1.65 to 385.71. Similar ranges in soil gradation were found for each

soil type. The vast ranges in the gradation within a given soil type is the direct result of
the glaciations and the glacial deposits in the State of Michigan.

Finally, the details of the USCS, the AASHTO and the MDOT soil classification

systems are included in Appendix A.

86

4.5 Cyclic Load Triaxial Test Results

Recall that, at most sample locations the soil was subjected to pocket vane shear
tester and pocket penetrometer (see section 4.3). After the two tests, bag samples were
collected and then transported to Michigan State University Geotechnical Laboratory. In
the laboratory, most soil samples were tested to determine their natural moisture contents,

Table 4.2 Number of samples per soil type

 

 

 

 

 

 

 

 

 

 

USCS AASHTO Classification
Soil Number of Soil Number of
classification samples classification samples
SP 20 A-l-a 2
SM 16 A— l -b 2
CL 8 A—2-4 21
ML 2 A-2-6 3
SC 18 A—3 25
SC—SM 7 A-4 10
SP-SM 8 A-6 12
SG 2 A-7-6 6

 

 

 

 

 

 

grain size distributions, Atterberg limits, and resilient modulus using cyclic load triaxial
tests.

As stated in Chapter 3, all cyclic load tests were conducted using one confining
pressure of 7.5 psi. In each test, after applying the confining pressure, the soil samples
were subjected to a 10 psi cyclic axial stress and their deformations were recorded at 5
intervals (cycle number 100, 200, 500, 800, and 1000). Each interval consisted of three
consecutive load cycles, for example, cycle number 100 consists of the sample
deformations at cycles numbered 99, 100, and 101. For each load cycle within a given
interval, the resilient modulus was calculated. Afterward, the average resilient modulus

from the three consecutive cycles was determined. After terminating the 10 psi axial

87

E02413 fig-03*“. 02-01 :1§<0§:1437¥Efl

[,_,,_ : L L___-

100.0
90.0 p ‘
80.0 e—fij—W
70.0 fl ‘ 1‘
60.0 0
50.0
40.0
. 30.0
20.0
10.0

 

Percent passing by weight

. HM
0.0 v» ‘ — **A_r

10.000 1.000 0.100 0.010
Particle size diameter (m) J

 

 

 

Figure 4.2 Typical particle size distribution curves
stress test at load cycle number 1001, the number of load cycle was reset and the test was
started again at the new axial stress level of 15 psi. The new test, and the calculation of
the resilient modulus, was conducted in the same manner as that of the 10 psi axial stress
test. For each axial stress level, Table 3.5 provides a list of the sample designation
number, the sample parameters (moisture content, saturation, and dry unit weight), and
the average resilient modulus values, which were calculated as the average of the average
resilient modulus values at the three intervals 500, 800, and 1000 load cycles.
Effects of Axial Stress Level on MR Values - Figure 4.3 depicts the resilient modulus
values obtained at 10 and 15 psi cyclic axial stress levels. The line of equality between

the two set of MR values is also shown in the figure. The data in the figure indicates that

88

the MR values, in general, decrease insignificantly with increasing cyclic axial stress
levels, indicating slight non-linearity. This observation was expected because as the axial
cyclic stress increases the strain in the sample also increases. This observation agrees
with that reported by Young and Baladi (1977). It should be noted that the above

observation does not necessarily disagree with bulk stress model stated in the M-E PDG

 

80,000
70,000
60,000
50,000
40,000
30,000
20,000
10,000

 

 

MR at 15 psi cyclic axial stress (psi)

 

0 —rr —

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000
MR at 10 psi cyclic axial stress (psi)

 

 

 

Figure 4.3 Resilient moduli at 10 and 15 psi cyclic axial stresses
(see Equation 2.1 in Chapter 2). In this model, as the axial stress increases, the bulk
stress (the sum of the axial stress and twice the confining pressure) and the octahedral
shear stresses increase. The effect of the latter is typically greater than that of the former.
Since k3 in Equation 2.1 is negative, it implies increasing the octahedral stress causes
decreases in the MR values.
Effects of the Sample Variables on MR Values - Given the standard pavement cross-
section used in the State of Michigan (the roadbed soil is located between 18 to 36 inches

below the pavement surface), the roadbed soil is typically subjected to 4 to 7 psi vertical

89

stresses due to an 18,000 pounds single axle load. Therefore, the effects of the sample
variables on MR values are discussed for the axial stress level of 10 psi only. In addition,
the effects of the sample variables on the subgrade MR values were studied in a two step
procedure. In the first step, the soils were divided into six groups according to their
USCS classification listed below.

1. Poorly graded sand (SP)

2. Silty sand (SM)

3. Clayey sand (SC), low plasticity clay (CL), and low plasticity silt (ML)

4. Poorly graded sand — silty sand (SP—SM)

5. Clayey sand — silty sand (SC—SM)

6. Gravelly sand (SG)

In the second step, univariate and multivariate analyses were conducted to
determine the relationships (if any) between the resilient modulus values and the sample
parameters. Results of these analyses for each of the six soil types are discussed below.
4.5.1 Poorly Graded Sand (SP)

Table 4.3 lists the locations, the USCS and the AASHTO soil classifications, and
the sample designation number of twenty disturbed soil samples that were collected from
various clusters and areas throughout the State of Michigan. Results of the soil
classification conducted in step 1 of the analyses indicated that nineteen samples can be
designated as A-3 and one sample as A-l-b according to the AASHTO soil classification
system. On the other hand, all twenty samples were classified as SP according to the
USCS. Please note that, in Table 4.3, twelve subgrade samples are labeled SP1 and eight

samples are labeled SP2. The reason for that is stated in section 4.5.1.1; the effects of

90

grain size on the resilient modulus of SP soils. Nevertheless, in step 2 of the analyses,
results of the cyclic load triaxial tests were used to calculate the resilient modulus of all
SP samples. Univariate and multivariate analyses were then conducted to determine the
relationships between the sample variables and the resilient modulus of the subgrade
soils. Results of these analyses are presented in the next two subsections. Results of the
cyclic load triaxial tests of three samples (the shaded samples in Table 4.3) were used to
verify the statistical models and are presented in section 4.5.1.3.
4.5.1.1 Univariate Analyses

In the univariate analyses, the effects of each of several sample variables on the
MR values of SP soils were studied. These sample variables include: the moisture
contents of the samples, the dry unit weight after laboratory compaction, and the grain
sizes. The discussion of the effects of each variable is presented below.
Effect of Sample Grain Size — To study the effects of sample gradation on the MR
values of SP soils, sieve analyses were conducted and grain size distribution curves were
plotted to determine the coefficients of curvatures and uniformity. Observation of the
grain size distribution curves indicate that the twenty subgrade samples can be divided
into two categories according to the slope of the gradation curve between the percent
passing sieves number 40 and number 200 as shown in Figure 4.4. Soils having the steep
curves are labeled SP1 whereas the others SP2 (see Table 4.3). In general, SP1 soils have
lower coarse sand contents and higher fine sand contents than the SP2 soils.

The effects of the sample variables of both SP1 and SP2 soils on their MR values
were studied by plotting the MR values versus each of the following grain size

parameters:

91

0 Percent passing sieves 200, 100, 40, 20, 10, 4, and 3/8 inch

0 The Coefficients of curvature and uniformity

0 The average particle size at ten, thirty, and sixty percent passing (D10 , D30, and D60,
respectively)

0 The coarse sand content (percent passing sieve number 4 — percent passing sieve
number 40) and the fine sand contents (percent passing sieve number 40 — percent
passing sieve number 200)

The data in most figures were scattered all over the board indicating insignificant
or no relationship between the grain size data and the MR values. An example plot
between the percent passing sieve number 10 and MR values is shown in Figure 4.5. The
scenario however, was drastically different when the data were separated into two groups

SP1 and SP2 as shown in Figure 4.6. The data showed moderate degree of correlation

(R2 for SP1 soils of 0.61 and for SP2 soils of 0. l 5). Given this observation, the particle

size parameters will be included in the multivariate analyses for the SP1 and SP2 models.
Similar, but stronger scenario was found when the MR values of the SP soils were plotted
against the dry density of the samples. This correlation is presented next.

Effect of Sample Dry Unit Weight - The effect of the dry unit weight on the MR for SP
soils was studied by testing one soil sample compacted at three dry unit weights and the
same water content using different compaction effort (vibrating times of l, 2, and 3
minutes). The three dry unit weights were 104.1 , 106.6, and 109.8 pounds per cubic feet.
After compaction, the three samples were subjected to cyclic load tests as stated in

section 4.5. Results of the cyclic load tests were used to calculate the resilient modulus

92

 

 

 

 

 

 

 

 

 

 

 

 

 

 

as m% 83E 8330 2E .8 as,» ea o2 A848 238-2
Sm m% BA A85 .8 58m ea 82 A84: 38.:
gm 3.. 82$ co €02 sue on A34: 3.8-:
«mm 3.. 8+? 83% A53 34.82
«mm m% 8m 2 co MO 582 use 92 A388 2.22.:
gm 3.. a: 2 83% A88: 3-88-2
Sm m% BA 8 8 $0 ssz soc 8m A848 38-:
«Boo o BSEBE Eo mEB : Bo 3“ u
Em 3% awe MAX 3 Saw Samoa? asomse 08 A388 $5-:
Em m% 2% as A co 68m ea m8 A308 .3 m TD
Em m% 3 £8.82 83 .6 $38 we 02 code .3 2-:
m% 3% 803880 .8 Sam ea 8? Amie 980.2
m% E 8:2 2 .6 Ham 2:: n: ASE: 358.:
3. N8 2 co co £52 we a: .A8. 88 288- A

 

m-<

 

2 an .8 225.2 ea 8m
2; E88 2::

garganaag

 

 

 

AB. 3 m- as. X
:.AA.o- we 2. 3.3

A8- 08 5.8... .D

 

 

     

  

 

 

 

Em m-< QEE co 0: EB mo “83 Bow N2 ANA: S 3%moA
confine
womb OHEm<< A5380:

 

 

 

aonmqmfioc 295m

 

 

2A8 gamma—um mm mo coumooq m4. 2an

93

of the samples. Figure 4.7 shows the MR plotted as a function of the dry unit weight. As
it can be seen from the figure, higher dry unit weights cause higher MR The observations
were expected and have been reported by other researchers (Maher et al. 2000).

Effect of the Sample Moisture Contents — To study the effects of moisture contents on
the MR values of SP soils, the soil sample with the lowest natural water content (0.2
percent) was selected for testing. This selection allowed the addition of water to increase
the water content of the sample from 0.2 percent to 5.3 and to 11.5 percent. The three
water contents were selected such that the highest degree of saturation of the samples will
be less than eighty percent. The eighty percent saturation level may cause liquefaction
and a total loss of shearing resistance (Richart et a1 1970). The three selected water
contents correspond to degrees of saturation of 1, 26.4, and 59.3 percent, respectively.
For each of the three water contents a soil sample was compacted using the same static
load and vibrating table described in section 3.5.4 of Chapter 3. It should be noted that
all three samples were compacted using the same compaction effort.

After compaction, the sample was subjected to a cyclic load test as stated in
section 4.5. Results of the cyclic load tests were used to calculate the resilient modulus
of the sample. Figure 4.10 depicts the resilient modulus of the three soil samples plotted
as a function of their water contents. The figure shows that for the given range of
moisture contents, their effect on MR is insignificant. This was expected because, in
general, the strength and stiffness of sand soils are only slightly affected by the water
content of the soils. At low water contents, the capillary between the sand particles
slightly increases the normal stress and hence, the friction between particles. As the

moisture content increases, the lubrication between the sand particles increases, which

94

.248 mm .5,“ 8E8 couspwaflu 0N6 880 Al. oSME

 

AEEV ONE 380

 

 

 

«6-3+ moéolql onol‘l moéolfll vo-mo+ 84.0 .0 L

ed

odm

0.9.

ode

0.3

063

iqfiiom Aq fiuissed moored

 

 

95

 

60,000

‘ 50,000

 

40,000

30,000

20,000

 

Resilient modulus (psi‘

10,000

 

0
70.0 80.0 90.0 100.0

Percent passing sieve number 10 by weight

 

 

 

Figure 4.5 Resilient modulus versus the percent passing sieve number 10 for SP soils

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l osp1 osm
~—- 2
60’000 y=~58.216(PP10) +9796.6(PP10)- 371595
5;: 50,000 -~ R2=0.6123
3 . u
g 40,000
2%
a 30,000
a
.3 20,000
.3 2
32 10,000 -_. MR = 39.366(PP10) — 7477.6(PP10) + 377914___c
2
0 -_. R =0.1526
70.0 80.0 90.0 100.0

Percent passing the number 10 sieve

 

 

 

Figure 4.6 Resilient modulus versus the percent passing sieve number 10 for SP1 & SP2

96

 

 

 

33,000 ~
31,000
29,000 -

 

 

4 MR = 337.53(l;d)2 - 70133(§d) + 4E+O6 if
i l :

 

 

 

 

27,000

4

2
R=l
‘/

 

 

25,000

 

23,000

 

21,000

Resilient modulus (psi)

 

19,000
17,000

 

 

15,000
103

106 107 108 109 110
Dry unit weight (pcf)

 

 

 

Figure 4.7 Resilient modulus versus the dry unit weights for one SP soil sample

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50,000 1 | .
l
1 1 l '1 °
A 40 000 -__ ‘ ‘ ie ';
.._. ’ . : O
:8; . ; l I f
m l ‘ 0 fi'
-;1,’ 30,000 '~ 1 -3 ° ,
3 l l l .0 l l
E l 1 .0 . i O
,3 20,000 . I g
E g 1
‘3‘ 10000 Al 9 ' 2
’ l MR=—77.83( d) +18284( d)-1E+06
2
l R =0.4332
O l T r fir 1
95.0 100.0 105.0 110.0 115.0 120.0

 

Dry unit weight (pct)

125.0

 

Figure 4.8 Resilient modulus versus the dry unit weight for 20 SP soil samples

97

 

 

 

.SP1 OSP2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50,000 - . l
2
MR = —20.777( d) + 6191.3( (1) - 399259 e 1
9 40,000 R2 = 0.852 1 I. '
é ‘ 1 l ’( :
S l 9 l
3 30,000 1 . 1
"U l ‘
O l
g 20 000 i 0% O
Q) ’ 1 l
a) i l
a: 5.4932
10,000 lo , l MR = 1E-07( d)
l l | R2 = 0.5019
0 1 i 1 l i
95.0 100.0 105.0 110.0 115.0 120.0 125.0

Dry unit weight (pci)

 

 

 

Figure 4.9 Resilient modulus versus the dry unit weight for SP1 and SP2 soil samples
overcomes the capillary effect and the strength begins to decrease. When the sample is at
80 percent saturation level or higher, the pore water pressure increases during shearing
and the effective stress decreases causing decreases in the angle of internal friction. At or
near saturation, the strength drops to zero and the sand liquefies. Similar results were
reported by (Young and Baladi 1977, Holtz and Kovacs 1981, and Richart et al 1970).
4.5.1.2 Multivariate Analyses

Multivariate analyses were conducted to study the combined effects of several
independent sample variables of the SP (SVSP) soils on the dependent variable MR of
those soils. The term SVSP was divided into two terms; SVSPl and SVSP2 to express

the two SP soil groups; SP1 and SP2. During the multivariate analyses:

98

0 Various models were used in an attempt to maximize the value of the coefficient of

determination (R2).

 

 

34,000

 

32,000 + l
30,000 1 11*
1

 

 

 

 

 

 

 

28,000 - -

l

 

26,000 -

1 4
i

 

l
1
24,000 ~ 4 ‘ i
' 2
l Wy=9.3102x -289.64x+31518

 

Resilient modulus (psi)

 

22,000 - 2
R =1

 

 

 

 

l

1
20,000 I l
0 2 4 6 8 10 12
Moisture content (%)

 

 

 

 

Figure 4.10 Resilient modulus versus the moisture content of one SP soil sample
0 Special care was taken to:
0 Ensure that the resulting equation satisfies the known trends between each of the
independent variable and the dependent variable MR.
0 Avoid any significant co-linearity between the independent variables.
0 Decrease the number of independent variables in the equation.
The multivariate analyses yielded Equations 4.2 and 4.3 for sets 1 and 2 SP soils,

respectively.

MR = 89.825(SVSP12"’437), R2 z 0.78 5......“
MR = 0.8295(SVSP23'6006); R2 z 0.81 Equation 43

99

1.15

SVSP1= 1 7.

When, 1.5 0.25 )0-5
(1)4 — 1’40

 

71.35 * P—0.l

d 200

SVSP2 = 0,
(P415 _ 1343.25) -

 

Yd = dry unit weight (pct)
P4, P40, and P200 = percent passing sieves numbers 4, 40, and 200,

respectively
Examination of Equations 4.2 and 4.3 indicates that the resilient modulus values

of the SP1 and SP2 soil groups are a function of the dry unit weight of the soil and the

a 0) . . .
parameter (P4 — P40B) , which represents the coarse sand content 1n the 5011. Further,

the resilient modulus values of SP2 soil group are also a function of the percent passing
sieve number 200. Hence, the data and results of the multivariate analyses reflect the
shape of the gradation curves of SP1 and SP2 soils shown in Figure 4.4. The shapes of
the gradation curves imply that the mechanistic behavior of SP1 and SP2 soils are not the
same. The SP1 soils are deficient in the coarse sand content; hence, coarse sand particles
are floating in the fine sand matrix. For SP2 soils, the opposite is true. The SP2 soils
contain relatively large amounts of coarse sands; hence, the coarse sand particles are
likely in contact with each others while the fine sand particles are filling the voids
between the coarse particles.

Figure 4.11 shows the MR values of SP1 and SP2 soils plotted against the sample

variables; SVSP] and SVSP2, respectively. The equations of the best fit lines and the

100

coefficients of determination are also shown in the figure. Examination of the figure

indicates that for SP1 and SP2 soils, higher values of SVSPI and SVSP2 produce higher

MR values. This was expected because higher dry unit weights imply denser particle

packing, higher relative density, higher fi‘iction, higher stiffness, and hence, higher MR

values.

It should be noted that numerous trials were made to unify the two models.

Various univariate analyses that included a range of other sample variables were tried

without any success. When the locations of SP1 and SP2 soils were studied, it was clear

that all SP2 soils are located in the eastern half of the State of Michigan while SP1 soils

in the western half. The possible differences between the two soils include:

The two soils have different origin, the SP2 soils were deposited as the glacial lobe,
which was advanced along Lake Huron, retreated. The SP1 soils, on the other hand,
were deposited when the glacial lobe, which was advanced along Lake Michigan
trough, was retreated.

The SP1 soils were deposited by gently flowing melted water (it contains higher
percent of fine sand) while the SP2 soils were deposited by relatively faster moving
melted water.

The differences in the soil origin and deposition may have created different angularity
of the coarse materials. Unfortunately, the soil angularity was not measured due to
lack of proper equipment nor was it a part of this study.

The fine materials (passing sieve number 200) of the SP1 and SP2 soils are different.
The fine materials of the SP1 soils is mainly silt while the fine materials of the SP2

soils are a mix of clay and silt.

101

Equations 4.2 and 4.3 apply to samples with dry unit weight values ranging from
100.9 to 120.3 pcf and percent passing sieve number 4 values between 84.5 and 100.0
percent while the percent passing sieve number 40 is between 36.8 and 97.2 percent. The
range in percent passing sieve number 200 is 0.5 to 4.7 percent. This range is the entire
allowable range of fine contents of SP soils. The use of the two equations outside the

stated ranges may yield high and unrealistic MR values.

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I SVSP2 El SVSP] A Verification
50,000 47,—? —— 7 — 7 A . “a .
7 1 ‘ ‘ 7 A“ 4—
A 1 I
'5 40 0 0 1X» . . 1 1 = l
3 , 0 1D 1 1 1 1 fl-
? 5"" ” 1 . 1 ‘
3 30,000 ~ 1 1 -. #1 / 1
8 ;m_j1_. 1 1 _
E 1 " 1 1
*a' 20,000 » « . 1 ._ .
i”) 2 9437* *“JMR — 0 8295 SVSP2 36°06
27. MR = 89.825(SVSP1) ‘ 1 - - ( )
3; 10,000 2 - R2208106
R = 0.7759 , 7 1.-- -
0. SE=1637 . fi‘ . 81373412 .
6 8 10 12 14 16 18 20 22
SVSP] and SVSP2
Figure 4.11 Resilient modulus versus SVSPl and SVSP2
4.5.1.3 Validation

In order to check the validity of Equations 4.2 and 4.3, the two SP1 and the one
SP2 samples highlighted in Table 4.3, which were not included in the development of the
equations, were subjected to grain size analysis, and cyclic load triaxial tests. The test
results were used to calculate the resilient modulus values of the three samples and the
sample variables (SVSPl and.SVSP2). The data are shown in Figure 4.11 as open

triangles. It can be seen that the data for the three samples are very close to the best fit

102

curve. Further, the two SVSPl values for SP1 samples and the SVSP2 value for SP2
sample were used in Equations 4.2 and 4.3, respectively, and the resilient modulus of the

three samples were predicted. Figure 4.12 shows the measured and the predicted MR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

36,000 . . , .
1 1 L1ne of equahty I
E ' 1 1 1
3 32,000 1 '1 1 1 41 1
3'3 1 1 A1 .LOA ' 1
E 1 1 a ‘
.5 28,000 1 1 1 I I
a 1 1 1
£83 ' 1 1 1
153 24,000 12_ A 1 1 1 L 1
0 1
"‘ 9 1 1
1 .
i3: 1 1 T 1 1
20,000 1 1 1 1 .
20,000 24,000 28,000 32,000 36,000
Laboratory measured resilient modulus (psi)

 

 

 

Figure 4.12 Predicted resilient modulus values for the validation points

values of the three samples. The straight line in the figure is the line of equality between
the predicted and the measured MR values. As can be seen from the figure, the predicted
MR values are almost equal to the laboratory measured values. Hence, one can conclude
that the developed models (Equations 4.2 and 4.3) are relatively accurate and can be used
to estimate the resilient modulus of the soils based on knowledge of the dry unit weight
of the soils and their grain size distribution.
4.5.2 Silty Sand (SM)

Table 4.4 lists the locations, the USCS and the AASHTO soil classifications, and
the sample designation number of sixteen disturbed soil samples that were collected from

various clusters and areas throughout the State of Michigan. The commonality between

103

the sixteen samples is that all of them were classified in step 1 of the analyses as silty
sand (SM) soils according to the USCS. Recall that the USCS specifies that SM soils
may contain anywhere between 12 and 49.9 percent passing sieve number 200 (Holtz and
Kovacs 1981). Hence, the fine materials play a major role in the mechanistic behavior
and the resilient modulus (MR) values of the soil. In step 2 of the analyses, univariate
and multivariate analyses were conducted to develop correlations between the test sample
variables and their resilient modulus values. The results are presented and discussed
below. It should be noted that the three samples highlighted in Table 4.4 were not
included in the development of the correlation equations between the sample variables
and their resilient modulus values. They were tested and the test data were used to verify
the developed models.
4.5.2.1 Univariate Analysis

In the univariate analyses, the effects of each of several sample variables on the
MR values of SM soils were studied. These sample variables include: the moisture
contents of the samples, the degrees of saturation, the liquid limits, the dry unit weight
after compaction, and the grain sizes. The effects of each variable on the resilient
modulus values are presented and discussion below.
Effect of the Sample Moisture Contents — For SM soils, it was hypothesized that
because of the high range of fine contents, the water contents should play a major role in
determining the elastic response of the soil to the applied loads. Figure 4.13 shows the
MR values of thirteen SM soil samples (the three shaded samples in Table 4.4 are not
included) plotted against the samples moisture contents. As it was expected, the figure

shows increases in the sample moisture contents cause significant decreases in the MR

104

Table 4.4 Locations of SM subgrade soils

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample number Location AASHTO USCS
U-002-E (02-01) 385 feet East of M-45 A-4 SM
U-002-E (03-03) 200 feet East of M-l 17 A-2-4 SM
M-065-S (04-04) 160 feet South of Elm Hwy A—2-4 SM
U-131-N (05_01) 200 feet South of Michigan Fisheries Visitor A—2-4 SM

. Center
M-06l-E (07-0 6) 420 feet East of lefi hand turn on M-61 (off A-2-4 SM
US-127)
M-061-E (08-02) 165 feet West of Hockaday A-2-4 SM
M-044-E (09-07) Station 137+10 A-2-4 SM
M-024-S (09-09) 20 feet North of Burley Rd A-2-4 SM

I—069-N (10-04) 150 feet North of Island Hwy A-2-4 SM
I-069-N (10-05) 100 feet North of Five Points Hwy A-2-4 SM
I-096-W (10-09) 140 feet West of Dietz Rd A-2-4 SM
U-012-E (12-04) 100 feet East of Emarld Rd A-2-4 SM
I-094-W (12-06) 53 feet West of Mt Hope Rd A-2-4 SM
M-024-S (13-01) 250 feet North of Best Rd A-4 SM
M-053-S (15-02) 300 feet South of M-46 A-2-4 SM

650 feet South of Thom son Rd 1 mile
M-Ol9-S (1507) South of M1542 A-2-4 SM

 

values. This observation tends to validate the hypothesis stated above. Similar results
were also reported by many researchers including (Maher et. al 2000, George 2000, and
2003). One observation is important to note herein is that the effect of moisture content
on the resilient modulus values of SM soils is much higher than that for the SP soils
reported in the previous section. This is mainly due to the much higher fine contents in
the SM soils compared to the fine contents of the SP soils (less than 5 percent). Since the
water content is strongly correlated to the MR values of SP soils, it will be included in the

multivariate analyses.

105

 

 

 

35,000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1 1 -0.069(MC)
,9 _
1 1 1 MR—31285e
Q 30,000 1 a ,_
32,. 1 1 R —0.7295
25,000 .' -
i=3 \\ 1 ‘
g 20,000 , . ; 4 1
E L \ i i
g 15,000 1 . \ 1
E 10,000 1 1 1 ’ 1
g ' ' 1' \
5,000 1 {a ; 1 -
0 1 1 4 J

 

 

 

0.0 5.0 10.0 15.0 20.0 25.0

Moisture content (%)

 

 

 

Figure 4.13 Resilient modulus versus water contents of the SM soil samples
Effect of Sample Dry Unit Weight - The effect of the test sample dry unit weight on the
MR values of SM soils were studied by plotting the MR values as a function of the
sample dry unit weight as shown in Figure 4.14. As it was expected, the figure shows a
weak correlation between the dry unit weight and the MR values of the test samples. The
main reason for the weak correlation is that the water contents of the test samples vary
from about 3 to about 23 percent. Such variation in the water content, (when examined in
perspective of the compaction curve) covers both the wet and the dry side of the curve.
This implies that two soil samples having the same dry unit weight value may have two
significantly different water contents. One is located on the wet side of the optimum
moisture content and the other on the dry side. Test samples compacted on the dry side
of optimum would have higher strength and stiffness and display a more brittle behavior
than those compacted wet of optimum. The latter would have lower strength, higher

plastic deformation and softer behavior under loads. The differences in the behavior are

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35,000 I 1 I
2.782
__ 1 1 0

A 30,000 _MR — 0.0327( d) 1 1 1 1
g 25 000 R2 = 0.428 1 1 1 1
m 3 1 1 ' '
g 1 1 1 1 1 /
a , 1 - 1 .
.2 o 1 1 .
a 10,000 1 1 1 ' 1 ,
34’ 1 1 1 1' 1 1 1

5,000 4 1 I I 1 ' I

0 . 1 #1 1 1 1

 

 

1
90.0 95.0 100.0 105.0 110.0 115.0 120.0 125.0 130.0
Dry unit weight (pci)

 

 

 

Figure 4.14 Resilient modulus versus dry unit weight of 13 SM soil samples
directly related to differences in the degrees of lubrication caused by the water and the
particle arrangement ion the soil. The soil particles of a soil sample compacted on the
dry side of optimum, tend to stay in a flocculated arrangement whereas on the wet side of
optimum, they are dispersed (they line up), (Holtz and Kovacs 1981).

The above discussion implies that the true effect of the dry unit weight on MR
values cannot be separated from the effect of the water content unless the latter is held
constant and the former is changed using different compaction effort. In the multivariate
analyses presented in the next subsection, the effect of dry unit weight on the MR values
were analyzed in conjunction with the effect of water content of the test samples.

Effect of the Sample Degree of Saturation — For each test sample, after the conclusion
of the cyclic load test, the sample moisture content and dry unit weights were determined

and the degree of saturation (S) was calculated using Equation 4.4. Please note that for

107

all SM soil samples, a typical value of the specific gravity of the solid (Gs) of 2.7 was

assumed and used in Equation 4.4.

* =1:
5: GS (MC/100) yd *100

* _ E uation 4.4
Gs yw Yd q

 

Where,
S = degree of saturation (%)

MC = moisture content (%)

GS = specific gravity of the soil solid = 2.7
Yd = dry unit weight of the sample (pci)

Yw = unit weight of water = 62.4 pcf

Figure 4.15 shows the MR values plotted against the degree of saturation of the
test samples. As it was expected and reported by Maher et al. (2000), the MR values
decrease significantly with increasing degrees of saturation. One may argue that the data
in Figure 4.15 is repetitive and are the same as the data in Figure 4.13, hence, Figure 4.15
can be eliminated. In reality, the water content of a soil sample is an independent
variable whereas the degree of saturation is a function of the dry unit weight and the
water content of the soil. Hence, the data in Figure 4.15 show the combined effects of the
dry unit weight and the water content of the test samples on their MR values. The degree
of saturation will be included with other sample variables in the multivariate analyses to

determine their combined effect on MR.

108

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35,000 1 7 00101(S)
O ' = - .
A 30,000 1 1 MR 226425e
'5 1 1 1 R =0.2479
§ 25,000 1 I 1 I
3 1
'5 N 1 I 1
E 20,000 .1 X 1 N r1
H 15,000 A .
E 1 ° 1 N
'77. 10,000 1 1 1 .
a: I .. 1
5,000 1 1 1 I g
1 1
0 , f 1 1
0.00 20.00 40.00 60.00 80.00 100.00

Degree of saturation (%)

 

 

 

Figure 4.15 Resilient modulus versus the degree of saturation of SM soil samples
Effect of the Sample Liquid Limit - For each SM soil sample, the Atterberg limits for
all materials passing sieve number 40 were determined in order to classify the type of
fine materials (silt or clay). The plastic limit tests showed 15 soil samples can be
classified as non-plastic (the plastic limit test failed repeatedly). The plastic limit test
was successful for only one soil sample and the plastic limit of the soil was very low.
Hence, the effects of the plastic limits and plasticity index on the MR values were not
analyzed. However, the effects of the liquid on the MR values of the soils were analyzed.
Figure 4.16 depicts the influence of the liquid limits on the MR values of SM soils. The
data in the figure indicate that the MR values of SM soils having higher liquid limits are
lower than those having lower liquid limits. Such observation was expected and has been
reported by many researchers for various soil types including silty and clayey sands, silt,

and clay (Gudishala 2004). Given the strong correlation between the liquid limit of the

109

material passing sieve number 40 and the soils MR values, the liquid limit data were
included in the multivariate analyses presented in the next subsection.
Effect of Sample Grain Size — Because of high fine contents, all SM soils were
subjected to wet sieving and hydrometer analyses to determine their grain size
distribution. The effects of sample gradation on MR values were assessed through the
following gradation parameters:
0 Percent passing sieves 200, 100, 40, 20, 10, 4 and 3/8 inch
0 The coefficients of curvature and uniformity
0 Average particle size at ten, thirty and sixty percent passing (D10, D30, and D60,

respectively)
The effects of each gradation parameter on the MR values were analyzed by plotting the
MR values of the soil samples as a function of that parameter. Only three gradation
parameters, the average particle size at 10, 30, and 60 percent passing showed minor
correlation to MR values, the others showed no correlation. Figure 4.17 shows the
correlation between the average particle size at thirty percent passing and the MR values.
Based on these observations, the average particle size at 10, 30, and 60 percent passing
were included in the multivariate analyses, which are presented in the next subsection.
4.5.2.2 Multivariate Analysis

Multivariate analyses were conducted to study the combined effects of several

sample variables on the dependent variable MR of SM soils. During the analyses:

0 Various models were used in an attempt to maximize the value of the coefficient of

. . 2
determination (R ).

110

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35,000 I 1
1 I , 1 MR=~18654Ln(LL)+69571
J. 1 2
.3.; 30,000 I 1 1 R =0.5264
3‘ 25,000 . 1 1 1+ . 1
.3 1 1 \1 ' 1 1
g 20,000 1 I . . . 1 1
E 1 1 1 1
.5 15,000 I I I \ :
8 9 1 T 1 1 1
m 1 D
5,000 L 1 1 e 1 \
0 1 1 1 1 1 1
0 5 10 15 20 25 30 35
Liquid limit (%)

 

 

 

Figure 4.16 Resilient modulus versus the liquid limits of SM soils

0 Special care was taken to:

0 Ensure that the resulting equation satisfies the known trends between each of the

independent variable and the dependent variable MR.

0 Avoid any significant co-linearity between the independent variables.

0 Minimize the number of independent variables in the equation.
Results of the analyses yielded two models having relatively high R2 values. The first
model is based on two sample variables of the SM (SVSM) soils (the dry unit weight and

the degree of saturation) as stated in Equation 4.5.

MR = 0.0303(SVSM)4"325 Equation 4.5

 

SVSM = 71"

Where, S 0.15

W = dry unit weight (pcf)

111

S = degree of saturation (%)

 

35,000 I 1

I .
' 1

30,000 1 1 2
1

 

 

)

Sl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9’ 25 000 4 1 1-
% 20,000 i“ 4 -
E ’ / 1 1 1
- g
*a 15,000 V 1 I I j
313’ 10 000 O J 1 4 1 1
a ’ 1 ‘ 1
. 0.842 P30
ad 5 000 5%; 1 1 1 MR=136126 ( )
’ 1 V 1 R2 — 0 2621
0 1 1 . —' .

0.000 0.200 0.400 0.600 0.800 1.000 1.200

Average particle size at 30 percent passing

 

 

 

Figure 4.17 Resilient modulus versus the average particle size at thirty percent passing
Although the two variables are some how co-linear (both are a function of the

water content of the soil), the interpretation of the model agrees with most literature. To
illustrate, consider the data in Figure 4.18, in which the resilient modulus is graphed as a
function of the SVSM It can be seen from the figure and from Equation 4.5 that increases
in the dry unit weight cause increases in the SVSM values and hence, increases in the
MR values. Further, increases in the degree of saturation causes decreases in the SVSM
and the MR values. That is the MR value can be increased by either increasing the dry
unit weight (i.e., higher compaction effort) or by decreasing the degree of saturation or by
combination thereof. The reason is that as the dry unit weight of the sample increases,
the relative density increases and the particle to particle contact in the sample increases
causing higher internal fiiction and hence, higher stiffness (Perloff and Baron 1976). On

the other hand, decreasing degree of saturation implies decreasing moisture contents and

112

decreasing the degree of lubrication between the soil particles. This causes increases in
the soil internal friction, soil stiffness, and MR values. Similar results were also reported

by Maher et al. (2000).

jModel 1336521353

 

 

 

 

 

 

 

 

 

 

 

 

 

35,000 4 1
1MR=0.0303(SVSM)4'1325 1 I o 1
a: 30,000 7 2 —1——«—~-—#7¥~
a 1 R =0.7592 I .
3 25,000 SE=1482 7"”“171—13“ '—
:: 7, __ .m 1
E 20,000 , . 7 7 .
.. 15,000 a
E
a 10,000»
0
‘1‘ 5,0004
0 4
20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.00
SVSM

 

 

, ‘__.___,J

 

Figure 4.18 Resilient modulus versus the sample variable for SM (SVSM) subgrade
The two important points that should be noted herein are:

1. The data for the three open symbols in Figure 4.18 are those of the three samples used
to verify Equation 4.5. They are discussed in the next subsection.

2. The test data used in support of Equation 4.5 have the following ranges: degrees of
saturation from 18.8 percent to 93.9 percent and dry unit weight from 94.6 to 128.8
pounds per cubic foot. The use of the equation outside these two ranges is not
recommended.

Nevertheless, the multivariate analyses of the SM soils yielded a second model
based on two independent variables; the moisture contents (MC) of the test samples and

the liquid limits (LL) of the soils passing sieve number 40. In this study, the two

113

independent variables were combined into one parameter, which was named. the moisture

index (MI) as stated in Equation 4.6.

MR = 45 722exp(—0.0258* MI) Equation 4.6

Where,

M1 = Moisture Index: LL"l + MCI-25

Figure 4.19 depicts the resilient modulus values of the thirteen SM soils plotted against
the moisture index. Inspection of the figure indicates that, in general, higher MI values
produce lower MR values. That is, increasing either the moisture content or the liquid
limit of the soils causes increases in the MI values and hence, decreases in the MR values
of the soils. These observations were expected because higher MI values due to higher
moisture contents cause sofiening of the SM soils and hence lower resilient modulus.
Likewise, soils having higher liquid limits tend to be more plastic and higher sofienin g

potential due to changes in their water contents.

 

 

0 Model B Verification

 

 

 

U)

Resilient modulus (p

35 ,000
30,000
25,000
20,000
15,000
10,000
5,000
0

21 .00

31.00

41.00

1
1

 

51.00

-0.0258(M 1)

MR = 45722e

R2 = 0.8847
SE = 1584

61.00 71.00 81.00

Moisture index

 

 

 

Figure 4.19 Resilient modulus versus the moisture index of the test sample

114

Equation 4.6 allows the user to estimate the MR values of SM subgrade soils
using the results of two simple tests: the natural water content of the entire soil and the
liquid limit of the particles passing sieve number 40. It should be noted that, in the field,
the natural moisture content of a soil varies from one season to another whereas its liquid
limit is constant. Hence, if the LL of a given subgrade soil is known, one needs only to
determine the moisture content of the soil only before using Equation 4.6. It is important
to note that the data used to generate Equation 4.6 include a range in moisture contents
from 3.9 to 23.7 percent and a range in liquid limits from 11 to 30 percent. The use of
Equation 4.6 outside the two ranges is not recommended.
4.5.2.3 Verification

The three shaded SM soil samples in Table 4.4 were not included in the
development of the two SM soil models (Equations 4.5 and 4.6). The three samples were
subjected to wet size analysis, Atterberg limit tests, and cyclic load triaxial tests to
determine the physical parameters and the MR values of the three soils. After the
laboratory tests were completed, the data for the three SM soils were plotted in Figures
4.18 and 4.19 as open symbols. It can be seen that the open symbols in both figures are
located in the vicinity of the best fit curves.

Finally, Equations 4.5 and 4.6 were used to calculate the resilient modulus values
of the three SM soils based on their parameters (dry unit weight and degree of saturation
and moisture contents and the liquid limits of the soils). The calculated and the
laboratory measured resilient modulus values are plotted in Figures 4.20 and 4.21. As
can be seen from the figures, the data for both equations are located close to the line of

equality. Based on this observation, one may conclude that the two models presented in

115

Equation 4.5 and 4.6 are accurate and can be used to estimate the resilient modulus
values of SM soils based on the soil parameters. Such parameters can be obtained from

simple tests such as water content, liquid limit and dry unit weight tests.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 25,000 . I ,
g 1 1 I 1 .
3 20,000 1 Line of equality VI
“:3
1: .
o o
a 15,000 1 , \ 1
1:; 1 . 1
2+: 1 L
§ 10,000 I I I
E 1 1 1 d
.2 5,000 , I I
8 1 »
H 1 1 1
m /1 1 L
O 1 1 1 I
0 5,000 10,000 15,000 20,000 25 ,000
Laboratory resilient modulus (psi)

 

 

 

Figure 4.20 Measured and calculated resilient modulus values, Equation 4.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 25,000 I I I I .
.171 1 I} .1 .
f”? 20 000 1 1 1 1
2 ’ 1 . . o
g 1 L1ne of equality I
E 15,000 1 I e .
5 1 1 1
:5: I . ' I
'53 10,000 I 1
8 1
B 1/11 1
,§ 5,000 1 1
“d 1 1 1
8 /1 1 ' 1
a” 0 AI 1 I
0 5,000 10,000 15,000 20,000 25,000
Laboratory resilient modulus (psi)

 

 

 

Figure 4.21 Measured and calculated resilient modulus values, Equation 4.6

116

4.5.3 Clayey Sand (SC), Low Plasticity Clay (CL), and Low Plasticity Silt (ML)

Table 4.5 lists the locations, the USCS and the AASHTO soil classifications,
Atterberg limits, and the sample designation number of twenty two disturbed and six
Shelby tube (marked with S in the table) soil samples that were collected from various
clusters and areas throughout the State of Michigan. According to the USCS, eighteen
samples consists of clayey sand (SC), eight clay (CL), and two low plasticity silt (ML)
samples. According to the USCS SC soils may contain anywhere between 12 and 49.9
percent by weight fine materials and the plasticity index and liquid limit of the material
passing sieve number 40 plot above the A-line on the plasticity chart. Clay (CL) soils
contain more than 50 percent by weight passing sieve number 200 and the plasticity
index and liquid limit of the soil plot above the A-line on the plasticity chart. Finally, the
ML soils contain more than 50 percent by weight passing sieve number 200 and the
plasticity index and liquid limit data plot below the A-line on the plasticity chart (Holtz
and Kovacs 1981). All soil samples listed in Table 4.5 were subjected to wet sieve and
hydrometer analyses, Atterberg limit tests, and cyclic load tests.

The reason that the three types of soils are housed in Table 4.5 is that, alter the
completion of the cyclic load tests, the resilient modulus values of the samples were
plotted against the sample moisture contents. The three soil types showed the same
relationship between the MR values and the sample moisture contents.

A total of 16 cyclic load tests were conducted on SC soils, 1] tests on undisturbed
soil samples and 5 tests on disturbed soil samples. Nine cyclic load tests were conducted
on CL soils, five tests on undisturbed soil samples and four on disturbed samples. Finally,

both ML soil samples were subjected to cyclic load triaxial tests.

117

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0m v< 2 E 5 85: aeozco :52 as at Goa : 2-2-2
8 8-3. 8 S E 88885 .8 cam 8e Swz m 63.: 32.:
em 3.. a «m Em 880: 8 :83 as 28 9.3: 3-38;
8 3. S X“ 8+8 8:3 $3: WE:
em 3. 2 9. 2 38m .8 :88 8e 821 A81: 38-:
mm 8-3.. 2 z. 2 $28.28 :88 8e 9. A81: 3::
Um 8-8% a a. E 8:55 8 58m 8e 8 m :3: 32:
um i. 2 mm 2 88828 883 8a Rm Goa: 3.29:
em «.3. E 8 E 888% .8 “am 8e 8w 2 To a m- §-E
0m 3. 2 3 E Seem 8 mam 83 OS 83: mag;
om 8.3, 2 mm 3 88 2&3 825.8 :83 8a 2.8. SE: 3.83
om v< E 3 SH @0388? 8 88m 8e o8 $988 was;
Um o-~-< 2 mm 2 82m .8 58m 8a m: m 333 was;
Um 3. 2 mm 3 38.382 8 :83 8e 9; 98-on 388-:
mm e-< E E E 8&3, 8 €02 8e 8 m $38 2-5-:
Um 8-3. 2 mm 2 88853 .8 Suez ea 2: 9.38 2-5-:
Um 3. : mm 2-2.8 88% 8e Omm @088 >288
Um 2}. mm mm mom 8888 one A888 2.33
:8: 8E
0:85 2:5 388% 8958
wow: Emmi 8:83 BB 38% academic
SEE wcofiouz 295m

 

 

 

 

 

 

 

28m 28825 d2 28 .16 .8 .8 8:83 3 2.5

118

 

 

<2

3

E 8:288 8 8mm 88 88
8728 88B 88 $2

 

_ @088 ”.7831
A888 3.88.:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88... 88: 88:58:81: 88 R
2 «N 8 888m .8 88m 88 com $0-3 908-2
15 8% cm mm E 88828 8E 88 28 808: 18:08.:
16 8-3. a a. E 8:8: .8 883 88 Rm 83.: 3.88.:
:o 8-3. a .3. E 888883 .8 EB 88 on m 83: 3+8:
:0 3. E E 8+8 8:88 m $0-3 @292
.6 8.8.. 2 mm 2% 88888 88 E 28...: 888E Goa c 1883-:
.6 88 3 mm E :88 808: .8888 883 88 88 A888 3-20-:
5 8% E 3 8E 888880 8 88m 88 88 :38 @292
:8: :8:
mUwD 05mm: 03.35 8:55 nonmooq 335.8 855:: 038mm
£8: 938838 . BB >£osm

 

 

 

 

P850 m4“ 2an

119

In step 2 of the analyses, univariate and multivariate analyses were conducted
simultaneously on the three soil types to study the effects of the sample variables on their
MR values. Results of the analyses are presented and discussed below.
4.5.3.1 Univariate Analyses

In the univariate analyses, the effects of each of several sample variables on the
MR values of SC, CL, and ML soils were studied. These sample variables include: the
moisture contents of the samples, the degrees of saturation, the dry unit weight after
compaction, and the grain sizes. Results of the analyses are presented and discussed
below.

Effect of Sample Moisture Contents — As is the case for the SM soils, because of the
high fine contents of SC, CL, and. ML soils, it was hypothesized that the water contents
of the samples would play a major role in determining the elastic response of the soil to
the applied loads. Figure 4.22 shows the MR values plotted against the samples moisture
contents. As it was hypothesized and expected, the figure shows increases in the sample
moisture content cause significant decreases in the MR values. Similar results were also
reported by many researchers including (Maher et. a1 2000, George 2000, and 2003).
Further, the data in Figure 4.22 also show that the three soil types have similar, if not the
same, relationship between the sample moisture contents and the MR values. Hence, the
moisture content or the degree of saturation will be considered in the multivariate
analyses.

Effect of the Degree of Saturation — At the conclusion of the cyclic load test, the

sample moisture content and dry unit weight were determined and the degree of

120

 

 

OSC

DCL

AML

 

 

100,000
90,000
80,000

—2.0312

MR = 3E+06(MC)
R2 = 0.6279

70,000
60,000
50,000
40,000
30,000
20,000
10,000

0

Resilient modulus (psi)

 

1 O o

0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0

Moisture content (%)

 

 

 

Figure 4.22 Resilient modulus versus the moisture contents of the samples

saturation (S) was calculated using Equation 4.4. It should be noted that for all test

samples, atypical value of the specific gravity of the solid (Gs) of 2.7 was assumed and

used in Equation 4.4.

G, *(MC/100)* yd *100

S :
a: _ E uation 4.4
Gs yw yd Cl

 

Where,
S = degree of saturation (%)

MC = moisture content (%)

GS = specific gravity of the soil solid
Yd = dry unit weight of the sample (pct)

'Yw = unit weight of water = 62.4 pcf

121

Figure 4.23 shows the MR values plotted against the degree of saturation of the test

samples. As it was expected and reported by Maher et al. (2000), the MR values

 

 

 

 

OSC EJCL AML

 

80,000
0.0501(5)

)

70,000 MR = 650486e
' R2 = 0.8782

81

(p
O\
O
O
O
O

9

50,000
40,000
30,000
20,000
10,000

0

40.00 50.00 60.00 70.00 80.00 90.00 100.00

Saturation (%)

1
1

1 1 o

Resilient modulus

 

 

 

 

Figure 4.23 Resilient modulus versus degree of saturation for SC, CL, and ML soils
decrease with increasing degree of saturation. The difference between this observation
and the previous one regarding the sample water content is that the degree of saturation is
a function of both the water content of the sample and its dry unit weight. Said
functionality caused the coefficient of determination to increase from about 0.63 in
Figure 4.22 to about 0.88 in Figure 4.23. Therefore, the degree of saturation will be
included in the multivariate analyses.

Effect of Sample Dry Unit Weight - The effect of the test sample dry unit weight on the
MR values of SC, CL, and ML soils was studied by plotting the MR values as a fimction
of the sample dry unit weight as shown in Figure 4.24. As it was expected, the figure

shows a very weak correlation between the dry unit weight and the MR values of the test

sample. The main reason for the weak correlation is that the water contents of the test

122

samples vary from about 6.7 to 30.4 percent. Relative to the compaction curve, this
variation causes some test samples to be on the wet side while others on the dry side of

optimum. This implies that two compacted samples having the same dry unit weight may

 

 

OSC EICL AML

 

 

 

80,000
70,000 1

, MR = 0.0106( (1)
50,000 1 R2 = 0.0608
40,000 I '
30,000
20,000
10,000

 

 

1.
3.0244

 

)

Sl

 

(p
O\
O
O
O
O

 

 

 

 

1.

 

 

 

 

Resilient modulus

 

 

 

 

1

:_A __.__.0._.: i A _

9E

 

O . O . P . I

l f

100 105 110 115 120 125 130
Dry unit weight (pci)

O

 

 

 

-11- ' —-41-< -———1- >————-——>—_.—.—41>-»

 

D
1
1

O
\O
O
31

 

 

 

Figure 4.24 Resilient modulus versus dry unit weight for SC, CL, and ML soils
or may not have the same water content. Soil samples compacted dry of optimum would
have brittle behavior, higher strength, higher stiffness and higher MR values than samples
compacted wet of optimum. The latter would have lower strength, higher plastic
deformation and sofier behavior under loads. In addition, when soil samples are
compacted on the dry side of optimum, the soil particles tend to stay in a fiocculated
arrangement. Samples compacted wet of optimum; the particles tend to disperse (line up)
due to the extra water lubrication (Holtz and Kovacs 1981). Nevertheless, the dry unit
weight will be included in the multivariate analyses to determine if the dry unit weight

interacts with other variables to significantly affect the MR values.

123

Effect of Sample Grain Size — Because of high fine contents, all SC, CL, and ML soils
were subjected to wet sieving and hydrometer analyses to determine their grain size
distributions. The effects of sample gradation on MR values were assessed through the
following gradation parameters:
0 Percent passing sieves 200, 100, 40, 20, 10, 4, and 3/8 inch
0 The coefficients of curvature and uniformity
0 Average particle size at ten, thirty, and sixty percent passing (D10 , D30, and D60,
respectively)
The effects of each gradation parameter on the MR values were analyzed by plotting the
MR values of the soil samples as a function of that parameter. However, as it was
expected, none of the variables showed a good correlation to the MR values.
4.5.3.2 Multivariate Analysis
Multivariate analyses were conducted to study the combined effects of several
independent sample variables on the dependent variable MR of SC, CL, and ML soils.
During the analyses:
0 Various models were used in an attempt to maximize the value of the coefficient of
determination (R2).
0 Special care was taken to:
0 Ensure that the resulting equation satisfies the known trends between each of the
independent variable and the dependent variable MR.
0 Avoid any significant co-linearity between the independent variables.

0 Minimize the number of independent variables in the equation.

124

Results of the analyses yielded models having relatively high R2 values. However, none.
of the models produced a better correlation than the degree of saturation (S) alone.
Therefore, it is recommended that the MR values of SC, CL, and ML soils be predicted
using the degree of saturation (S) alone through Equation 4.7.

MR 2 650486 exp— 0.0501(S) Equation 4.7

The data used to develop Equation 4.7 have saturation values ranging from 43.2 to 99.9
percent with a corresponding range of dry unit weight from 92.5 to 128.5 pcf and
moisture content range from 5.7 to 30.4 percent. The use of Equation 4.7 outside those
ranges of values is not recommended.

Figure 4.25 shows the MR values plotted against the degree of saturation. The
figure consists of two sets of data. The first set (solid symbols) is the data that was used
to develop the model (Equation 4.7) with R2 value of about 0.88. The second set (open
symbols) is the data that was used to validate the developed model. The latter data were
obtained from testing three disturbed and three undisturbed soil samples. The sequential
procedure used to test the samples is presented in the next subsection.

One important point should be noted herein is that the data used to generate
Equation 4.7 were obtained from both disturbed and undisturbed samples. Hence, the
equation applies equally to both types of samples. The implication of this is that, the
elastic behavior of laboratory compacted samples is equivalent to that of undisturbed

samples provided both samples have the same degree of saturation.

125

[__‘A7 —i 777 #777 7 : : ,:,W: ,:,7WI7 #77" fi"
0 MiodICIiIEIVerificatiorfl

 

   

 

 

 

 
  

 

 

 

 

80’000 v_i 17* 1V 7 if 0 0501(3)
A 70,000 :1- : ——«~r—r = 04 '
E. 60 000 I 7 T :7 #k !‘ MR 6: 866
8 50000 1; ° 1 1 # 1:12:05???
E 40,000 t—__...\. 1 .1 1 —~*_
8 30,000 1‘51 ,1 1
E 20,000- A —1 7*:
8‘ 10,000 id—r-f—iavrr—D
b
0 W WWWW ,,__IWW WWW WW
40 50 60 70 80 90 100

 

 

 

Figure 4.25 Resilient modulus versus degree of saturation

4.5.3.3 Validation

As stated above, six additional soil samples (three disturbed samples and three
undisturbed soil samples) were tested to verify the model presented in Equation 4.7. The
three disturbed soil samples were allowed to dry from their natural contents of 14.4, 25.4,
and 21.9 percent to 10.3, 11.3 and 18.8 percent, respectively. Afier drying, the soils were
compacted using standard proctor and standard compaction mold. The compacted soil
was then extracted from the compaction mold, trimmed to the size of the test sample and
then subjected to cyclic load triaxial tests. When the test was terminated, the water
content and the dry unit weight of each sample were then measured and its degree of
saturation was calculated.

The natural water contents of the three undisturbed Shelby tube samples that were
tested for verification purposes were 12.3, 18.4 and 1 1.2 percent. One test sample was

extracted from each of the three Shelby tubes. Two test samples were subjected to cyclic

126

load triaxial tests at their natural water contents of 12.3 and 1 1.2 percent. The third
sample was dried at room temperature for two days then it was sealed in a plastic bag for
one week to even up the moisture in the sample and then it was tested. After the cyclic
load triaxial test was terminated the moisture content of each sample and its dry unit
weight were determined.

For the six validation samples, results of the cyclic load triaxial test data were
used to calculate the resilient modulus of the soil. The data are shown in Figure 4.25 by
the open symbols. After measuring the test sample water content and dry unit weight, the
data were used in Equation 4.7 to estimate the resilient modulus values of the soils.
Figure 4.26 shows the laboratory measured resilient modulus values plotted against the
MR values calculated using Equation 4.7 and the degree of saturation of the test samples.
It can be seen from Figure 4.26 that all six data points are located near the line of

equality. Hence, the six data points validate the accuracy of Equation 4.7.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

70,000 _7——' 1 ‘ 1
Q 60 000 1 1 1 1 1 1'
§ 1 1 1 1 9 1/1
s 1 ' ' 1 i
3 50,000 . . . f I
'8 L1ne of equality 1 1

|
E 40,000 I I 1 I
‘5 ‘ ,
:8 30,000 . ’ \ . 1 1
8 . 1 i
'5 20,000 I . . I .
81’ ‘ 1 1 1 1
53% 10,000 . 1 1 I I
. 1
O i 1 T1 T1 r r
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000
Laboratory measured resilient modulus (psi)

 

 

Figure 4.26 Laboratory measured and calculated MR values for SC, CL, and ML soils

127

 

4.5.4 Poorly graded sand -— silty sand (SP-SM)

Table 4.6 lists the locations, the USCS and the AASHTO soil classifications, and
the sample designation number of eight disturbed soil samples that were collected from
various clusters and areas throughout the State of Michigan. The commonality between
the eight samples is that all of them were classified in step 1 of the analyses as SP-SM
according to the USCS. SP-SM soils may contain anywhere between 5 and 12 percent
passing sieve number 200 materials and a plasticity index less than 4 (Holtz and Kovacs
1981). In step 2 of the analyses, univariate and multivariate analyses were conducted and
are discussed below.
4.5.4.1 Univariate Analysis
In the univariate analyses, the effects of each of several sample variables on the MR
values of SP-SM soils were studied. These sample variables include: the moisture
contents of the samples, the dry unit weight after compaction, and the grain sizes.
Discussion of the effects of each variable is presented below.

Table 4.6 Locations of SP-SM subgrade soils

 

 

 

 

 

 

 

 

 

Sample number Location AASHTO USCS
M-028-W (02-03) ~2000 feet East of M-35 A-3 SP-SM
U-002—E (03-01) 400 feet East of Hwy 13 A-3 SP-SM
M-028-W (03—03) 500 feet West of Basnau Rd A-2—4 SP-SM
I-196-N (06-05) 1 10 feet North of Schmuhl Rd A-3 SP-SM
I-069-N (10-01) 75 feet North of Base Line Hwy A-l-b SP-SM
I-069-N (1 1-01) 160 feet North of mile marker 42 A-3 SP-SM
I-094-W (12-03) 36 feet West of bridge after exit 135 A-3 SP-SM
U-023-N (13-07) 60 feet North of Sherman A-3 SP-SM

 

 

 

 

 

 

Effect of the Sample Moisture Contents — Because of the narrow range in fine contents,

it is expected that the water contents would have minimal effects on the elastic response

128

of the soil to the applied loads. Figure 4.27 shows the MR values of all SP-SM soil.
samples plotted against the samples moisture contents. The figure shows that the sample
moisture content has no effect on the MR values. This result was not expected and it
contradicts findings by many researchers including (Maher et. al 2000, George 2000, and
2003) who stated that the MR value decreases with increasing moisture contents. One
possible explanation of the above result is that the effects of moisture contents on the MR
values interact with other variables that are not included in the equation. This issue is

addressed in the multivariate analyses subsection.

 

 

35,000

 

__ _.. .4

30,000

 

fi——— _._1_. _

25,000 1

W —4+—-——— ____.I__.__ __ _ _____.,

 

._ WQWW.-- WW W

20,000 5 ;

 

 

15,000 I

 

10,000

Resilient modulus (psi)

_-L WW W_

 

 

 

 

‘ -' ‘_ "'11 ”k“

1 1
1 1
1 1

I

—o___‘ .—

 

0
0.0 2.0 4.0 6.0 8.0 10.0 12.0

Moisture content (%)

 

 

 

Figure 4.27 Resilient modulus versus moisture content for SP-SM soils
Effect of Sample Dry Unit Weight - The effect of the test sample dry unit weights on
the MR values of SP-SM soils was studied by plotting the MR values as a firnction of the
sample dry unit weights as shown in Figure 4.28. As it was expected, the figure shows a

weak correlation between the dry unit weights and the MR values of the test samples.

129

The main reason for the weak correlation is that it is possible for two test samples to have
the same dry unit weight but significantly different elastic behavior under load. This
scenario is certain if one sample was compacted dry of optimum and the second wet of
optimum. A sample compacted dry of optimum has higher strength and stiffness and
displays more brittle behavior than the one compacted wet of optimum. The latter would
have lower strength, higher plastic deformation and softer behavior under loads (Holtz
and Kovacs 1981). For the SP-SM soils, the water contents of the test samples varied
from about 2 to about 11 percent. Such range in the water content extends from the dry
side to the wet side of the optimum moisture content on the compaction curve. To
overcome the problem, the dry unit weight was included in the multivariate analyses to
determine whether or not it interacts with other variables to significantly affect the MR

values.

 

35,000

MR = 6299.86
-. 30,000 R2 = 0.074

0.0104(cd) I

51)

25,000

20,000

Resilient modulus (p

15,000

 

10,000
100.0 105.0 110.0 115.0 120.0 125.0

Dry unit weight (pcf)

 

 

 

Figure 4.28 Resilient modulus versus the dry unit weight of the test samples

130

Effect of Sample Grain Size — Because of low fine contents of SP-SM soils, all soil
samples were subjected to dry sieving to determine the grain size distribution shown in
Figure 4.29. Examination of the figure indicates that the SP-SM sub grade soil samples
contain variable amount of fine and coarse sands. Since the fine and coarse sand contents
are co-linear or dependent, both variables should not be included in the analyses. Hence,
the effect of either the fine or coarse sand contents on the MR values should be included
in the analyses. The effects of sample gradation on MR values were also assessed through
the following gradation parameters:

0 Percent passing sieves 200, 100, 40, 20, 10, 4 and 3/8 inch

0 The coefficients of curvature and uniformity

- Average particle size at ten, thirty and sixty percent passing (D10 , D30, and D60,

respectively)

 

100.0

80.0

60.0

40.0

20.0

Percent passing by weight

 

0.0
10.00 1.00 0.10 0.01

Grain size (mm)

 

 

 

Figure 4.29 Eight gradation curves of the SP-SM subgrade samples

131

The effects of each gradation parameter on the MR values were analyzed by graphing the
MR values of the soil samples as a function of that parameter. Few gradation parameters;
the percent passing sieve number 200, the percent fine sand content, and the coefficients
of uniformity and curvature showed minor correlation to MR values. For example, the
effect of the percent passing sieve number 200 and the MR values is shown in Figure
4.30. Although the data in the figure shows that the percent passing sieve number 200
has insignificant effects on the MR values, it also show that increasing the percent fine

materials (passing sieve number 200) causes decreases in the MR values.

 

35,000 _
33,000 MR = 40843e

31,000 R2=0.1415
29,000

27,000
25,000
23,000
21,000
19,000
17,000
15,000

 

-0.1026(P200)

Resilient modulus (psi)

 

5.5 6.0 6.5 7.0 7.5 8.0 8.5

Percent passing sieve number 200

 

 

 

Figure 4.30 Resilient modulus versus percent passing sieve number 200 for SP-SM soils
4.5.4.2 Multivariate Analysis
Multivariate analyses were conducted to study the combined effects of several
independent sample variables on the dependent variable MR of SP-SM soils. During the

analyses:

132

0 Various models were used in an attempt to maximize the value of the coefficient of
determination (R2).
0 Special care was taken to:
0 Ensure that the resulting equation satisfies the known trends between each of the
independent variable and the dependent variable MR.
0 Avoid any significant co-linearity between the independent variables.
0 Minimize the number of independent variables in the equation.
Results of the analyses yielded several different models with coefficient of determination
values from 0.62 to about 0.93. For example, the model shown in Figure 4.31 and

Equation 4.8 has a coefficient of determination of slightly higher than 0.93.

MR 21357.5(SVSP— SM)2 — 6145.3(SVSP— SM) + 23613

Equation 4.8

Where,

_ _ 73.95
SVSP SM _ 107 D3066XD12025XD§041XS0103)

SVSP—SM = Sample variable for the SP-SM soils

 

Yd = dry unit weight of the test sample (pcf)

D30, D10, D60 = the particle diameter at 30, 10 and 60 percent passing (m)

S = degree of saturation

133

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40,000 L I
MR = 1357.5(SVSP-SM)2 - 6145.3(SVSP-SM) + 23613 1
'78 2
3 R = 0.9313
32 30,000 _ 1 WI 1 ,
O 1 L Wt 1
8 1 ' . 1 1 1
_: 20,000 F . / 1 1 1
'8 1 0 1
ad 0 1 1 , 1 1
1 1 1 1 1
10,000 1 1 1 . 1 1 1
2.00 2.50 3.00 3.50 4.00 4.50 5.00 .5
SVSP-SM

 

6.00

 

Figure 4.31 Resilient modulus versus the sample variable model, SP-SM soils

The model was not accepted although the value of the coefficient of determination

is relatively high. Three reasons can be cited for rejection; first the model has too many

variable for a sample size of 8, second, the model interpretation (higher particle size yield

lower modulus) cannot be physically supported, and third, the three particle sizes in the

denominator are not truly independent variable, which makes the model very sensitive to

small changes in the values of the variables.

Several other models were also rejected for similar reasons. Finally the model

presented in Figure 4.32 and expressed in Equation 4.9 was accepted although the value

of the coefficient of determination is moderate (0.74).

MR = 1749.6exp0.0054(SVSP - SM)

Where,

134

Equation 4.9

 

1.75

Yd
SVSP - SM =
MCO'S + LL06 + (P40 _ P200 )001

 

Yd = dry unit weight (pcf)
LL = liquid limit
MC = moisture content (%)

P40, P 200 = percent passing sieves number 40 and 200, respectively

The term (P40 _ P200) expresses the percent fine sand content in the soils.

Examination of the data in Figure 4.32 and Equation 4.9 indicates that increases in the
values of the dry unit weight cause increases in the MR values whereas increases in the
values of either the moisture content, liquid limit, or the percent fine sand content cause
decreases in the resilient modulus values. .

It is important to note that the data used in support of Equation 4.9 includes soil
samples having dry unit weights range from 105.5 to 121.6 pcf, percent passing sieve
number 40 from 43.2 to 92.6, percent passing sieve number 200 from 5.7 to 8, liquid
limit from 13 to 21 and moisture content from 2.0 to 11.4 percent.

4.5.5 Clayey sand — silty sand (SC-SM)

Table 4.7 lists the locations, the USCS and the AASHTO soil classifications, and
the sample designation number of seven disturbed soil samples that were collected fiom
various clusters and areas throughout the State of Michigan. The commonality between

the seven samples is that all of them were classified in step 1 of the analyses as SC-SM

135

 

 

 

 

 

 

 

 

 

 

35,000 I
MR = 1 7 4 9. 6e0.0054(svs1>-SM) I
.,, 30,000 -1 *
3 R2=0.7431 1
5 SE: 1414 /
:s 25 000 —~
'0 ’ I '
g I 1. 1 ./. .
5 20,000 1 . 1 1
’5'} ~I,/I .1 I
E . 1 1
15,000 . 1 . ~
10,000 1 . 1 1

 

 

 

400.00 420.00 440.00 460.00 480.00 500.00

SVSP-SM

520.00

540.00

 

Figure 4.32 Resilient modulus versus SVSP-SM soils

according to the USCS. SC-SM soils may contain anywhere between 12 and 49.9

percent passing sieve number 200 and a plasticity index between 4 and 7 (Holtz and

Kovacs 1981 ). Hence, the fine materials may play a major role in the mechanistic

behavior of the soil. In step 2 of the analyses, univariate and multivariate analyses were

conducted and are discussed below.

Table 4.7 Locations of SC-SM subgrade soils

 

 

 

 

 

 

 

 

 

 

 

 

Sample number Location AASHTO USCS
M-068-W (04-02) 180 feet West of US-23 A-2-4 SC-SM
M-032-W (04-05) 220 feet East of Herron Rd A-4 SC-SM

I—075-N (08-06) 80 feet North of bridge after exit 195 A-2-4 SC-SM

I-096-W (09-02) 141 feet West of Morse Lake Ave A-2-4 SC-SM
M-060-W (11_03) 135 feet West of Southbound 1—69 A-2-4 S C-SM

overpass

I-069-S (1 1—05) 95 feet South of Bridge alter exit 10 A-4 SC-SM

I-094-W (12-01) 95 feet West of 29 Mile Rd A-2—4 SC-SM

 

136

 

 

4.5.5.1 Univariate Analysis

In the univariate analyses, the effects of each of several sample variables on the
MR values of SC—SM soils were studied. These sample variables include: the moisture
contents of the samples, the degrees of saturation, the liquid limits, and the grain sizes.
The effects of each variable are presented and discussed below.
Effect of the Sample Moisture Contents — Because of the high range in fine contents,
water content may play a significant role in determining the elastic response of the soil to
the applied loads. Figure 4.33 shows the MR values of all seven SC-SM soil samples
plotted against the samples moisture contents. As it was expected, the figure shows
increases in the sample moisture contents cause decreases in the MR values. Similar
results were also reported by many researchers including (Maher et. a1 2000, George
2000, and 2003). The low value of the coefficient of determination of about 0.25 implies
that MR values cannot be explained accurately by the moisture content alone. Therefore,
it will be included with other sample variables in the multivariate analyses to determine
their combined effect.
Effect of the Degree of Saturation — At the conclusion of the cyclic load test, the
sample moisture content and dry unit weight were determined and the degree of
saturation (S) was calculated using Equation 4.4 assuming a specific gravity of the solids

of2.7.

G, *(MC/100)* y, *100

S :
* _ Equation 4.4
Gs 7w 7d

 

137

 

33,000
31,000
29,000
27,000
25,000
23,000
21,000
19,000
17,000
15,000

Resilient modulus (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e 1 .. MR = 2855860113 650%)
W R2 = 0.2458
1‘ 1 Q ’ .
\ I
T ‘
’ 1
. \
1 1 .9 . 1
1 1 1
-_ 1 . 1 0
0 2 4 6 8 10

Moisture content (%)

 

Figure 4.33 Resilient modulus versus moisture content for SC-SM soils

Where,

S = degree of saturation (%)

MC = moisture content (%)

GS = specific gravity of the soil solid

"Yd = dry unit weight of the sample (pcf)

'Yw = unit weight of water = 62.4 pcf

Figure 4.34 shows the MR values plotted against the degree of saturation of the test

samples. The data in the figure indicates an insignificant correlation between the MR

values and the degree of saturation although increasing degree of saturation causes

decreases in the MR values as it was reported by Maher et al. (2000). When the data and

the values of the coefficients of determination of Figure 4.33 are compared to those in

Figure 4.34, it becomes clear that the effects of moisture contents on the MR values can

138

 

be better expressed using the water content. The reason is that the water content is an
independent variable whereas the degree of saturation is a function of both the water
contents and the dry unit weights. Figure 4.35 depicts the MR values plotted against the
dry unit weights of the test samples. It can be seen that correlation between them is
insignificant although, as it was expected, increasing the values of the dry unit weight
cause increases in the MR values. The reason for the insignificant correlation is that the
dry unit weight of the test sample is a function of its water content. Further, it is possible
for two test samples having the same dry unit weight to have drastically different
mechanistic behavior. The scenario is possible provided that the two samples were
compacted at two different water contents on the opposite sides of the optimum moisture
content (Holtz and Kovacs 1981).

Effect of the Sample Liquid Limit - For each SC-SM soil sample, the Atterberg limits
for all materials passing sieve number 40 were determined in order to classify the type of
fine materials (silt or clay). The test results indicate that for all samples the liquid limit
varied from 15 to 22, the plastic limit from 10 to 15 and the plasticity index from 4 to 7.
The effects of the liquid limit on the MR values of the soils were analyzed. Figure 4.36
depicts the influence of the liquid limits on the MR values of SC-SM soils. The data in
the figure indicate that the MR values of SC-SM soils decrease as the liquid limit of the
material passing sieve number 40 increases. This observation was expected and has been
reported by many researchers for various soil types including silty and clayey sands, silt,
and clay (Gudishala 2004). Therefore, the liquid limit will be included in the multivariate

analyses.

139

 

 

Resilient modulus (psi)

33,000
31,000
29,000
27,000
25,000
23,000
21,000
19,000
17,000
15,000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1 .1 _
e . 1 MR = 25235e 0'0023(S)

1 1 R2 = 0.0863
1 1 9 ° 1
1 1 1

\ . 1

. 1
1
1 'r
1 1 1 1
1 1 1 1 o

0 20 40 6O 80 100

Saturation (%)

 

Figure 4.34 Resilient modulus versus saturation for SC-SM soils

 

 

Resilient modulus (psi)

33,000
31,000

29,000 ‘

27,000
25,000
23,000
21 ,000
19,000
17,000
15,000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4 V . . 1 1 1
8MR=14042e0'004(+d) e 1 1 1
R2=0.0407 1 1
1 1 1 1 «3 . 1
1 1 , .
1 1 1
_J 1
Q. 1 1 1 1
1 1
1 1 1 1 1
1 1 1
1 1 1 1 ° 1
104 109 114 1 124 129 134

Dry unit weight (pcf)

 

Figure 4.35 Resilient modulus versus the dry unit weight for SC-SM soils

140

 

 

 

33,000 1 1 x
31,000 1 ' . MR = 462626
29,000 1 1 R2 = 0.1467
27,000 r
25,000
23,000
21,000
19,000
17,000 . ,
15,000 - a 1 . 1 1

15 16 17 18 19 20 21 22 23

Liquid limit (%)

-0.0402(LL)

Resilient modulus (psi)

 

 

 

 

 

Figure 4.36 Resilient modulus versus liquid limit for SC-SM soils

Effect of Sample Grain Size — Because of the relatively high fine contents, all SC-SM
soils were subjected to wet sieving and hydrometer analyses to determine their grain size
distribution. The effects of sample gradation on the MR values were assessed through
the following gradation parameters:
0 Percent passing sieves 200, 100, 40, 20, 10, 4, and 3/8 inch
0 The coefficients of curvature and uniformity
0 The average particle size at ten, thirty, and sixty percent passing (D10 , D30, and D60,

respectively)

The effects of each gradation parameter on the MR values were analyzed by
plotting the MR values of the soil samples as a function of that parameter. All gradation
parameters showed minor correlations to the MR values. Despite this, several attempts
were made to include gradation variables in the multivariate analyses as presented in the

next subsection.

141

4.5.5.2 Multivariate Analyses

Multivariate analyses were conducted to study the combined effects of several
independent sample variables on the dependent variable MR of SC-SM soils. Diu'ing the
analyses:

0 Various models were used in an attempt to maximize the value of the coefficient of
determination (R2).
0 Special care was taken to:
0 Ensure that the resulting equation satisfies the known trends between each of the
independent variable and the dependent variable MR.
0 Avoid any significant co-linearity between the independent variables.
0 Minimize the number of independent variables in the equation.

Results of the analyses yielded a model having a relatively high R2 value. The
model is based on three sample variables of the SC-SM (SVSC-SM) soils (the water
content, liquid limit, and the coefficient of uniformity) as stated in Equation 4.10. Figure
4.37 depicts the resilient modulus values plotted as a function of the SVSC-SM. It can be

seen from the figure that higher SVSC-SM values yield lower MR values.

MR =39638exp—0.0037(SVSC-SM) Equation 4.10

Where,

SVSC " SM : C32 * (LLLIS + MC”)

LL = liquid limit (%)

MC = degree of saturation (%)

Cu = coefficient of uniformity

142

The liquid limit and coefficient of uniformity are constant for a given soil.

Therefore, lower SVSC-SM can be obtained by decreasing the moisture content of the

test sample. As the moisture content of the sample increases, the lubrication between the

sand particles increases, thus reducing the MR values (Perloff and Baron 1976). The

term (LL1 '15 + MC1'3) can be thought of as the moisture index of the SC-SM subgrade

soils.

 

Resilient modulus (ps:

 

33,000
31,000 . , MR = 396386

R2 = 0.9284
SE = 2042

-0.0037(SVSC-SM)

29,000
27,000
25,000
23,000
21,000
19,000
17,000
15,000

 

50 100 150 200 250
SVSC-SM

 

Figure 4.37 Resilient modulus versus SVSC-SM soils

It is important to note that the data used in support of Equation 4.10 includes soil

samples having liquid limits values ranging from 15 to 22, coefficients of uniformity

between 14.74 and 270, and moisture contents from 1.2 to 9.2 percent. The use of the

equation outside these ranges is not recommended.

143

 

4.5.6 Gravely Sand (SG)

Table 4.8 lists the locations, the USCS and the AASHTO soil classifications, and
the sample designation number of two disturbed soil samples that were collected from
various clusters and areas throughout the State of Michigan. The two samples were
classified in step 1 of the analyses as SG. However, due to the large gravel particles
cyclic load tests could not be performed on these samples. The AASHTO standard test
procedure that was used requires that the diameter of the sample be at least four times
larger than the largest particle. For this to be satisfied the sample diameter would have to
about 9 inches. Because of the limited number of samples and because this type of
subgrade materials is very much limited to small areas in the State of Michigan, no

further analyses were conducted on the two samples.

Table 4.8 Locations of SP-SM subgrade soils

 

 

 

Sample number Location AASHTO USCS
M-O33-S (05-05) 750 feet South of Peters Rd A-l-a SG
U-012-E (12-07) 120 feet West of Person Hwy A-l-a SG

 

 

 

 

 

 

144

4.6 Climatic Damage Models

The State of Michigan is located in the AASHTO wet-freeze region. The average
annual rainfall and snowfall in the State varies from one location to another. In the
Lansing area, the average annual rainfall is about 32-inch and the average annual snow
fall is about 56-inch. Further, the frost depth varies from about 7-feet in the Upper
Peninsula to about 3-feet in the Lower Peninsula. These climatic data affect the behavior
of the paving materials and roadbed soils. Because of the variability of the climatic
conditions, the resilient modulus of any given soil is dynamic in nature and it changes
seasonally with changing water content and below and above freezing temperatures.

For most pavement structure, the subgrade material is the weakest part of the
structure. The 1993 AASHTO Design Guide calls for protecting the subgrade soils under
flexible pavements by providing adequate structural number. For rigid pavement, the
modulus of subgrade reaction is typically integrated with that of the base layer to yield a
composite modulus of subgrade reaction. In addition, the 1993 AASHTO Design Guide
includes damage assessment scenario that must be used to obtain the design modulus of
the roadbed soils. Finally, the Guide does not provide pavement layer thicknesses based
on fi'ost penetration. That is the guide does not provide any protection to the roadbed soils
against freezing. The new M-EPDG also includes climatic model to reduce the given
subgrade modulus and hence, account for the seasonal damage.

For any pavement structure, the amount of water in the subgrade materials is a
function of many variables including:

0 The duration and frequency of the rainfall

o The geometry of the pavement and the flow regime of the adjacent ground

145

o The conditions and permeability of the surface layers and the number of unsealed
cracks.

0 The permeability of the subgrade materials

0 The type of drainage system installed in the pavement structure

0 The initial crown of the subgrade soils created during rolling

0 The elevation of the ground water table

Based on the above list, a given subgrade soil located in the State of Michigan could
be saturated during spring thaw season, during extended periods of rainfall (more than
few hours), or during frequent rainfall. Based on this scenario, one can conclude that the
subgrade soil is highly likely saturated during spring thaw, and it could be occasionally
saturated during the summer and fall seasons. The subgrade soil is likely frozen during
the late winter season.

Given the variability of the degree of saturation of the subgrade soil during the
pavement service life and the corresponding variation in its resilient modulus, the
question become what subgrade resilient modulus should be used in the design of new
pavement or the rehabilitation of existing pavements. The answer depends on the type of
damage model used in the analysis.

In general, two types of damage models are available; empirical and mechanistic-
empirical. An example of the former is the 1993 AASHTO Design Guide. Examples of
the latter is the fatigue and rut models in the MICHPAVE computer program, in VESYS
computer program, in the Asphalt Institute pavement design, and in almost every other
mechanistic-empirical computer program and the 1993 AASHTO Design Guide damage

model presented in Equation 4.10

146

Uf 21.18 *108 * MI{"2'32 Equation 4.10

Where

Uf = damage factor
MR = resilient modulus of the subgrade soils

In this study, the reduction factor to reduce the resilient modulus value from its

summer and fall season values was obtained using the following procedure:

For each of the eight soil types except the SP, the proper correlation equation was
used to calculate the resilient modulus at water content near the optimum water
content of a standard proctor compaction. The correlation equation was also used to
calculate the resilient modulus at a water content corresponding to near soil saturation
(degree of saturation between 93 and 99 percent).

For the SP soils, the reduction factor between the resilient modulus at the optimum
water content and that near saturation was obtained from Holtz and Kovacs (1981).
The damage factor of each of the eight soil types was calculated as the ratio of the
two modulus values of the first two steps. The average damage factor of all eight soil
types was then calculated as shown in Table 4.9.

A standard pavement cross-section of 12 inch subbase, 6 inch base and 7 inch asphalt
layer was used to calculate the service life of the pavement in terms of ESALs using
the MICHPAVE computer program. The same pavement cross-section was analyzed
twice, once for subgrade modulus value of 16,000 psi (an average modulus near the
optimum water content) and once for an average modulus value near saturation (4000

psi). Results of the analysis yielded two expected pavement lives in term of ESALs.

147

The ratio of the expected ESAL at the high modulus was then divided by the ratio of
ESAL at the low modulus. The results yielded a damage factor of about 8.

0 Similar procedure was used in the 1993 AASHTO design guide, the damage factor
was about 7.

0 The average damage factors obtained from the 1993 AASHTO design procedure, the
MICHPAVE computer program and the correlation equations was then calculated and
is listed in Table 4.9.

0 For each soil type, the average resilient modulus of all test samples was then
calculated and listed in Table 4.10.

o The recommended subgrade design resilient modulus values for all eight soil types
are then calculated by dividing the average resilient modulus per soil type by the
average damage factor listed in Table 4.9. The values of the recommended design
resilient modulus are listed in Table 4.10.

It should be noted that the recommended design resilient modulus values listed in Table

4.10 are to be used in the M-EPDG design level III. For design level II, the correlation

equations should be used. And for design level I, FWD tests should be conducted and the

design modulus value backealculated and adjusted using the damage factor. The FWD

procedure is not a part of this thesis. It is addressed in the final report to MDOT.

148

 

 

 

 

 

 

 

 

oonfi $0.3 Em
mmm€ wmmdm Em
movd 53.: Em
_No.m SEN Emdm
mmm.m www.mm $5-0m
_vw.m nwwdm 1:2 JD .0m
5 02 m> 52 ms: 06
M252w2m—00 A. oowfihhmx mUmD

 

 

 

 

m BK: 2wmm0o Dam m-E 20.2 329» 2220022 “20:68 amen 2.2 2an

 

09$ 20m 02928 05 .20 305 028.20 323200 0222022 2208502 223 00002m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

n w n n 2020012 0962200 0wm20><
6 83. $0.6 202% 2 cm 2 New
8 23 won: w Sam 2 cm : Em
<2 <2 <2 m comm 84H: 0 :an mm om N. $6 :2
m ooo \t ooo Z ooo mm on 2 w $6-2m
m oood ooo.m_ 03.2 mm 2 m.w Emém
w omvé onwfi mvmdm om mm .2 1:2 JD .Um
22222220
20200.2 m><mIUHE Ohmm<< 22022003 oonw 3-3. no-3 23m 0220002 282
202005320
09230 («0 0 “28 2028228 wUmD
owfio>< 208$ 0w02§Q 02 w2€202m0200 3% 228200 2026.3
33 22200.2 2202602

 

 

20222200 2080.2 0wwEwQ 0.2 033.

149

CHAPTER 5

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

5.1 Summary

The resilient modulus of the roadbed soil plays a crucial role in the design of
pavement systems. Currently, MDOT uses different procedures to determine the MR.
Most of these procedures are applicable to M-E PDG level 3 design. Therefore, a
consistent, uniform and implementable procedure that meet the requirements of M-E
PDG design Levels 1, 2, and 3. To do this, the State of Michigan was divided into fifteen
clusters where the physical and engineering characteristics of the soil are similar. The
clusters were then divided into 99 areas to narrow down the ranges of the engineering and
physical characteristics of the soils. Disturbed subgrade soil samples were collected from
seventy five areas. At some locations where the disturbed samples were obtained,
pocket penetrometer and vane shear tester were used to determine the penetration and
shearing resistances of the soils, respectively.

The undisturbed soil samples were then tested to determine their moisture
contents, grain size distributions, Atterberg limits (when applicable) and resilient
modulus using cyclic load triaxial tests. Correlation equations (see Table 5.1) were then
developed to estimate the MR values of the roadbed soil based on the results of the
moisture content, Atterberg limits, dry unit weight, and grain size distribution data.

5.2 Conclusions
Based on the field and laboratory investigations and the data analyses, the

following conclusions were drawn:

150

. Michigan Department of Transportation’s (MDOT) current procedure for determining

the MR values is not consistent between the regions.

. Grain size distribution on soils with more than 10 percent passing sieve number 200

should be performed by wet sieving.

. The subgrade soils in Michigan were classified based on the USCS and the AASHTO

soil classification systems.

. In most cases, subgrade soils having similar elastic behavior under loads tend to fall

within one classification (designation) of the USCS but within several AASHTO
classification groups. Hence, based on the elastic behavior of the subgrade soils in
Michigan, the USCS produces much better soil grouping than the AASHTO
classification system.

. Most of the subgrade soils in the State of Michigan can be divided into the following
eight soil types:

0 Poorly graded sand (SP)

0 Silty sand (SM)

0 Clayey sand (SC)

0 Poorly graded sand —- silty sand (SP-SM)

0 Clayey sand — silty sand (SC-SM)

0 Gravelly sand (SG)

0 Low plasticity clay (CL)

0 Low plasticity silt (ML).

. For SC and CL subgrade soils the resilient modulus values obtained from testing

undisturbed Shelby tube samples were compatible to those from disturbed samples.

151

10.

11.

5.3

Good correlation equations relating the resilient modulus to some of the physical
characteristics of the soils were developed and are summarized in Table 5.1.
The correlation equations listed in Table 5.1. satisfy the M-E PDG requirements for
level 2 design.
An average damage factor was calculated using the correlation equations, the
MICHPAVE computer program and the 1993 AASHTO design procedure. The
damage factor accounts for the seasonal effects on the resilient modulus values of the
subgrade soils in the State of Michigan.
Average design resilient modulus values for the eight soil types were developed and
are listed in Table 4.10 and presented in Figures 5.1 through 5.4.
The design resilient modulus values in Figures 5.1 through 5.4 satisfy the M-E PDG
requirements for level 111 design.

Recommendations

Based on the results and conclusions of this study, it is strongly recommended

that:

The subgrade soils in the nineteen areas where neither disturbed nor undisturbed soil
samples were obtained be subjected to the full testing schemes presented in Chapter 3
of this thesis.

Nondestructive deflection tests be conducted at the network level as to fill the gaps in
the existing FWD data. Ideally, the tests should be spaced at 500 feet interval.

The deflection data collected at the network level be used to backealculate the
resilient modulus of the various pavement layers to meet the requirements of M-E

PDG level 1 design.

152

MDOT to implement the findings of this study by adopting the data presented in
Figures 5.1 through 5.4 for M-E PDG level III design.

MDOT uses the correlation models presented in Table 5.1 for M-E PDG level II
design.

The seasonal damage factors developed in this study be implemented by all regions

when using the 1993 AASHTO design guide.

153

 

22220.22: .20 22020222000 0 :0 .2022 2.3 n .583 .20 EH03 222 n 3». K.N u 020m 022 .20 2222» 2.20on n 60 “2.28200 0.222022 u
02 J22: 0220: H 15 Ago 2022228 0 m noom 02m 62 .2 208222 m0>2m mammmm 2200202 H 8mm .82 .42 $03 EwBB “.22 2.6 H 20>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.. 2m - 0m>m 680.0
A202 + and * 20 M 2m -Um>m A: mammoow n 22 a m 22.8
. A22 I and + EUQ + 202
so 2. M 2m - 2m>m 2m- 252832693: 0 22 m a zmam
Q..—
Q 1 6K * 4b
. n 5882 198 23mm 0 22 mm 2 . .22
OOHuw 2R*AOO~\DV<V* G M4 AUUm
2.202 + 25 u 22 82223012226824 1 22
. W 2 : 2m
2 o .
.2 n 2m>m m32296208 o u 22
mo
E l .2 m Em
.22 ._. .22 1 N22 .....2N2m>moma~m.o n 22
ov v
. A _ 2 1 . no
mm o m _ . l
j1 0 2622.22 22222522 mm 1 22 w 6 :2
w:
2022.23 03223 20222200 02820022 mmom2onfimzo mom:

 

 

 

 

 

023 20m 2000 20.2 202223 0322002220 22222225 mm 032%

154

«252282 BED 06 .20 202.202 228mm? 05 20.2 839» 9220022 2w2m00 m .052 $02 WE 0022022220002 2m 32me

 

       

28 mm? 1 22 .Em - I
28 285 u 22 .2m2m 1 5
24mm 1 22 .22 1 fl

28 m8.“ 1 2 .2m 1 D
28 3mm 122 .20 - m

 

 

 

155

Esmfiaom 225.3 05 20 28:8 5323 2: 52 2029 23258 @622 m _o>2 Cam m-E vowcoEEooom N.m Esmi

 

       

    
  

V 1 .
.
£2235? , gull...
Her. .51.” 0......

xFIIU'I.
.1. 1...\$..1.....

 

an 83 u 22 2% 1 E,

an :22 u 22 .8 1 B
28 23 u 22 2% 1 I
an 83 u 22 .2mom 1 I
an a? n 22 .2m2m 1 a
an Sta 1 22 .22 1 D

    

 

.r"

 
 

 

 

 

156

32552522 203012 222.20 newton 52.2.8: 2: 20.2 mos—22> 222608 :3va m 2052 Cam m12 82222088832 m.m Emmi

 

 

 

 

 

 

   

 

 

 

   

 

 

 

 

1.111 . t 3.. «t ..(Ilfl
U +.. 1 17.1 ......) 15.1.11... 1 1 _ .1. .1 . 1. 1
221.111.111.112” .. ._ 2 1 «w a 22,. .
n .3.. is 2 ice. 2 1 J1? . 2 W” 11mm11HVWM1|11hm .qo IE
..11 22.. I. .. . .15... 91 10.1 .. . . . x
B 1w 1...: _. 1 222.282.2121 111222.12 1E
«.6 .....111... .2. r1 .1 1 1 ...n .0. «mm .1 21mm :wwwmé 21.0%!
.... 19¢ (:4 1.11.13...“ 1. 1 1.911 2... O 11 . 1 1.. 1
111.1 3 21.1 11211.2 . 1:. . «1 ”.2. 1 12222221221221 21.122.21.11
_ 131.1 1 11.11 1:22.... .4 1.22.22.12.21... 1?: 1. . 22.22222 2m...om I
1 .1 1.111.111.1211 1 1 12222222112112 2222 ...-
1 1 .111... 1...... 1.1.2 2w 1 1 .
222.2,. .11. J11. .1. .11. .., 222.212 1
.1 3. 12¢. .12.. .11 1.
u .111..fl . O
¢1 1
1 .1. .22
.....L .9 .1
3:
9+2;
1.. 1
.H o _.
2. .194 .w‘
2. 1 o
J 1 1... 3:
1 11.11.11... :3
x 2. . 2
141121!!! 13111111114111: Prim
>3
a:
111.51%... .2. (WM 112N151: .vnu
{V 121W!
Ill-4.!) ._
n. .....h I‘ll. 1 11.

 

157

2255522 203012
2:120 28an 2:222:38 on... 120.2 8323 9222608 22.222.22me m 3222 00m m1: 2202222088832 226 2322i

2222 22.2 1 m2 .222m 1

2222 22.2.2 1 222 .5 1m 2222 2812 um: 12:22 1 a.
“M: 2 I H l

158

«2.230 1:
1 nu:

 

APPENDIX A

Soil Classification Systems

159

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cam fivfimmE 222212822222
228 232230 £20m bum E28 @2222 20>me 232020 8 bum w 2.2 _ 2253:2228 2amo¢2cw2m
. . . . 051.2 325892212 .
mo 825 :2sz
228m
EE 58 SE 288 28222 SE 2 ES 2 288 525 m2 28:20 28:2 28222222355923
2 Z : 02 02 . . O2 3 o . . .
58 ES 23:2 £8 28222 SE :2 288 SE :2 28:2 - 1 1 22823522212
:2 :2 ov :2 ow . ov . ow . . . .
s2. 12222 222222
mmvd mammsm 22282.2
mo 85221280822220
EE ES 58 SE 2222: 5.18 288 288 288 288 238 2228 80m .36 SE de
cm mm 0m 0m cm mm mm mm mm 3 mm 2
1 1 - - - 1 1 1 1 ES 2228 22228 A .o 2.288 .
2m om om ov 2:. o
288
- - - - - - - - - - - 22221227222222.2812
om
”mafia
:28qu £632.25“. 952m
0 m
I I I I I I I I I I “I I
411222 1n1< o1< m1< v1< m m < _ c N < 1 m ~< 7v m< m-< 9. 2 < 2222 2222821222236 @2280
211.22 N1< 21.222
2.22222 22.8.0 22222222 2x222 .
22222: 8025 2222128228 22201226 CS: 3.0 0 mafia $2 8 Xmmv 22223228 $322.80 252282122820 32250
A502 mom>oM E222 N231: 8822.222 222282122320 :8 093m: 2.< 03221—1

160

 

 

2:

2222222 222222212
222

ow

om

_

 

 

 

 

 

 

 

 

v1m1<

 

 

 

 

 

 

 

 

 

 

_
m1m1<
1111 1 11 1| l I m1< 1|.
1- 1 I | 1 11 1 1| 1.1141112111111111.
1 11 [111 1 1 11 11 h1m1< \
4

 

 

 

 

v1<

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1111111111 2-322.111
02-1221: 11 11 11 1111
/ 1 1 11211 11!- 1111 11.1 1

1111‘ i 1 1-1421 IL [I '1 21‘1 111r|11

 

 

 

 

 

 

 

 

 

 

 

 

1
O
v—4

1
O
N

1
O
m

I
O
V

1
O
ln

1
O
\O

 

 

O
[\

XQPU! mouseld

 

 

2222 22252 22222 N235 22222222 2222222 2222222222222 Emmi. 2.22 222222222

161

 

 

ma: 2%:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

xapu; Kinsmen

 

 

ASE mom>oM Ea Nzomv tuna 36:me m.ow§&mmmo N4 053m

162

 

 

 

 

 

 

 

 

m 2 um 823550 .3 2 :52 EN
35% 2 2 32»:on 02m 2 2 32%on 21 d v 8 w 2 a. o N no .m 2 5:2 .2 a
1.23pm 2 :8 as .m v 8 1.0mém 2 :8 2: .m v 8 .2 2 2 023528 “B 2 522 am
vgoADommEEi: vgoAsomeBJQ .1amvoowSvmaodmcofirmé
20-20 2 2 023550 8-20 2 2 025220
20-30 2 =8 05 .m v 8 60-30 2 =8 05 “w v 8 2 522 2m 22:2 u 5:2 2m
w 2 a v N no no 2 5:2 t 2 w 2 a. v m :0 d 2 $22 .2 .2 2% 252982 .822 330 2:22

 

 

 

 

 
 

__, __,

+

a grout Em A 2828an >30 fl 8.», a a
8% 3% corona Em A 20:08.2 >30 fl _

 

 

 

 

 

_ $2 a 2m 523$ _ gm v $2 A 2;

 

 

 

 

v“ cowomt Em + >20 2 A

 

 

 

 

_1 m 2 852 «L 45% _ a aonomb 23m A 28:08.2 E>8m fl TE 0 m2 532 “L 42620

_ 3589388 mgom _

 

 

 

 

8:32-25 2 :8 TI_ 02 I EadeAmosEan 2.8 8< T'— 8» _

A52 moa>oM 98 £05 :oumoEmmflo :8 852m 358 mUmD m.< 83mE

163

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Em 20 >20 Em
$2 m2 :8 2: 0:920 20 $6 8.2 m2
3 m2 532 EN :8 05 ME 3 532
2223.22 33 2m 222522 22:
oz 1 a we.»
o\oomA 15 3 A
_ 2 2 $22 1 .25 _ ~ 0 2 5:2 i .220 _
r _ O 2 532 1 £8 2590
a . E532
cocofim . o . A Tag m a
22m A 822E 220 2 a .o \omn o v E H. 2;.
_ wofifimécm m2 mom _
vofifimémhwoo 2 com a ES 2106 v 323th mo Ram 2< I 8% _

 

 

:we 8262 23 £2: 8282220 :8 822m 26 mom: 5‘ 2:22

Table A2 Possible AASHTO soil classifications per USCS group (Holtz and Kovacs

 

 

 

 

 

 

 

 

 

 

 

1981)
EEOC; Possible AASHTO classification
SP A-3, A-l-b, A-l-a, A-2—4, A-2-5, A-2-6, A-2-7
SG A-3, A-l-b, A-l-a, A-2-4, A-2-5, A-2-6, A-2-8
SM A-l-b, A-2-4, A-2-5, A-2-7, A-2-6, A-4, A-S, A-6, A-7-5, A-7-6
SC A-2—6, A-2—7, A-2-4, A-2—6, A-4, A-7-6, A-7-5
CL A-6, A-7-6, A-4
ML A-4, A-5, A-6, A-7-5
SP-SM A-3, A-l-a, A-l -b, A-2-4, A-2—4, A-2-6, A-2-7
SC-SM A-4, A-S, A-2-4, A-2-5, A-l-a, A-l-b

 

Table A3 Possible USCS classification per AASHTO group (Holtz and Kovacs 1981)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AASHTO group Possible USCS classification
A-l-a GW, GP, SW, SP, GM, SM
A-l-b SW, SP, GM, SM, GP
A-3 SP, SW, GP
A-2-4 GM, SM, GC, SC, GW, GP, SW, SP
A-2-5 GM, SM, GW, GP, SW, SP
A-2-6 GC, SC, GM, SM, GW, GP, SW, SP
A-2-7 GM, GC, SM, SC, GW, GP, SW, SP
A-4 ML, OL, CL, SM, SC, GM, GC
A-5 OH, MH, ML, OL, SM, GM
A-6 CL, ML, OL, SC, GC, GM, SM
A-7-5 OH, MH, ML, OL, CH, GM, SM, GC, SC
A-7—6 CH, CL, ML, OL, SC, OH, MH, GC, GM, SM

 

 

165

 

 

APPENDIX B

Laboratory resilient modulus results

166

Table 8.1 Laboratory resilient modulus results

 

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soil Type ‘ 10 15
Sample number “61:3,; Average Average Average Avémge MR Average Average Avgage Average MR
cyclic load deformation resilient (p51? atslggd cyclic deformati re5111ent (p51) at load
AASHTO USCS (lbs) (mils) modulus (psi) 80502:: 100’0 load (lbs) on (mils) magi“ Sggciifisggo
100 31.6 2.304 35,043 49.0 3.740 31,266
200 32.1 2.202 36,823 50.3 3.774 31,862
M—045-S (01-01) A—6 CL 500 32.2 2.262 36,639 36,543 50.1 3.663 31,747 31,503
800 32.5 2.205 37,056 50.1 3.817 31,297
1000 32.8 2.227 35,934 50.4 3.872 31,465
100 32.5 3.729 13,894 50.3 5.850 12,872
200 32.9 3.592 14,285 50.1 5.727 13,150
U-002—E (02—01) A—4 SM 500 32.7 3.442 15,044 15,352 50.4 5.551 13,686 13,818
800 32.7 3.325 15,708 50.4 5.496 13,826
1000 33.3 3.415 15,305 49.9 5.364 13,942
100 32.0 1.741 48,422 50.7 2.777 45,310
200 32.5 1.650 50,092 51.0 2,801 44,090
M—028-W (02-02) A—4 ML 500 32.7 1.569 53,892 53,824 51.3 2.969 42,510 41,516
800 32.7 1.600 53,350 51.3 3.047 41,331
1000 33.0 1.598 54,230 51.3 3.087 40,707
100 33.9 2.675 19,996 51.4 4.042 16,997
200 33.8 2.698 20,013 51.4 3.956 16,510
M—028—W (02—03) A—l —b :1}; 500 33.7 2.821 19,057 19,195 52.6 3.873 17,649 17,845
800 33.8 2.796 19,502 51.7 3.733 17,942
1000 34.0 2.792 19,025 51.5 3.774 17,945
100 32.8 2.499 31,653 50.0 3.944 29,991
200 32.8 2.471 33,225 49.8 3.855 30,881
U-002—E (02-04) A—4 ML 500 33.7 2.322 36,319 37,012 50.0 3.724 31,614 33,191
800 33.1 2.219 36,874 50.1 3.560 33,569
1000 33.1 2.207 37,843 50.5 3.516 34,390
100 33.3 2.393 22,822 51.0 4.295 18,193
200 33.9 2.412 23,466 50.2 4.135 18,644
U—002-E (03—01) A-3 311:4 500 33.9 2.441 23,426 22,830 51.6 4.005 19,685 19,629
800 34.1 2.522 22,465 52.0 4.114 19,323
1000 34.6 2.560 22,598 51.7 3.990 19,880
167

 

 

 

 

Table B.1 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cyclic stress (psi)
8011 Type 11) 15
Sample number Cyde Average Avera re Average Average MR Average Average Average Average MR
number cyclic defonnaEtDion resilient (psi) at load cyclic deforma resilient (psi) at load
AASHTO USCS load (mils) modulus cycles 500, load tion modulus cycles 500,
(lbs) (p51) 800 and 1000 (lbs) (mils) (p81) 800 and 1000
100 33.1 2.429 22,556 50.3 3.861 20,301
200 33.3 2.428 22,706 50.9 3.783 21,157
M—028-W (03—02) A—3 SP 500 34.1 2.460 23,286 23,003 51.3 3.731 21,613 22,536
800 33.9 2.374 23,167 51.3 3.644 22,811
1000 33.8 2.483 22,555 52.0 3.582 23,185
100 32.7 3.357 15,294 49.8 5.258 14,150
200 33.4 3.283 16,085 50.3 5.113 14,855
M—028—W (03-03) A—3 3:]; 500 33.0 3.059 16,876 16,911 50.5 4.774 15,866 15,956
800 33.8 3.094 16,885 50.9 4.876 15,840
1000 33.9 3.175 16,971 50.9 4.722 16,162
100 31.7 3.230 15,364 50.3 5.163 14,590
200 32.2 3.208 15,857 50.5 5.056 15,302
U-002-E (03—03) A—2-4 SM 500 32.7 3.167 16,240 15,984 50.3 4.973 15,412 15,833
800 32.7 3.251 15,919 51.1 4.873 15,966
1000 32.6 3.286 15,793 51.0 4.793 16,120
100 33.8 2.198 25,827 51.4 3.419 24,035
200 33.8 2.255 25,821 51.9 3.423 24,209
1—075—N (03—04) A—3 SP 500 34.0 2.203 25,887 26,140 51.7 3.456 24,040 24,401
800 34.3 2.156 26,592 52.3 3.402 24,474
1000 34.2 2.263 25,940 52.3 3.348 24,689
100 33.6 2.390 23,852 51.5 3.757 21,584
200 33.9 2.392 24,136 51.7 3.775 21,768
U—023-S (04—01) A—3 SP 500 33.7 2.472 23,456 23,060 51.5 3.807 21,526 21,735
800 33.8 2.440 22,395 51.7 3.700 21,852
1000 33.7 2.508 23,330 51.7 3.814 21,828
100 34.1 2.034 29,159 51.3 3.595 22,151
200 33.0 2.006 28,338 52.3 3.505 23,435
M-068—W (04—02) A-2—4 :3 500 34.1 1.883 30,987 30,958 52.1 3.377 24,481 24,764
800 34.4 1.861 30,960 51.6 3.383 24,598
1000 34.6 1.980 30,927 51.9 3.299 25,212

 

 

 

 

 

 

 

 

 

 

 

 

168

 

 

 

 

Table B.1 (cont‘d)

 

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soil Type 10 15
Sample number £131]:in Average Average Average Average MR Average Average Average Average MR
cyclic deformation resilient (ps1) at load cyclic deformation resrlient (p31) at load
AASHTO USCS load (mils) modulus cycles 500, load (mils) modulus cycles 500, 800
(lbs) (ps1) 800 and 1000 (lbs) (psr) and 1000

100 29.4 5.139 8,572 46.6 8.078 8,440
200 30.7 4.944 9,491 47.1 7.723 8,966

M-068—W (04—03) A-3 SP 500 31.3 4.879 9,725 9,979 48.6 7.368 9,822 10,013
800 31.7 4.685 10,215 48.8 7.052 10,308
1000 31.6 4.806 9,996 48.5 7.166 9,910
100 31.7 4.463 10,722 48.7 6.850 10,728
200 31.5 4.383 10,903 49.7 6.524 11,210

M-065—S (04—04) A—2—4 SM 500 32.1 4.101 11.945 1 1,943 50.0 6.427 11,637 11,909
800 32.3 4.157 11,833 50.3 6.268 11,995
1000 32.5 4.190 12.050 49.6 6.038 12,096
100 32.5 2.880 17.806 50.5 4.474 16,758
200 32.3 2.805 17.979 50.7 4.478 17,269

M—032-W (04-05) A—4 2E4— 500 32.9 2.739 18L9i 19,255 50.6 4.283 18,002 18,161
800 33.2 2.725 193% 50.9 4.267 18,269
1000 33.1 2.708 19,546 50.7 4.208 18,211
100 32.1 2.263 23L6_9_ 51.4 3.311 24,414
200 32.2 2.284 23fi91_ 51.3 3.281 24,858

U—131—N (05—01) A—2-4 SM 500 33.3 2.257 24LW4_8# 24,651 51.4 3.239 26,201 25,604
800 33.6 2.314 Elli 50.9 3.247 25,241
1000 33.8 2.266 24,456 51.9 3.298 25,370
100 34.1 1.757 35,1i5__ 52.6 2.846 29,879
200 34.3 1.708 36,388 51.9 2.861 29,921

U-127—N (05—04) A—3 SP 500 34.5 1.687 37,843 37,158 52.9 2.866 30,325 29,949
800 35.3 1.727 36:138— 51.9 2.901 29,588
1000 35.2 1.766 37,194 52.6 2.850 29,935
100 30.6 3.211 22,916 48.9 4.652 24,395
200 31.3 3.095 24,442 49.0 4.593 25,076

M—072-W (05—06) A—6 SC 500 31.7 2.941 26,104 26,492 49.8 4.403 26,622 27,193
800 31.9 2.888 27,204 49.8 4.321 27,183
1000 31.2 2.958 26,168 50.2 4.252 27,774

 

 

 

 

 

 

 

 

 

 

 

 

169

 

 

 

 

 

 

 

 

Table 31 (cont’d)

 

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 10 15
8011 Type C cle
Sample number nur>iib e r Ag??? Average 12:12:? ?V§i;aiel(l::§ Average Average Average Average MR (psi)
load deformation modulus cycPTes 500 800 cyclic load deformation resilient at load cycles 500,
AASHTO USCS (lbs) (mils) (psi) and 1060 (lbs) (mils) modulus (psi) 800 and 1000

100 33.8 2.004 27,798 52.1 3.079 26,746
200 33.9 2.055 29,008 52.7 3.027 27,574

M—l 32—N (06-01) A—3 SP 500 34.1 1.984 30,874 31,741 52.6 2.977 28,591 28,997
800 34.4 1.960 31,899 53.2 2.989 29,341
1000 34.0 1.857 32,449 52.0 2.917 29,059
100 33.4 1.616 26,603 51.6 2.882 26,762
200 33.3 1.535 28,072 51.9 2.855 28,300

l-075—N (06—02) A-3 SP 500 33.8 1.461 32,068 32,450 52.3 2.552 31,485 31,187
800 34.5 1.499 32,023 52.1 2.508 31,026
1000 34.4 1.417 33,260 52.7 2.450 31,049
100 34.1 1.933 30,572 52.5 2.814 29,633
200 33.8 1.884 31,084 52.1 2.832 29,692

U—031—N(06—03) A—3 SP 500 34.6 1.808 32,659 31,867 52.6 2.897 29,123 29,636
800 35.0 1.908 31,347 52.9 2.861 30,084
1000 35.4 1.937 31,594 52.6 2.842 29,701
100 33.5 2.456 22,276 51.9 3.650 21,630
200 33.7 2.428 23,097 51.7 3.641 21,694

1—196-N (06—05) A—2—4 :3; 500 34.2 2.386 23,190 23,030 51.5 3.618 22,017 21,985
800 33.6 2.395 22,525 51.7 3.675 21,801
1000 33.9 2.371 23,375 52.1 3.637 22,136
100 33.6 2.112 26,636 51.2 3.135 25,897
200 33.8 2.001 29,046 51.6 3.012 27,403

M—020—W (07—02) A—3 SP 500 34.3 1.969 30,442 31,489 51.8 2.891 28,918 31,766
800 34.0 1.902 31,795 52.5 2.649 33,296
1000 34.1 1.893 32,230 52.7 2.550 33,084
100 32.8 2.029 27,192 50.6 3.326 23,445
200 33.6 2.032 28,722 50.5 3.436 23,201

M—020-W (07—02) A—3 SP 500 33.8 1.969 30,147 30,272 51.7 3.361 25,035 24,896
800 34.0 1.971 30,271 51.7 3.315 24,950
1000 33.9 1.965 30,399 51.6 3.408 24,702

 

 

 

 

 

 

 

 

170

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 13.1 (cont’d)
Cyclic stress (psi)
Soil Type Cycle A I; T I: A 15
load deformation modulus c C] 500 cyclic load deformation dulu 1e 500
, y es , mo 5 eye 3 ,
AASHTO USCS (lbs) (““1” (p111) 800 and 1000 (lbs) (”1115) (psi) 800 and 1000
100 33.7 2.016 28.992 51.2 3.080 25,785
200 33.9 2.064 28,380 51.3 3.000 27,443
M—OZO-W (07-02) A—3 SP 500 33.9 1.961 29,237 29,446 51.4 2.948 28,001 28,593
800 34.0 2.042 29,31 51.6 2.893 28,741
1000 34.3 1.984 29.743 51.7 2.938 29,037
100 33.2 1.883 19,618 51.0 3.025 16,754
200 33.1 1.884 194$ 51.0 2.949 18,301
M-020-W (07—02) A—3 SP 500 33.6 1.901 20,0_lO_ 19,693 52.1 2.779 19,917 20,257
800 33.8 1.922 19,731 52.0 2.722 20,529
1 1000 33.6 1.876 19,334 51.7 2.723 20,325
‘, 100 31.7 2.516 20,fi?;l__ 49.5 4.178 17,679
1, 200 32.8 2.321 22,926 50.3 3.959 19,876
M—OZO—W (07—02) A—3 SP 500 33.1 2.329 24,360 24,320 51.4 3.364 23,953 24,552
800 33.3 2.344 24,091 51.3 3.310 24,799
1000 33.8 2.284 24,508 51.4 3.307 24,904
100 33.9 1.963 29,497— 52.6 3.106 27,496
200 34.0 1.935 30,737__ 52.2 2.999 27,797
M—020-E (07—03) A-3 SP 500 34.4 1.881 32,158 32,696 51.9 2.992 28,321 28,182
800 34.3 1.941 31,9517 52.3 3.019 27,995 1
1000 34.2 1.810 33,972 51.9 3.081 28,230 1
100 42.5 11.329 3,466 72.2 15.047 4,432 1
200 43.7 10.944 3,698 73.5 7.386 4,716 ‘-
U—127-N (07—05) A—6 SC 500 44.3 10.593 3,897 3,984 75.1 6.798 5,246 5,481
800 44.4 10.260 4,015 75.4 6.602 5,455
1000 44.2 10.170 4,041 75.4 12.596 5,742
100 52.0 2.068 47,427 82.9 3.013 49,924
200 51.7 1.926 50,103 82.3 2.944 50,430
U—127—N (07—05) A—6 SC 500 52.8 1.878 53,735 54,737 81.6 2.806 51,951 53,030
800 51.5 1.860 54,842 82.5 2.821 53,516
1000 52.3 1.860 55,634 82.0 2.761 53,623 7

 

 

 

 

 

 

 

 

 

 

 

 

 

 

171

 

 

Table 8.1 (cont’d)

 

    
  
    

 

 

 

 

  
  
  
  
  

 

 

 

 

 

 

 

 

 

  
  
  
  
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
  
 
    
  
  
   
  
 
   
  
  
  
     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cyclic stress (psi)
Soil Type Cycle 10 . 15
Sample number number Average Average 2:31:15): Average MR Avefage Average 5V81age Average MR
# cyclic deformation psi) at load cyclic deformation 1es111ent (p51) at load
AASHTO uses load (lbs) (mils) 111611111115 Cydcs 500’ load (mils) mOd‘lluS CYCIeS 500,
(ps1) 800 and 1000 (lbs) (p51) 800 and 1000
100 46.3 6.202 7,133 1 75.8 10.783 6,642
200 46.7 6.185 7,244 76.7 10.605 6,765
U—127—N (07-05) A—6 SC 500 46.7 6.248 7,223 7,323 77.2 10.557 6,875 6,925
800 47.2 6.196 7,353 77.5 10.534 6,936
1000 47.9 6.195 7,395 77.9 10.543 6,965
100 44.7 9.544 4,319 74.4 14.108 4,852
200 45.8 9.458 4,487 75.7 13.880 5,081
U—127—N (07-05) A—6 SC 500 46.2 9.290 4,651 4,713 75.8 13.294 5,339 5,358
800 46.5 9.278 4,718 75.7 13.131 5,338
1000 46.4 9.113 4,770 76.4 13.119 5,398
100 30.4 2.658 31,474 47.9 4.349 28,993
200 30.9 2.535 33,999 48.7 4.299 30,151
U—127—N (07—05) A-6 SC 500 31.5 2.333 36,628 36,054 49.1 4.290 27,523 27,729
800 31.6 2.285 36,290 49.0 4.221 27,851
1000 32.5 2.231 35,243 50.2 4.333 27,814
100 28.0 7.229 10,631 41.8 12.878 11,660
200 28.6 7.191 10,855 42.3 12.581 11,834 1'
M-061—E (07—06) A-2—4 SM 500 29.3 6.855 __ 11,362 11,483 43.2 12.189 13,155 12,907 i
800 29.6 6.630 ”11,709 44.2 11.636 12,831 x
1000 28.8 6.826 H 11,377 44.2 11.476 12,736
100 33.1 1.937 29,999 51.2 2.807 30,104
200 32.9 1.864 ”30,417 51.7 2.743 30,517
M-061—E (08-02) A-2-4 SM 500 33.8 1.897 32,344 32,231 52.1 2.713 31,551 31,763
800 33.8 1.930 __32,106 51.9 2.701 31,719
1000 34.0 1.875 32,242 52.3 2.728 32,020
100 41.4 10.662 3,592 70.9 14.499 4,516
200 42.3 10.310 : 3,829 71.7 13.972 4,771
U-010-W (08-04) A—6 SC 500 42.9 9.824 __ 4,076 4,134 72.8 13.214 5,099 5,268
800 43.2 9.691 4,164 73.7 12.850 5,323
1000 42.9 9.608 4,163 73.8 12.675 5,382

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 13.1 (conl’d)

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soil Type C ‘1 10 15
Sample number nuiib; ACverage Average Average Average MR Average Average Average Average MR
ycllc deformation 16511lent (ps1) at load cycllc deformation regihent (ps1) at load
load . modulus c ele‘ 500 1 d . d 1 1e 500
(111113) 1, y 5 7 0a (m1ls) m0 u us eye 5 ,
AASHTO USCS (165) (p81) 800 and 1000 (lbs) (p31) 800 and 1000
100 43.2 7.759 5,212 71.7 13.897 4,831
200 44.4 7.585 5,439 72.2 13.671 4,937
U—010—W (08-04) A—6 SC 500 45.6 7.430 5,791 5,873 73.0 13.508 5,049 5,106
800 45.9 7.266 5,946 73.6 13.342 5,130
1000 46.1 7.252 5,884 73.4 13.283 5,138
100 32.4 3.493 14,265 51.1 4.992 15,346
SC 200 32.9 3.499 14,932 (#509 4.807 15,758
1—075—N (08—06) A—2—4 SM 500 33.3 3.406 15,448 15,798 51.4 4.785 16,290 16,577
800 33.3 3.297 15,986 51.2 4.676 16,606
1000 33.1 3.263 15,960 51.3 4.594 16,836
100 33.8 2.123 26,881 52.3 3.220 26,018
200 33.9 2.126 27,797 52.8 3.183 26,982
U—131-S(09—01) A—3 SP 500 34.5 2.127 29,155 28,793 52.8 3.200 27,204 27,732
800 34.3 2.052 __29,674 52.6 3.078 27,868
1000 34.6 2.152 27,550 53.2 3.081 28,124
100 33.5 2.694 _20,127 52.0 4.254 18,607
200 33.8 2.624 “21,049 51.6 4.205 18,942
1—096—W (09-02) A—2-4 STE/i 500 33.7 2.566 - 21,588 22,163 51.5 4.164 19,295 19,597
800 34.0 2.530 _22,509 52.0 4.085 19,756
1000 33.8 2.473 22,392 51.2 3.999 19,740
100 33.9 1.997 #28,736 51.1 3.381 23,393
200 34.3 2.019 29,283 51.9 3.341 24,843
U—131-S (09-03) A-3 SP 500 34.9 1.990 30,848 30,368 52.1 3.006 27,525 28,022
800 34.0 1.948 30,648 52.4 3.010 27,615
1000 34.4 1.983 29,608 52.6 2.953 28,925
100 34.0 2.036 36,835 52.5 2.817 33,838
200 33.9 2.070 37,497 52.0 2.833 34,449
U—131—S(09-05) A-l—b SP 500 34.5 1.978 38,818 38,498 52.5 2.776 35,389 35,390
800 34.9 1.943 ’38,902 52.5 2.796 35,440
1000 34.8 2.007 37,773 52.5 2.856 35,340

 

 

 

 

 

 

173

 

 

 

 

Table 13.1 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cyclic stress (psi)
$611 '1pr Cycle A 10 _, 15
Sample number number Cverlage Average Avellage Average MR Average Average Average Average MR (psi)
yc1c deformation 163116111 11351) at load CyChC deformation resflient at load c cl 500
load . ‘ modulus cycles 500. load . modulus y es ’
AASHTO uses (lbs) (““19 (psi) 800 and 1000 (lbs) (mils) (psi) 800 and 1000
100 33.4 2.943 17,969 51.4 4.275 18,433
200 33.5 2.928 18,261 52.0 4.175 19,068
M—044-E (09—07) A-2—4 SM 500 33.9 2.948 18,534 18,434 51.6 4.086 19,511 19,654
800 33.7 2.841 18,744 51.6 4.043 19,663
1000 34.0 3.046 18,023 51.9 4.041 19,788
100 33.1 3.415 15,148 50.4 5.279 14,639
200 33.1 3.421 15,097 50.5 5.143 15,004
M—024-S (09-09) A—2—4 SM 500 33.6 3.458 15,204 15,156 50.6 4.889 15,786 15,854
800 32.9 3.427 14,945 50.8 4.853 15,880
1000 33.3 3.378 15,318 51.0 4.891 15,897
100 34.0 1.985 29,263 52.2 3.321 25,428
200 34.1 2.074 29,172 52.2 3.248 25,709
1-069—E (09-10) A—3 SP 500 34.4 2.071 28,746 28,663 52.3 3.163 26,255 26,095
800 34.3 2.079 29,232 51.9 3.231 25,802
1000 34.8 2.160 28,012 52.4 3.192 26,227
100 33.5 3.483 14,917 51.2 4.732 16,473
200 33.5 3.542 ifl 51.1 4.694 16,512
1—069-N (10—01) A-3 :11; 500 33.1 3.344 15,551 15,873 51.3 4.485 17,528 17,394 i
800 33.8 3.457 15,312 50.8 4.533 17,144 :1
1000 34.0 3.162 16.756 51.7 4.526 17,509 i
100 32.9 2.180 41,549 50.1 3.266 40,469 i
200 33.4 2.210 ___42,02_ 50.5 3.275 41,776
1—096—W (10-03) A-2—6 SC 500 33.8 2.175 43,219 43,824 51.5 3.167 42,767 37,712
800 34.1 2.196 44.499 49.2 3.779 34,806
1000 34.1 2.203 43,754 48.7 2.686 35,563
100 33.4 2.846 18,890 51.4 4.400 18,049
200 33.5 2.771 19,221 51.8 4.330 18,293
1—069—N (10-04) A—2—4 SM 500 33.5 2.839 19,530 19,190 51.0 4.232 18,653 18,963
800 33.4 2.862 19,049 51.5 4.107 18,952
1000 33.9 2.802 18,990 51.6 4.076 19,284

 

 

 

 

 

 

 

 

 

 

 

174

 

 

 

 

Table 13.1 (cont’d)

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 10 15
8011 Type C cle
Sample number nugiaber Acverage Average Average Average MR Average Average Average Average MR
yclic . 10s1lient (ps1) at load cyclic . res111ent (p51) at load
load deformation modulus cycles 500, load deformation modulus cycles 500
AASHTO USCS (lbs) (mils) (psi) 800 and 1000 (lbs) (mus) (psi) 800 and 1000
100 25.4 8.607 4,273 37.6 15.895 3,469
200 25.9 8.326 4,542 41.3 12.494 4,766
1—069—N (10—05) A—2-4 SM 500 27.0 7.715 5,123 5,295 43.1 11.611 5,377 5,646
800 27.2 7.630 5,241 43.7 11.104 5,712
1000 27.6 7.381 5,521 43.8 11.027 5,850
100 30.1 5.580 8,027 47.9 7.602 9,168
200 30.9 5.176 8,832 48.1 7.430 9,504
1-096—W (10—09) A—2—4 SM 500 31.2 5.002 9,361 9,518 49.5 6.721 10,908 11,394
800 31.0 4.917 9,419 49.5 6.363 11,495
1000 31.6 4.875 9,775 49.8 6.307 11,778
100 32.9 1.119 30,534 51.7 1.595 29,788
7% 33.8 1.119 32,960 52.0 1.567 30,484
1—069—N (1 1—01) A—3 :1}; 400 34.0 1.063 30,406 30,733 52.0 2.994 28,203 28,147
800 34.0 1.119 30,967 52.2 2.974 28,154
1000 34.8 1.119 30,827 51.9 2.995 28,083
100 33.6 1.694 36,073 51.2 3.205 25,752
200 33.3 1.627 37,965 52.2 3.144 26,314
1—094-W(11—02) A—3 SP 500 34.1 1.487 45,141 44,521 52.8 3.102 27,442 27,372
800 32.6 1.432 42,908 51.9 3.136 26,857
1000 34.1 1.453 45,513 52.6 3.020 27,817
100 31.9 2.614 19,615 50.2 4.354 17,216
200 31.3 2.561 19,255 50.7 4.426 17,262
M—060—W (11—03) A—2—4 2115/1— 500 32.0 2.553 19,808 19,812 50.8 4.481 16,817 16,639
800 32.4 2.561 19,861 50.7 4.560 16,601
1000 32.2 2.563 19,768 50.9 4.669 16,498
100 33.7 2.252 23,451 52.3 3.358 24,923
200 33.8 2.220 24,393 52.6 3.317 25,291
I—069-S (1 1—05) A—4 STE/I— 500 34.0 2.095 25,903 27,303 52.5 3.274 25,489 25,645
I 800 34.2 2.014 26,908 51.6 3.267 25,632
1000 34.3 1.931 29,098 52.2 3.245 25,814

 

 

 

 

 

 

 

 

 

 

 

 

175

 

 

 

 

 

Table 8.1 (cont’d)

 

1 Sample number

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 10 15
8011 Type
CYCIC Average Averaoe Average Average MR Average A Average Average MR
number cyclic deformation resilient (psi) at load cyclic defd/rfrriiftijon resilient (psi) at load
load . modulus cycles 500, load . modulus cycles 500,
AASHTO USCS (lbs) (”1115) (psi) 800 and 1000 (lbs) (mus) (psi) 800 and 1000
100 33.3 1.963 28,024 51.3 3.519 22,990
200 32.5 2.028 27,129 52.0 3.594 23,024
1-094—W (12—01) A—2—4 :54— 500 34.1 2.041 28,985 27,636 51.8 3.438 23,554 23,872
800 33.6 2.129 26,615 52.0 3.414 24,001
1000 33.9 2.113 27,308 52.4 3.534 24,060
100 33.6 2.783 19,527 50.8 4.851 15,566
200 33.6 2.766 19,827 50.9 4.881 15,796
1-094-W (12—03) A—3 :1; 500 33.4 2.814 18,886 18,139 50.4 4.789 16,090 15,977
800 33.9 3.021 17,820 50.8 4.831 15,893
1000 34.1 3.066 17,711 50.8 4.798 15,947
100 33.3 2.862 18,848 50.8 4.340 18,416
200 33.3 2.930 19,047 51.3 4.323 18,683
U—012—E (12—04) A-2—4 if; 500 33.6 2.781 19,237 19,234 51.2 4.264 18,191 18,343
800 34.1 2.881 19,210 51.4 4.312 18,324
1000 34.1 2.766 19,255 51.3 4.266 18,515
100 33.9 2.675 19,996 51.4 4.042 19,797
200 33.8 2.698 20,013 51.4 3.956 20,110
1—094—W (12-06) A—2—4 SM 500 33.7 2.821 19,357 19,425 52.6 3.873 21,249 21,382
800 33.8 2.796 19,802 51.7 3.733 21,552
1000 34.0 2.792 19,115 51.5 3.774 21,346
100 34.4 3.172 17,093 51.9 5.000 15,746
200 34.0 3.101 17,359 50.0 4.846 15,814
M-024—S (13—01) A—4 SM 500 34.8 3.149 17,853 17,950 51.5 4.878 16,213 16,175
800 34.6 3.049 17,891 51.6 4.844 16,042
1000 35.1 3.052 18,106 51.8 4.844 16,271
100 33.6 2.042 28,216 51.7 3.362 23,959
200 33.5 2.112 27,648 51.9 3.351 24,234
M—059—W (13-02) A—3 SP 500 33.9 2.186 26,464 24,863 52.1 3.478 23,699 23,810
800 33.9 2.368 24,623 51.7 3.453 23,882
1000 33.8 2.439 23,502 51.9 3.436 23,849

 

176

 

 

 

 

Table 8.1 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cyclic stress (psi)
Soil Type 10 15
Sample number CYCIC Average Avera re Average Average MR Average Avera e Average Average MR
number cyclic 1”. regilient (psi) at load cyclic g. re51lient (psi) at load
load deformation modulus cycles 500, load deformation modulus cycles 500
AASHTO USCS (lbs) (““1” (psi) 800 and 1000 (lbs) (“1115) (psi) 800 and 1000
100 32.8 2.693 20,399 51.2 4.441 17,533
200 32.5 2.615 20,565 50.7 4.353 17,856
1—094-W (13—04) A—3 SP 500 33.2 2.592 21,384 21,470 50.8 4.245 18,355 18,859
800 33.5 2.589 21,598 50.8 4.133 18,836
1000 33.4 2.648 21,427 51.2 4.040 19,387
100 34.0 2.515 22,197 51.6 3.907 20,649
200 33.8 2.443 23,009 51.4 3.846 20,641
U—023—N (13-07) A-3 :11; 500 34.0 2.573 22,214 22,629 51.2 3.961 20,201 20,593
800 33.9 2.428 22,768 52.4 3.910 20,900
1000 34.7 2.477 22,904 52.2 3.952 20,678
100 29.7 4.124 16,710 45.5 6.531 16,006
200 30.1 4.182 16,898 46.1 6.637 15,991
M—010—E (13—08) A—6 CL 500 30.2 4.256 16,855 17,012 46.3 6.562 16,218 16,345
800 30.7 4.226 16,995 46.5 6.433 16,417
1000 30.5 4.202 17,186 46.6 6.492 16,399
100 48.6 3.375 14,374 77.7 7.441 9,934
200 49.6 3.334 15,053 78.1 7.453 9,867
M-OlO—E (13—08) A—6 CL 500 49.6 3.271 15,423 15,561 78.2 7.743 9,627 9,553
800 49.8 3.258 15,631 78.2 7.779 9,528 1
1000 49.8 3.257 15,629 78.3 7.849 9,504 }
100 51.5 1.331 31,968 83.3 1.796 36,929 ,1
200 51.8 1.291 36,534 82.8 1.731 38,488
M-010—E(13—08) A-6 CL 500 52.7 1.218 43,564 44,641 82.5 1.808 40,155 41,989
800 52.3 1.211 45,089 82.2 1.629 42,399
1000 51.8 1.152 45,271 82.4 1.793 43,414
100 39.6 7.658 8,407 46.6 8.078 8,440
200 41.3 7.157 9,399 47.1 7.723 8,966
M—OlO-E (13-08) A-6 CL 500 42.2 6.971 9,004 9,713 48.6 7.368 8,822 8,280
800 43.5 6.473 9,818 48.8 7.052 8,108
1000 44.0 6.376 10,317 48.5 7.166 7,910

 

 

 

 

 

 

 

 

 

 

 

 

 

177

 

 

 

  

 

  
  
  
  
  

 

 

 

 

 
  
 
 

 

 

 

 

 

 

 

 

 

 
  
 
 

 

 

 

 

 

 

 

 

 

 

 
  
 
 

 

 

 

 

 

 

 

 

 

 

 
  
 
 

 

 

 

 

 

 

 

 

 

 

 

 
  
 
 
 
 
 
 
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table B.1 (cont’d)
Cyclic stress (psi)
Soil Type 10 15
Sample number “(5:3); Acverlage Average A3velraxge Average MR Average Average Average Average MR
yc lC . rcSI lent (p51) at load . . res1llent (p81) at load
load deformation modulus cycles 500 CyChC load deformation modulus cycles 500
AASHTO USCS (lbs) (““1” (psi) 800 and 1000 (lbs) (“15) (psi) 800 and 1000
100 30.3 2.283 11,369 48.9 3.899 14,813
200 31.0 2.259 13,560 48.7 3.804 15,893
1—075—S (14—01) A—7—6 SC 500 31.3 2.074 17,389 18,221 48.9 3.721 16,938 17,842
800 31.7 1.971 18,449 49.6 3.586 18,253
1000 32.0 2.067 18,825 49.6 3.573 18,336
100 51.3 2.172 32,901 82.3 2.390 32,808
200 51.2 1.994 36,098 82.8 2.299 31,287
1-075—S (14—01) A—7—6 SC 500 52.3 1.815 31,799 32,510 82.7 2.045 29,226 29,860
800 52.3 1.481 32,377 82.4 1.858 30,295
1000 51.7 1.417 33,354 82.3 1.668 30,060
100 35.0 10.936 5,114 61.1 14.155 6,907
200 35.9 9.896 5,982 61.6 13.624 7,285
1—075—S (14—01) I A—7—6 SC 500 36.6 9.349 7,441 7,187 62.5 12.617 7,928 8,386
800 36.7 8.808 7,284 63.5 12.002 8,545
1000 37.3 8.616 6,835 64.1 11.896 8,685
100 32.7 2.313 21,994 51.7 3.526 19,968
200 33.0 2.337 22,227 52.0 3.540 20,296
U—024—S (14-04) A-3 SP 500 33.5 2.305 22,633 22,765 52.3 3.336 21,874 21,913
800 33.6 2.300 22,813 52.1 3.396 21,707
1000 33.6 2.297 22,849 52.4 3.329 22,159
100 42.5 10.621 3,715 65.6 19.895 3,023
200 42.7 10.593 3,736 65.4 19.906 3,021
M—153—E (14—06) A—7—6 SC 500 42.8 10.681 3,717 3,732 66.4 20.023 3,036 3,015
800 42.9 10.628 3,745 66.4 20.120 3,014 -\.~
1000 43.1 10.729 3,733 66.0 20.123 2,995
100 29.2 13.397 3,483 65.6 19.995 4,023
200 30.3 12.818 3,798 65.4 19.996 3,821
M—153—E (14—06) A—7-6 SC 500 32.1 12.062 4,285 4,430 66.4 20.017 3,936 3,915
800 32.8 11.875 4,471 66.4 21.120 3,814
1000 33.0 11.708 4,535 66.0 21.123 3,995
178

 

 

 

 
 
   

 

 

 

 

  
  
   
  
  

 

 

 

 

 

 

 

 

 

  
  
  
  
  

 

 

 

 

 

 

 

 

 

  
  
  
  
  

 

 

 

 

 

 

 

 

 

  
  
  
  
  

 

 

 

 

 

 

 

 

 

 

 

 

  
  
  
  
  
  
  
  
  
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

Table B.l (cont’d)
Cyclic stress (psi)
Soil Type 10 15
Sample number Cycle Average A I T1 Average Average MR Average Average Average MR
‘ number cyclic d fwragc resilient (psi) at load cyclic Average resilient (psi) at load
load e ormation modulus cycles 500, load deformation modulus cycles 500,
AASHTO USC S (lbs) (”“1” (ps1) 800 and 1000 (lbs) (“1115) (psi) 800 and 1000
100 33.8 2.378 38,348 51.2 3.050 42,427
200 34.4 2.254 39,970 51.5 2.964 42,728
M-153—E (14—06) A—7—6 SC 500 34.5 2.223 40,365 40,902 51.4 3.004 43,684 44,483
800 34.0 2.119 41,453 51.5 2.965 44,394
1000 33.7 2.120 40,889 52.2 2.876 45,372
100 33.6 2.237 25,772 51.5 3.727 21,646
200 33.6 2.150 25,870 51.7 3.731 21,643
M—053—S (14—07) A—3 SP 500 33.9 2.249 26,465 25,738 51.9 3.688 22,217 22,296
800 34.0 2.258 25,493 52.0 3.646 22,403
1000 33.7 2.315 25,255 51.8 3.622 22,268
100 49.1 5.308 8,870 77.2 9.347 7,782
200 49.1 5.217 9,211 77.2 9.253 7,846
1—094—W (14—09) A—7—6 CL 500 49.2 4.966 9,690 9,955 77.1 9.107 7,995 8,080
800 48.8 4.777 9,943 77.8 9.010 8,089
1000 49.2 4.675 10,234 77.8 8.918 8,156
100 51.2 2.114 45,953 81.9 2.609 57,985
200 51.1 1.853 52,917 82.8 2.466 61,580
1—094—W (14—09) A—7—6 CL 500 52.2 1.602 67,009 73,344 82.5 2.327 67,663 70,094
800 52.7 1.426 75,719 82.2 2.190 70,504
1000 51.2 1.383 77,304 81.8 2.205 72,116
100 33.0 1.604 53,229 50.7 2.211 57,722
200 32.7 1.585 55,517 51.7 2.228 59,950
1-094—W (14—09) A—7—6 CL 500 33.8 1.530 60,326 60,217 51.7 2.152 60,448 60,303
800 34.1 1.462 60,280 51.8 2.104 60,142
1000 34.2 1.459 60,046 52.3 2.044 60,318
100 33.3 2.923 18,400 51.1 4.424 18,022
200 33.2 2.914 18,486 51.0 4.470 18,018
M—053—S (15—02) A—2—4 SM 500 33.4 2.921 18,171 18,342 51.4 4.471 17,918 18,060
800 33.4 2.963 18,372 51.3 4.416 18,113
1000 33.4 2.894 18,483 51.0 4.372 18,149
179

 

 

Table B.1 (cont’d)

 

 

 

Cyclic stress (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soil Type C cle 10 15
Sample number number A‘s/”121136 Average AJ/filiagc Average MR Average Average 9V81age Average MR (psi)
yc It deformation “51 lent (pm) at load CyChC deformation 1651116111 at load cycles 500
load . modulus cycles 500. load . modulus ’
AASHTO USC S (lbs) (““1” (psi) 800 and 1000 (lbs) (“1115) (psi) 800 and 1000
100 34.6 1.494 65,657 51.5 2.170 60,204
200 34.2 1.492 65,191 51.9 2.192 61,455
M-090-E (15—04) A—4 CL 500 34.6 1.487 67,087 67,841 51.7 2.159 61,666 62,065
800 34.6 1.510 68,335 51.7 2.128 62,105
1000 34.5 1.398 68,102 52.0 2.212 62,423
100 34.0 1.585 37,971 52.6 2.503 35,506
200 34.0 1.601 38,716 52.2 2.445 35,369
M—025—S (15-05) A—3 SP 500 34.1 1.588 39,705 40,152 51.7 2.500 35,195 35,481
800 34.9 1.643 40,506 52.3 2.468 35,680
1000 35.0 1.595 40,246 52.0 2.437 35,567
100 34.3 2.740 19,702 51.3 4.328 18,630
200 35.7 2.770 20,960 51.9 4.203 18,904
M—019—S (15—07) A—2—4 SM 500 35.0 2.584 21,859 22,233 51.7 4.118 19,310 19,500
800 35.2 2.539 22,379 51.4 4.096 19,441
1000 34.6 2.572 22,462 53.2 4.183 19,750

 

 

 

 

 

 

 

 

 

180

 

 

 

References

1982 Quaternary Geology Map of Michigan. “Department of Natural Resources” (1982)
[Online] available http://www.deq.statemi.us/documents/deq—ogs-gimdl-GGQGM.pdf,
January 5, 2007.

AASHTO. (1993). "American Association of State Highway and Transportation
Officials, Guide for Design of Pavement Structures." Washington, D. C.

Baladi, G.Y. and Boker, TD. (1978). “Resilient Characteristics of Michigan
Cohesionless Roadbed Soils in Correlation to the Soil Support Values.” Final
Report of Research Conducted under Research Grant 75-1 6 79, Michigan State
University, East Lansing, Michigan.

Boateng-Poku, Y., and Drumm, E. C. (1989). "Hyperbolic Model for the Resilient
Modulus Response of Fine—Grained Subgrade Soil." Resilient Moduli of Soils
ASCE Geotechnical Special Publication No. 24, 1-14.

Coree, B., Ceylan, H. and Harrington, D., Implementing the mechanistic empirical
pavement design guide. Technical Report, IHRB Project TR-509, 2005 (Center
for Transportation Research and Education, Iowa State University).

Das, Braja. M. (2004). Principles of Foundation Engineering, Thompson Brooks/Cole,
Pacific Grove, CA 93950.

Dawson, Tyler A. (2008). "Backcalculated Subgrade Resilient Modulus Design Values
for the State of Michigan," MS. Thesis, Michigan State University, East Lansing,
draft.

De‘hlen, G. L. (1969). "The Effect of Non-Linear Material in the Behavior of Pavements
Subjected to Traffic Loads," Ph.D. Thesis, University of California, Berkley.

P inn, F. N., Nair, K., and Monismith, C. L. "Applicationof Theory in the Design of
Asphalt Pavements." 3 rd Proceedings International Conference on the Structural
Design of Asphalt Pavements, University of Michigan, Ann Arbor, Michigan,
392-409.

George, K. P. (2003). "Falling Weight Deflectometer for Estimating Subgrade Resilient
Moduli." FH WA/MS-DOT -RD-03-1 5 3 , The Mississippi Department of
Transportation.

George, K. P. (2004). "Prediction of Resilient Modulus from Soil Index Properties."
FH WA/MS-DOT -RD-04-1 72, The Mississippi Department Of Transportation.

George, K. P., Bajracharya, M., and Stubstad, R. (2004). "Subgrade characterization
employing the falling weight deflectometer." Transportation Research
Record(1869), 73.

181

Goitom, T. (1981). "Characteristics of Michigan Cohesive Subgrade Soils under Cyclic
Loading," Ph.D. Thesis, Michigan State University, East Lansing.

Groeger, J. L., Rada, G. R., and Lopez, A. “AASHTO T307-Background and
Discussion,” Resilient Modulus Testing for Pavement Components, AST M S T P
143 7, 2003.

Gudishala, Ravindra. (2004). "Development of Resilient Modulus Prediction Models for
Base and Subgrade Pavement Layers From In Situ Devices Test Results," MS
Thesis, Louisiana State University, Baton Rouge.

Han, Yuh-Puu, Petry, T. M., and Richardson, D. N. (2006). "Resilient Modulus
Estimation System for F ine-Grained Soils." Transportation Research
Record(1967), 69-77.

Hardcastle, J. H. (1992). "Subgrade Resilient Modulus for Idaho Pavements." FH WA
Report No. RP] I O-d, Idaho Transportation Department.

Harichandran, R. S., Ramon, C. M., and Baladi, G. Y. (1994). "MICHBACK user's
manual." Department of Civil and Environmental Engineering, Michigan State
University, East Lansing, Michigan.

Holtz, Robert D., Kovacs, William D. (1981). An Introduction to Geotechnical
Engineering, Prentice Hall, Upper Saddle River, NJ 07458.

Huang, Y. H. (2004). Pavement Analysis and Design, Pearson Prentice Hall, Upper
Saddle River, NJ 07458.

J anoo, V., Irwin, L., and Haehnel, R. (2003). "Pavement Subgrade Performance Study."
ERDC/CRREL T R-03-5 , US Army Corps of Engineers.

Janoo, V. C., Jr., J. J. B., Durell, G. D., and Jr., C. E. S. (1999). "Resilient Modulus for
New Hampshire Subgrade Soils for Use in Mechanistic AASHTO Design."
Special Report 99-14, US Army Corps of Engineers.

Kathleen, T. H., Carlos, E. C., Samuel, H. C., and Robert, P. E. (2001). "Rehabilitation
Strategies for Highway Pavements." NCHRP Web Document 35 (Project CI-38):
Contractor '3 Final Report, NCHRP.

Kathleen, T. H., and Crovetti, J. A. (2000). "LTPP Data AnalysiszRelative Performance
of Jointed Plain Concrete Pavement with Sealed and Unsealed Joints." NCHRP
Web Document 32 (Project SP20-50[2]): Contractor ’3 Final Report, NCHRP.

Lentz, R. W. (1979). "Permanent Deformation of Cohesionless Subgrade Material under
Cyclic Loading," Ph.D. Thesis, Michigan State University, East Lansing.

Li, John Chien-Chung. (1979). "Dynamic Properties of Frozen Granular Soils," Ph.D.
Thesis, Michigan State University, East Lansing.

182

 

Lie, Cheng., and Evett, J. B. (2008). Soils and Foundation 7111 ed. Pearson Prentice Hall,
Upper Saddle River, NJ 07458.

Maher, A., Bennert, T., Gucunski, N., and Walter J. Papp, J. (2000). "Resilient Modulus
Properties of New Jersey Subgrade Soils." FH WA NJ 2000-01, FHWA.

Maher, M. H., J ., P. J. W., and Gucunski, N. (1996). "Measurement of Soil Resilient
Properties Using Non-contacting Proximity Sensors." Transportation Research
Record(1548), 16-23.

Marr W.A., Hankour R., Werden SK. (2003). "A Fully Automated Computer Controlled
Resilient Modulus Testing System." AST M ST P 143 7, 141-151.

Nazarian, S., and Feliberti, M. (1993). "Methodology for Resilient Modulus Testing of
Cohesionless Subgrades." Transportation Research Record(1406), 108-115.

NCHRP (2004). Guide for Mechanistic-Empirical Design of New and Rehabilitated
Pavement Structures. Final Report. National Cooperative Highway Research
Program (N CHRP) Project 1-37A, Washington, DC.

NHI Course No 131008, “Techniques for Pavement Rehabilation”, Reference manual,
US Department of Transportation, Federal Highway Administration, National
Highway Institute (N HI), publication number FHWA NHI-98-033, revised
August 1998.

Pezo, R.F., G. Claros, and W.R. Hudson. “An Effective Resilient Modulus Test for
Subgrades and Nongranular Subbase Material,” Presented at 71St Annual Meeting
of the Transportation Research Board, Washington, DC, 1992.

Pezo, R., and Hudson, W. R. (1994). "Prediction Models of Resilient Modulus for
Nongranular Materials." Geotechnical Testing Journal, 17(3), 349-355.

Pezo, R. F., Kim, D. S., Stokoe 11, K. H., and Hudson, W. R. (1991). "A Reliable
Resilient Modulus Testing System." Transportation Research Record (1307), 90—
98.

Pierce, L. M. (1999). "Development of a Computer Program for the Determination of the
AREA Value and Subgrade Modulus using Dynatest FWD." Washington State
Department of Transportation.

Ping, W. V., Yang, Z., and Gao, Z. (2002). "Field and Laboratory Determination of
Granular Subgrade Moduli." Journal of Performance of Constructed Facilities,
Vol. 16(No. 4), 149-159.

Prozzi J .A and F. Hong (2006). “Seasonal Time Series Models to Support Traffic Input
Data for Mechanistic-Empirical Design Guide.” Transportation Research Record
(1947), 175 — 184.

183

 

Quintus, H. V., and Killingsworth, B. (1998). "Analyses Relating to Pavement Material
Characteriztions and Their Effects on Pavement Performance." FH WA -RD-9 7-
085, FHWA.

Rada, G., and Witczak, M. W. (1981). "Comprehensive Evaluation of Laboratory
Resilient Moduli Results for Granular Material." Transportation Research
Record(810), 23-33.

Rahim, A., and George, K. P. (2003). "Falling weight deflectometer for estimating
subgrade elastic moduli." Journal of Transportation Engineering, 129(1), 100.

Richart, F. E. Jr., Hall. J. R. Jr., and Woods, RD. (1970). Vibrations of Soils and
Foundations, Prentice-Hall International Series, Englewood Cliffs, NJ.

Seed, H. B., Chan, C. K., and Lee, C. E. "Resilient Characteristics of Subgrade Soils and
Their Relation to Fatigue Failures in Asphalt Pavements." Proceedings

International Conference on the Structural Design of Asphalt Pavements,
University of Michigan, Ann Arbor, Michigan, 61 1-636.

Standard Specifications for Transportation Materials and Methods of Sampling and
Testing, AASHTO. Washington, D. C., 2001.

Stubstad, R. N., Lukanen, E. 0., Tayabji, S. D., and Clevenson, M. L. (2002). "LTPP
Data Analysis: Feasibility of Using FWD Deflection Data to Characterize
Pavement Construction Quality." NCHRP’s Project 20—5 0(9), Final Report,
NCHRP, NCHRP's Project 20-50(9), Final Report.

Sukumaran, B., Kyatham, V., Shah, A., Sheth, D. (2002). “Suitability of Using California
Bearing Ratio Test to Predict Resilient Modulus.” Presented for the Federal
Aviation Administration Airport Transfer Conference.

Svasdisant, Tunwin (2003). "Analyses of Top-Down Cracking in Rubblized and Flexible
Pavements," Ph.D. Thesis, Michigan State University, East Lansing.

Thompson, M. R., and Robnett, Q. L. (1976), "Final Report, Resilient Properties of
Subgrade Soils." Illinois Cooperative Highway and Transportation Serial No. 160,
University of Illinois, Urbana-Champaign, Illinois

Thompson, M. R., and Robnett, Q. L. (1979). " Resilient Properties of Subgrade Soils."
Transportation Engineering Journal of ASCE, 105(TE1), 71-89.

United States Department of Agriculture. (1992). “Soil Survey of Ingham County,
Michigan. National Cooperative Soil Survey.

Web Soil Survey. “Natural Resources Conservation Services” [Online] available
http://websoilsurvey.nrcs.usda.gov/app/WebSoi1Survey.aspx, January 23, 2007.

184

Yau, A., and Quintus, H. V. (2002), "Study of LTPP Laboratory Resilient Modulus Test
Data and Response Characteristics" FHWA—RD-02-05I, FHW A

Yoder, EJ. (1959). Principles of Pavement Design, John Wiley & Sons, Inc., New York,
NY.

Young, M. A., and Baladi, G. Y. (1977), "Repeated Load Triaxial Testing State of the
Art" Final Report of Research Conducted under Research Grant 75-1 6 79,
Michigan State University, East Lansing, Michigan.

185

 

111111111111111111111111111

250

     

1