Final Research Report
for
RESILIENT MODULUS OF GRANULAR SOILS WITH FINES CONTENTS
Prepared by
Nien-Yin chang, Ph.D., P.E. Hsien-Hsiang Chiang, Ph.D.
and Lieu-Ching Jiang, Doctoral Candidate
for
Colorado Department of Transportation and
Federal Highways Administration
under
Contract No. COOT 91-379
Expansive Soil Research Center Department of Civil Engineering university of Colorado at Denver
Denver, Colorado 80204-5300
May 1994
The contents of this report reflect the views of
authors who are responsible for the facts and the
accuracy of the data presented herein. The
contents do not necessarily reflect the official
views of the Colorado Department of Transportation
or the Federal Highway Administration. This report
does not consti~ute a standard, specification, or
regulation.
i
Technical Rep0l1, Documentation Page
1. Report No. 2. GoVel"lllllellt Accessioll No.
CDOT-DTD-R-95-9
4. Tille and Subtitle
Resilient Modulus of Granular Soils With Fines Contents
7. Author(s)
3. Recipient's Catalog No.
5. RepOl·t Date
Apri11995
6. Perfol"luing Organization Code
8. Ped"orming Organization Rpt.No.
Nien-Yin Chang, Hsien-Hsiang Chian.l~;, Lieu-Ching Jiang CDOT-DTD-R-95-9
9. Pcrrol"luing O.·galli7.ation Namc anel Add.·css
Colorado Department of Transportation
4201 E. Arkansas Ave. Denver Colorado 80222
12. Sponsol"ing Agcncy Namc and Addl'css
Colorado DepaJ1ment of Transportation
4201 E. Arkansas Ave. Denver, Colorado 80222
15. Supplcll1cnhll")' Notes
10. WOI'k Unit No. (TRAYS)
11. Contract 01' GI'ant No.
13. Type or Rpt. and Pel'ioel COVCI'Cel
Final Reoort 14. SI)Onsm'ing AgenL'Y Code
Prepared in Cooperation with the U.S. Department of Transportation Federal Highway Administration
16. Ahslrad
This study examined the relationship of resiJient modulus with the stabilometer R-values, and with the index properties of Colorado soils.
A total of 39 resilient modulus tests were conducted, 20 in this test program and 19 in an earlier test program. Soil types ranged from A-I-a
to A-7-6 with valying amounts of fines. Regression analyses were
performed to formulate the functional relationship lIsing the least square fit.
Two regression equations were formulated to relate resilient modulus to R-value (one linear and one nonlinear), and four linear regression equations to relate
resilient modulus to percent fines content, plasticity index, uniformity coefficient, and mean grain size. In general, the values of the Correlation coefficients are small
llnd care should be taken when using these equations, particularly for granular soil samples with less than 30% of fines.
Implcmcnt"t ion:
It is recommended to abandon the effort to formulate a relationship between resilinet modulus and R-vaJue because of the inability of R-value jn reflecting dynamic behavior of soils.
l7. Kcy Words
Resilient Modulus,
R -vallle,I ndex properties
19.5ccmity Classif. (reIWI'I)
Unclassified
20.SeclII·ity CIassif. (page)
Unclassified
18. Distl"ihutiol1 Stalell1cl1t
No Restrictions: This report is
available to the public through
the National Technical Info. Selvice. Springfield, VA 22161
21. No. of l'ilges
ii
EXECUTIVE SUMMARY
After several decades of development, the design of pavement
using theory of elastic layers has matured to an extent that its
introduction for general acceptance by pavement engineers becomes
feasible. This led to the adoption of the elastic layer theory
in pavement design by the American Association of state Highway
and Transportation Officials (AASHTO) in its "AASHTO Guide for
Design of Pavement structures - 1986. 11 The use of elastic layer
theory involves two elastic material parameters, namely resilient
modulus and Poisson's ratio. The major difficulties in the
implementation of the 1986 AASHTO Guide include: 1) the lack of
proper/affordable testing equipment, 2) the lack of
understanding of the physical meaning of resilient modulus, and
3) site (or soil type) dependency of resilient modulus. Thus,
each state is advised to develop its own design curve and/or
design equation.
Objectives of this research project are three fold: · 1) to
determine the resilient modulus and the stabilometer R value of
soils of different index properties, 2) to formulate the
functional relationship between the resilient modulus and R
value, and between the resilient modulus and different index
properties of soils, and 3) to provide detailed test procedures
for resilient modulus. Because of the recent nature of the
implementation of resilient modulus in pavement design, its data
base is extremely small compared to the data base for the
iii
stabilometer R-value. Thus, the functional relationship between
the resilient modulus and the R-value with a large data base
would seem desirable. Once the equation is formulated., the
resilient modulus of a subgrade soil can be evaluated for a R
value that can be easily determined. However, the R value does
not reflect the dynamic property of soils that is required in the
design of pavement. It would be extremely desirable to determine
the resilient modulus of soils and then relate the resilient
modulu.s to their index properties.
The resilient modulus is a dynamic property of soil.
Previous studies indicate that the dynamic properties of soils
are strongly related to the index properties of soils. These
index properties include percent fines content, gradation
characteristics and Atterberg I s limits. Thus, the resilient
modulus can be formulated as a function of index properties of
soils. Given the functional relationship, a resilient modulus
can be determined using the index properties that can be easily
determined.
This study aims to provide the functional relationship where
the resilient modulus is related to the stabilometer R value
and/or index properties of Colorado soils. The statistical
sample size is, however, too small to obtain a good functional
relationship because of the large number of influencing factors
involved. To form a bigger statistical data base, the results
from the twenty resilient modulus tests from this research and
the nineteen tests conducted earlier for the Colorado Department
iv
of Transportation are merged. Test results are found to be quite
scatter, particularly for the granular soils. This is believed
to be caused by a large number of factors affecting the dynamic
behavior and properties of soils.
Both linear and nonlinear regression analyses were performed
to relate the resilient modulus to the stabilometer R value.
Both linear and nonlinear regression equations give nearly the
same result as the one recommended by Yeh and Su (1989). The
correlation relationship between the resilient modulus and the R
value is weak. This weak correlation between MR and R values
could be because of the inability of the stabilometer R value in
reflecting the dynamic properties of soils.
Linear regression analyses are also performed to relate the
resilient modulus to each of the following index properties of
soils: mean grain size 050, uniformity coefficient, Cu, percent
fines content, FC (passing u. s. standard sieve #200), and
plastic index, PI. The resilient modulus was found to be weakly
correlated to these index properties in a descending order of the
strength of correlation: % fines content, FC, plasticity index,
PI, uniformity coefficient, CU, and mean grain size, 050.
This study reveals the fact that the many factors affect the
value of resilient modulus. These factors include density (or
void ratio), moisture content, percent fines content, Atterberg's
limits of fines, uniformity and mean grain size of coarse grain
soils with grain size greater than US Standard Sieve #200,
conf ining stress, and cyclic stress amplitude and the number of
repeti tion, etc.
v
To formulate an equation for an effecti ve
evaluation of the resilient modulus, it is recommended to conduct
a systematic resilient modulus test program involving soils with
a wide range of values for each index property. To facilitate
technology transfer, it is recommended to hold resilient modulus
training sessions for material and pavement design engineers
throughout the state aiming at an eventual transfer the resilient
modulus technology including the method of testing and the
analysis of test results to the Colorado Department of
Transportation.
1
I. INTRODUCTION
A highway pavement system includes pavement and subgrade
materials, and, sometimes, base (and subbase) course. Its design
naturally requires the mechanical properties of pavement
materials and all base course and subgrade materials.
Conventionally these mechanical properties are lumped in one
single parameter, namely the stabilometer R-value or CBR. The
continual distress of pavement designed using R-value indicates
the shortcoming of using R-value in the pavement design.
Besides, in the last three decades, the development of the
elastic layer theory for pavement design has matured to an extent
that its implementation in pavement design becomes feasible.
This leads to the adoption of the 1986 AASHTO Guide for Design of
Pavement structures.
In the elastic layer theory, the material is assumed to be
isotropic linear elas.tic with two independent material
parameters, namely Poisson's ratio and Young's modulus. In the
repetitive loading environment, the resilient Young's modulus (or
resilient modulus) of subgrade varies with the number of
repetition, amplitude of repetitive loading, confining pressure,
and index properties of soils including density, void ratio,
moisture content, gradation characteristics of soils, amount of
fines passing #200 sieve and their Atterberg's limits. Major
difficulties for the implementation of this design guideline in
the state of Colorado include: 1) the lack of an appropriate test
2
apparatus, 2) many factors influencing the resilient modulus of
soils, and 3) the lack of a technology transfer program. To
conduct cyclic triaxial tests for resilient modulus determination
requires both a closed-loop hydro-electric universal testing
machine and good analytical skills. In this study, twenty cyclic
triaxial tests were conducted for different soils. Because of
the variability of soils from one state to the other, AASHTO
recommends each state to develop its own formula for resilient
modulus. Many factors affect the resilient modulus of soils.
This makes the task of formulating resilient modulus complex.
Objectives of this research project are three fold: 1} to
determine the resilient modulus and the stabilometer R value of
soils of different index properties, 2) to formulate the
functional relationship between the resilient modulus andR
value, and between the resilient modulus and its influencing
factors, and 3) to provide detailed test procedures for resilient
modulus. These influencing factors include gradation charac-
teristics (uniformity coefficient, effective and mean diameters
and coefficient of curvature), fines content and plasticity,
density, moisture content, cyclic stress amplitude and repetition
and confining pressure, etc.
Regression equations were formulated to relate resilient
modulus, individually, to R values, mean diameter, uniformity
coefficient, %fines content and plasticity index. However, the
number· of resilient modulus tests is too small and the
influencing factors too many to provide equations for effective
3
evaluation of resilient modulus in terms of some easily
determinable index properties of soils. Additional research is
needed for accomplishing the effort in characterizing the
resilient modulus of Colorado soils. This should involve a well
designed experimental program to provide a sufficient data base
for the statistical modeling.
4
II. LITERATURE REVIEW
11.1 AASHTO Pavement Design Guide
Based on the information provided by the AASHTO Road Test
(AASHTO, 1962) conducted between 1958 and 1961, AASHTO published
"AASHTO Interim Guide for Design of Pavement structures - 1972"
(AASHTO, 1972). This design guide takes into account of the soil
support by incorporating the soil support value and strength
coefficients in the design procedures. The soil support value
and strength coefficients are determined from the California
Bearing Ratio (CBR) and the stabilometer R-value, respectively.
The Interim Guide had served its main objectives for many years
without serious problems. The state highway agencies were
generally satisfied with the Interim Guide but acknowledged that
some improvements could be made.
After many years under the Interim Guide, the AASHTO Design
Commi ttee recommended that some revisions and additions were
required to incorporate the information developed since 1972.
This effort resulted in "AASHTO Guide for Design of Pavement
structures - 1986 {AASHTO, 1986)." This revised design guide,
while retaining the basic algorithms developed from the AASHTO
Road Test as used in the Interim Guide, adopts the resilient
modulus of soils for characterizing soil support and assigning
layer coefficients. The AASHTO test method T-274 was recommended
as the definitive test for evaluating the resilient modulus of
subgrade soils.
5
Additionally, the concept of reliability was
introduced to permit a designer to use the concept of risk
analysis for various classes of roadways.
II.2 stabilometer R-Value
The R-value (resistance value) is a parameter representing
the resistance to the horizontal deformation of a soil under
compression at a given density and moisture content. This
parameter is an indication of the ability of soil to carry a
load. The better can a subgrade soil resist horizontal
deformation under a traffic load, the less surface pavement
material is required to carry the design traffic load.
The R-value is derived from the result of a test (AASHTO T-
190) conducted in a Hveem stabilometer as shown in Figure II.1.
A cylindrical specimen with 4 inches (10.16 cm) in diameter and
2.5 inches (6.35 cm) in height is enclosed in a membrane. As a
vertical load of 2000 lb (8896 N) is applied over the full face
of the specimen to produce a pressure, Pv ' the resulting
horizontal pressure, Ph' is read. The vertical load is then
reduced to 1000 lb (4448 N), and the horizontal pressure is
adjusted to 5 psi (34.5 kPa) with a displacement pump. By
turning the pump handle to raise the horizontal pressure from 5
HEAD Of TESTING MACHINE
FOLLOWER fOR APPL YING LOAD TO SPECIMEN
NOT TO SCAL E
AIR CHAMBER MANUALLY OPERATE
f"""'~~ ~ SCREW TYPE PUMP ... ,,#, ." !O'¥' """"" :;"""" ~I
~j lIaUID UNDER SMALL r, -= j INITIAL PRESSURE .... flEXI BLE DIAPHRAGM
PLATEN OF TESTING MACHINE
NOTE-The specimen is given lateral support by the nexible sidewaU. which transmits horizontal pressure to the liquid. The magnitude of the pressure can be read on the gauge.
Figure II.l Schematic diagram of Hveem stabilometer.
0'1
7
to 100 psi (34.5 to 689.5 kPa), the number of turns is recorded
as the turn displacement, 0, of the specimen. The Stabilometer
R-value is then determined using the equation below:
R • 100 ~ 100 ,
(1)
where Pv is usually 160 psi (1103.2 kPa) .
The R-value provides the information of relative quality of
subgrade soils and empirically relates to the field performance
of pavement materials. It has been correlated with CBR, soil
classification, and other properties of different soil types.
The R-value test, while being time and cost effective, does not
have a sound theoretical base and it does not reflect the dynamic
behavior and properties of soils. Its development and use are
mainly based on trial and error, previous experience and
observations, and engineering judgments. The R-value test is
static in nature and irrespective of the dynamic load repetition
under actual traffic.
II.3 Resilient Modulus
A pavement structure is designed to sustain millions of
repeated wheel loads during its service life and it is more
realistic to obtain and use in pavement design the repetitive
8
load property of pavement materials. The resilient modulus, MR,
in AASHTO T292 and T294 is an elastic rebound stress-strain
relationship. It measures the elastic rebound stiffness of
flexible pavement materials, base courses and subgrades under
repeated loading.
At a point under investigation, each moving wheel imparts a
dynamic load pulse to all layers of a pavement system. It is
followed by a period of zero dynamic load, relaxation period,
before the next moving wheel arrives that causes the dynamic load
cycle to be repeated. Therefore, the resilient modulus is
determined from a repetitive pulse load triaxial compression test
(AASHTO T-274) with a suggested pulse duration of 0.1 second
followed by a 1.9 second rest period.
The test is conducted on a cylindrical specimen with 2.8 or
4 inches (7.11 or 10.16 cm) in diameter, depending on particle
size, and at a desired density and moisture content. Before
starting a test sequence, a series of axial load repetitions
under various deviator stresses is applied for sample
conditioning. The test is then conducted by varying the deviator
stress and confining pressure, respectively, and by applying 200
repetitions under each load condition. The decreasing load and
recovered deformation at the 200th repetition of each load
condition are recorded for calculating the unloading stress and
recovered strain. The resilient modulus at certain stress
condition is defined as the secant modulus of the 200th unloading
curve and is expressed as
M • R
,
9
(2)
where 0d is deviator stress and Er is recovered axial strain at
the 200th repetition.
The resilient modulus test closely simulates the pavement
materials at different depths under various traffic loads in
field. It provides a fundamental dynamic stress-strain property
of materials in a flexible pavement system that governs its
response under a traffic load. However, the test needs
sophisticated and expensive equipment, which many state highway
agencies do not have. Based on the AASHTO procedure, the
required testing time to determine a modulus is at least 2.5
hours for cohesive soils and 4.5 hours for granular materials,
excluding the time for sample preparation and test setup. For
lack of equipment and being very time consuming, Yeh and su
(1989) indicated that it would be impractical to attempt a large
scale testing program for investigating every aspect of resilient
modulus of all types of Colorado soils. The revised guide
provided correlations of MR with CBR, R-value, and other soil
properties for some types of soil. Nevertheless, The AASHTO
still recommended that each state should develop its own
correlations due to the geographic dependency of resilient
modulus.
10
II.4 Factors A~~ectinq Resilient Modulus
A number of factors affect the resilient modulus of subgrade
soils. Previous studies by Seed, et ale (1967), TRB (1975),
Thompson and Robnett (1976), Thornton and Elliott (1986), Elliott
and Thornton (1987), etc. have shown the influence of various
factors on resilient modulus. Among these factors are grain size
distribution, plasticity, density, moisture content, compaction
method, freeze-thaw cycle, confining pressure, and deviator
stress. The effect of these factors is discussed below.
11.4.1. Index properties
The grain size distribution, fines content, liquid limit,
plasticity index, and group index may influence the dynamic
behavior of subgrade soils. Thompson and Robnett (1976)
performed a detailed study on the effect of these properties of
Illinois soils on their resilient moduli. However, this study
did not find any significant correlation between resilient
modulus and any single soil property.
11.4.2 Moisture content and Density
The influence of moisture content has been found to be
significant in some studies. As illustrated in Figure 11.2,
Thompson and Robnett (1976) presented the variations of resilient
modulus of the AASHTO Road Test subgrade. with six different
moisture contents ranging from o.~% below to 1.9% above optimum,
the resilient moduli decrease with the increase of moisture
11
16
AASHO'
Symbol w". MR (ksi)
14\ D 0-0.8 9.29 0 0-0.5 6.07 £ 0+0.4 3.51
• 0+0.5 3.27
12 X 0+ 1.3 2.89 + 0+ 1.9 2.4\3
,-... .... g Compacted y = 95". T-99
~
:;E 10 .
III ::J ::J
'"0 0 ~
-c cu -III 6 cu
.~
4
10 15 20 25 30
R~p~ated Deviator Stress. 0'0' psi
Figure II.2 Variation of resilient modulus with moisture contents ranging from 0.8% below to 1.9% above optimum (from Thompson and Robnett, 1976).
12
content. Robnett and Thompson (1976) studied the relationship
between resilient modulus and moisture content of two fine
grained soils and AASHTO Road Test subqrade. Figure II.3 shows
that the resilient modulus evidently decreases as the compaction
moisture content increases. Thompson and Robnett (1976) also
reported the influence of moisture content in terms of the degree
of saturation. As plotted in Figure II.4, the general trend of
resilient modulus decreases significantly with the increase of
soil saturation. Therefore, the selection of an appropriate and
representative moisture content for design, and the control of
moisture content during sample preparation and field construction
can be crucial.
Although Figure II.4 reveals the difference in the
relationships between resilient modulus and soil saturation for
95% and 100% of compaction, the degree of saturation reflects the
combined effect of moisture content and density. Robnett and
Thompson (1973) conducted tests on two different cohesive soils
at 1.7% and 2.5% over the optimum moisture content, respectively.
Each sample was compacted to 95% and 100% of the standard proctor
density. The results as presented in Figure II.5 indicates that
the higher density results in higher resilient modulus but the
difference in modulus is small. The investigation performed by
Elliott and Thornton (1987) also concluded that resilient moduli
of two Arkansas cohesive soils were not affected significantly by
the variation of density.
14~-----'------~--~--~-------r------,
12
DrlJ . en
6 /)]~e,. & a. 0 10 II ~. a 'il-b "0 - ~d'.
<:( 8 ~ q, ,..... 0(0 .-
~ '-' % I>:
/-. ::E 6 ~ - 0 en ~ 0 ::J
'S7J'~ ::J -0 0 0
02,<h :2: 4 :l"Q
C ~ (U
0 en Q)
0: 2
oL------L------L-----~~--~~----~ -1.0 Optimum +1.0 +2.0 +3.0 +4.0
(AASHO T-93)
Compact ion Moisture Content. ~o
Figure II.3 Relationship between resilient modulus and moisture content of two fine-grained soils and AASHO Road Test subgrade (from Robnett and Thompson, 1976).
1 3
i .~ u .. ... .. .. ~ a .. 3 ~
"a o :e c .. = ·iii ..
16
I~
12
10
8
6
a: ~
2
o
a
", a 0 ., rfJ
0 0 ", 0
\..,0 0 ~o ~ 0
~ P 0' 0 0 0 0
t> '" 0
~ ;> 0
~ 0 o 0 0 (
~ 1'\ " 0 "' R ..
rrA.~ o 0 0 0
0 0
.... LC o 0 Q,
0
/' 0
~ '0
00
o 0 0 0
o 0 n ~ 00 r---. 95 .,.
p OB ERi" 3~334 s,1 J 0 0 0 0 0
o 0 0 O~ R "0.64 Sit = 2.69 ksi 0 0 '0.
0 0
8 p -lOp 0 0 p 0 o 00
80
-0
60 70 80
"10 Saluration. S r
a /0'0.,. AASH~0-T99 o 95". AASHTO - T 99
1
0 1100 .,.
M R= 45.2 - 0 . 428 S
11'"0' n7.6 ''=~' '" ' -
a
1'" a'" a
, ° I' []o a
"" P o ~ ~
a 00 ~
O,~ 0" 000 0"":: 0
~ooo -"<
0 ~ 0
90 100
Figure II.4 Relationship between resilient modulus and degree of saturation at 95% and 100% of compaction (from Thompson and Robnett, 1976).
14
12
FlonaClon B, Optimum Moisture + 1.7 '7'0
10-- ... 95'7'0 AASHO T-99 Density 10
~IOO'7oAASHO T-99 Density
Wisconsinan Loam Till Optimum Moisture + 2.5'70
0---095'70 AASHO T-99 Density 8 0---0 100'70 AASHO T -99 Dens i t y
,-.. .... ~ NOTE: 0·0 3
c.: :a 6
." :::J
:::J "0 0 ::E
-C 4 cu
." cu cr
2
°O~------~---------ILO--------~15~-------2~O---------2~5--------~30
Figure 11.5
Applied Deviator Stress, ps i
Effect of density on resilient modulus of specimens compacted at 95% and 100% of AASHTO T-99 density (from Robnett and Thompson, 1973):
15
16
Some studies demonstrated that the method of compaction
affects the resilient modulus as well, The results of test using
static or kneading compaction method are compared in Figure 11.6.
Both Robnett and Thompson (1973) and Elliott and Thornton (1987)
found that the specimen made by using static compaction has a
higher resilient modulus than that using kneading compaction.
Generally, the kneading compaction produces soil structures
similar to that under field compaction, and generates more
consistent test results. Thus, the standard resilient modulus
test procedure CAASHTO T-274) specifies kneading compaction
(AASHTO T-99) as the method for sample preparation.
11.4.3 Climate
The variation in resilient modulus throughout a year is
expected because of the seasonal moisture changes of subgrade
soil. Based on the analysis of deflection measurements taken
during the AASHO Road Test, the seasonal variation in subgrade
resilient modulus is shown in Figure 11.7. The modulus has the
lowest value in the spring and the highest value in the winter
coinciding with freezing and thawing seasons. The spring thaw
modulus of roadbed is typically 10% to 30% of the summer modulus
while the frozen roadbed modulus is typically two orders of
magnitude greater than the summer modulus. As a result, the
roadbeds are softer in the spring than at other time of the year.
,..... ..... g
1 CII ::J
::J "0 0
:::E -.::::: CII
en CI)
a:
12~-------.--------r--------r--------r--------r--------'
Fayette B (95 "1. Density)
q .-til Kneading Compaction Opt. +2.5 "10
\ ...... Stalic Compaction Opl.-t-2.5"10 IO~----~~~-------;---------1
8
6
"'
Wisconsinan Loam Till (100 "I. Density)
0--0 KneadinQ Compaction Opt.-t-2.8"1o
0--0 Sialic Compaction Opl.+2.6"l0 ~--~~-r~------r--------;
NOTE: a l • 0
---2~----~~----~------~~----~-------t-------;
°OL-------~L-------~IO--------~1~5--------~2~O------~2~5~------~30
App l ied Deviator Stress, psi
Figure II.6 Variation of resilient modulus of specimens made by using static and kneading compaction methods (from Robnett and Thompson, 1973).
17
A &
-. .... ~ 7 '-"
i 6 .. .= ::J
"0 S 0 ~ Q C , ~ 4
, , .- , .. , • a: • s "0 0 ... 01 ..a ::J 2 en ,
\ • , ,
0 Jan
?\ . , I
, \
I , b.
I I I I
, . b'
Time of Year
~ I' I ~ • 1 : ,
I • • • . , I
f'J , • • I • • •
6 1958 o 1959 c 1960
(Only Time Periods in Which 4 or More SeclIOM Wen Available are Plotted)
1 8
Figure II. 7 Seasonal variation of subgrade resilient modulus during the AASHO Road Test.
19
It has been recognized that the freeze-thaw cycle has a
major impact on the resilient modulus of subgrade soil. Robnett
and Thompson (1976) investigated the effect of freeze-thaw cycles
on the resilient modulus of fine-grained soils. Figure 11.8
reveals that the first freeze-thaw cycle caused a dramatic
reduction in the resilient modulus and the subsequent cycles
caused additional minor reductions. Elliott and Thornton (1987)
also reported that one freeze-thaw cycle significantly reduced
the resilient modulus of three Arkansas soils.
II.4.4 stress states
The effect of stress state on resilient modulus of subgrade
soil has been studied extensively in the past two decades.
Howard and Lottman (1977) conducted resilient modulus tests on
four Idaho soils, including two fine sands, silt, and silty clay,
under various stress conditions. The test results demonstrated
that all four soils showed an increase in resilient modulus as
deviator stress decreases, and resilient modulus exhibited a
drastic increase for deviator stress less than 0.75 psi (5.17
kPa). It was also found that the resilient modulus of granular
soils was predominantly dependent on confining pressure. The
resilient modulus of soils, however, was not so sensitive to the
variation of confining pressure, rather was more a function of
deviator stress. Elliott and Thornton (1987) confirmed that in
15
TAMA B
14
12 jNUmbef or Freeze - Thaw ,-." Cycles ..... ~ '-" 0
r:r: 10 ~
.. II)
:J :J 8
"0 0 ~
- 6 c u .-en u
Q: 4
2
°0~-------4~------~8---------1~2--------~16---------J20
Repeated Deviator Stress t uo' psi
Figure 11.8 Effect of number of freeze-thaw cycles on resilient modulus of fine-grained soil (from Robnett and Thompson, 1976).
20
21
no case was the effect of confining pressure of major significant
for cohesive soils.
Because the resilient modulus varies with axial load and
confining pressure, the resilient modulus has usually been
plotted against deviator stress and confining pressure. Based on
the results of early resilient modulus studies, Thornton and
Elliott (1986) concluded that the modulus of granular base
materials had a positive linear relationship with the sum of
principal stresses on a log-log plot. As for cohesive soil, the
test results are reported in an arithmetic plot of resilient
modulus versus deviator stress at each confining pressure.
22
III. USE OF RESILIENT KODULUS IN PAVEMENT DESIGN
The surface deflection of a pavement results from the
accumulation of load induced strain within the pavement and
subgrade with the subgrade being a major contributor. In the
AASHO Road Test (AASHO, 1962), 60% to 80% of the deflection
measured at the surface was found to develop within the subgrade.
Thus, the resilient modulus of subgrade is a maj or factor
governing the surface deflection and the performance of flexible
pavement and the procedure for its evaluation is included in the
revised AASHTO pavement design procedure (AASHTO, 1986). In this
procedure, resilient modulus is used in determining an effective
resilient modulus as a direct input in pavement design and in
selecting layer coefficients in determining layer thicknesses.
The effective resilient modulus of a roadbed soil is an
average modulus weighted by relative damage and adjusted for
seasonal variations. The following steps are involved in
determining the effective resilient modulus:
1. Perform laboratory tests to develop a relationship
between resilient modulus, MR, and moisture content.
2. Estimate the seasonal variations of moisture content in
subgrade.
3. Determine the monthly or bimonthly resilient modulus for
a year from the above relationship.
23
4. Select a relative damage, u f , value for each seasonal
resilient modulus based on the MR-U f scale provided by the design
guide.
5. Calculate the average of relative damage values for the
year.
6. Select effective resilient modulus corresponding to the
average relative damage from the MR-U f scale.
The effective resilient modulus which accounts for the
combined effect of temperature, moisture, and seasonal damage is
then used in design.
The revised guide provides a design chart as shown in Figure
111.1 for the flexible pavement design. Reliability and overall
standard deviation are introduced ·at the beginning of the design
procedure. This uncertainty factor accounts for the combined
effect of the variation in all the design variables so that the
designer no longer need to use conservative estimates for the
other input parameters. The traffic factor is then considered by
entering the cumulative expected l8-kip equivalent single axle
load (ESAL) during the design period. In terms of effective
roadbed soil resilient modulus, the material properties and
environmental effects are then included in the design. By
incorporating the loss in design serviceability, which is the
change between initial and terminal serviceability indexes, the
nomograph leads to a structural number. "structural number" is
a parameter used in the design of layered pavement structure.
It relates to the thickness of each layer by a layer coefficient.
tOD:>RAFH SOINmz
10910 4.2 - 1.5 f
t. PSI ~ 1091:'18 .. ZR*So+ 9.36*I0910 (SN+1) - 0.20 + + 2.32*I091h - 8.07
1094
1.
99.9
_ 199
~ ~ It:
~ :g .~ 0; It:
10
70
60
50
Figure III.1
~l _i c~
0.40 + --519 (SN+l) •
Design Serviceability Lass, 6P51 ) .!!!. I!
L V I
~ .; -$~ o.~ =-=8 1._ IID--Ill. 0: 0'" 1- 0 ... ~ .. - .. D-E .. :::<t .n.! I
EKample'
WII : 5 K 10'
R=95 %
50 = 0.35
-.;; ~~ ... ::E .. ... -..... o " 0-a:~ .. 0 .?: ::E uc .f .! ... = w :
a:
M" & 5000 psi
6PSI = 1.9
Solution: SN = 5.0
...
~ /~~
LV ~ ,.
/ V ~ ~
/V ~ ~ / / /.
/ ~ 05/ /; ~ 1.'/ /~ til I.i',
2.0 % 'I 7 6 5 4 3 2
Deslg" Slruclural Number, SN
Design chart for flexible pavements based on using mean values for each input (form AASHTO, 1986).
I\J
"'"
25
The revised guide employs layer coefficients, ai' to express
the empirical relationship between structural number and layer
thicknesses, 0i' as formulated in the following equation:
(3)
Based on the resilient modulus of each pavement material, the
layer coefficients are individually estimated for asphalt
concrete pavement, base course,. and subbase layer. Once the
structural number and layer coefficients are determined, the
thickness of each layer can be calculated from the above
equation. This equation does not provide a unique solution
because many combinations of layer thicknesses all satisfy the
design load-carrying capacity.
26
IV. TESTING PROGRAM, PROCEDURES, AND RESULTS
IV.l Testing Program
Twenty resilient modulus tests were performed in this
research program. Initially, twenty two bags of row soils were
provided by the Materials Laboratory of Colorado Department of
Transportation (COOT). The bag soils were first sieved through
a stack of sieves including u.s. standard sieves #4, #8, #20,
#40, #60, #100, #200 and bottom pan. After sieving the soils,
the amount of fines passing #200 sieve was found to be sufficient
for preparing only eight samples, Nos. 1 to 8, of a desired
gradation characteristics for this test program. with the COOT's
agreement, additional soils were carefully selected and delivered
to the Geotechnical and structural Laboratory at the university
of Colorado at Denver (UCD) for preparing the other twelve
samples, Nos. 9 to 20.
In order to test a broad spectrum of soil types for this
research program, twenty soil samples were prepared of different
gradation characteristics with the mean diameter (050) ranging
from 0.0053 inches (0.135 rom) to 0.187 inches (4.75 rom), and the
coefficient of uniformity, Cu ' from 6.3 to 75.4, and the percent
fines content by weight from 6% to 32%. However, all samples
except Nos. 17 and 19 do not have measurable plasticity index and
their classification ranges from A-1-a to A-2-4 with the AASHTO
classification. The gradation curves of twenty soil samples are
27
reported in figures in Appendix A, and the AASHTO classifica
tion, gradation characteristics, and plasticity index, PI, of
each sample are included in Table IV.1.
Soil samples were mixed for compaction in the Geotechnical
and Structural Laboratory at the University of Colorado at Denver
(UCO). The compaction (AASHTO T-99) and R-value test (AASHTO T-
190) were carried out in the Materials Laboratory at the Colorado
Department of Transportation (COOT). A mechanical kneading
compactor was used to compact all soil samples and the correction
was then made for rock fraction. Results of the compaction test
including maximum dry density, Yd , and optimum moisture content,
~oPt' provided by the COOT are summarized on Table IV.1. The R
value of tested samples ranges from 48 to 81, and are also listed
on Table IV.l.
The cyclic triaxial test system at UCO was used in this
resilient modulus test program. An MTS-810 series closed-loop
electro-hydraulic universal testing machine with a capacity of
200 kips (890 kN). Figure IV.l shows the test system, which
consists of a stiff load frame, MicroConsole Model 458.20
providing the closed-loop control of the servo electro-hydraulic
system, and data acquisition apparatus. Other system components
include LVOT, load cell, actuator, servovalves, hydraulic
pressure supply, and hydraulic service manifold.
28
Table IV.1 Soil classification, gradation characteristics, and laboratory test results of each soil sample.
No Class. D50 Cu -#200 PI Yd cuopt R MR & G.I. (rom) (%) (pcf) (%) (ksi)
1 A-2-4 (0) 4.75 13.6 6 NP 132.6 5.7 78 13.6
2 A-1-a 4.75 22.8 6 NP 133.0 5.9 81 10.3
3 A-1-b 1.8 18.0 8 NP 128.0 8.0 72 28.8
4 A-1~b 1.6 25.5 8 NP 129.3 7.0 75 29.0
5 A-1-b 1.5 38.6 9 NP 130.6 6.6 78 13.0
6 A-1-b 1.3 57.7 13 NP 129.2 6.8 77 15.4
7 A-1-b 0.35 9.3 14 NP 122.7 9.6 78 10.0
8 A-1-b 0.54 16.0 14 NP 124.4 9.0 81 12.4
9 A-1-b(0) 1.9 41.3 10 NP 121.7 8.3 64 10.3
10 A-1-b(0) 1.68 11.2 2 NP 110.2 10.8 71 16.7
11 A-1-b(0) 0.63 11.1 . 9 NP 117.6 8.4 72 13.6
12 A-2-4(0) 0.14 7.0 28 NP 116.3 12.8 55 10.6
13 A-2-4(0) 0.135 9.3 32 NP 118.0 10.4 67 13.0
14 A-2-4(0) 0.25 9.5 18 NP 118.5 11. 7 73 9.6
15 A-2:"4CO) 0.24 6.3 15 NP 118.7 9.45 81 13.0
16 A-2-4(0) 0.4 25.0 22 NP 116.9 13.4 50 6.8
17 A-2-4 (0) 0.147 11.5 32 4 123.5 10.1 62 20.0
18 A-1-aCO) 2.5 75.4 11 NP 133.3 6.1 77 11.0
19 A-2-4(0) 0.3 30.5 27 7 129.1 8.2 48 12.6
20 A-1-a(0) 2.24 55.4 8 NP 132.0 5.8 81 12.0
NP: Non-plastic
Figure IV.l Cyclic triaxial test system MTS-801 at the UCD. I\)
\0
30
V.2 Procedures of Resilient Hodulus Test
Twenty specimens for resilient modulus tests were prepared
and compacted. A large size triaxial tests were used because
samples contained large particles of over 1 inch (25.4 rom). The
cylindrical sample dimensions were 6-inch (15.24-cm) in diameter
and 12-inch (30.48-cm) in height. Soils samples were first cured
at the optimum moisture content for 24 hours before testing. The
specimen was then compacted in a 6-inch (15.24-cm) diameter mold
by the standard Proctor method. To maintain the specimen
uniformity and to achieve the maximum dry density, samples were
compacted in eight 1.5-inch (3.81-cm) lifts. All specimens were
consolidated in the triaxial chamber under a confining pressure
for 24 hours. The procedure for sample preparation for resilient
modulus test is detailed in Appendix B.
Because of the non-plastic nature of soils used in this
research, the standard procedures for determining resilient
modulus of granular soils as described in AASHTO T-274 were
followed. During the test, 200 repetitions of axial load are
applied under each load condition. Each axial load repetition
comprises a O.l-second load pulse followed by a 1.9-second zero
load period.
The resilient modulus test for granular soils begins the
sample conditioning under a cyclic load: 200 repetitions of each
deviator stress, ad' of 5 and 10 psi (34.48 and 68.95 kPa) under
a confining pressure, 0c' of 5 psi (34.48 kPa), then, 200
repetitions of cyclic ad at 10 and 15 psi (68.95 and 103.43 kPa) ,
31
respecitvely, under a a c of 10 psi (68.95 kPa), and, finally,
another 200 repetitions of ad at 1"5 and 20 psi (103.43 and 137.90
kPa) are applied under a a c of 15 psi (103.43 kPa). This load
condi tioning eliminates the effects of the interval between
compaction and loading, initial loading versus reloading, and
initially imperfect contact between the end platens and the
specimen.
In the resilient modulus test, each sample was tested under
all combinations of a c of 20, 15, 10, 5, to 1 psi (137.90,
103.43, 68.95, 34.48, to 6.90 kPa) , and ad from 1, 2, 5, 10, 15,
to 20 psi (6.90, 13.79, 34.48, 68.95, 103.43, to 137.90 kPa).
Load and deformation were recorded throughout the test. The
recorded load and deformation for the 200th repetition of each
load condition are used in determining the MR under each stress
condition. The confining pressures, deviator stresses, and the
number of repetitions used during the sample conditioning and
testing are summarized on Table IV.2. The detailed procedure for
resilient modulus test is presented in Appendix c.
IV.3 Results of Resilient Modulus Test
The resilient modulus was calculated from the recorded axial
load and recovered deformation after complete unloading under
each load condition. since the samples tested in this research
were granular soils, the resilient moduli are reported in two
forms:an arithmetic plot of resilient moduli versus deviator
Table IV.2 Confining pressures, deviator stresses, and number of repetitions used in resilient modulus test for granular soils.
Confining Deviator Repetitions Pressure stress at
(psi) (psi) Each Load
5 5, 10 200 Condi-tioning 10 10, 15 200
15 15, 20 200
20 1, 2, 5, 10, 15, 20 199 + 1
15 1, 2, 5, 10, 15, 20 199 + 1
Testing 10 1, 2, 5, 10, 15 199 + 1
5 1, 2, 5, 10, 15 199 + 1
1 1, 2, 5, 1'0 199 + 1
32
33
stresses at various confining pressures, as presented in
Appendix D and a log-log plot of resilient moduli versus the sum
of principal stresses, e, as presented in Appendix E.
The arithmetic plots exhibit an increase in MR as O"d
increases from 5 psi (34.48 kPa), although the trend grows vaguer
at O"d below 5 psi (34.48 kPa). These plots illustrate the strong
dependency of ~ on O"c. This dependent relationship between the
resilient modulus and confining pressure for granular soils
agrees with the findings of previous research. The test results
also demonstrate a strong linear relationship between MR and e on
a log-log plot. The equation of least-square line relating MR
and e for each sample is presented at the bottom of each log-log
plot in Appendix E.
Yeh and Su (1989) concluded in their research report that
the MR under the stress condition of O"d of 6 psi (41.37 kPa) and
o"c of 3 psi (20.69 kPa) is the most appropriate value for
adoption in pavement design. From the arithmetic plots shown in
Appendix D, the ~ under such stress condition was determined for
each sample and was included in Table IV.1. These values of
resilient modulus are used in developing the functional
relationship between the MR values and R values and index
properties of soils. To enlarge the database for the statistical
modeling of MR , the results of the nineteen samples from the
previous test program (Yeh and Su, 1989) with soil types ranging
from A-1-b to A-7-6 per AASHTO classification are included in
this study. The AASHTO classification, fines content, plasticity
34
index, PI, maximum dry density, Ydl and optimum moisture content,
Wopt' resilient modulus values and R values of these samples are
included in Table IV.3.
35
Table IV.3 Soil classification of soil samples and laboratory test results from previous test program (From Yeh and Su, 1989).
No Class. -#200 PI Yd Cl)opt R MR & G.I. (%) (pcf) (%) (ksi)
21 A-7-6(17) 69 28 108.0 17.8 6 3.5
22 A-7-6(2) 42 22 108.5 17.1 15 4.2
23 A-6(11) 69 19 109.1 16.4 11 -4.6
24 A-6(2) 44 13 110.8 15.2 30 8.4
25 A-4(1) 42 10 119.0 11.6 26 7.8
26 A-2-6(1) 25 14 119.9 11.1 39 10.5
27 A-2-4(0) 20 10 116.2 12.1 41 6.4
28 A-1-b(0) 25 4 130.0 7.2 34 8.5
29 A-2-4(0) 23 9 115.4 13.5 37 11.2
30 A-4 (3) 57 9 114.4 14.4 37 7.2
31 A-2-4(0) 34 8 116.7 12.8 42 10.3
32 A-4(0) 44 6 116.2 12.1 39 7.7
33 A-4(0) 36 9 117.9 16.0 40 6.8
34 A-4(0) 48 1 120.3 11.3 70 8.7
35 A-2-4 (0) 16 NP 119.9 10.9 77 8.6
36 A-1-b(0) 10 NP 118.2 6.2 79 15.5
37 A-1-b(0) 10 NP 127.7 8.0 62 11.0
38 A-1-b(0) 17 3 120.9 11.3 72 8.7
39 A-1-beO) 9 NP 129.9 8.5 80 21.9
NP: Non-plastic
36
V.l STATISTICAL MODELING
V.l Introduction
Descriptive statistics and inferential statistics are the
two major branches in statistics. The descriptive statistics
deals with summary and description of data.
statistics concerns with analysis of sample
The inferential
data to make
inferences about a large set of data - a population, from which
the sample is selected. Experimental research in engineering
involves the use of experimental data - a statistical sample, to
infer the nature of some conceptual population that characterizes
a phenomenon of interest to the .experimenter. One of the most
important application of inferential statistics in engineering
involves estimating the mean value of a response variable or
predicting some future value of the response variable based on
the knowledge of a set of related independent variables. A
relationship used to relate a dependent (response) variable toa
set of independent variables is generally referred to as a
regression model or a statistical model (Mendenhall and Sincich,
1991) • The regression modeling was used in this study to
formulate the functionial relationship between the resilient
modulus and their influencing parameters.
V.2 Reqression Analysis
Regression equation relates a dependent variable, Y, to a
set of independent variables, Xi' where i is a nonzero integer.
37
In this study, the dependent variable, resilient modulus, MR , is
related to the independent variable, stabilometer R-value and
index properties of soils including uniformity coefficient, Cu '
mean diameter,Dso ' % fines content, FC, and plasticity index, PI.
The least-squares approach is used to determine the best estimate
of a regression equation. In general, the least-squares method
chooses the best-fitting model which minimizes the sum of squares
of the distances between the observed responses and those
predicted by the fitted model. Once a regression model is
obtained, it is desirable to test the contribution of each
independent variable involved in predicting the response variable
so that the model may be refined.
V.3 Regression Hodel between ~ and Stabilometer R-Value
Over a period of eight years, UCD has conducted resilient
modulus tests on 39 different samples under the CDOT sponsorship.
The data base from the results of the resilient modulus test on
these samples is used in the statistical modeling. The type of
soils tested ranges from clay to granular soils with various
percentages of fines of minus #200 sieve. Besides the R-value,
the gradation characteristics, including coefficient of
uniformity, coefficient of curvature, and percent fines content
by weight, are included as independent variables in the analysis.
However, only 20 out of 39 samples have the information on
gradation characteristics.
38
The regression analysis was performed to formulate the
functional relationship between MR and R-value and index
properties. A commercially available statistical graphics system,
"STATGRAPHICS" developed by statistical Graphics corporation was
used. As shown in Figure V.1, data points of MR versus Rare
quite scattered, particularly at the R-value above 60. Thus, it
is extremely difficult to use one regression curve to represent
the complete population of data points. After many experiments,
the "best" regression equation is:
log MR - O.llB + 0.517 . log R , (4)
or
M _ 1. 312 . R 0.517. R (5)
The above statistical formulation gives the intercept of
0.118, the slope of 0.517, the R2 of 42.79% and the standard
error of estimate of 0.155. The value of R2 is much less than
desired. This is mainly due to the scatter of data points.
Figure V.1 shows the regression curve, upper and lower bound
curves for one and two standard deviations, respectively. The
predicted resilient modulus, MRP ' is plotted against MR obtained
the laboratory test in Figure V.2. The linear comparison between
them has a R2 value of 25.11%. Again this low R2 value reflects
the scatter of data points.
34.67
21.88
13.80 -.. .... ~ '-"
Pi:
:E 8.71
5.50
3.47
Figure V.l
.j ......... ...... ..... . } ...... .. ... ......... + ......... · · ··········1····· ···· ············i········ ·· ···/' ····· ~ ..
• • • 1 )/// .
: : : : /. /: 4~"""""""""'" .~ •• ••••• - •• ••••• ······"f··········· ..... .... . ~ ...... ,oJ " ......... ~ .... .......... .. )".~ ..
• : • /,,/ 0 • /// :
! ! l / 0 i / ~ ~ ~ / ~/ /. :
iU J~~>j0.UJ.:U~:/" /> ~ , .: .............. /.. .... : ....... ... ............ ~ ......... ,/. .. --:. .... : ....... ( ............ /) ...... .. ............. i..
~ / ~ : /00;/ . ;.
: /: ;/:" / : / : /. r, · . / L: r ~ // / o ~ 0
· . / /:0
..... /
/
:...-...-; /
... ... ?; ... I. .. ............. ~ .............. -/-. :':'oi ...... .......... ..... ~ .. · ...-. ." .......... ~ ........ -:.- ........... .
0 /
/
;/ /
/; / / ; /
..... . /. . . . .~ •••.. 0 •..• _ .•••••••• ~ ••••• ~ _, •••••••• ••••• ~ ••• 0_ •• •• '" •••••••••• ~ •• ••••••• _ •• ••••••••• !, ........ _ .... .. ..... ~ ..
5 10 20 40 80 160
R-value
Scatter plot of log ~ versus log R with regression curve, upper and lower bound curves for one and two standard deviations, respectively.
39
13 . 3
11. 3
'1.3
7.3
5 . 3
3.3
Figure V.2
. . . -... -.. ... ... ... ;: . -. . . --....... .. - _. ~ . .. -............ ".
[] :[] [] [] []
.:. [] [] []
[] [] []
[]
[] 0:
0 0 0
[]
0
• ••••• •• •••••• • •• "!,C' ••••••••••••••• ..... ..... -- .. .. .. ;. ..... ...... ...... ; ..... . .. ..... __ ... : .. · . · . · . []
· . · . · . · . · . · . · . · . · . : 0 · . · . · . · . · . o · . · . · . · . [] · . · . · . · . . . . . .~ ................. -: .. ............ .. . -:' ................. ~ ..... ............ -:: -... .. ....... ... .. ; .. ... ... . _ .. .. ... ":-' · . . . . . . : : :0 : : : : • • 0 • - • .
~ ~ [] 0 :0 ~ ~ ~ ~ : : 0 : 0 : • : : ., . .. . . . . . .. . . . : : D: : : . , . . . .. . . , .. . . ' . . . . , . . . . . • • D . • • . , . .' . , . .. .. -i- o .............. . .. ~ ••• 0 •••• 0 •••• •• •• i··· · 0 •••• •••• 'o ••• ~. 0 ••• 0' , _ ••• • •••• ~ ••••• 0 •• •• •• _oo • •• ~ • • o . o . 0 0 ••••• •••• :' .
: : [] : : : : : : : : : : : : : : : : : : : · . . . . . . · . . . . . -· . . . . . -; ~ ~ ; ~ i ~ : : : : : : : · . . . . . . · . . . . . . · . . . . . . · . . . . . . · . . . . . . · . . . . . .
.\ •••• ••• ••••••• D •• j. ••••••••••••••••• \ •••••••••••••••••• i •••••.. ... . . . • •. . ..:; . .. ..•. .. .••• . .... ; . ···· . · .• ····· .••. .: ••
· . · . ., . , .: ............ ~ .... : ................. : ................... : ....... . .. ... .. .. ; ...... .... ... ..... : .. .. .... ...... "':"
o 5 10 15 20 2S 30
comparison of predicted ~ wi th 1\ obtained from laboratory test.
40
41
Linear regreeeson analyses were performed to formulate the
functional relationships between the resilient modulus and R
value and each of the following index properties: percent
fines content, FC, plasticity index, PI, uniformity coefficient
, Cu ' and mean grain size, D50 • Results of these linear
regression are shown in Figures V.3 to 7. Values of R2 for these
regression analyses are quite small.
The above-mentioned regression functions are less than ideal
because of the scatter of data points. The scatterness of data
points is caused by the fact that many factors can affect the
value of resilient modulus and one single independent parameter
is simply insufficient as a predictor. Instead, all significant
influencing factors should be included in the formulation of the
regression equation as independent parameters. These factors
include gradation characteristics, liquid limit, plasticity
index, confining pressure, deviator stress amplitude, number of
loading cycles, etc. This will require additional tests with
well documented index properties and test conditions.
30
25
~20 (/)
~15 c::: ~ 10
5
0
MRvsR f).f).
Rearession output: Constant 3.20046 1
Std Err of Y Est 4.659746 R SQuared 0.32965 No. of Observations 39 .... __ ._._._._. . ......... _._ ............... _._._.6 ... _ ............... _ ............. _ .... _._ ......... f). ._._ ............... _._._ .....•......•.. -.-._._ ........... __ . Degrees of Freedom 37
X Coefficient(s) 0.14543 Std Err of Coef. 0.034094
_ .. _. __ ._ .......... _ ...... _._._ ............ -._._ ... _ .... --............. -...... ~ ... -.-.-.-.~ ............ -.-.-._ •..........•.. _.- ._._ ..... -
f). - ~~ -._._ .... _ ...... _ .... _._ ............... __ ... _._ ...... -.-.--...... - .... -.-.. -......... --.--.---::~ ... ...c:.._ ..... ~ .. ···--··LS·-·-···~·-·A.·-·-····-····-·-······· .. ·········-._ ............... _._.
f). f). f). .
6.
0 20 40 60 80 100 R
Figure V.3 Linear Regression Equation of Mr vs. R-value
~
N
30
25 ~20 (/)
c15 Cl:: ~ 10
5
0 0
~ M~vs-#200
Recression OutDut: Constent Std Err of Y Est R Squared No. of Observations ....................•.............. ~ ..•......................................................•.............. ································6····················· .. _ .• Decrees of Freedom
f:::, ~ Coefficient(s) Std Err of Coef.
f:::, f:::, f:::,L->~f:::,
- ................ ~ ............ ~ ........ 6 ........... zs: ... _._ ....... _ .... A ..... _-=' •.•.... ~. ~ . f:::,
f:::,L:::. L:::.
f:::,
10 20 30 40 - #200 (0/0)
50
-0.18359 0.04375
60
Figure V.4 Linear Regression Equation of M vs. Fines Content r
16.00795 4.684633 0.322471
39 37
70
of>, w
30
25
:=-20 en ~15 0:: ~ 10
5
0 0
MRVS PI·
Rearession Output: Constant . 13.53223 Std Err of Y Est 4.847377
............................................ & .............................................................................................................................................................. .. . R Squared 0.274578 No. of Observations 39 Degrees of Freedom 37
. ~ Coefficlent(s} -0.42362 Std Err of Coef. 0.113198
6. 6.
~ .... __ .... _ ...... _ .. _ ..... _ ... _ ............. _.6. ......................... __ ............ _ .... _ .... _ .......... _ ...... _ ................ _ .... _ ...... __ ._ ...... _ ..................... _ .. _. __ . __
6.'6. 6.
5 10 15 PI
20 25 30
Figure V.S Linear Regression Equation of M vs. Plasticity Index r .;:.
01:>0
30
25 ..---.. --~ 20 "'-"
0:: 15 ~
10
5 0
6. 6. MRvs. CU
Rearession Outout Constant 14.n62 std Err of Y Est 5.918286 R Squared 0.00889 No. of Observations 20
.. __ ..... _ .......... _._._._ ............. & .. _ ............ _.-._._._ ............ _._.-._ ............. -._.-._ ....... __ ._._._._._ ............ __ .-._ .... -........ _.-.... _. Degrees of Freedom 18
IX Coefficient(s) -0.02793
----.:-.-.. -.----;----;-.;;-----~-:-------~---6.6. 6. 6. 6
6 6 6 6
Stet Err of C~ef. 0.069501
_._._ .............. _Q_ .. ~" ... _._._ .......... _._._._._ ........... _ ... .6. ... __ .... _._ .... _ ....... _ ... _. __ ._._ ...... _._._._ ..... _6 ._. ___ ............ _ ._._ ............ _._._._ ......... " ... _._ ........ _ .. _. __ . __ ....... _ ... _._._._. __ ..... _ .. _. __ ._. __ ..... __ ._.
6
10 20 30 40 Cu
50 60 70 80
Figure V.6 Linear Regression Equation of M vs. C r u
.r::.. lJ1
30
25 ..-.... .-~ 20 ...........
0::: 15 ~
10
5
6. 6. MRvs. D50
Regression Output: Constant 13.79n Std Err of Y Est 5.937036 R Squared 0.002601 No. of Observations 20 Dearees of Freedom 18
I .... ;6 .................................................................................... _ ........................................................................................... _ .. .
IX Coefticient(s) 0.21162 Std Err of Coef. 0.976782
6. _ ..... .................... _ .......... A .......... ..... - ... _ .................. - ... . ..... __ .... _. 6. ....................... . .... . . • .~ 1:::,.
" "'" " -" -.,,--z;"'---:-------------------------"----------<:---------------------------------------- ------1:::,.
0 1 2 3 4 5 D50 (mm)
Figure V.7 Linear Regression Equation of Mr vs. DS O oj:>.
0'1
47
VI. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
VI. 1 Summary
The AASHTO Guide for Design of Pavement structures - 1986
uses the theory of elastic layers in the design of pavement,
which, in turn, requires the resilient modulus, MRf of pavement
materials, base course and subgrade. However, its implementation
has met some technical difficulties resulted from 1) the lack of
and the complexi ty of testing equipment, 2) the scarci ty of
resilient modulus data base, 3) the soil type dependency of
resilient modulus, and 4) many influencing factors of resilient
modulus.
Because of the recent nature of its implementation, the Mr
data base is extremely small. Conversely, the StabilometerR
value has been in use in the pavement design for decades and has
a very large data base. To remove this shortcoming and to take
the advantage of the large R-value data base, numerous attempts
have been made to establish the correlation between Mr and R
value and to establish their functional relationship. The state
of Colorado has also invested in this endeavor through the
sponsorship of the resilient modulus research at the Univeristy
of Coloradoat Denver by the Colorado Department of
Transportation. A total of 39 resilient modulus tests were
conducted, 20 in this test program and 19 in an earlier test
program.
48
Regression analyses were performed to formulate the
functional relationship using the least square fit. Two
regression equations were formulated to relate the resilient
modulus to R-value (one linear and one nonlinear) and four linear
regression equations relate the resilient modulus to percent
fines content, plasticity index, uniformity coefficient, and mean
grain size. These equaitons , however, have small values of
coefficient of representation called R square. The use of the
equation relating the resilient modulus to the R-value can give
the value of resilient modulus smaller than that obtained from
the in-situ deflectometer test on subgrade and base course.
V~.2 Conclusions
From the data base of 39 samples with the soil type ranging
from A-1-a to A-7-6, two regression equations, one linear and one
nonlinear, were formulated in which the resilient modulus was
wri tten as a function of the stabilometer R-value. These
equations do not represent significant improvement over the one
provided in the study by Yeh and su (1989).
Four linear equations were also formulated to express the
resilient modulus in terms of fines content, plasticity index,
uniformity coefficient and mean grain size. The values of the
coefficients of representation are also small. This implies any
one single independent parameter is ineffective as a predictor.
The correlation is weak for samples of granular soils
containing varying amounts of fines of up to around 30%. For
49
granular soils with less than 30% fines, the index properties of
granular constituencies may have an important influence on the
resilient modulus of soils. These influencing factors include
gradation characteristics, relative density and grain shape of
the granular soils and percent fines content and fines
c6nsistency. These factors are responsible for the scatter of MR
values at the stabilometer R-value greater than 60 and the poor
functional relationship.
The inability of the Stabilometer R-value to realistically
reflect the engineering properties of granular soils with less
than 30% fines has also contributed to its poor functional
relationship to resilient modulus.
To appropriately evaluate the resilient modulus of soils
requires a sophisticated universal testing machine capable of
simulating the pulsating load from traffic, and continual
monitoring of lateral deformation, vertical deformation and the
amplitude of pulsating load, etc.
To strengthen the equation, the multiple regressioin analysis
is needed to formulate an equation expressing the resilient
modulus in terms of a number of significant influencing factors.
VI.3 Recommendations
1. Use the MR vs. R-value equations with caution, particularly
for granular soil samples with less than 30% of fines.
50
2. Conduct a systematic research to study the influence of
various factors affecting the resilient modulus of soils.
For soils containing less than 30% fines of minus #200
sieve, the factors should include: gradation characteristics,
relative density, grain shape, fines contents and plasticity,
degree of saturation, confining pressure, cyclic stress
amplitude, and number of cycles of repeated stress. For
clayey soils, the effect of density, moisture content, liqid
limit and plastic index should investigated.
3. When formulating the regression equation for MR, it would be
more reasonable to separate the granular (or nonplastic)
soils from the clayey (or plastic) soils because of the
different factors affecting the resilient modulus of these
two types of soils, as discussed above.
4. Mutiple regression analysis should be performed to formulate
the resilient modulus in terms of all significant
influencing factors. With the combined effort of Items 2, 3
and 4, the authors of this report believe that a good set of
re9ression equations can be formulated to effectively
evaluate the resilient modulus in terms of factors that can
be easily determined.
5. It is recommended to abandon the effort to formulate a
functional relationship between resilient modulus and
stabilometer R-value because of the inability of R-value in
reflecting the dynamic behavior/properties of soils.
51
6. Develop design charts for the selection of resilient modulus
with given index properties of soils.
7. Hold training sessions in different regions of COOT, cities
and counties in Colorado for the purpose of effecti ve
technology transfer.
8. To strengthen the equation, the multiple regressioin analysis
is needed to formulate an equation expressing 'the resilient
modulus in terms of a number of significant influencing
factors.
52
REFERENCES
AASHO, "The AASHO Road Test, Report 5 - Pavement Research," Special Report 61E, Highway Research Board, 1962.
AASHTO, "AASHTO Interim Guide for Design of Pavement Structures -1972," American Association of state Highway and Transportation Officials, 1972.
AASHTO, "AASHTO Guide for Design of Pavement Structures - 1986," American Association of State Highway and Transportation Officials, 1986.
Elliott, R.P. and Thornton, S.I., "Interim Report - Resilient Properties of Arkansas Subgrades," Report No. UAF-AHTRC-86-002, Arkansas Highway and Transportation Research Center, University of Arkansas, Fayetteville, Arkansas, 1987.
Howard, H.R. and Lottman, R.P., "A Test Procedure and Analysis of Resilient Modulus for Highway Soils," Proceedings of the 15th Annual Engineering Geology and Soils Engineering Symposium, Pocatello, Idaho, 1977, pp. 243-256.
Mendenhall, W. and sincich, T., " statistics for Engineering and the Sciences," Third Edition, Dellen Publishing Company, San F~ancisco, 1991.
Seed, H.B., Mitry, F.G., Monismith, C.L., and Chan, C.K., "Factors Influencing the Resilient Deformations of Untreated Aggregated Base in Two-Layer Pavements Subjected to Repeated Loading," Highway Research Record, No. 190, National Research Council, 1967.
Robnett, Q.C. and Thompson, M.R., "Interim Report - Resilient Properties of Subgrade Soils, Phase I - Development of Testing Procedure," Transportation Engineering Series, No. 5, University of Illinois, Urbana, Illinois, 1973.
Robnett, Q.C. and Thompson, M.R., "Effects of Lime Treatment on the Resilient Behavior of Fine-Grained Soils, II Transportation Research Record, No. 560, Transportation Research Board, 1976.
Thompson, M.R. and Robnett, Q.C., "Final Report - Resilient Properties of Subgrade Soils," Transportation Engineering Series, No. 14, University of Illinois, Urbana, Illinois, 1976.
53
Thornton, S.I. and Elliott, R.P., "Resilient Modulus - What Is It ?" Proceedings of the 3 7th Annual Highway Geology Symposium, Montana Department of Highways, Helena, Montana, 1986, pp. 267-282.
TRB, "Test Procedures for Characterizing Dynamic Stress-strain Properties of Pavement Materials," Special Report 162, Transportation Research Board, National Research council, 1975.
Yeh, S.T. and Su, C.K., "Final Report - Resilient Properties of Colorado Soils," Report No. CDOH-DH-SM-89-9, Colorado Department of Highways, Denver, Colorado, 1989.
54
APPENDIX A
GRADATION CURVES OF SOIL SAMPLES
· o z Q)
0..
E o
U1
I 0-. « 0:::: C)
W N U1
W --.J U I-0:::: « D....
r-"
I~ ~ ~
-=" ..... r-..
.......... ~
"-" ~ ~
I I J I I I I I I I I I
00 ~ 09 09 Ot Ol
J8 11DWS +U8JJ8d
..1 J
~ C
I I .1
n
..
o o ~
E .. E '--"
Of
Q)
r-N
..
n
N
~
o 00
(f)
Q)
o .-
55
. o z Q)
0.-
E o
if)
oo~
, ~
"'" " ......... "-...
~
~
l I
I I I I I I I I I I I I I
oe 09 o? OZ
J8 11DWS tU8JJ8d
" \ \ ~
\ ~
;:J_
I I I
n
..
•
..
..
..
..
..
o o .....
0 ..... r---.
E E
'---""
ill ...... N .-
if)
ill U
+J L
0 ~o... 0
,..... o
0 0
56
· o z (])
0...
E o
Ul
I 0.<C 0::: C)
W N Ul
W -.-J U r 0::: <C 0.-
1\ :-... ......
I I I
OO~
-
:-.... i"-.,. ~
~
" ~ "'-~
" .'\. '" ~ ~
I I I I I I I I I
OB 09 Ot OZ
J 8 1IDW S tU 8 JJ 8 d
n
N
-- n
N
n
" ~ N
..
.. I I I
o o .....
8,..--.... E E
'-.-/
OJ ..... N .-
(f)
OJ
U -+-l L
0 '":(l 0
..... o
0 0
57
· o z Q)
0..
E o
(f)
I (L
<C 0:::: C)
W N (f)
W --.l U ~ 0:::: <C CL
~
I I I
OOL
'" " " "'-"" ~ -~
"-'\.
"-)g
'\ ~
~ !:I-
I I I I I I I I I ~ I I
09 09 at oz J8 11 DWS tU8JJ8d
..
..
o o .....
o ..... ,..--.....
E .. E
..
..
'--"
Q)
..... N if)
Q)
U .. 4-..1
..
..
L
o '":[L o
..... o
oei
58
· o z Q)
0...
E o
(J)
W N (J)
W ~
U ln:::: « (L
. -
, ~
"-
" '" -~ '\
~
~ "-.'\ ~
~ 'S,
\ !;L
r--
J I I .l ~ ~ ~ i ~ J ~ ~ ~ ~ ~
OO ~ OB 09 Ot Ol
J8 11 DWS +U 8 JJ 8 d
..
N
o o .--
E .. E
N
..
..
N
"-../
.--N
.--o
if)
Q)
o .-
0 0
59
CD
. o z Q)
0..
E a
(f)
"-~
['\,. .'1.
"-"-
r----" ""-I\,
_\
\ -.!:II.
"' ~ f'\
m \ , " ~
I I I I I I I I I I 1.1
DOL OB 09 Ot Ol
J 8 11DWS tU8JJ 8 d
.I:.!.
11 i
..
..
..
..
o o .-
o ...-r---.
Q) .... N .-
(f)
Q)
u N -+-'
L
n
N
a ~~ o
,..-o
0 0
60
· o z ill
D-
E o
U1
I CL « ~ (9
W N U1
W .-1 U I~ « D-
\ Iw,...
'\. ~
I I I
OO~
"" ~ ~
" '& ~
"" ~ ~
I I I I J I I I L
OB 09 at O~
J8 11DW S t U8 :JJ8d
!:L
~I J_
..
..
a a ~
E .. E ..
..
'--'"
ill .... N .-
U1
ill U .-... -+--'
..
..
L o
--:D-a
.-a
O ci
.. 61
. o z (l)
D-
E o
(f)
I (L
« ~ C)
W N (f)
W .-J U ln:::: « 0....
(
...... ........
"-"'\
I I I
DOL
-
. - .
"-~
~ ......
........
" '&
"" .~ ~
-
I I I II I I I 1
DB 09 Ot OZ
J8 11DWS t U8 :JJ8 d
..
...
..
...
..
..
.j;
-- ..
...
1 I I
o o .--
0 ..... ,.--.....
E E
'--'
(l) .... N .-
if)
ill U
-4-J L
0 ~D-0
.--o
D ei
62
m .
o z Q)
0..
E o
(f)
I (L
« 0:::: C)
W N (f)
W _J U r 0:::: « (L
~.
(
~ '--.
I I I
aa ~
.. -
..
o o ......
a ...... ~
.. -E ........
i .........
............
~
'" .. E
""'--"
""-~
i'...
.. Q)
...... N ~
......
" .-(f)
.........
"l>1 ..
Q)
\
\ \
U .. ...J-I
L
0 "":(L 0
c:.J --n
-.
.. I I I t I I I I I I I I ......
o 08 09 at OG o ci
J81 1DWS tU88J8d
6 3
o
. o z ill
0..
E o
Ul
I CL <{ n:::: C)
w N Ul
W _.-J U ln:::: <{ (L
-.
l\
" r---
I I I
00 ~
"-
~ ~
"'-.... ~
~ ~
......
"' .......
. ..1.
I I I I I I I I I
DB 09 Ot Ol
J8 11DWS tU8JJ8d
');,
\ \
. \
I I I
c:1
..
..
o o .-
E .. E
..
..
ill .-N .-
(()
Q)
U .. -+-'
L
..
..
o --:D-o
.-o
0 0
64
· o z Q)
0...
E o
if)
I CL « 0::: C)
W N if)
W --.J U I-0::: « D....
-
cy--- -
-~ ~
............
~ -........
"'-............
i"& ~
"" "
..
-- ..
..
..
..
o o .-
~...--..
E E
"'--'
ill .- N .-
if)
ill U
+-' L
~ 0 "":(L 0
~
n
--r-I
.1-------- .. I I I 1 I I I 11 J1 I i J1 .-
o OOL OB 09 Ot Ol Od
J 8 11 DWS f U8 :JJ8d
65
. o z Q)
0...
E o
(f)
--.
.-
cr-- _.-
l \
\ \ \ A ~ ~ "~
-- ~
---. -- --
i I I i i i J 1 i i .i 1
OO L OB 09 Ot Ol
J8 11DWS tU8JJ8d
I i i
..
..
..
..
..
..
..
N
o o .-
~~
E E
"--/
Q) .... N .-
(f)
Q)
U -4--1 L
0 '":0..... 0
.... o
0 0
. 66
· o Z
Q)
0...
E o
(f)
I (L
« 0:::: C)
W N U)
W --.J U ~ 0:::: « D-
Ll.
~. \
\ \..
I I I
OO ~
-
- ..
~ _\.
I'I:l
'" ~ ~
....... tI
I I I I I I I I I
OB 09 at Ol
J81 1DWS tU88J8d
I I I
n
Of
o o ~
E n E
Of
..
n
Of
'--'"
Q) .... N
.... o
.-(f)
Q)
U
0 0
6 7
· o z Q)
0...
E o
U1
I CL « cr:: (9
W f'-J (f)
W ---.J U ln::: « D-
--- - .-
- -
y
_\ \
~ "'-
........
1---- ."& ~ ~
~ - . I~
-
J. J J J I I I 1.1 L ..1 t
DOL 08 09 ot OZ
J8 11 DWS +U8JJ8d
-.
.l
o o .,....
---.-
-
.11
o ..... ~
E n E
n
"--'"
Q)
...... N .-U1
Q)
U 0-
... --I-' L
..
...
o '":(L o
.,.... o
0 0
68
L[)
. o z Q)
0...
E o
(f)
I (L
<C 0::: ()
W N (f)
W -.J U ~ 0::: <C (L
f-.
~
oo~
\ \ ~
\. ~
" "-"-
)g. -I'.... ~
~
I I I I I I I I J J J J
De 09 Ot 0,
J8 11 DWS tU8JJ8d
~-
J J J
n
..
o o -
E n E
..
..
'---"
Q)
.-N (f)
Q)
U .. ~
L
..
..
o ~(L o
.-o
00
69
ill
. o z Q)
0..
E o
(f)
I (L
« 0::: C)
W N (f)
W ---1 U I-0::: « 0_
• -
11 ! \ -."
\ \ ~
~ ~
r--"- ...... , ..........
~
"-"-
"'~ ~ c:J
I I J I I I I I I I I I
DOL 09 09 ot Ol
J8 11 DWS "fU8JJ8d
I I I
..
..
o o .....
o ..... ,..--....
E .. E
..
..
Q) ..... N .-
(f)
Q)
U .. +-'
L
..
..
o r:o.... o
..... o
00
70
· o z Q)
CL
E o
(f)
I (L <{ n::: C)
w N (f)
W --1 U ln::: « (L
c
\
OOL
.--
\ \ "( -\ \ \
'\ \ \ k
"'" " ~ ~
- CJ
I I I I I I I ·1 I I II
08 09 ot Ol
J8 11 DWS tU8 :JJ8d
l J I
..
o o r-
E .. E
..
'--""
Q) ..... N .-
(f)
Q)
U .. -+-'
L
..
..
o r:(L o
.-o
Od
71
· o Z
ill 0..
E o
if)
I 0.« ~ C)
W N (f)
W --1 U ~ 0::: « Q-
--_.
"'-~
~-~
" "-........
r\.
"'" ""-"'" ~
'-"-"-
"\t
1\
\ eJ
I I I I I I II I I I I I I I
DOL 09 09 ot OG
J8 11 DWS iU8JJ8d
..
..
..
..
o o .-
ill ..-N
.-o
.-if)
ill U
00
72
m
. o z Q)
0...
E a
if)
I D« n::: C)
w N (f)
W --l U ln::: « n...
.---
c ---r--
~ --" " '\.
"---~ -- -"-
~
\ \. &
'" "" "'-'" .~.
.J J 1 .J ..l I J _lL J1 I
DOL DB 09 Ot Ol
J8 11 DWS lU8JJ8d
..l .J I
..
...
...
1ft
o o .....
Q)
.... N .-if)
(l)
u .-... -4--'
L
..
N
a r:(L o
.... o
0 0
7 3
o N .
o Z
Q)
0..
E o
if)
:r: (l.
« n:::: 19
w N (f)
W -1 U I~ « CL
f--.
"-
-. I I I
DOL
--t-. --I-.
ae...
~ "' .'\.
'\.
r'-.. .. -'\.
~ I'\.
"-~ ~
.. - ~
..
II
---- ..
II
..
1\
a a r-
er---.. E E
'--'"
Q)
..... N .-en ())
U
\ -
COl +-' L-
0 ~D-.. a
CI
..
- N
I I I I I I I I I I I I r-a
De 09 ot Ol O ci
J8 11 0WS lUdJJ8d
74
75
APPENDIX B
SAMPLE PREPARATION PROCEDURES
B.l preparing soil
1. Mix sieved soils with different particle sizes in designed amounts to obtain a designated gradation curve.
2. Prepare a soil mass not less than 30 pounds (13.6 kg), W" for a 6-inch diameter specimen, and determine its moisture content, (d,.
3. Determine the weight of water, Ww' required to reach the optimum moisture content, (dopt.
w. = tor (1)
4. Add the water Ww to the soil mass in small amounts and mix thoroughly after each addition.
5. Place the mixture in Ziploc plastic bags. Seal the bags and store them in a humidity room for 24 hours.
B.2 Initial Measurements
1. Carefully check the pressure system, triaxial cell, and rubber membrane under pressure for any possible leakage.
2. Measure the thickness of rubber membrane. Take four measurements, two at the top and two at the bottom. Take an average for double thickness, dm•
3. Obtain two dry porous stones and two filter paper discs. The filter paper should not be larger than the porous stone.
4. Place two porous stones, two filter papers and loading cap on base pedestal. Measure the height of the stack at three different locations. Take an average as an initial height, hi' without specimen.
5. Remove the loading cap, filter papers and porous stones.
B.3 Weighing Soil for Specimen
1. Obtain the soil prepared in Step B.1, and check the moisture content which should maintain at (dopt. Adjust moisture content to (dopt if it is necessary.
2. Measure the inner diameter of split compaction mold. The expected diameter of specimen, Do' is subtracting the double thickness of membrane, dm, determined in Step B.2.2 from this
76
measurement.
3. Measure the total height of mold sitting on triaxial cell base. The expected height of specimen, ho' is subtracting the initial height, hi' determined in step B. 2 . 4 from this measurement. The ho snould be at least two times D.
4. Determine the expected volume of compacted specimen, Va' to be prepared.
1 4
(2)
5. Determine the total weight of the soil, Wt , to be compacted.
W t = (1 + (,.) opt ) • Yd' VO' (3)
where Yd is the maximum dry density.
6. Specimen is to be compacted in eight layers. Divide the soil mass into eight equal portions. Seal each portion of soil in a Ziploc bag.
B.4 Mountinq split compaction Mold
1. Apply a thin layer of grease sealant around the side of base pedestal.
2. Place a porous stone and a filter paper on the base pedestal.
3. Place the bottom of rubber membrane over the base pedestal, and smooth it around the side of base pedestal.
4. Fix the rubber membrane in place with three O-rings.
5. Apply a thin layer of grease sealant to the splitting surfaces of mold.
6. Tighten the split mold on base pedestal, and draw the rubber membrane up through the mold.
7. stretch the top of rubber membrane over the rim of mold.
8. Apply a vacuum of 10 psi (68.95 kPa) to the inside of mold, and remove membrane wrinkles.
9. Measure the distance from the bottom filter paper to the rim of mold.
77
B.5 compacting specimen
1. carefully place a bag of soil prepared in step B.3.6 in the mold, and level the surface of soil.
2. compact the soil in the mold with a Proctor hammer until the thickness of the first layer is slightly over 1/8 of ho determined in step B. 3 . 3 . The thickness of each layer compacted should be progressively smaller than the previous layer so that the compaction effort would be more uniform throughout the specimen. Thus, the thickness of each layer varies from slightly over ho/8 for the first layer to hol8 for the top layer.
3. scarify the surface of the compacted layer.
4. Carefully place the next bag of soil in the mold, and repeat Steps B.5.2 and B.5.3 for each new lift.
5 . Level and smooth the final surface .
B.6 Removing Mold
1. Place a filter paper and a porous stone on the top of compacted specimen.
2. Apply a thin layer of grease sealant around the side of loading cap, and place it on top of the porous stone.
3. Check the level of the loading cap in two opposite directions. The maximum allowable tilt is 0.2% of D.
4. Carefully unfold the top of rubber membrane, and smooth it around the side of loading cap.
5. Fix the rubber membrane in place with three O-rings which already hung on the pressure line connecting to the loading cap.
6. with vacuum still in the mold, apply a vacuum of 5 psi (34.48 kPa) to the top of specimen.
7. Release the vacuum from the mold.
8. Carefully open and remove the split mold, and check the specimen for any irregularity on the surface.
B.7 Final Measurements
1. Measure the sample diameter including rubber membrane at
78
three different heights. Take an average as a final diameter, Df • The actual diameter of specimen, D, is subtracting the double thickness of membrane, ~, determined in step B. 2.2 from Df •
2. Measure the total height to the top of loading cap at three different locations. Take an average as a final height, h t , wi th specimen. The actual height of specimen, h, 1S
subtracting the initial height, hi' determined in step B.2.4 from h f •
3. Determine the actual volume of specimen, v,
v 1 4
• 1t • D2 • h.
4. Determine the unit weight of specimen, Ytl
(4)
(5)
where Wt is total weight of soil determined in step B.3.5.
5. Check any leakage around rubber membrane and fittings by spreading some water on them. If necessary, carefully install a second rubber membrane or reinstall fitting to stop leakage.
B.8 Mounting Triaxial Chamber
1. Clean the contact surface of triaxial cell base, and apply a thin layer of grease sealant.
2. Apply a thin layer of grease to the large O-ring, and place it on the triaxial cell base.
3. Assure the loading ram is fully lifted and locked in place.
4. Clean the bottom surface of the triaxial chamber, and apply a thin layer of grease sealant.
5. Open the valve at the top of chamber to the air.
6. Carefully place the chamber on the cell base, and lock it in place by a rim locking band.
7. Gently lower the loading ram to check if the specimen is properly centered. Then, raise and lock the loading ram back in place.
79
B.9 Applying confining Pressure
1. with the chamber top valve open, fill the chamber with water, and leave about 0.5 inches short from the chamber top for an air pocket.
2. Turn off the water, disconnect the water supply line from the cell, and then close the chamber top valve.
3. with the pressure valves at the cell base closed, connect an empty confining pressure line at atmospheric pressure to the chamber top valve, and an empty drainage line at atmospheric pressure to the bottom of specimen.
4. Slowly release the vacuum from the specimen, and gradually raise the confining pressure to 5 psi (34.48 kPa) by slowly opening the chamber top valve.
5. Connect the transducer to the triaxial cell base.
6. Calibrate the transducer.
7. The specimen is ready for resilient modulus test after cured over night.
80
81
APPENDIX C
TEST PROCEDURES
C.1 Testing setup
1. Turn on the material testing system MTS-810 and the hydraulic pressure system.
2. setup the MicroConsole and data recording system as required.
3. with the drainage valve at the bottom of specimen closed and the confining pressure valve open, carefully place the triaxial cellon the platform of the loading machine.
4. Center the triaxial cell, and raise loading piston to couple the loading ram of triaxial cell with load cell. The testing setup at this step is shown in Figure C.l.
5. Apply an axial load about 5 psi (34.48 kPa) which slightly over compensates the chamber pressure to lower the loading ram slowly until it contacts the loading cap.
6. Open the drainage valve to dissipate the excess pore water pressure caused by possible disturbance in the steps C.1.3 through C.1.5.
C.2 conditioning
1. Set the confining pressure to 5 psi (34.48 kPa) and apply 200 repetitions of an axial deviator stress of 5 psi (34.48 kPa). Then, apply 200 repetitions of an axial deviator stress at 10 psi (68.95 kPa).
2. Set the confining pressure to 10 psi (68.95 kPa) and apply 200 repetitions of an axial deviator stress of 10 psi (68.95 kPa) • Then, apply 200 repetitions of an axial deviator stress at 15 psi (103.43 kPa).
3. Set the confining pressure to 15 psi (103.43 kPa) and apply 200 repetitions of an axial deviator stress of 15 psi (103.43 kPa) . Then, apply 200 repetitions of an axial deviator stress at 20 psi (137.90 kPa).
4. with the drainage valve open, let the specimen sit for a few minutes to fully dissipate any possible excess pore water pressure.
C.3 Recorded Resilient Modulus Test
1. Rebalancing the measuring devices and recording system.
2. Begin the recorded testing by increasing the confining pressure to 20 psi (137.90 kPa).
82
8 3
Figure C.l Triaxial cell is on the loading machine and coupled with load cell ready for test.
-84
3. Apply 200 repetitions of a deviator stress of 1 psi (6.895 kPa) and record the vertical recovered deformations for the 200th repetition.
4. Repeat step C. 3 • 3 for deviator stress levels of 2, 5, 10, 15, and 20 psi (13.79, 34.48, 68.95, 103.43, and 137.90 kPa) and continue to record vertical recovered deformations for each 200th repetition.
5. Reduce the confining pressure to 15 psi (103.43 kPa), and repeat step C.3.3 for deviator stress levels of 1, 2, 5, 10, 15, and 20 psi (6.895, 13.79, 34.48,68.95,103.43, and 137.90 kPa).
6.
7.
8.
Reduce repeat and 15
Reduce repeat and 15
Reduce repeat 10 psi
the confining pressure to 10 psi (68.95 kPa) , and step C.3.3 for deviator stress levels of 1, 2, 5, 10, psi (6.895, 13.79, 34.48, 68.95, and 103.43 kPa).
the confining pressure to 5 psi (34.48 kPa) , and step C.3.3 for deviator stress levels of 1, 2, 5, 10, psi (6.895, 13.79, 34.48, 68.95, and 103.43 kPa).
the confining pressure to 1 psi (6.895 kPa) , and step C.3.3 for deviator stress levels of 1, 2, 5, and (6.895, 13.79, 34.48, and 68.95 kPa).
9. stop the loading after 200 repetitions of the last deviator stress level or when specimen fails.
C.4 post-Testing
1. Reduce the sitting load set in step C.1.5, and slowly lower the loading piston to fully raise the loading ram.
2. Disconnect the loading ram from load cell, and lock the loading ram in place.
3. Reduce the chamber pressure to zero, and disconnect the transducer and pressure lines.
4. Remove the triaxial cell from the loading machine, and dismantle the cell.
5. Turn off the hydraulic pressure system and the material testing system.
6. Use a part of specimen to determine the moisture content after test.
85
APPENDIX D
ARI~TIC PLOT OF TEST RESULTS
RESILIENT MODULUS
:[ r-. 30 .-m ~ '--"
a::: 20 2
10
-x-
--+-*- -+---
./ e a--a
Sample No. 1
---x-
--+
__ x
GeeeE)
)He X*'f(
4AAA ..
+++++
a c = 20 psi 15 psi 10 psi
5 psi 1 psi
o ~l~~~~ __ ~~ __ ~~~~~~ __ ~-L __ ~~ __ ~~~ __ ~~ __ ~~ __ ~~~~~
o 5 10 15 20 25
Deviator Stress (psi)
co 0"1
87
It) N
N '-'ij'ij'ij'ij [a.a.a.a. olt)OIt)~ . N ........
0 n Z " b 0
N Q)
UBI - / 0...
E / ,,-... .-0 /
(/)
0.. (f) t
It)'-'" .... \ (/)
en (f) \ \ <U
L..
\ +-' :J \ U1 ---1 \ t o L.. :J x .- 0
\ +-' 0 \ 0
0 > 2 \ <U
0
~ z It)
W ---1
L _I -(f) W et:: I I I 1 I I I J I I I I I 0
0 0 0 0 0 0 It) .q- ...., N ....
(!S>\) HV'J
88
n I '-'-iil~ .- J) en en en ~Q.Q.Q.Q.
. 010010
-- I 0
N.- .....
Z II u , b 0
N OJ
UBI -0..
E ,-... .-0 en
(f) D-10"'-/
+ ..-\ en I \ (f)
(l,) (f) , L..
:J \ -+-'
I (/)
--1
J, ::::J r o L..
..- 0
0 -+-'
I 0
0 > :2 I
Q)
0
1-- I \ Z , + 10
W \
--1 -(f) ~ -W 0::: _LJ--L-L-J I I 1 I I 1.-1 I I I I a
0 0 0 0 0 0 10 v 1'1 N .....
(!S>I) ~V'J
RESILIENT MODULUS Sample No. 4
::r ,--..30 .-m ~ '--"
20 0.::
:2 ~ 10 E
.
- --- -
-+ - -+
G&eee )H(X_
& ......... .a +++++ ae&BEJ
ac == 20 psi 15 psi 10 psi
5 psi 1 psi
o ~I~~~~ __ ~-L __ ~-L~~~-J __ ~-L __ L--L __ ~~~ __ ~~ __ ~~ __ L--L~~~
o 5 10 15 20 25
Deviator Stress (psi)
00 \.0
90
10 N
L() ._.;;.ij.ij9ij s.o.o.o.o.
0 100 11') .... . N ........
0 Z II
u
If b 0
N Q)
\ UUI -0..
E \ r-.
0 Cf)
(Jl Q.
~ + 10'--"
\ ....
\ Cf) Cf)
if) Q) L.
:=:> +-' (f)
---1 \ :J • 1
o L. .... 0
0 +-' 0
0 > 2
Q)
0
I- l. I
Z 1 w '\ -1 -(f) W 0:: I I I I I I I I I I J I
0 0 0 0 0 II) v t') ('II ....
(!s>!) ~~
RESILIENT MODULUS Sample No. 6 50 ~
40
'\ ~ --ok --tc
-:; 30 t-.::L -'--'"
-+ -.-+-et:: 20 ~ ~ -+--
~ B
10
I oC 1
o 5
-El
L I I I
Gee&EI O'c = 20 psi - )CofH( 1 5 psi
10 psi 5 psi 1 psi
10 15 25
Deviator Stress (psi)
~ I-'
92
III C'I
'" "ij"ij "ij "ij'ij a. a. a.. a.. a..
• 0 100111 ....
0 C'I .... ....
Z II u
b
Q) x
- HBI \
0
0..
C'I
E \
".-...
0 (f)
VJ
)(
C-
\ 1 Ill""""" ....
\ \ VJ VJ
(f)
=:J \ \
OJ
~
L
\ ....,
:::J
(/)
0 t o L
..... 0 ...., c
0 2
> OJ
0
l-Z t w ~
(f) -W - ---n::: 0 0 0 ....
~~
93
I() N
OC) .-'g;'g;'g;'g; [a.a.a.a. o I() 0 I() or-
0 Nor- .-
Z II u
b 0 N
<V
UBI 0..
E ,,-.... (J) 0 0.. U)
+ I()"--" r-
Ul
\ (J) Q)
(f) L
\ -+-' =:l (f)
~ ;. ~ o L
=:l .- 0 I -+-' 0 I c .-0 I >
I Q) 2
I 0
J- ! z W ~ / ---.J -U) W n:: I I I I I LL I I I I L...L...LL I I 0 ..
0 0 0 0 0 0 I() v I') N .-
(!s>j) ~V'J
RESILIENT MODULUS Sample No. 9 50 ~I--------------------------------------------------------------~
40
~30 .-en .:Y. '--" ! ........
(t: 20 r- .; ........ )f-::;;; r,,- ..-----Is lao ....6.
10 +-+- ........ - ......
-+
-£I
t3- B
G&&eQ a c = 20 psi )H( x.....c 1 5 psi A666~ 10 p~ +++++ 5 psi Ge&ee 1 psi
tc
o ~I ~~~~~~~--~~~~--~~~~--~~~~--~~~~~~~~ o 5 10 15 20 25
Deviator Stress (psi)
1.0 ~
50
40
,-... 30
en ..:::£ '--'
RESILIENT MODULUS Sample No. 1 0
e
~ O'a = 20 psi _x_ 15 psi &6AAA 10 psi +++++ 5 psi
1 p:si
e --Q -_ ....... ~
A
- - -+
o ~1~~~~ __ ~-L __ ~-L~~~~ __ 4--4 __ ~-L __ ~~~ __ ~~ __ ~-L __ ~~~~~
o 5 10 15 20 25
Deviator Stress (psi)
\0 U1
50
40
,--... 30
en ..::x -....-
D:: 20 :2
10
RESILIENT MODULUS Sample No. 11
Gee&EI Uc - 20 psi )He X*tC 1 5 psi A666A 10 p~ +++++ 5 psi
1 psi
~
--* -*
- -+ +-+- -+-
+-
--IiI ~ a
o ~ I I I I I I I I I I I I I o 5 to 15 20
Deviator Stress (psi) 25
1.0 m
97
LO
N N
~ "ij",;"ij";;";; 0.0.0.0.0.
. 0 10010 ..-
0 N"-"-
Z II
b OJ
UBI , 0
-0...
N
E \
r--.
0 (f)
rn , Q.
+ It)'--'"
I -1
(f) I
rn
:J I
rn Q)
\ L ......., (f)
----1 :J 1 0
t o L
I .,.... 0 .......,
0 a :2
.-> (])
0
~ I Z w
+ I
LO
----1 \
-(f) -~\-<
I
w +-0:::: 0
0 0 ..-
(!S>\) ~~
RESILIENT MODULUS Sample No. 13 50 r------------------------------------------------------------,
40
GG&eQ a c = 20 psi - x*'C 1 5 psi .666. 10 p~ +++++ 5 psi Geee-EI 1 psi
~30 ~ '-/ ~ ~ 20 L-
Q e
....x -.-.c _ .-.c
/"
~ - -,..- --+-- -+ ~
10 ~-- w
o I ..J
o 5 10 15 20 25 Deviator Stress (psi)
\.C OJ
RESILIENT MODULUS Sample No. 14 50 ~------------------------------------------------
40
,,-... 30
en ~ '--'
0::: 20 ::E
10 ~ .......... -
~
a--
--+ -+
-EI
G&eeQ )H(X_
!l6666
+++++ [H!66EI
ac = 20 psi 15 psi 10 psi
5 psi 1 psi
_x
o ~i ~~~~~-L~ __ ~~~~ __ ~~-L~ __ ~~~~~~~~~~~~~
o 5 10 15 20 25
Deviator Stress (psi)
~ ~
50
40
".--... 30
CIl ~ -....-
~ 20 L
10
RESILIENT MODULUS
_.-ac .-ac
""It--
-+ .........
~ .... ----+ I!I .......£I
I
Sample No. 15
G6&eQ JHCX_
+++++ IHt£Ht6
Uc = 20 psi 15 psi 10 psi
5 psi 1 psi
-
o !~~~~~ __ ~~~~ __ ~~-L~ __ ~~~~~~~~~~~~~-L~
o 5 10 15 20 25
Deviator stress (psi)
~ a a
RESILIENT MODULUS Sample No. 1 6 50 ~--------------------------------------------------------~
40
....... 3J en
.Y "'-J
a:::: 20 ~
10
- -- --..... -
G&e&e _ X-tHC
+++++ GaeeG
ac == 20 psi 15 psi 10 psi
5 psi 1 psi
---o I I I I I I I I I I I I
o 5 10 15 20 25
Deviator Stress (psi)
I-' o I-'
50
40
r--. 30 .-Cfl ~ "'--'"
0:: 20 ::2
10
RESILIENT MODULUS Samp le No . 1 7
~ "-* --+-- -of- -+
Ge6&E)
)H()(~
AAAA6
+++++
.-.c
~=W~ 15 p~ 10 p~
5 p~ 1 p~
o I~~~~ __ ~~~~~~~~ __ ~~-L~ __ ~~~ __ ~~-L~ __ .~~~~
o 5 10 15 20 25
Deviator stress (psi)
t-' o I\J
RESILIENT MODULUS Sample No. 18 50
40 r-
. \
C' 30 ~ \ &-----e-----e----m \ e ~ G
'--' ~ ~ e __ x- ___ x- - - _x- -
_ -x
0:= 20 2
6 ~
10[:""'-+- - -f- - - - -+- - - ---+ GeeeQ C1e = 20 psi
15 psi _XofHC
.666616. 10 psi +++++ 5 psi I3-&S9S 1 psi
o ~'~~~~~~~ __ ~~~~ __ ~~-L~ __ ~~~~~~~~
o 5 10 15 20
Deviator stress (psi) 25
..... a w
RESILIENT MODULUS Sample No. 1 9 50
40
030~ en
.:::L. ""--'"
~ 20 ~
10
" I \ \
- -lit-" - *"" ot(
_--4-+-+- - +- - - - +- -
'- -e B Ge&&E)
IHC XotHC
AAAA6
+++++ IHt&&EI
O'c = 20 psi 15 psi 10 psi
5 psi 1 psi
o I L
o 5 10 15 20 Deviator Stress (psi)
25
...... o ~
RESILIENT MODULUS Sample No. 20
50 F
40
_x ........ x- _ _ -x-- -- -
-+ ---+-'+- --
G&e&e a c = 20 psi 10 ~ Ii... ------- -
)He X_ 15 psi AAAAA 10 psi +++++ 5 psi IHHHH!.I 1 psi
at I I I I a 5 10 15 20 25
Deviator Stress (psi)
I--' o U1
106
APPENDI% E
LOG-LOG PLOT OF TEST RESULTS
100
I
2
,....... en
.:::, 10
0::
~ 5
2
"" .,.../
2
RESILIENT MODULUS Sample No. 1
0
0 ~ ~ ..",r
~ ~
:,.- .....
~ ~ ~
" .......,
~ ~ "
1 10 100 Sum of Principal Stresses, 9 (psi)
MR = 4.988 X 8 0.461
1 07
100
5
2
--.. rn ~ 10 ~
II::
::2 II
2
1
/
RESILIENT MODULUS Sampl e No. 2
0 QCP
~ ~ pt
k<' ~ ~ ~
:>
A ,/
..Air"'"
./ ."
./ V
2 5 2
". ~i-'
10 100 Sum of Principal Stresses, e (psi)
MR = 2.977 X 8 0.550
108
100
2
~
CI) ..Y. 10 -..-
0:: :2
5
2
2
RESILIENT MODULUS Sample No. 3
0
~ 0 n :nO u u ~u
0 0 0
I
I 2
~c
I
10 100 Sum of Principal Stresses, e (psi)
MR = 27.502 X e 0.061
109
100
a
2
~
rJ) ..Y. 10 -...-
0:: 2
5
RESILI ENT MODULUS Sample No. 4
0 ... b
~ 0 g po ----t-'lp 0 ~ ~ ~
~
2 a
~o .... ~
1 10 100 Sum of Principal Stresses, 8 (psi)
MR = 21.089 X 8 0.156
110
.--U)
100
6 10
1
5
2
5
2
/' V'
RESILIENT MODULUS Sample No. 5
90l
~ ~
o~
~~ ~O 0
~~
.JI' Jt'
...... "" v
~
2 5 2 5
'" ).0' "
10 100 Sum of Principal Stresses, e (psi)
MR = 3.568 X e 0.550
III
100
~
rn o 10
a: ~
1
II
2
...-,.....
5
2
RESILIENT MODULUS Sample No. 6
u P Igj ~
~ ~ b .........:
... .:::0-0 I""
"I..- J
...... i--" 0
~ ~ C ~
~ .....
z II 2 II
~ i""
1 10 100 Sum of Principal Stresses, e (psi)
MR = 7.278 X 8 0.374
112
100
II
2
---en ~ 10
II:::
2 II
2
r---
/""
RESILIENT MODULUS Sample No. 7
D
0 ~
~ rr ~o ~~ 0
" .... ~
.,'" ~
2 2
~ t'" ~
10 100 Sum of Principal Stresses, 8 (psi)
MR = 3.910 X e 0.449
113
100
II
2
..-(I)
25., 10
It: I--"'" 2
II
2
--
RESILIENT MODULU S Sample No. 8
.., ~ :>
0
~ ...-~
~~ ClfO I'"'
0
,1--~
.... 0 ~
~ .......
2 II 2
~
ro it'"
10 100 Sum of Principal Stresses, 8 (psi)
6.345 X e 0.359
114
100
....-rn
.:::f. 10 -...-
a:: :=!!
~-
5
2
~.
5
V
2
RESILIENT MODULUS Sample No. 9
p -
I ~ ~ 0 . ;ao1
~ ~O ~
v
....-! I"
./ ./
./ .....
2 5 2 5
~ ...
~
1 10 100 Sum of Principal Stresses, 8 (psi)
MR = 3.208 X 8 0.523
115
100
,-... en -=.. 10
II:::
~
5
2
~
5
2
2
RESILIENT MODULU S Sample No . 1 0
0 ~
o 0 ~O I=>
O~ ....
1 10 100 Sum of Principal Stresses , 8 (psi)
MR = 1 3. 207 X 8 0.139
116
100
5
2
---rn ~ 10
a:;
~ I' 5
2
1
...........
RESILIENT MODULUS Sample No. 11
~ AI ~
~ ~ II) P
..... .... ~ ~ ~
...-
2 :I 2 :I
~
10 100 Sum of Principal Stresses, 8 (psi)
MR = 5.5 24 X 8 0.385
117
100
5
2
----rn o 10
0: ~
5
2
----
2
RESILIENT MODULU S Sample No. 12
~
0 00
.Qo1 ~
I--~ ~ iTo
0 -f-'" f--'" ~ n -
I 2
~ .,.. .... 1--1-'
10 100 Sum of Principal Stresses, e (psi)
6.935 x e 0.304
118
100
5
2
.......... CIl
c., 10
a:: ~
5
2
...... --
2
RESILIENT MODULU S Sampl e No . 13
0
~ ~ ,... ......
~~ foQ ~. P p
"... ,; .... 0
--~ -
5 2
~ I'-" ....
10 100 Sum of Principal Stresses, 8 (psi)
MR = 6 .866 X e 0.333
119
100
'2
,........,. en ~ 10 '-"
a::: ~
5 ......-
2
RESILIENT MODULUS Sampl e No. 14
b
~ ~ ~
~ p
-..... ~
~ .... ........... 1"'"
2 5 '2
..oc tl... i""' .... r-i"""
10 100 Sum of Principal Stresses, 8 (psi)
4.339 X 8 0.374
100
5
2
,...-.. CI) ~ 10 ---
a:: ::?:
5
2
--
RESILIENT MODULU S Sample No. 15
~
0 ~ ~ ,~ '"' ~O ~~ ~ ... , ~ 0
... ~ ~ ---
1 5 2 5
i-" ,....
10 100 Sum of Principal Stresses, 8 (psi)
MR = 6.544 X 8 0.346
121
100
5
2
,......... en ~ 10 '-"
a:: ~
5
2
2
RESILIENT MODULUS Sample No. 1 6
p
.n O 0
"-
~
'" 0
5 2
10 100 Sum of Principal Stresses, e (psi)
MR = 11.192 X 8 -0.073
122
100
5
~
.r--...
CII ~ 10 '--'"
0::
:2 5
----
RESILIENT MODULUS Sampl e No. 1 7
- p 0
oW-! p-~( ~ 8 fIJ'
I-
~ -. ~ p
0
~Ot l--
10 100 Sum of Principal Stresses, 8 (psi)
M R = 1 5. 2 1 1 X 8 0.162
123
100
:I
,.---...
Ul .::{. 10 '-'
0::
.2 5
2
/:
RESILIENT MODULU S Sampl e No. 1 8
ro n.
~ ~
..........
~ ~u
0
100' ..... 00
~
./ /
/
~
./ "
2 2 5
~ .....
"
10 100 Sum of Principa l Stresses , 8 (psi)
2 .880 X 8 0.610
124
100
2
----.. en
.:::, 10
a:: ..... ~
5
2
",.,.
RESILIENT MODULU S Sampl e No. 19
n
1'-'0
a.,.... ~ ~
~~ ~o :'> :>
~ ~~
v ~ ,
0 ,. -,.,.
2 2
~
f ~
10 100 Sum of Principa l Stresses, 8 (psi)
6.565 X 8 0.348
" 125
100
2
,,--.... Cf)
~ 10 "--'"
0::
~ 5
2
./ ~
2
RESILIENT MODULU S Sample No. 20
~ ~ D"I 11 ">
~~ ~U p-
..... ~ ...... 0
..... ~
~ L'
o~ ~ ....
10 100 Sum of Principal Stresses, 8 (psi)
MR = 4.196 X 8 0.490
126