EXAMPLE PROBLEM ILLUSTRATING THE

APPLICATION OF THE NATIONAL LIME ASSOCIATION

MIXTURE DESIGN AND TESTING PROTOCOL (MDTP)
TO ASCERTAIN ENGINEERING PROPERTIES OF
LIME-TREATED SUBGRADES FOR MECHANISTIC
PAVEMENT DESIGN/ANALYSIS

By: Dallas N. Little and F. A. M. Shafee Yusuf

September 2001

This document presents an example of mechanistic design and analysis

using a mixture design and testing protocol (Reference 1) to address
structural properties of lime-treated subgrade, subbase, and base layers
 TABLE OF CONTENTS
Page
PURPOSE ............................................................................................................................. 1
INTRODUCTION
OUTLINE OF MIXTURE DESIGN AND TESTING PROTOCOL (MDTP) ..................... 2
Step 1: Soil Classification and Assessment of Suitability for
Lime Stabilization ................................................................................................. 2
Step 2: Eades and Grim pH Test to Determine Appropriate Optimum
Lime Content ......................................................................................................... 3
Step 3: Moisture/Density Relationship for Lime-Treated Soil, and
Moisture Sensitivity and Strength Gain Following Accelerated Curing ............... 3
Step 4: Compressive Strength Testing to Determine the Unconfined Compressive
Strength (UCS) of Lime-Soil Mixtures ................................................................. 4
Step 5: Resilient Modulus Test to Determine the Resilient Properties .................... 4
Step 6: Tube Suction Test (TST) to Evaluate Moisture Sensitivity .......................... 4
LABORATORY TESTING .................................................................................................. 4
Eades and Grim pH Test............................................................................................ 4
Atterberg Limits Test................................................................................................. 5
Unconfined Compressive Strength Test .................................................................... 5
Rapid Triaxial Tester (RaTT) Resilient Modulus Test.............................................. 6
Tube Suction Tests/Swell Test .................................................................................. 6
FIELD TESTING .................................................................................................................. 7
Ground Penetrating Radar (GPR).............................................................................. 7
Dynamic Cone Penetrometer (DCP) ......................................................................... 8
Falling Weight Deflectometer (FWD)....................................................................... 8
COMPARISON OF LABORATORY AND FIELD DATA................................................. 9
Comparison of Laboratory UCS with Field DCP...................................................... 9
Comparison of Laboratory Resilient Moduli with Field
FWD Back-Calculations........................................................................................ 9
Comparison of Laboratory Dielectric Value to Field GPR ....................................... 10

i
 TABLE OF CONTENTS (cont’d)
Page

MECHANISTIC ANALYSIS ............................................................................................... 11

Resistance of HMA and LTS Layers to Flexural Fatigue.......................................... 11
Permanent Deformation Potential ............................................................................. 12
RECOMMENDED ENGINEERING PROPERTIES FOR MECHANISTIC DESIGN....... 14

REFERENCES ...................................................................................................................... 15

APPENDIX: OTHER TESTS TO ASSESS LTS DURABILITY ....................................... 17

Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray (EDX) Test... 17
X-ray Diffraction (XRD) Test ................................................................................... 17
Analysis of Pozzolanic Development........................................................................ 18

ii
 LIST OF FIGURES

Figure Page

1 Relationship Between Dry and Soaked Unconfined Compressive Strength

Of Unstabilized and Stabilized Mississippi Soils.............................................................. 19

2 Comparison of Resilient Moduli of Unstabilized and Stabilized Mississippi

Soils for Dry and Soaked Conditions ................................................................................ 19

3 Comparison of Dielectric Value, Change in Moisture Content and Swell for

Unstabilized and Stabilized Mississippi Soils after 8-days of Capillary Soak................... 20

4 XRD Patterns of Soil (As-received Native Soil; Immediately After Mixing with Lime;
and After Curing for 7-days at 400C). ................................................................................... 21

5 Unstabilized and Stabilized Soil Samples after 12-hours of Soaking ................................... 22

iii
 LIST OF TABLES
Table Page

1 Unconfined Compressive Strength and Resilient Moduli of Unstabilized and

Stabilized Mississippi Soils............................................................................................... 23

2 GPR, FWD and DCP Results for LTS and Unstabilized Subgrade Soils for
Four Mississippi Pavements ............................................................................................... 24

3 Mechanistic Analysis for Mississippi Pavements With and Without the LTS Layer ........... 24

4 Recommended Design Values for Mechanistic Analysis...................................................... 24

iv
PURPOSE
This document presents an example of mechanistic design and analysis using a
mix design and testing protocol (Reference 1). More specifically, it addresses the
structural properties of lime-treated subgrade, subbase, and base layers through
mechanistic design.

INTRODUCTON
This design example is based on three Mississippi Department of Transportation
(MDOT) widening projects on state highways in Kemper, Lowndes, and Washington
counties. In all three, the lime-treated subgrade (LTS) layers are being evaluated as
structural components in the design. The design options considered are
– using hot mix asphalt (HMA) directly over the LTS and
– using HMA with a lime-fly ash stabilized base over the LTS.
Both options have previously been used successfully in other parts of the State.
In fact, MDOT has used lime stabilization in clay soils for over 30 years to
improve performance and as a platform for pavement structures. A large percentage of
Mississippi’s land area contains soils that respond well to lime stabilization—as
evidenced by improved engineering properties and long-term performance as roadway
sublayers.
This example presents three elements of the mixture design and testing protocol
(MDTP):
– Laboratory Testing,
– Field Testing, and
– Mechanistic Analysis.
First, laboratory testing of soils is conducted using the procedures described in
Reference 1. The soil samples evaluated were determined to be representative of the
subgrade soils to be stabilized based on county soil reports and boring samples.
Also presented is field testing of existing pavements with LTS layers in the same
soil series as those in the widening project. The engineering properties of the newly
lime-treated soils are compared to in situ properties of lime stabilized soils that have been
in service for 15 to 20 years. The field pavements selected for this comparison were

1
 – US highway 45N (17 years of service) in Kemper County,
– US highway 82W (20 years of service) in Lowndes County,
– US highways 61N (15 years of service) in Washington County, and
– 82E (20 years of service) in Washington County.
In addition, the structural effectiveness of the LTS layer is analyzed by a
mechanistic analysis. The properties of the LTS layer (as determined by in situ testing)
are used in a layered elastic model to calculate maximum tensile flexural stress in the
HMA and compressive strain at the top of the subgrade. These values are then placed in
the appropriate transfer function to evaluate the structural effectiveness of the LTS layer.
Specifically, the projected lives of the pavements with and without the LTS layer are
compared, in terms of equivalent single axle loads (ESALs).
Finally, although not part of the MDTP, the quality of the pozzolanic reaction in
the laboratory-cured LTS samples is confirmed by x-ray diffraction (XRD) and scanning
electron microscopy (SEM)—see Appendix.

OUTLINE OF MIXTURE DESIGN AND TESTING PROTOCOL (MDTP)

The MTDP (1,2) is designed to produce a mixture that possesses the desired
structural properties and durability in a pavement layer. The procedure measures
engineering properties that are critical to the performance of the lime-stabilized mixture
as a structural layer in a pavement system. The MDTP is comprised of the following
steps:
Step 1: Soil Classification and Assessment of Suitability for Lime Stabilization
Determine whether the soil has at least 25% passing the 75-µm (micron) sieve and
has a plasticity index (PI) of at least 10 (3). The reactivity of lime with soil is predicated
on the type and the amount of clay minerals present in the soil. Soils with organic
contents exceeding 1% by weight are difficult to stabilize or may require uneconomical
quantities of lime in stabilization. The screening criteria also limit soluble sulfates to less
than 0.3% by weight in a 10:1 water-to-soil solution (4). High sulfate concentrations can
cause deleterious reactions among lime, soil minerals, sulfate ions, and water. This can
lead to loss of stability and heave.

2
Step 2: Eades and Grim pH Test to Determine Approximate Optimum Lime
Content
Perform the Eades and Grim pH test (ASTM D 6276) to determine lime demand.
This test identifies the lime content required to satisfy immediate lime-soil reactions and
still provide significant residual calcium and a high system pH (about 12.4 at 250C). This
is necessary to provide proper conditions for the long-term pozzolanic reaction that is
responsible for strength and stiffness development.
Step 3: Moisture/Density Relationship for Lime-Treated Soil, and Moisture
Sensitivity and Strength Gain Following Accelerated Curing
Determine the moisture/density relationship according to the appropriate protocol
defined by the user agency, i.e., AASHTO T-99, T-180; ASTM D 698, D 1557; Texas
Method 113A, etc. A modified compaction effort (or some reasonable percentage thereof,
e.g., 95% of AASHTO T-180) is recommended for fabricating the samples. This level of
compaction is usually achievable with conventional field equipment.
Samples are prepared for strength testing and moisture sensitivity testing at
optimum moisture content with a tolerance of ±1%. All samples are cured for 7 days at
400C in sealed plastic bags (in order to retain sufficient moisture during curing). Previous
studies have demonstrated that this period of accelerated cure promotes a strength that is
representative of a long-term cure [13]. Furthermore, 400C represents high temperatures
that can be experienced in the field.
Following curing, the samples are subjected to capillary soak for 24-hours
(depending on soil plasticity) prior to strength testing. The capillary soak protocol
consists of placing the sample wrapped in a wet absorptive fabric on a porous stone in
water. The water level reaches the top of the porous stone, so that both the porous stone
and the fabric wrap are in contact with water throughout the capillary soak process.
Extensive laboratory testing has demonstrated that untreated clayey soils will typically
degrade to a compressive strength of less than 70 kPa (about 10 psi) following capillary
soak (1,2). Hence the capillary soak moisture-conditioning phase is considered an
effective method of assessing moisture damage potential.

3
Step 4: Compressive Strength Testing to Determine the Unconfined Compressive
Strength (UCS) of Lime-Soil Mixture
Unconfined compressive strength is determined using ASTM D 5102. UCS can
be used to approximate design parameters such as flexural strength, deformation potential
and stiffness (resilient modulus) when these data are not available:
– Tensile strength can be conservatively estimated as 10 percent of the UCS,
and flexural strength can be conservatively estimated to be twice the tensile
strength or approximately 20 percent of UCS. (2)
– Several correlations have been developed between UCS and resilient modulus.
One of the most conservative for lime-stabilized fine-granular subgrades was
developed by Thompson (5) in which the resilient modulus ER = 0.124 (UCS)
+ 9.98, where UCS is presented in psi and ER in ksi. Comparisons in this
example have shown this relationship to be overly conservative by
approximately 50 percent.
Step 5: Resilient Modulus Test to Determine the Resilient Properties
Resilient properties define the ability of the lime soil mixture to distribute
load/pressure developed under heavy wheel loads so that those wheel loads will not over-
stress the weaker pavement layers. Resilient modulus is determined using AASHTO T
294-94 or a rapid triaxial tester (RaTT) (which can be used instead of the more time-
consuming and material-intensive AASHTO T 294-94 test) (1).
Step 6: Tube Suction Test (TST) to Evaluate Moisture Sensitivity
The dielectric value (DV), measured by a Tube Suction Test (6), is a measure of
how much moisture a base/subbase/subgrade will absorb through capillary rise and the
state of bonding of the absorbed moisture. Low dielectric values indicate the presence of
tightly absorbed and well-arranged water molecules. Scullion and Saarenketo (6) have
established DV selection criteria for subbase layers. The results of this test are used to
assess resistance of the stabilized material to moisture damage.

LABORATORY TESTING

All the in situ materials are oven dried for 24-hours, before they are pulverized to
make samples for laboratory testing.

4
Eades and Grim pH Test
Dry soils are screened through a No. 40 sieve. All the soils are tested with 0, 2, 3,
4, 5, 6 and 7% of hydrated lime in accordance with ASTM D 6276. Special attention is
given to maintain the room temperature at 250C, as pH of lime-soil mixture is
temperature dependent.
The natural soil from Kemper County (US 45N) had a pH of 4.77. All other soils
had pHs between 6.5 and 7.6. According to the Eades and Grim pH test results, 5%, 6%,
6%, and 4% of lime were selected as optimum lime content for the soils of pavements
from US 61N, US 82E, US 82W and US 45N, respectively.
Atterberg Limits Test
Liquid limits, plastic limits and plasticity indices are determined on all
unstabilized soils following ASTM D 4318. After the pH test, the Atterberg limits test is
performed again on all the soils with optimum lime content. Soil-water-lime mixtures are
allowed to mellow for one-hour before testing.
Atterberg limits measured on unstabilized and stabilized soils from the four
pavements demonstrated the ability of lime treatment to reduce plasticity and improve
workability:
– US 61N, PI was reduced from 29 to 15;
– US 82E, PI was reduced from 32 to 16;
– US 82W, PI was reduced from 28 to 9; and
– US 45N, PI was reduced from 17 to 9.
This represents a substantial PI reduction after only one-hour of mellowing. Continued
plasticity reduction occurred with curing (24-hours) due to early pozzolanic reaction.
Plasticity indices were reduced to 10, 5, 5, and 6, respectively.
Unconfined Compressive Strength Test
UCS test specimens are prepared for unstabilized and stabilized soils at three
moisture contents that bracket the optimum moisture contents for each soil and lime-soil.
For stabilized materials, lime is mixed with the dry soil at the optimum lime percentage
determined by the pH test. After the soil-lime mixtures are thoroughly mixed with water,
they are placed in plastic zip-lock bags for one hour. After the mellowing period,

5
specimens are compacted according to ASTM D 1557 to produce 64-mm diameter and
127-mm high samples.
Specimens are divided into two sets for curing, each consisting of three
specimens. One set of specimens is cured for 7-days at 400C and the other set is cured for
30-days at 250C. All the specimens are cured in the zip-lock bags. Samples are subjected
to 24-hour capillary soak prior to compressive strength testing. An identical set of
replicate samples is tested without capillary soak to evaluate the effect of moisture
conditioning.
The unconfined compressive strengths for unstabilized and stabilized specimens
in the dry condition and after 24-hour capillary soak are shown in Table 1. All the data
show substantial strength increase due to lime stabilization. The relationships between
dry and soaked compressive strength for unstabilized and stabilized soils are shown in
Figure 1.
Rapid Triaxial Tester (RaTT) Resilient Modulus Test
Two sets of test specimens (dry and capillary soaked) are prepared with
unstabilized soils and stabilized soils in accordance with the Feed Back Controlled RaTT
Stress Stage Resilient Modulus Test (Test No: 042) that embodies the requirements of the
AASHTO T 294-94 specification. The design of the RaTT hardware allows the testing of
cylindrically shaped specimens of nominal 150-mm diameter and 150-mm height. After a
mellowing period, specimens are molded with an automatic compactor at optimum
moisture content. After the curing period of 7-days at 400C, all specimens are tested to
measure the resilient modulus at dry condition (not subjected to moisture conditioning).
One set of specimens is subjected to a 24-hour capillary soak before determination of
resilient modulus. These are referred to as soaked samples. Specimens made of
unstabilized soils (except US 45N) swelled and cracked significantly during capillary
soak, preventing them from being able to be tested in the RaTT device.
The RaTT (Test No. 042) resilient modulus testing protocol for subgrade soils
consists of a conditioning period followed by determining resilient moduli at various
deviatoric stresses (ranging from 14 to 69 kPa) and for confining pressures of 41, 28 and
14 kPa, respectively. A deviatoric stress of 41 kPa is typical within the subgrade and was
used to identify a single resilient modulus. The resilient moduli at 41 kPa deviatoric

6
stresses for unstabilized and stabilized soils are summarized in Tables 1. Figure 2
compares these resilient moduli for different soils.
Tube Suction Test/Swell Test
Specimens prepared for RaTT resilient modulus test are used for tube suction
testing. Specimens are dried at 400C for 4-days before they are placed on porous stones
with a deionized water level reaching almost the top of the porous stones. During the
capillary soak, a dielectric probe is used to measure the surface DV of the compacted
samples. Swell potential for each sample is also monitored. Capillary soak is continued
for 10-days, or until the DV achieved an ultimate or asymptotic value.
Comparisons of DV, moisture content, and swell potential for the four Mississippi
untreated and lime-treated soil are shown in Figure 3. Dielectric values greater than 16
indicate the presence of substantial “free” moisture and are considered poor in terms of
moisture damage potential. A DV below 10 is considered excellent (6). However, these
criteria were developed for untreated aggregate/soil systems. The effect of higher cation
concentration (electrolyte concentration) on osmotic section and correspondingly DV is
not precisely known. The existing criteria are considered conservative for treated soils.

FIELD TESTING
In this study, field tests were performed to evaluate in-situ properties of MDOT
pavements with LTS.
Ground Penetrating Radar (GPR)
GPR operates by transmitting short pulses of electromagnetic energy from an
antenna into the pavement. These pulses are reflected back to a receiving antenna. The
reflected energy is collected and displayed as a waveform. The large peaks of the
waveform are the energy reflected from the surface and interfaces between layers. The
time interval between peaks is the time the radar wave needs to travel from the top of the
layer to the bottom and back (twice the layer thickness). The speed with which the
electromagnetic radar wave travels in a particular layer is related to the dielectric constant
of that layer. In pavements, the parameter that most influences the dielectric properties of
materials is the moisture content (6).

7
 The layer thicknesses as determined by GPR are used to validate actual layer
thicknesses, which are in turn used to more accurately back-calculate layer moduli.
Dielectric values from GPR data are used to assess layer moisture sensitivity of the layers
and durability.
Software called COLORMAP was used to analyze the radar signals. This
software measures the amplitude and time delay of each radar trace received and applies
the signal processing to calculate layer dielectric constants and layer thicknesses (7).
Layer thicknesses and dielectric constants for all the MDOT pavements are summarized
in Table 2.
Dynamic Cone Penetrometer (DCP)
DCP provides a log of resistance to penetration under an impact load that has
been effectively correlated to in situ California Bearing Ratio (CBR) and modulus by the
U.S. Army Corps of Engineers (8). This process provides not only in situ strength and
stiffness data but also a log of the thickness of various paving layers to be used together
with GPR determined thicknesses in modulus back calculations using FWD data.
DCP tests were performed in three locations in each pavement. The average
values of these tests for each pavement are presented in Table 2.
Falling Weight Deflectometer (FWD)
FWD delivers a transient force impulse to the pavement surface that simulates a
moving wheel load in both magnitude and duration. By varying the amount of weight and
the height of drop, different impulse forces can be generated. The deflection data
obtained from the FWD testing are used for evaluation of the in-situ stiffness of
individual pavement layers (9).
A program called MODULUS, developed at Texas A&M University (Texas
Transportation Institute), was used to determine layer moduli by back-calculation
method. MODULUS stores many generated deflection basins and corresponding moduli
values in a database for a given layer configuration. When a measured deflection basin is
analyzed, the database is screened, and interpolations are used to find a deflection basin
that best represents the measured basin (10). Although the moduli values for all the layers
of the pavements were back-calculated from the field data, only the LTS and subgrade
moduli are shown in Table 2.

8
COMPARISON OF LABORATORY AND FIELD DATA
The following laboratory test properties to field-test properties comparisons were
made:
(1) Laboratory unconfined compressive strengths (UCS) were compared to
field dynamic cone penetrometer (DCP) strength measurements.
(2) Laboratory-determined resilient moduli of the lime-stabilized soils
were compared to back-calculated resilient moduli based on Falling
Weight Deflectometer (FWD) deflection basins.
(3) Laboratory assessment of moisture sensitivity and durability of the
lime-treated soils based on the dielectric value measurements and swell
tests were compared to field Ground Penetration Radar (GPR)
measurements.
Comparison of Laboratory UCS with Field DCP
From Figure 1, it is apparent that soaked UCS tests for all Mississippi LTSs are
substantially higher than their untreated counterparts. In fact, the stabilized-to-
unstabilized strength ratios shown in Table 1 are all greater than 45. Furthermore, the
value of UCS of the soaked soils cured 7-days at 400C are all above about 1,700 kPa
(about 250 psi) which are considered by Thompson (5) as values acceptable for
assignment of structural significance as a base layer. These values correlate well with
field DCP values (Table 2). These values show that the back-calculated CBR of the LTS
layers are between 12 and 33 times the strength of the in situ natural subgrade.
Comparison of Laboratory Resilient Moduli with Field FWD Back-Calculations
Figure 2 summarizes laboratory determined resilient moduli for the four
Mississippi LTSs. There is a striking difference between the resilient moduli for
stabilized and unstabilized soils following soaking. For US 45N, the stabilized layer is
6.7 times greater than its unstabilized counterpart. The other unstabilized samples could
not be tested due to disintegration during soaking. Furthermore, the soaked resilient
moduli values for the four treated soils are all above about 200 MPa or about 30,000 psi.
This is the level normally correlated with structural significance, e.g., AASHTO 1986
(11) assigns a structural layer coefficient of 0.14 to subbases with a resilient modulus of

9
200 MPa. It should also be noted that the soaked resilient moduli of the LTS layers for
US 61N, US 82W, and US 45N are considerably higher than 200 MPa, ranging from
about 260 MPa to about 415 MPa. These laboratory measured moduli are considerably
lower than the FWD back-calculated moduli, but are consistent with the back-calculated
values in the sense that both lab and field moduli of the LTS measurements are
considerably higher than their unstabilized counterparts. Note that the back-calculated
moduli are all above about 425 MPa and that the ratios of LTS to natural subgrade
moduli for the field conditions at the time of testing demonstrate that the stiffnesses of
the LTS layers are between 4.4 and 20.7 times greater than the stiffnesses of the natural
subgrade.
It is often difficult to develop unique layer moduli based on back-calculation
techniques in pavement with multiple layers. However, in this analysis the subgrade
modulus was assigned based on the deflection the sensor located approximately 225 cm
from the point of load impact (Sensor #7). The modulus of the HMA surface was
assigned based on the mixture properties, the impact properties of the falling weight and
the HMA temperature at the time of testing. These data were placed in Witczak’s
equation (12) to estimate the HMA modulus. Since the HMA and subgrade moduli were
already determined, the modulus of the LTS was the only unknown in each back-
calculation.
Comparison of Laboratory Dielectric Value to Field GPR
Figure 3 shows DV results of unstabilized and stabilized soils from the four
Mississippi sites. These samples are allowed to imbibe water through capillary suction.
Figure 5 clearly demonstrates the differences between unstabilized soils and their
stabilized counterparts. For moderately plastic clay soils (US 82E and US 82W), the
differences are more obvious as the stabilized samples are intact after 12-hours of soak;
whereas the unstabilized samples have swelled significantly and have cracked,
particularly at the horizontal interface separating compaction lifts. The correlation
between visual evidence of moisture uptake and DVs is not as good in these soils as for
samples from US 61N and US 45N. This is mainly due to the interruption of flow due to
horizontal fracture at compaction zone interfaces. However, as shown in Figure 5, the

10
degree of moisture absorption and swelling is substantially and consistently higher in the
unstabilized than in the stabilized samples for all of the four pavement soils.
Field DVs derived from the GPR are consistent with the laboratory results as all
the field DVs are well below the threshold value of 16 which separates an acceptable
pavement base or subbase from a marginal or poor quality base or subbase. In fact, the
DVs shown in Table 2 indicate a sublayer of good quality in terms of its ability to resist
moisture damage effects. Considering that these are stabilized clay layers, this is a
significant statement.

MECHANISTIC ANALYSIS
Mechanistic analysis was performed for the Mississippi pavements with and
without considering the LTS layer. ELSYM5, a layered elastic computer program, was
used to determine four parameters:
(1) tensile strain (εt) at the bottom of the hot mix asphalt (HMA) layer
which is related to fatigue cracking in the HMA layer,
(2) vertical compressive strain (εv) at the top of the subgrade which is
related to deep layer rutting,
(3) flexural stress (σt) at the bottom of the LTS layer which is related to
fatigue within the LTS, and
(4) vertical compressive stress (σv) at the top of the subgrade which is also
related to deep layer rutting due to the accumulation of deformation
within the subgrade.
The pavement structure in the layered elastic model was based on GPR data, which
defined layer thicknesses and the back-calculation of moduli values (E) of different layers
from FWD data (Table 2). These values are presented in Table 3.
Resistance of HMA and LTS Layers to Flexural Fatigue
The allowable number of load repetitions (Nf) to control fatigue cracking in the
HMA layer for the pavements with and without the LTS layer were calculated using the
well-known transfer function (15):

N f = 0.0796(ε t )
−3.291 −0.854
E (1)

11
 A typical fatigue-response relationship for LTS layers between the ratio of load-
induced flexural stress to flexural strength (stress ratio, S) and the number of load
applications to failure (N) for different pavements was developed by Thompson and
Figueroa (16):
S = 0.923 − 0.058 log N (2)

This relationship was used to assess fatigue damage due to load-induced flexural stresses
in the LTSs.
The flexural stress is approximately 0.20 to 0.23 times the UCS (13). The four
Mississippi DOT pavements were respectively modeled using thicknesses verified by
GPR testing and moduli from FWD back-calculations. Pavement design lives were
calculated from layered elastic analysis and the results are cataloged in Table 3. The
stress ratio was calculated as the maximum flexural stress induced in the LTS divided by
0.29 x UCS. This value was then used to determine N in equation 2.
Permanent Deformation Potential
The allowable number of load repetitions (Nd) to control permanent deformation
for the pavements with and without LTS layer were calculated using the familiar transfer
function (15):
N d = 1.365 × 10 −9 (ε v )
−4.477
(3)

The effect of the LTS on deep-layer rutting potential can also be assessed by
comparing the vertical stress at the top of the subgrade to the compressive strength of the
subgrade layer. The analysis of many different UCS tests (17) has demonstrated that the
stress-strain plot derived during the UCS tests becomes non-linear at about one-half the
stress required for failure. In other words, at stress levels below about 0.5 times the UCS,
strains are recoverable, but at values of vertical compressive stress at the top of the
subgrade that exceed about 0.5 times the UCS, some permanent strain occurs that
accumulates with loading cycles. This can result in deep-layer or subgrade rutting.
Therefore, the vertical compressive stress at the top of the natural subgrade was
computed for each pavement and compared to the soaked subgrade strength to assess the
potential for the accumulation of permanent strain or deep-layer rutting.

12
 The mechanistic analysis is summarized in Table 3. This analysis is the result of a
layered elastic determination of four important parameters (εt, εv, σt , and σv) as previously
discussed. Based on these parameters and the appropriate transfer functions or related
damage models, relative performance predictions for the four Mississippi DOT
pavements can be made. The HMA thicknesses are so thick for the pavements analyzed
that the predicted life in terms of equivalent 80kN (18,000 pound) single axle loads
(ESALs) are very large. Therefore, the effect of the LTS was judged in terms of the
percentage increase in pavement life (increase in number of ESALs the pavement is able
to carry during its design life). These increases are summarized as follows:
1. Based on the parameter εt, which is the controlling design parameter in each case,
the pavements with LTS layers show an increase in life of
• 900% for US 61N,
• 30,000% for US 82E,
• 3,000% for US 82W, and
• 2,000% for US 45N.
2. Based on the parameter σt, which together with the flexural strength of the LTS
defines the ability of the LTS to resist flexural fatigue, each LTS layer is capable
of easily supporting the ESALs in the design life used in item 1.
3. Based on stress induced at the top of the subgrade, σv, the LTS reduces stress by
• 47% for US 61N,
• 55% for US 82E,
• 48% for US 82W, and
• 61% for US 45N.
Because the subgrade UCSs are so weak after soaking, the stress reductions from
lime treatment significantly reduce the potential for accumulation of permanent
strain at the top of the subgrade (see Table 3).

13
RECOMMENDED ENGINEERING PROPERTIES FOR MECHANISTIC
DESIGN

The MDTP generates unconfined compressive strengths and resilient moduli of

lime-treated pavement materials following moisture conditioning (as described in
Reference 1). The values produced by this testing protocol for these example pavements
are summarized in Table 4. These design values are realistic and reasonably
conservative.

In this design example, laboratory values were validated through field testing.
This field testing was performed in the spring of the year (March through April), which is
typically a wet period for the sections involved. Therefore, although field conditions can
never be controlled, the field values should also be conservative.

This example validates the MDTP. It confirms that structural values to be used in
mechanistic design can be derived from practical laboratory test procedures.

14
REFERENCES

1. Little, D.N., Evaluation of Structural Properties of Lime Stabilized Soils and

Aggregates, Volume 3: Mixture Design and Testing Protocol for Lime Stabilized
Soils, National Lime Association, 2000 (http://www.lime.org/SOIL3.PDF).
2. Little, D.N., Evaluation of Structural Properties of Lime Stabilized Soils and
Aggregates, Volume 1: Summary of Findings, National Lime Association, 1999
(http://www.lime.org/SOIL.PDF).
3. Currin, D.D., J.J. Allen, and D.N. Little, Validation of Soil Stabilization Index
System with Manual Development, Report No. FJSRL-TR-76-0006, Frank J.
Seiler Research Laboratory, United States Air Force Academy, Colorado, 1976.
4. Petry, T.M. and D.N. Little, Recent Developments in Sulfate-Induced Heave in
Treated Expansive Clays, Second Interagency Symposium on Stabilization of
Soils and Other Materials, US Army Corps of Engineers, Metairie, Louisiana,
1992. [or National Lime Association, Guidelines for Stabilization of Soils
Containing Sulfates (Austin White Lime, Chemical Lime, Texas Lime), August,
2000 (http://www.lime.org/sulfate.pdf).]
5. Thompson, M.R., Soil Stabilization for Pavement Systems – State of the Art,
Technical Report, Construction Engineering Research Laboratory, Champaign,
Illinois, 1970.
6. Scullion, T. and T. Saarenketo, Using Suction and Dielectric Measurements as
Performance Indicators for Aggregate Base Materials, Transportation Research
Record 1577, TRB, National Research Council, Washington, D.C., 1996.
7. Scullion, T., Y. Chen, and C.L. Lau, COLORMAP-User’s Manual with Case
Studies, Research Report 1341-1, Texas Transportation Institute, Texas A&M
University, College Station, Texas, 1995.
8. Webster, S.L., R.H. Grau and T.P. Williams, Description and Application of Dual
Mass Dynamic Cone Penetrometer, Instruction Report GL-92-3, US Army Corps
of Engineers Waterways Experiment Station, Geotechnical Laboratory, 3909
Halls Ferry Road, Vicksburg, Mississippi, 1992.

15
9. Huang, H.Y., Pavement Analysis and Design, Prentice Hall, Englewood Cliffs,
New Jersey, 1993.
10. Uzan, J., T. Scullion, C.H. Michalak, M. Paredes, and R.L. Lytton, A
Microcomputer Based Procedure for Back-calculating Layer Moduli from FWD
Data, Research Report 1123-1, Texas Transportation Institute, Highway Materials
Division, College Station, Texas, 1988.
11. AASHTO Pavement Guide for the Design of Pavement Structures, 1986.
12. Witczak, M. W., and R.E. Root, “Summary of Complex Modulus Laboratory Test
Procedures and Results,” STP 561, American Society for Testing and materials,
pp. 67 – 94, 1974
13. Little, D.N., Handbook for Stabilization of Pavement Subgrades and Base
Courses with Lime, Kendall/Hunt Publishing Company, Iowa, 1995.
14. Eades, J.L., F.P. Nichols, and R.F. Grim, Formation of New Minerals with Lime
Stabilization as Proven by Field Experiments in Virginia, Highway Research
Board Bulletin No. 335, 1962.
15. Asphalt Institute, Research and Development of the Asphalt Institute’s Thickness
Design Manual (MS-1), 9th Ed., Research Report 82-2, 1982.
16. Thompson, M.R. and J.L. Figueroa, Mechanistic Thickness-Design Procedure for
Soil-Lime Layers, Transportation Research Record 754, National Academy of
Sciences, Washington, D.C., 1980.
17. Zollinger, D., S. Kadiyala, and R. Smith, Review of Alternative Pavement Design
for Area A of Taxiway B and C Construction, Prepared for Dallas/Fort Worth
International Airport, Report No. 4, 100-1, Texas Transportation Institute, 1989.

16
APPENDIX: OTHER TESTS TO ASSESS LTS DURABILITY

Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray (EDX) Test
SEM/EDX analysis was performed at the Electron Microscopy Center of Texas
A&M University. Sample preparation included mounting samples on carbon double-stick
tape on aluminum stubs. The samples were coated with 300 Å of gold/palladium using a
hammer I Sputter Coater.
The microstructure of the natural soils contained significant amounts of typical
thin, platy clay structures that form the dominant feature in these micrographs. The pores
in the matrix range from a few micrometers to 20µm in size. EDX analysis of several of
these plates suggests that the clay is predominantly composed of aluminosilicates, with
detectable amounts of potassium and iron in the lattice. The iron content was found to
vary from one location to another. Calcium was either absent or present in a negligible
amount.
The microstructures of the counterpart lime-stabilized samples were distinctly
different. A much denser matrix with fewer micropores was observed in these samples.
The presence of calcium (from lime) was detected on the clay particles. Additionally,
calcium-silicate-hydrate (C-S-H) was identified in the matrix. The EDX analysis
indicates that the pozzolanic reaction had already converted some of the clay minerals to
C-S-H after a 7-day cure period (13). An elevated curing temperature used for these
experiments most likely accelerated this reaction. The densification of the matrix is due
to the pozzolanic reaction. The net effect was a significant increase in strength after 7-
days.
X-ray Diffraction (XRD) Test
XRD analyses were performed at the Geology Department of Texas A&M
University. Sample preparation included grinding of material. Approximately 1 gm of the
clay size fraction was applied to a slide with acetone.
In order to explain the pozzolanic reaction that takes place in the lime-stabilized
soil samples, X-ray diffraction (XRD) analysis was performed on samples of the as-
received native soils as soon as they were mixed with lime, and on the lime-stabilized soil
after 7-days of curing at 400C. Figure 4 (top), which represents the XRD pattern of the
native soil from US 61N, clearly indicates that it is composed mainly of quartz, feldspars

17
and clay minerals. In contrast, the characteristic peaks of hydrated lime or Ca(OH)2 are
visible in the XRD pattern of the same soil sample immediately after it was mixed with
the lime, Figure 4 (middle). Reduction in the amount of hydrated lime after this lime-
stabilized soil was cured for 400C is evident from the decrease in the peak heights of
Ca(OH)2, shown in Figure 4 (bottom). This decrease was calculated by peak integration
and found to be 35% less than the original value. The Ca(OH)2 is expected to have been
used to form C-S-H; however, formation of C-S-H cannot be identified in the XRD
pattern because of its amorphous nature.
Analysis of Pozzolanic Development
Each of the four soils (representing the four Mississippi LTS layers) is amenable
to lime stabilization based on fines content and plasticity requirements. Furthermore, the
Eades and Grim pH test results indicate that sufficient lime was used for stabilization.
This indicates not only that enough lime was present for a pozzolanic reaction but also
that the pH would remain high for a substantial period of time. Eades et al. (14)
demonstrated that as long as the pH remains above about 10.5, the pozzolanically driven
strength gain is likely to continue as is the concomitant autogenous healing. The SEM,
EDX and XRD previously discussed further substantiate that a denser soil-lime matrix
has been established that supports pozzolanic strength gain. The level of compressive
strength gain and resilient modulus improvement, following capillary soak, indirectly
validates this, as well.

18
 3500
3308 3473 Unstabilized Samples
Stabilized Samples
3000
2734
2620

2500
2175 2349
1987 1993
1885 1930
UCCS (kPa)

2000
1678 1700

1500

1000

500

31 42 1 16
0
Curing Dry Wet Dry Wet Dry Wet Dry Wet
Condition US 61 N US 82 E US 82 W US 45 N

FIGURE 1 Relationship Between Dry and Soaked Unconfined Compressive Strength of

Unstabilized and Stabilized Mississippi Soils.

600
Unstabilized Samples
516 517
Stabilized Samples
500

415
Resilient Modulus (MPa)

399
400 373
353 367

294
300
257 260
234
201
200

100
55

0
Curing Dry Wet Dry Wet Dry Wet Dry Wet
Condition US 61 N US 82 E US 82 W US 45 N

FIGURE 2 Comparison of Resilient Moduli of Unstabilized and Stabilized Mississippi

Soils for Dry and Soaked Conditions.

19
 50
45 Dielectric Value
45
Change in Moisture Content (%)
39
40 Swell (%)

35

30 28
24
25
20 19
18
20
16
15
12 12 13
13 10
9 8
10
6 7 6 6
5 4
5 4
3 3
0
Unstab. Stab. Unstab. Stab. Unstab. Stab. Unstab. Stab.
US 61 N US 82 E US 82 W US 45 N

FIGURE 3 Comparison of Dielectric Value, Change in Moisture Content and Swell for
Unstabilized and Stabilized Mississippi Soils after 8-days of Capillary Soak.

20
FIGURE 4 XRD Patterns of Soil, where C = Clay Minerals, Q = Quartz, F = Feldspar
Minerals and CH = Ca(OH)2 for: (top) As-received Native Soil; (middle)
Immediately After Mixing with Lime (note the CH peaks at 22 of 180 and
340); and (bottom) After Curing for 7-days at 400C.

21
FIGURE 5 Unstabilized and Stabilized Soil Samples of (top) US82W and US82E;
(bottom) US45N and US61N after 12-hours of Soaking.

22
TABLE 1 Unconfined Compressive Strength and Resilient Moduli of Unstabilized and Stabilized
Mississippi Soils (NT = Not Tested).
Soil I.D. Curing Unstabilized Soil Stabilized Soil Ratio (Stabilized
Condition Individual Average Individual Average / Unstabilized)
Unconfined Compressive Strength (kPa) of Dry Specimens
US 61 N 7-days @ 400C 2342 2175 3056 3308 1.52
Washington Co. 2008 3559
US 82 E 7-days @ 400C 1912 1993 3661 3473 1.74
Washington Co. 2074 3284
US 82 W 7-days @ 400C 2604 2620 2349 2134 0.81
Lowndes Co. 2636 1919
US 45 N 7-days @ 400C 1740 1700 3008 2734 1.61
Kemper Co. 1660 2460
Unconfined Compressive Strength (kPa) of Soaked Specimens
30-days @ 250C 2 14 1585 1704 125
10 1461
US 61 N 29 2067
0
Washington Co. 7-days @ 40 C 25 31 1827 1987 63
32 2030
37 2104
30-days @ 250C 20 8 1249 1506 188
2 1650
US 82 E 2 1620
0
Washington Co. 7-days @ 40 C 41 42 2080 1885 45
58 1803
26 1773
30-days @ 250C 1 1 1198 1356 1017
1 1298
US 82 W 2 1573
Lowndes Co. 7-days @ 400C 1 1 1729 1678 1678
1 1650
1 1654
0
30-days @ 25 C 4 8 1158 1445 188
6 1486
US 45 N 13 1690
Kemper Co. 7-days @ 400C 15 16 1788 1930 118
26 2072
8 1931
Resilient Modulus (MPa) of Dry and Soaked Specimens
US 61 N Dry 288 294 530 516 1.76
Washington Co. 300 502
Wet NT 415 415
US 82 E Dry 217 257 377 353 1.37
Washington Co. 297 329
Wet NT 201 201
US 82 W Dry 252 234 404 399 1.70
Lowndes Co. 216 393
Wet NT 260 260
US 45 N Dry 340 373 520 517 1.39
Kemper Co. 405 514
Wet 55 55 367 367 6.67

23
TABLE 2 GPR, FWD and DCP Results for LTS and Unstabilized Subgrade Soils for
Four Mississippi Pavements.

Pavement GPR Results DCP Results FWD Results

Layer Dielectric Layer Subgrade LTS Ratio Subgrade LTS Ratio
Thickness Constant Thickness CBR CBR (Stabilized/ Moduli Moduli (Stabilized/
(mm) (mm) Unstabilized) (MPa) (MPa) Unstabilized)
US 61 N HMA: 250 9 - 13 LTS: 125 15 500 33.33 97 425 4.38
Washington LTS: 150
County
US 82 E HMA: 325 6-8 LTS: 125 12 150 12.50 119 2466 20.72
Washington LTS: 150
County
US 82 W HMA: 363 7 - 10 LTS: 150 4 47 11.75 123 1350 10.98
Lowndes LTS: 150
County
US 45 N HMA: 250 No LTS: 275 10 133 13.30 125 1482 11.86
Kemper LTS: 250 Data
County

TABLE 3 Mechanistic Analysis for Mississippi Pavements With and Without

Considering the LTS Layer.

Soil Pavement UCCS E (HMA) E (LTS) E (Subgrade) εt εv σt σv N Nf Nd

ID Structure (kPa) (MPa) (MPa) (MPa) (mm) (mm) (kPa) (kPa) (no.) (no.) (no.)
US61N w LTS 1987 1151 425 97 0.0040 0.0062 91 40 6E+12 9E+06 2E+07
w/o LTS 31 1151 97 0.0070 0.0086 449 76 0E+00 1E+06 5E+06
US82E w LTS 1885 1627 2466 119 0.0006 0.0017 208 20 2E+08 3E+09 6E+09
w/o LTS 42 1627 119 0.0034 0.0041 314 44 0E+00 1E+07 1E+08
US82W w LTS 1678 1434 1350 123 0.0011 0.0021 126 21 5E+10 6E+08 3E+09
w/o LTS 1 1434 123 0.0031 0.0037 245 40 0E+00 2E+07 2E+08
US45N w LTS 1930 4420 1482 125 0.0011 0.0013 70 15 3E+13 2E+08 2E+10
w/o LTS 16 4420 125 0.0024 0.0033 680 38 0E+00 1E+07 3E+08

TABLE 4 Recommended Design Values for Mechanistic Analysis.

Material Unconfined Compressive Resilient Modulus, MPa
Strength, kPa
US 61 N 1900 400
US 82 E 1600 200
US 82 W 1600 250
US 45 N 1900 300

24