Seasonal Resilient Modulus Inputs for Tennessee Soils and Their Effects on Asphalt Pavement Performance

Transcription

1 Seasonal Resilient Modulus Inputs for Tennessee Soils and Their Effects on Asphalt Pavement Performance Changjun Zhou czhou@utk.edu Baoshan Huang (Corresponding author) bhuang@utk.edu Eric Drumm edrumm@utk.edu Xiang Shu xshu@utk.edu Qiao Dong qdong@utk.edu Sampson Udeh Sampson.Udeh@tn.gov Word Count: (text: ; figures & tables: ) Department of Civil and Environmental Engineering, University of Tennessee, Knoxville, TN -00, Phone: () -, Fax: () - Department of Biosystems Engineering and Soil Science, University of Tennessee, Knoxville, TN -, Phone: ()-, Fax: ()-0 Design Division, Tennessee Department of Transportation, Nashville, TN -0, Phone: ()-, Fax: ()- TRB 0 Annual Meeting

2 0 0 Abstract The soil subgrade, which supports the above pavement layers and traffic, should be stiff enough to maintain the integrity of the pavement structures and the smoothness of the pavement surfaces. The resilient modulus, as a property of subgrade stiffness, is an important input in the AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG). In the MEPEG input level, the generalized universal model is used to represent the resilient modulus, and coefficients for this model are required. The change to this model has raised the interest of states converting old resilient modulus test data, which may have been presented in terms of other models, into the general universal model as implemented in MEPDG. Based on cyclic triaxial load test data of clayey soils in Tennessee (TN), coefficients of the generalized model were regressed. Also the coefficients were regressed from soil physical properties, which can be utilized as an alternate time-saving and economical method to obtain soil resilient modulus. The coefficients were obtained at different post-compaction water contents, to allow the estimation of pavement response under natural seasonal variation of subgrade water content. Two typical pavement sections, I-0 Knox and SR- Washington, were evaluated for pavement performance utilizing a multiple layered software Wealsea.0 and the version. MEPDG software. The results showed that because moisture variation significantly affected the subgrade resilient modulus, without including these effects with appropriate coefficients for the generalized model the full pavement performance is not captured. Keywords: subgrade, resilient modulus, MEPDG, generalized model, coefficients, seasonal moisture variation TRB 0 Annual Meeting

3 INTRODUCTION Mechanical-empirical based pavement design models use resilient modulus of soils to evaluate dynamic response and fatigue behavior of pavement materials under vehicle loading. The AASHTO guide for design of flexible pavement () suggests the use of resilient modulus ( ) for characterizing the subgrade soil. is defined as deviator stress divided by the resilient or recoverable axial strain, due to cyclic axial stress. Currently, MEPDG allows the use of AASHTO T0 test standard () or the NCHRP - A procedure () to evaluate resilient modulus of soil. The two methods are almost the same except in the tolerance of moisture and density between replicate samples. In repeated load triaxial test, a combination of three confining stresses (.,., and.kpa for subgrade soil) applied on cylindrical specimens simulates overburden pressure and applied wheel load. And a series of load pulses (.,.,.,., and.kpa for subgrade soil) are applied with a distinct rest period on cylindrical specimens of soil, simulating the stresses from multiple wheels moving over the pavement. In the field, subgrade soil at different depth experiences varied bulk stresses, dependent on the stiffness, thickness, and other factors of the pavement overlayers. Therefore, this test standard still needs improvement. In-situ tests are preferred as long as reliable correlation could be reached, due to the complexity and time consuming of triaxial tests in the laboratory. Factors including stress state, soil type and its structure, moisture content, density, gradation, etc, are usually named when analyzing of soil. California Bearing Ratio (CBR) is commonly used to estimate (). The new Mechanical-Empirical Pavement Design Guide Level input may determine from CBR, R value, and layer coefficient respectively. In-situ apparatuses, such as field static plate bearing load test (,), falling weight deflectometer (,,) can be used to obtain field resilient modulus. Usually relationships between resilient modulus and CBR or other mechanical properties obtained in the field can be established. Relationships between the resilient modulus and the material stress state have been studied for decades. The K-θ model (0), generally for granular materials, does not consider shear stress and shear strain developed during loading. The K- model, proposed by Moossazadeh and Witczak (), is adequate for cohesive soils found at shallow depths. The universal model () covers the effects of shear, confining, and deviator stresses and gives a better explanation for the stress state of soils. Later the generalized model () was adopted in MEPDG software. After constants involved in the constitutive models are determined from laboratory tests, the soil resilient modulus under a specific stress state can be obtained. Generally, constants involved in the generalized model can be obtained from repeated load test results in the laboratory, as MEPDG input level, or by regressing the coefficients from soil physical properties (, (for the universal Model)), and both the effects of season and stress sensitivity can be considered. OBJECTIVE AND SCOPE The objective of the study was to establish a subgrade resilient modulus input database for the MEPDG in Tennessee, and to study the variation of soil resilient modulus due to moisture change and its effect on pavement performance. Coefficients of the generalized model were regressed from cyclic triaxial load test data of soils in Tennessee. Coefficients of the generalized model were regressed from soil physical properties and utilized in evaluating the seasonal variation of subgrade resilient modulus. And TRB 0 Annual Meeting

4 0 0 0 the influences of subgrade resilient modulus seasonal variation on pavement performance were studied. COEFFICIENTS OF THE UNIVERSAL MODEL VERSUS COEFFICIENTS OF THE GENERALIZED MODEL The MEPDG uses the generalized model, as shown in Equ., to calculate soil resilient modulus. M =k P + where: M = resilient modulus; θ= bulk stress; σ +σ +σ ; σ,σ,σ = principal stresses; τ = octahedral shear stress, (σ σ ) +(σ σ ) +(σ σ ) ; P =atmospheric pressure; k,k,k = regression constants. Soils from locations around Tennessee were collected and the physical properties and resilient modulus (in accordance with SHRP Protocol P ()) were tested in the University of Tennessee (,,). Among these soils, are silty soils and are clayey soils. The resilient moduli of the clayey soils were evaluated under three different moisture contents: optimum water content and two higher water contents. The results of the lab tests were represented by determining coefficients for the universal model (), which can be expressed as: M =k P The coefficients of the universal model and the generalized model were determined in the linear regression analysis based on the resilient modulus data from the laboratory (Table ). The coefficients of the first row in each soil were for optimum moisture content, i.e. for the representative resilient modulus in the MEPDG input level, while coefficients of the other rows can be used to predict resilient modulus of soils with higher water contents. The ratios of coefficients from the universal model to the corresponding ones from the generalized model were shown in Figure. It can be seen that there was almost no change of k, while k and k varied significantly. The variation of k was more scattered than k and k. It is obvious that coefficients from the universal model should not be adopted directly in the MEPDG software. Therefore, if agencies have the coefficients from the universal model but the original resilient modulus test data is missing, they can convert the universal model coefficients to those of the generalized model: firstly, obtain a series of resilient modulus from the universal model with these coefficients; secondly, obtain coefficients of the generalized model from the linear regression analysis with the calculated resilient moduli. () () TRB 0 Annual Meeting

5 TABLE Coefficients of the Generalized Model and the Universal Model for Tennessee Soils Location Crockett Co. Sta Shelby Co. Sta Roane Co. Sta Hamilton Co. Sta Roane Co. Sta Crockett Co. Sta Crockett Co. Sta 0 AASHTO Classification A- A- A- Moisture Dry Generalized Model Universal Model Content, Density, % g/cm k k k R k k k R A A- A- A TRB 0 Annual Meeting

6 TABLE Coefficients of the Generalized Model and the Universal Model for Tennessee Soils (Continued) Location White Co. Sta Giles Co. Sta 0 Knox Co. Sta 00 VanBuren Co. Sta Knox Co. Sta Rutledge Pike Knox Co. Sta 00 AASHTO Classification Moisture Content, % Dry Density, Generalized Model Universal Model g/cm k k k R k k k R A A-- A A A-- A TRB 0 Annual Meeting

7 Ku/Kg k u /k g 0.0 Ku/Kg Ku/Kg Sample Number 0 Figure the Ratios of Coefficients from the Universal model to the Corresponding Coefficients from the Generalized Model COEFFICIENTS OF THE GENERALIZED MODEL REGRESSED FROM PHYSICAL PROPERTIES The establishment of relationships between coefficients of the generalized model and soil physical properties provides a convenient and economical way to evaluate resilient modulus of a new soil as long as this soil is similar to the ones utilized in the regressions. Factors frequently in developing coefficients models are moisture content, degree of saturation, plasticity index, material passing the #00 sieve, and dry density. Based on sensitivity analysis, George (0) found that the most important input variable is judged to be sample moisture content, followed by materials passing #00 sieve, plastic index and sample density in that order. This is likely to vary for different soils and different stress conditions however. Drumm et al. () reported values of soil physical properties such as Atterberg limits, specific gravity, gradation, water content and dry density, which were introduced as independent variables and Log(k ), k, and k obtained from cyclic triaxial tests were dependent variables. A commercial statistical software SAS. was used to conduct the multiple linear regression on the universal model and the generalized model. clayey soils, i.e. A- and A-, and silty TRB 0 Annual Meeting

8 0 0 soils, i.e. A-, were regressed separately. As samples with three different water contents of each soil were included, the seasonal moisture variation of soils was considered. Since there are many independent variables, an ever-present danger is that of selecting a model that overfits the "training" data used in the fitting process, yielding a model with poor predictive performance. Using k-fold cross validation is one way to assess the predictive performance of the model. The PRESS statistic was used here to choose among models that were chosen based on entry and stay significance levels (both are 0., as defaulted). Regressed models were shown in Table. It can be seen from Table that factors including plastic limit, percentage of clay, percentage passing #00 sieve, specific gravity, liquid limit, optimum water content, maximum density, and water content significantly influence the resilient modulus of clayey soils, while specific gravity, water content, and percentage passing # sieve significantly influence the resilient modulus of silty soils. Also it can be seen from the regressed models, resilient modulus of soils decreases as the water content increases from optimum water content. It should be noted that only samples were utilized when developing the regressed coefficients for silty soils, and more confidence would be achieved if more samples were involved. Table Regressed models of coefficients from physical properties for soils in Tennessee Model R F Value Clayey Soils logk = PL Clay 0.00Passing#00 +.0SG.0 w /w 0.. k = LL 0.0Clay 0.. k = Clay.γ + 0.w.w +.0 w /w 0.. Silty Soils logk =.0.SG.00 w /w 0.. k = 0..SG 0.. k =. + 0.Passing# Note: Clay=percentage of clay in soil, %; SG=specific gravity; =optimal water content, %; = maximum density; = water content, %; #00= percentage passing #00 sieve; #= percentage passing # sieve. TRB 0 Annual Meeting

9 The regressed coefficients were compared with the experimental coefficients, as shown in Figure. It can be seen that the regressed models proved fairly good predictions on the coefficients of the generalized model. 0 Figure Experimental Coefficients Versus Regressed Coefficients in the Generalized Model Regressed Coefficients from Physical Properties A series of resilient moduli of each clayey soil, which were named regressed resilient moduli hereafter, were calculated from regressed coefficients, and the cyclic stresses and confining pressures in its corresponding cyclic triaxial tests. Figure shows the comparison between regressed resilient moduli and experimental resilient moduli of clayey soils. In general, the majority of regressed resilient moduli were close to the experimental resilient moduli. Therefore, the relationships for clayey soils in Table are valid and can be utilized as a time- saving and economical method to evaluate coefficients in the generalized model and the resilient moduli of clayey soils. log(k) k k Experimental Coefficients TRB 0 Annual Meeting

10 0 Regressed Mr (MPa) Experimental Mr (MPa) Figure Experimental Resilient Moduli versus Regressed Resilient Moduli SEASONAL CHANGE OF CLAYEY SOIL RESILIENT MODULI IN TN In pavement design, resilient modulus of soil under optimum water content (standard proctor) is usually adopted. However, resilient modulus of soil is related to the moisture (0,,). Zuo () selected four locations in TN and monitored moisture variation in subgrade. Among these four locations, the subgrade soil from Blount County is classified as A--, which is the same as the soil in Knox Co. Sta 00, and these counties are geograghically close and likely form the same parent similar climate. To evaluate the influence of moisture on soil resilient modulus, Knox Co. Sta 00 soil in Table was selected and assumed to experience the annual moisture variation, as shown in Figure, which was the change of moisture 0.m under the subgrade surface on the Blount County pavement site (). Water contents in the soil shown in Figure were higher than the optimum water content (.%) in Table. Coefficients of the generalized model for Knox Co. Sta 00 soil were determined from the regressed models in Table, and the results were also shown in Figure. It can be seen from Figure that log(k ) and k decrease while subgrade moisture increases, vice versa. As recommended by AASHTO T0, stress states including three confining pressure levels, i.e.,.,., and.kpa and five deviator stresses, i.e.,.,.,.,.,.kpa were applied to the Knox Sta.00 soil, as shown in Figure. As the subgrade depth increases, TRB 0 Annual Meeting

11 confining pressure increases while deviator stress decreases. As the horizontal distance increases from point of the traffic load, the deviator stress in soil decreases. Figure and Figure showed the seasonal change of soil resilient modulus. It is indicated when the water content is higher than the optimum one, there is a negative correlation between resilient modulus and water content. The variation of soil resilient modulus can be around 0MPa. It can also be seen that soil vertically under traffic loads displays smaller resilient modulus than those located deeper (Figure ) or horizontally farther away from traffic loads (Figure )...0 moistu re content ( %) log(k) k Time since 00/0 (month) Figure Annual Changes of Coefficients of the Generalized Model with Seasonal Moisture Variation in Subgrade on Knoxville Sta. 00 TRB 0 Annual Meeting

12 0 Traffic Loads Asphalt Surface Course Base Course Subbase Course σ =.kpa, σ d =.kpa σ =.kpa, σ d =.kpa σ =.kpa, σ d =.kpa Figure the Sketch of Subgrade Stress State.kPa,.kPa.kPa,.kPa σ =.kpa, σ d =.kpa σ =.kpa, σ d =.kpa σ =.kpa, σ d =.kpa Subgrade.kPa,.kPa.kPa,.kPa 0.0.kPa,.kPa moisture content. Resilient Modulus (MPa) Moisture Content (%) Time since //00 (month) Figure Annual Resilient Moduli Change of Clayey Soil due to Seasonal Moisture Variation TRB 0 Annual Meeting

13 .kpa,.kpa.kpa,.kpa.kpa,.kpa moisture content 0.0. Resilient Modulus (MPa) Moisture Content (%) 0 Figure Seasonal Variation of Soil Resilient Modulus at Different Depths INFLUENCE OF SOIL RESILIENT MODULUS ON FLEXIBLE PAVEMENT PERFORMANCE As shown above, the stiffness of the subgrade varies seasonally. This support to pavement structure would change the stress state in the pavement structure and influence pavement performance. Two typical pavement structures, interstate highway I-0 at Knoxville (I-0 Knox.) and state route (SR-) were selected to study the influence of soil resilient modulus variation on pavement performance. The details of pavement sections and material properties are listed in Table..0 0 Time since //00 (month) Table Pavement Structures and Material Properties Layers I-0 SR- Elastic Modulus (MPa) Poison's Ratio Asphalt Surface Course.. 0. Asphalt Base Course.. 0. Granular Base Subgrade - - Varied 0..0 TRB 0 Annual Meeting

14 0 0 A multiple elastic-layered software, Wealsea.0 was adopted to evaluate the tensile strain in the upper asphalt layer, the compressive strain on the top of subgrade, and the fatigue life. The default values of elastic moduli and Poison s ratio were used, and it was assumed that the subgrade soils under the two pavement sections had the same properties as the Knox Sta. 00 in Table. The deviator stress and confining pressure of subgrade change continuously horizontally and vertically, which contributes to varied resilient modulus at different locations. However, the analysis was simplified as taking only a series of representative resilient moduli under a. kpa deviator stress and a.kpa confining pressure, as shown in Figure. A transfer function developed at the University of Illinois using Mn/ROAD fatigue crack data was used in Wealsea.0 to predict fatigue life of asphalt pavement, as shown in Equ.. =. 0 ( ). () where: = number of repeated loads under current structural conditions before a fatigue crack will form; = maximum horizontal tensile strain at bottom of first layer caused by one pass of current wheel configuration, expressed in microstrain. Pavement responses from this model are shown in Figures -, and the fatigue lives corresponding to varied resilient moduli of subgrade through one year are shown in Figure0. It can be seen that as the subgrade resilient modulus decreases, the longitudinal tensile at the bottom of the first asphalt layer and the compressive strain on the top of subgrade increases, and the fatigue life decreases by approximately % compared to that with no water content variation. For comparison, a similar analysis was conducted for the thinner pavement section, SR-. For this thinner pavement section, a higher tensile strain was experienced at the bottom of the upper asphalt layer and higher compressive strain at the surface of the subgrade, which results in a lower fatigue life approximately % less that that predicted with no water content change. However, there is no evidence to show that the pavement responses of a thick pavement were less sensitive than the responses on a thin pavement to the variation of resilient modulus. Therefore, resilient modulus variation due to the seasonal moisture change in subgrade should be fully considered in both low and high highways. TRB 0 Annual Meeting

15 Longitudinal tensile strain at the bottom of asphalt layer (0 - ) Longitudinal tensile strain at the bottom of asphalt layer Resilient modulus 0 Time since //00 (month) a) I-0 Knox Resilient modulus (kpa) Longitudinal tensile strain at the bottom of asphalt layer (0 - ) b) SR- Washington Figure Longitudinal Tensile Strain at the Bottom of the First Asphalt Layer and Resilient Moduli through One Year Longitudinal tensile strain at the bottom of asphalt layer Resilient modulus Time since //00 (month) Resilient modulus (kpa) TRB 0 Annual Meeting

16 Longitudinal tensile strain at the bottom of asphalt layer (0 - ) Compressive strain at subgrade surface Resilient modulus Resilient modulus (kpa) Time since //00 (month) a) I-0 Knox. Longitudinal tensile strain at the bottom of asphalt layer (0 - ) Compressive strain at subgrade surface Resilient modulus Resilient modulus (kpa) Time since //00 (month) b) SR- Washington Figure Compressive Strain on the Top of Subgrade and Resilient moduli through One Year TRB 0 Annual Meeting

17 0 Fatigue Resilient modulus 0.0 Number of loads allowed (million) Resilient modulus (kpa) Time since //00 (month) a) I-0 Knox. Fatigue Resilient modulus Number of loads allowed (million) Time since //00 (month) b) SR- Washington Figure 0 Predicted Fatigue Lives with Varied Soil Resilient moduli through One Year Rutting development of the two sections was evaluated in the version. MEPDG software with an input level on subgrade material property and with input level with all other input factors on traffic, climate and material properties. The same traffic was applied on both two sections, with an initial 0 AADTT (average annual daily truck traffic). Rutting development of both pavement sections with a representative resilient modulus (MPa) of soil is shown in Figure, as well as rutting development under a series of resilient moduli (under a.kpa confining Resilient modulus (kpa) TRB 0 Annual Meeting

18 pressure and a.kpa deviator stress, as shown in Figure ) due to seasonal moisture variation in subgrade. AC under seasonal Mr SG under seasonal Mr Rutting with seasonal Mr (cm) Total under seasonal Mr SG under rep. Mr AC under rep. Mr Total under rep. Mr Year since built a) I-0 Knox. AC under seasonal Mr SG under seasonal Mr Rutting with seasonal Mr (cm) Total under seasonal Mr SG under rep. Mr AC under rep. Mr Total under rep. Mr Year since built b) SR- Washington Figure Rutting Development Comparison under Representative Mr and Seasonal Mr of Subgrade It is indicated that a relatively higher rutting depth developed if seasonal variation of soil resilient modulus is considered than only a representative resilient modulus is taken into account. It can also be seen that the decrease of subgrade stiffness makes the subgrade rutting deeper TRB 0 Annual Meeting

19 0 0 0 while the rutting of asphalt layers remain the same. And comparing the two pavement sections, it can be found that the increase latitudes of rutting depth due to the decrease of subgrade stiffness of the two pavement sections seem almost the same. Therefore, thick pavement layers could not effectively prevent the total rutting increase as long as the subgrade support is jeopardized by moisture increase. Since the pavement responses due to the variation of the subgrade resilient modulus vary largely through a year, it is recommended that the representative resilient modulus in input Level and the representative coefficients of the generalized model in input Level should be substituted by the seasonal resilient modulus and the seasonal coefficients of the generalized model, respectively, when those information are available. CONCLUSION AND RECOMMENDATIONS Based on the triaxial cyclic test results of Tennessee soils, coefficients of the generalized model and the universal model were regressed in multiple linear regressions and compared. The variation of soil resilient modulus due to seasonal moisture change and its effect on pavement performance were studied, and the following conclusions can be drawn: Relationships between coefficients of the generalized model and physical properties for clayey soils were developed and validated, which can be used as a time-saving and economical way to estimate the resilient modulus of clayey soils in TN. A negative correlation was found between soil resilient modulus and moisture content. Considerations of the seasonal variation of subgrade resilient modulus due to the moisture change will decrease estimations of pavement fatigue life and increase the rutting depth in subgrade. The fatigue life of both low volume and heavy volume pavements are greatly influenced by the subgrade resilient modulus reductions due to moisture change. The seasonal resilient modulus and the seasonal coefficients of the generalized model are recommended as level and level inputs for MEPDG software. ACKNOWLEDGEMENT This study was funded by the Tennessee Department of Transportation (TDOT). The authors would like to thank TDOT engineers for their help with data acquisition. The contents of this paper reflect the views of the authors, who are solely responsible for the facts and the accuracy of the data presented herein, and TRB 0 Annual Meeting

20 do not necessarily reflect the official views or policies of the TDOT, nor do the contents constitute a standard, specification, or regulation. TRB 0 Annual Meeting

21 0 0 0 References. American Association of State Highway and Transportation Officials. AASHTO Guide for Design of Pavement structures... American Association of State Highway and Transportation Officials, AASHTO T0. Determining the Resilient Modulus of Soils and Aggregate Materials.. NCHRP project -A, Appendix A, Procedure for Resilient Modulus of Unstabilized Aggregate Base and Subgrade Materials. NCHRP Transportation Research Board, 00.. Hopkins, T.C., Beckham, T.L., and Sun C. Resilient Modulus of Kentucky Soils. University of Kentucky, 00.. Ping, W.V. and Sheng, B. Developing Correlation Relationship between Modulus of Subgrade Reaction and Resilient Modulus for Florida Subgrade Soils. Transportation Research Record: Journal of the Transportation Research Board, No., 0, pp Ahn, J., Cote, B.M., Robinson, B., Gabr, M.A. and Borden, R.H. Inverse Analysis of Plate Load Tests to Assess Subgrade Resilient Modulus. Transportation Research Record: Journal of the Transportation Research Board, No. 0, 00, pp Mohammad, L.N., Gaspard, K., Herath, A., and Nazzal. M.D. Comparative Evaluation of Subgrade Resilient Modulus from Non-destructive, In-situ, and Laboratory Methods. Louisiana Transportation Research Center, 00.. Nazzal. M.D. and Mohammad, L.N. Estimation of Resilient Modulus of Subgrade Soils Using Falling Weight Deflectometer. Transportation Research Record: Journal of the Transportation Research Board, No. 0, 00, pp Dawson, T.A., Baladi, G.Y., Sessions, C.P., and Haider, S.W. Backcalculated and Laboratory-Measured Resilient Modulus Values. Transportation Research Record: Journal of the Transportation Research Board, No. 0, 00, pp Seed, H.B., Mitry, F.G., Monismith, C.L. and Chan, C. K. Prediction of pavementdeflection from laboratory repeated load tests. NCHRP Transportation Research Board Report,.. Moossazadeh, J. and Witczak, M.W. Prediction of Subgrade Moduli for Soil that Exhibits Nonlinear Behavior. Transportation Research Record: Journal of the Transportation Research Board, No. 0,, pp. -.. Uzan, J., Witczak, M.W., Scullion, T. & Lytton, R.L. Development and validation of Realistic Pavement Response Models. th International Conference on Asphalt Pavement, Nottingham, UK, Vol.,, pp. -0. TRB 0 Annual Meeting

22 Von Quintus, H. and Killingsworth, B. Analyses Relating to Pavement Material Characterizations and Their Effects on Pavement Performance. A Report to Federal Highway Administration, U.S. Department of Transportation,.. Hossain, M.S. Characterization of Subgrade Resilient Modulus for Virginia Soils and Its Correlation with the Results of Other Soil Tests. Virginia Transportation Research Council, 00.. Mohammad L. N., Huang, B., Puppala, A J., and Allen, A. Regression Model for Resilient Modulus of Subgrade Soils. Transportation Research Record: Journal of the Transportation Research Board, No.,, pp. -.. SHRP Protocol P-. Resilient Modulus of Granular Base\Subbase Materials and Subgrade Soils. Strategic Highway Research Program, ShRP-LAPP Laboratory Material Testing Guide,.. Drumm, E.C., Reeves, J.S., and Madgett, M.R. Design Handbook of Index Properties and Resilient Response. Final report to Tennessee Department of Transportation, University of Tennessee, Knoxville,.. Drumm, E.C., Li, Z., Reeves, J.S., and Madgett, M.R. Alternative Test Method for Resilient Modulus of Fine-Grained Subgrades. Geotechnical Testing Journal, Vol., No.,, pp. -.. Drumm, E.C., Reeves, J.S., Madgett M.R., and Trolinger W.D. Subgrade Resilient Modulus Correction for Saturation Effects. Journal of Geotechnical and Geoenvironmental Engineering, Vol., No.,, pp George, K.P. Prediction of Resilient Modulus from Soil Index Properties, Final Report Mississippi Department of Transportation Research Division, 00.. Figueroa, J.L. Monitoring and Analysis of Data Obtained from Moisture Temperature Recording Stations. Final Report to the Ohio Department of Transportation, Case Western Reserve University, 00.. Shalaby, A. Sensitivity of Sbugrade Resilient Modulus to Moisture Variation. Annual Conference of the Transportation Association of Canada, 00.. Zuo, G. Impacts of Environmental Factors on Flexible Pavements. Doctoral Dissertation of University of Tennessee, Knoxville, 00. TRB 0 Annual Meeting