U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-RD-03-049
Date: November 2005

Improving Pavements With Long-Term Pavement Performance: Products for Today and Tomorrow

Paper 4. Effect of Seasonal Moisture Variation on Subgrade Resilient Modulus

Hassan M. Salem 1

ABSTRACT

It is well known that environmental changes have severe effects on pavement performance. While an asphalt layer may be more sensitive to temperature, a soil or untreated pavement layer might be more affected by the change in moisture. This research aims at quantifying the effect of subgrade moisture variation, caused by environmental changes, on subgrade’s resilient modulus and including its effects in the design process for new and rehabilitated pavements. To achieve this objective, data representing different soil types in non-freeze zones at various Long-Term Pavement Performance Seasonal Monitoring Program (LTPP-SMP) sites were downloaded from the DataPave 3.0 software. The downloaded data were analyzed to establish the effect of subgrade moisture variation on subgrade’s resilient strength represented by the backcalculated elastic modulus. The analysis indicated that moisture in the subgrade layer is related to the precipitation intensity. The study also revealed that a Seasonal Adjustment Factor (SAF) could be used to shift the subgrade modulus from a normal season to another. The SAF is considered a key input in the mechanistic-based pavement design system. It allows the inclusion of the seasonal effects on the layer moduli for different seasons. In this paper, a method is presented for calculating the SAF for the subgrade soils. Using the collected data, regression analysis was performed and correlation equations were developed. These equations relate the backcalculated subgrade modulus to the subgrade moisture content and to other soil properties. The SAF relates the change in the moisture content to the change in the modulus value.

1Graduate Exchange Student, Department of Civil Engineering, University of Idaho, Moscow, ID 83844-1022 Phone: 208-885-6818, Fax: 208-885-6618, hassan@uidaho.edu

INTRODUCTION AND BACKGROUND

It is well known that high subgrade moisture content, with resulting decrease in subgrade strength and stiffness, is detrimental to roadway pavement performance. Establishing relationships between the highway pavement response and subgrade moisture variation is necessary for efficient pavement design. The backcalculated modulus, or the resilient modulus, of subgrade soil is the key parameter that is considered to represent pavement response. A brief summery of previous work discussing the effect of seasonal variations on soil resilient modulus is presented in the following subsections.

STUDY OBJECTIVE

The primary objective of this study is to examine the impacts of seasonal moisture variation on the subgrade resilient modulus.

MOISTURE EFFECTS ON SOIL RESILIENT MODULUS

Many researchers have investigated the influence of water content on the resilient modulus of fine grained soils. Seed, Chan, and Lee (1962) studied the influence of“natural” water content on the resilient modulus of undisturbed samples of the silty clay (CL) American Association of State Highway and Transportation Officials (AASHTO) Road Test subgrades soil. The positions of the test points showed that for this soil a decrease in water content of only 3 percent below the T99 optimum doubled the decrease of the resilient modulus (from about 34 megapascal (MPa) (4931.28 poundforce per square inch (psi)) to about 69 MPa (10007.6 psi)). Tests conducted on CL subgrade soil at the San Diego County Experimental Base Project by Jones and Witczak (1977) showed that as its compaction water content was increased from about 11 percent to about 20 percent, the resilient modulus varied from almost 275 MPa (3988.5 psi) to a low of about 52 MPa (7541.96 psi).

Carmichael and Stuart (1985) presented correlations relating resilient modulus to fine grained soil composition parameters. Using a database representing more than 250 soils (fine and coarse) and 3300 modulus test data points, they developed the following relationship:

MR = 37.431 - 0.4566 PI - 0.6179 w - 0.1424F
+ 0.1791CS - 0.3248 sd + 36.422CH + 17.097MH
(1)

Where MR is resilient modulus in kilopounds per square inch (ksi), PI is plasticity index in percent, w is water content in percent, F is percent passing sieve number 200, CS is the confining stress in psi, and sd is deviator stress in psi. The (CH) term is a material factor that is equal to one for soils classified as CH and is equal to zero for soils classified as inorganic silts (ML), elastic silts (MH), or clay (CL). MH is a material factor equal to 1 for soils classified as MH and equal to zero for soils classified as ML, CL, or CH.

The moisture sensitivity of coarse-grained materials depends on the amount and nature of their fine fraction. Clean gravels and sands classified as well graded gravel (GW), poorly graded gravel (GP), well-graded sand (SW), and poorly graded sand (SP) are not likely to exhibit moisture sensitivity because they lack a sufficient number of the small pores necessary to create significant suction-induced effective stresses, even at low water contents (Hicks and Monismith, 1971). Studies of coarse materials containing larger amounts of fines have shown that increasing degrees of saturation above about 80 to 85 percent can have a pronounced effect on resilient modulus. Rada and Witczak (1981) concluded that changes in water content of compacted aggregates and coarse soils could cause modulus decreases of up to 207 MPa (30,022.8 psi).

Several researchers have developed regression relationships between the resilient modulus of granular materials and water content. The general regression relationship for granular materials of Carmichael and Stewart (1985), stated previously as equation (1), contains a water content term that results in an 11.6 MPa (1682.45 psi) decrease in resilient modulus for each 1 percent increase in water content. Lary and Mahoney (1984) found regression relationships for resilient moduli of specific northwest aggregate base materials and predominantly coarse subgrade soils. The regression equations for the materials showed that if the initial modulus is on the order of 140 MPa (20,305.3 psi), a 1 percent increase in moisture content typically results in a resilient modulus decrease from about 4 to 11 MPa (580.15 psi to 1595.42 psi). A reasonable estimate for the influence of water content on reference resilient modulus of coarse soils would be about a 3.4 MPa (493.13 psi) decrease for each 1 percent moisture content increase for uniform or well graded coarse materials containing little or no nonplastic fines (GW, GP, SW, SP). That value would increase to about 3.8 MPa (551.14 psi) per 1 percent moisture content increase for sands and gravels containing substantial amounts of plastic fines silty gravels (GM), clayey gravels (GC), course grained sands (SM), clayey sands (SC).

TEMPERATURE EFFECTS ON SOIL RESILIENT MODULUS

Temperature has significant effects on soil resilient modulus. The penetration of freezing temperatures into moist pavement subgrade soils can cause more severe effects than the effects of any of the water content changes likely to occur as a result of seasonal variations in precipitation. Freezing of soil moisture can transform a soft subgrade into a material that, at the stress levels existing in pavements, is essentially rigid. Thawing of the same material can produce a softening effect such that the material has a resilient modulus that is only a fraction of its pre-freezing value for some time after thawing, (Hardcastle, 1992).

The effects of an annual cycle of freezing and thawing on the deflections of pavements having coarse and fine grained subgrades in Illinois and Minnesota were studied by Scrivner et al., (1969). The study showed that freezing results in sharp reductions in surface deflections while thawing produces immediate deflection increase for all of the pavements. The authors also found that pavement deflection changes could occur due to freezing of the structural layers alone, while the largest thaw induced deflection increases take place when there is deep frost penetration into the fine grained subgrade soils. Increases in deflection due to deep frost penetration and thawing of the coarse grained subgrade soil are smaller than those for fine grained soils. Because the sites included in this study are from the non-freeze zones, the effect of freeze/thaw is not considered in our study.

SEASONAL VARIATION AND SEASONAL ADJUSTMENT FACTORS

In a study on the LTPP data from site 48SA of a non-freeze zone, Ali and Parker, (1996) found that the backcalculated resilient moduli of both subgrade and asphalt concrete (AC) surface could be correlated to the month of the year in a sinusoidal function with reasonable accuracy.

Several research projects were conducted at the University of Idaho to study the effect of seasonal variations on pavement performance (Hardcastle, 1992; Al-Kandari, 1994; Bayomy et al., 1996, 1997; and Abo-Hashema et al., 2002). These projects recommended the use of the Falling Weight Deflectometer (FWD) to evaluate the pavement structure conditions, and provided initial values of subgrade soil resilient modulus for various climatic regions and soil types across the State of Idaho. Based on the study by Hardcastle (1992), Al-Kandari (1994) incorporated the SAF for subgrade soils in an environmental database derived from various climatic zones in Idaho. Abo-Hashema et al. (2002) suggested a general procedure for calculating the seasonal adjustment factor for subgrade soils, but they did not consider the effect of soil type (fine, coarse, plastic, and/or nonplastic). The previous projects demonstrated the need to establish a realistic SAF to be applicable to different soil types and environmental conditions. The SAF is a rationale adopted that incorporates the environmental effects in the design system to adjust subgrade modulus from one season to another.

The Washington State Department of Transportation (WSDOT) uses a mechanistic-empirical (M-E) system developed at the University of Washington and implemented in the computer program EVERPAVE 5.0 (1999). This program uses SAFs as key inputs by users and does not compute the SAF. The Minnesota Department of Transportation (MNDOT) uses an M-E flexible pavement thickness design that is implemented in the computer program ROADENT 4.0 by Timm, Birgisson, and Newcomb (2001). Because ROADENT does not have the SAF to adjust the resilient modulus from one season to another, the user must calculate and enter the resilient modulus values for each season.

APPROACH

The elastic modulus of a pavement layer represents the main parameter that reflects the materials’ structural adequacy; the study thus is focused on the determination of the seasonal impacts on the layer moduli, but is limited to the subgrade layer. The elastic modulus of a pavement layer changes in response to environmental changes such as variations in temperature for asphalt layers, and moisture variation for untreated layers such as base and subgrade. To account for the changes in subgrade elastic modulus, a multiplier can be used for adjusting subgrade resilient modulus from one season (reference) to another. This multiplier, Seasonal Adjustment Factor (SAF), is based on soil type and environmental conditions. This study aims at developing a concept for calculating the SAF for subgrade soil layer as well as a model relating the subgrade backcalculated modulus to its moisture content.

MOISTURE AND MODULUS DATA

To address the issue of environmental impacts on pavement at a national level, the Federal Highway Administration (FHWA) launched the SMP as a major component of the LTPP program (Rada et al., 1994). The data used in this study were downloaded from the DataPave 3.0 (2001), a software package that contains most data from the LTPP experiments. Seven different LTPP sites were considered in this study to represent different soil types (sites 35-1112, 28-1016, 48-1122, 48-1077, 13-1005, 48-4143 and 24-1634). The subgrade soils of the sites are sand, coarse silty sand, coarse clayey sand, fine sandy silt, fine sandy clay, clay, and silt, respectively. All selected sites are from the non-freezing zones except the last site, 24-1034, which is considered as a wet-freeze zone, where the minimum average monthly air temperature is 1.7 °C (35 °F), as shown in table 1. Site 24-1034 thus also could be considered to be from the non-freezing zones.

Detailed explanations for the selected sites are shown in table 1, which shows the site location, minimum average monthly air temperature, subgrade soil type, soil classification, soil sieve analysis, Atterberge limits, dry density, and optimum moisture content for each soil type in the selected sites. Downloaded data for each site included the backcalculated elastic moduli for subgrade soil and AC surfaces, the AC layer temperature, and both volumetric and gravimetric moisture content of subgrade soil at different time intervals. For the purpose of this study, only the analyses of the change in subgrade soil moduli with seasonal moisture variation were considered.

Table 1. LTPP site locations and subgrade soil characterization

Sites ID 35-1112 28-1016 48-1122 48-1077 13-1005 48-4143 24-1634
State
NM MS TX TX GA TX MD
Surface Type
Flexible Flexible Flexible Flexible Flexible Rigid Flexible
Minimum Monthly Air Temperature, Co
5.80 5.00 9.70 3.60 8.70 9.70 1.70
Soil Type
Coarse, poorly graded sand Coarse, silty sand Coarse, clayey sand Fine, sandy silt Fine, clayey sand Fine, clay Fine, silt
Soil Symbol
S SM-C SC-C SM-F SC-F C M
Unified Soil Classification
SP SM SC SM SC CL ML
AASHTO Soil Classification
A-3 A-2-4 A-2-6 A-4 A-6 A-7-6 A-4
% Passing # 4
100.00 92.00 99.00 94.00 - - 99.00
% Passing # 10
99.00 91.00 97.00 93.00 - - 98.00
% Passing # 40
94.00 85.00 75.00 87.00 - - 98.00
% Passing # 200
2.70 25.70 6.50 51.80 38.40 90.00 97.90
D60, mm
0.18 0.23 0.30 0.10 - - -
Liquid Limit, %
- 18.00 26.00 - 27.00 41.00 -
Plasticity Index, %
NP 3.00 12.00 NP 12.00 23.00 NP
Dry Density, gm/cm3
1.698 1.906 1.858 1.906 2.05 1.730 1.746
Optimum Moisture, %
12.00 13.00 8.00 10.00 10.00 15.00 12.00

DATA ANALYSIS

To study the effect of seasonal moisture variation on pavement performance, seven different LTPP sites were considered; they represent different soil types, as explained above. The primary data collected at these sites are the gravimetric moisture content (considered to be the main factor affecting subgrade soil strength) and the backcalculated elastic moduli for different subgrade soils at different seasons. Results are discussed and analyzed in the following subsections.

SEASONAL VARIATION OF MOISTURE AND MODULUS OF SUBGRADE SOIL

Moisture content of soils near the ground surface depends on a variety of climatic and physical factors, including soil type, temperature, precipitation, vegetation, and others. It is widely known that pavement subgrade soils not only experience temporary (seasonal) changes in moisture content but also undergo changes in their long-term average annual moisture content. This section discusses the seasonal variation in both moisture and modulus of subgrade soil.

Moisture and Modulus Variation with Time

Figures 1 through 4 show the relationship between both gravimetric moisture content and subgrade backcalculated modulus with time for different sites from the non-freeze zones. The figures indicate that both moisture content and backcalculated elastic moduli have almost a sinusoidal function with time. The figures also indicate that the backcalculated elastic modulus could be related to moisture content with an opposite fitness function, as the elastic modulus increases when the moisture decreases and vice versa. This behavior could be observed at all sites except site 28-1016, shown in figure 4, and also site 35-1112. The main reason for the different behavior of sites 28-1016 and 35-1112 regarding the direct proportional relationship between their moduli and moisture content is that the subgrade soils at both sites are noncohesive (silty sand and sand, respectively), and the field moisture content at both sites is below the optimum moisture content measured at the lab from a compaction test, as shown in table 1. In this case, increasing soil moisture results in increasing its modulus until the optimum moisture content is reached, then the modulus will reduce with further increase in moisture content. Figures 2 and 3 also indicate that both the maximum modulus values and minimum moisture values are measured through the summer season (July and August), while the minimum modulus values and maximum moisture values are measured through the winter (January and February).

Figure 1. Moisture content and elastic modulus versus season for clayey soil, site 48-1122

Figure 1. Moisture content and elastic modulus versus season for clayey soil, site 48-1122.

Figure 2. Moisture content and elastic modulus versus season for silty soil, site 24-1634

Figure 2. Moisture content and elastic modulus versus season for silty soil, site 24-1634

Figure 3. Moisture content and elastic modulus versus season for clayey soil, site 13-1005

Figure 3. Moisture content and elastic modulus versus season for clayey soil, site 13-1005

Figure 4. Moisture content and elastic modulus versus season for silty sand, site 28-1016

Figure 4. Moisture content and elastic modulus versus season for silty sand, site 28-1016

Relating Moisture Content to Average Precipitation

The average monthly precipitation and the measured moisture content at sites 28-1016, 24-1634, 13-1005 and 48-4143 are shown in figures 5 through 8, respectively. The figures indicate that the moisture increases when the average precipitation increases and decreases with decreasing precipitation. Therefore, the moisture content could be highly related to the average precipitation. Figure 9, for site 48-4143, shows that moisture content could be related to the average precipitation with a liner function having R2 value of 0.48. Similar relationships could be developed for the other sites. It should be noted that the average precipitation is not the only factor that affects subgrade moisture content. There are many other climatic and physical factors such as: soil type, temperature, precipitation, vegetation, and others. Witczak et al., (2000), are extensively studying the use of precipitation in moisture content predictions through the intergraded climatic model (EICM 2.6), for the development of the AASHTO 2002 Guide for the Design Of New And Rehabilitated Pavement Structures.

Figure 5. Moisture content and rainfall versus season for silty sand soil, site 28-1016

Figure 5. Moisture content and rainfall versus season for silty sand soil, site 28-1016

Figure 6. Moisture content and rainfall versus season for silty soil, site 24-1634

Figure 6. Moisture content and rainfall versus season for silty soil, site 24-1634

Figure 7. Moisture content and rainfall versus season for clayey soil, site 13-1005

Figure 7. Moisture content and rainfall versus season for clayey soil, site 13-1005

Figure 8. Moisture content and rainfall versus season for clayey soil, site 48-4143

Figure 8. Moisture content and rainfall versus season for clayey soil, site 48-4143

Figure 9. Moisture content versus rainfall for clayey soil,site 48-4143

Figure 9. Moisture content versus rainfall for clayey soil, site 48-4143


CORRELATING THE BACKCALCULATED ELASTIC MODULUS TO SUBGRADE SOIL MOISTURE AND OTHER SOIL PROPERTIES

Model Development for Plastic Soils

For the purpose of this study, multiple regression analysis is applied using the SAS program to relate the backcalculated elastic modulus to subgrade moisture content and such other soil properties as: Atterberge limits, percentage passing sieve # 200, D60. Data from three different LTPP sites (48-4143,13-1005 and 48-1122) are used in this analysis. The subgrade soils at these three sites are: clay, fine sandy clay, and coarse sandy clay, respectively. These sites were chosen because their soils have values for plasticity index (PI; plastic soil), while most of the other sites have nonplastic soils. The SAS program output of the regression analysis is shown in table 2, in which:

E = Backcalculated elastic modulus, MPa
E1 = Log (E)
X1 = Log (moisture content, percent)
X2 = 1/(moisture content, percent)
F = Percentage passing sieve # 200, %
PI = Plasticity index, percent

Table 2. The regression procedure using R-square selection method for three sites

SAS program output, dependent variable: E1
Number in Model R-Square C(p) BIC Root MSE Variables in Model
1 0.9767 176.6577 -844.4975 0.07112 PI
1 0.7840 2926.776 -491.2373 0.21651 F
1 0.6981 4151.479 -491.2373 0.25593 x1
1 0.5068 6880.904 -359.4122 0.32712 x2
2 0.9795 137.8704 -863.9304 0.06683 F PI
2 0.9768 177.2805 -844.1832 0.07120 x2 PI
2 0.9767 178.3979 -843.6575 0.07132 x1 PI
2 0.9022 1241.671 -617.3199 0.14614 x1 x2
3 0.9884 13.2697 -950.0654 0.05045 x1 F PI
3 0.9871 32.4110 -933.3682 0.05329 x1 F PI
3 0.9781 159.8558 -852.5841 0.06930 x1 x2 PI
3 0.9161 1045.302 -641.4009 0.13579 x1 x2 F
4 0.9891 5.0000 -957.7523 0.04902 x1 x2 F PI

Table 2 indicates that the logarithm of the backcalculated modulus (E1) could be related only to the logarithm of moisture content (X1) with a function having a coefficient of determination (R2) value of 0.698. However, when adding other soil properties, like PI and F to the model, a better model having R-square value of 0.989 could be achieved. The developed model is shown in equation (2).

Log(E) = C0 + C1 * Log(moisture) + C2 * (1/moisture) + C3 * F + C4 * PI             (2)

Where: E, F, and PI are as described before in table 2 and Co, C1, C2, C3 and C4 are model constants.

The estimated values of the model constants and the analysis of variance (ANOVA) for that model are shown in table 3. Table 3 also shows that the sum of squared errors (SSE) for that model is 0.372, the coefficient of variation is 0.859, and the adjusted R2 is 0.989. The last two columns in the bottom of table 3 contain results of the statistical test that evaluates the significance of each regression coefficient. The test results indicate that at a significance level of 95 percent, all estimated model parameters are significant (p-value is less than 0.05). The final model with its estimated constants is shown in equation (3).

Log(E) = 8.82 - 0.673 * Log(moisture) - 2.44 * (1/moisture) + 0.084 * F - 0.11 * PI             (3)

Table 3. Analysis of variance table and estimated model parameters

Analysis of Variance
Source DF Sum of Squares Mean Squares F-Value
Model 4 33.90986 8.47747 3528.46
Error 155 0.37240 0.00240  
Corrected Total 159 34.28227    
 
Root MSE 0.04902 R-Square
Dependent Mean 5.70630 Adj R-Sq
Coeff Var 0.85899  
Estimated Model Parameters
Variable DF Parameter Estimate Standard Error t Value
Intercept 1 8.81933 0.31794 27.74
X1 1 -0.67276 0.12405 -5.42
X2 1 -2.43912 0.76112 -3.20
F 1 0.00838 0.00066926 12.52
PI 1 -0.11065 0.00343 -32.28

Figures 10 through 12 show the model application on the data collected from sites 48-4143, 13-1005, and 48-1122, respectively. The figures indicate that the model fits the data very well and that the modulus decreases with increasing soil moisture even if the field moisture content is less than the optimum moisture content, as shown in figures 11 and 12 for sites 13-1005 and 48-1122, respectively. The reason is that the subgrade soils at both sites are cohesive soils (sandy clay). Therefore, when the moisture content decreases, the soil becomes harder and its modulus increases, while the modulus deceases when moisture increases. It should be noted that this model could be applied only for plastic (clayey) soils, as there is a term in the model for PI. For nonplastic soils, this model will be modified to account for soil properties other than PI (explained later in this paper).

Figure 10. Backcalculated modulus versus moisture content for clay soil, site 48-4143

145 MPa = 1 psi

Figure 10. Backcalculated modulus versus moisture content for
clay soil, site 48-4143

Figure 11. Backcalculated modulus versus moisture content for fine sandy clay soil, site 13-1005

145 MPa = 1 psi

Figure 11. Backcalculated modulus versus moisture content for fine
sandy clay soil, site 13-1005

Figure 12. Backcalculated modulus versus moisture content for coarse sandy clay soil, site 48-1122

145 MPa = 1 psi

Figure 12. Backcalculated modulus versus moisture content for coarse
sandy clay soil, site 48-1122


Model Development for Nonplastic Soils

As was described previously, the model shown in equation (3) could not be applied directly for non-plastic soils (sandy and/or silty soils), as there is a term in the model for PI. To generalize the previous model, it is simplified to the generic form shown in equation (4).

Log(E) = C * C1 * Log(moisture) + C2 * (1/moisture)             (4)
 where:

C* = Co + C3 * F + C4 * PI
= 8.82 + 0.0084 * F - 0.11* PI (for plastic soils)
C1 = - 0.673 (for plastic soils)
C2 = - 2.44 (for plastic soils)
E, F, and PI are as described before for table 2.

Several trials were made, using Solver in the Microsoft® Excel® program, to fit the moisture-modulus data of nonplastic soils to the model shown in equation (4). Data from sites 24-1634, 48-1077, and 35-1112 were used in these trials. The subgrade soils at these sites are silt, fine silty sand, and sand soils, respectively. The model constants C*, C1, and C2 were calculated for each site using the Solver program, and are shown in table 4. The model was found to fit the data, as shown in figures 13 through 15. The estimated R2 values for the previous figures are 0.72, 0.36, and 0.32 respectively. The higher R2 value was achieved with the silty soil (site 24-1634), while the lower R2 value was achieved with sand (site 35-1112).

Table 4. Estimated model constants for nonplastic soils

Site Type C* C1 C2
24-1634 Silt 30.16 -89.670 -6.910
48-1077 Fine sandy silt 12.40 -24.900 -2.080
35-1112 Sand 5.74 0.043 0.133

Regression analysis was applied using the SAS program to relate the model constants (C*, C1, and C2), shown in table 4, to other material properties of nonplastic soils like dry density, D60 and percentage passing sieve #200. It was found, with good accuracy (R2 = 1), that the model constants (C*, C1, and C2) could be related to the percentage passing sieve #200 (F) and D60, with a linear function. Therefore, the following equations were developed through regression analysis to estimate the model parameters for nonplastic soils.

C* = 90.63 -0.618 * F - 462.35 * D60  
C1 = -304.96 + 2.199 * F + 1661.50 * D60                 (5)
C2 = -20.14 + 0.135 * F + 110.58 * D60  

Figure 13. Backcalculated modulus versus moisture content for silty soil, site 28-1634

145 MPa = 1 psi

Figure 13. Backcalculated modulus versus moisture content for silty soil, site 28-1634


Figure 14. Backcalculated modulus versus moisture content for fine sandy silt, site 48-1077

145 MPa = 1 psi

Figure 14. Backcalculated modulus versus moisture content for fine sandy silt, site 48-1077


Figure 15. Backcalculated modulus versus moisture content for sandy soil, site 35-1112

145 MPa = 1 psi

Figure 15. Backcalculated modulus versus moisture content for sandy soil, site 35-1112

In conclusion, as figures 10 through 14 show the backcalculated elastic modulus decreases with increasing soil moisture content for non-freeze sites. Only figure 15 indicated that the elastic modulus increases with increasing moisture content. The reason for the direct relationship shown in figure 15 is the same as described for figure 4 (e.g., noncohesive soil). The soil in figure 4 is sand, and its field moisture is much less than the optimum moisture content for this soil, which is 12.0 percent as shown in table 1. Therefore, increasing moisture content in such cohesionless soils by increases their modulus until the optimum moisture is reached, then the modulus reduces. For cohesive soils (clayey soils), increases in subgrade moisture will be accompanied with a reduction in their moduli, whatever the soil moisture condition is, before or after the optimum, as discussed earlier.

It should be noted that the general model shown in equation (4), could be applied to any soil type. The model constants (C*, C1, and C2) could be estimated either from equation (4) in the case of plastic soils, or equation (5) in the case of nonplastic soils.

ESTIMATING SEASONAL ADJUSTMENT FACTORS

The SAF is considered to account for the changes in subgrade soil modulus due to seasonal change in subgrade moisture content. The SAF reflects both the subgrade soil class and the climate in which the pavement is located. For the purpose of developing a model to measure the subgrade modulus shift factor, regression analysis was applied to the moisture-modulus data of different soil types. The soil types that were included in this study are: silt (M), clay (C), silty sand (SM), fine sandy clay (F-SC) and coarse sandy clay (C-SC). The moisture-modulus data of the previous soil types were downloaded from sites 24-1634 (M), 48-4143 (C), 48-1077 (SM), 13-1005, and 28-1016 (F-SC).

The result of the regression analysis is shown in figure 16. The X-axis of figure 16 shows the moisture increase, which is the seasonal moisture divided by the minimum moisture content measured at this site through the year (usually summer moisture content). The Y-axis shows the modulus shift factor: the seasonal backcalculated modulus at any site divided by the maximum modulus measured throughout the year (usually the summer modulus or the modulus corresponding to the minimum seasonal moisture). The analysis of figure 16 indicates that the shift factor could be related to moisture increase with the model shown in equation (6).

SF = ko * M - k              (6)
 Where:

SF = Modulus shift factor (E Season/ Eo)
M = Moisture increase (MSeason/ Eo)
ko, k = Model constants, dependent on soil type
MSeason,ESeason = Moisture and modulus at any season, respectively Mo,Eo= Reference moisture and corresponding modulus, usually taken during summer or at the season having the minimum moisture throughout the year.

Figure 16. Estimated modulus shift factor for different soil types

Figure 16. Estimated modulus shift factor for different soil types

The model shown in equation (6) is simple and dimensionless, and was found to fit the data from most sites with reasonable accuracy (R2 ranges from 0.5 to 0.72), as shown in figure 16. The model constant (ko) was found to be 1.0 for all soils, as shown in the figure, while the constant (k) changes with soil type. The model constant (k) increases with increasing soil sensitivity to moisture increase. The figure indicates that constant (k) ranges from 0.32 for coarse sandy clay to 1.1 for clay and 1.32 for silty soils. The figure indicates also that if the moisture content is increased by 20 percent (factor 1.2), the modulus for silty soils reduces to 0.80 of its value. The corresponding modulus reductions for clay and sandy clay soils are 0.83 and 0.96, respectively. Therefore, the pure silty soils are more sensitive to moisture increase than pure clayey soil and all other soil types. Figure 16 also indicates that water sensitivity for sandy soils (sandy silt, fine sandy clay, and coarse sandy clay) is almost the same and is much less than that of pure silty or clayey soils. Equation (6) and/or figure 16 could be used to estimate the seasonal adjustment factor for a certain season by just knowing soil type and the expected moisture increase in that season with respect to the reference season

SUMMARY AND CONCLUSION

The data used in this study were downloaded from the LTPP-SMP database, DataPave 3.0 (2001). Through the analysis of these data it was found that both moisture and modulus follow almost a sinusoidal function with different months of the year. While the moisture increases during a season, the modulus decreases during the same season and vice versa. The minimum seasonal subgrade moisture content (higher modulus values) was observed during summer season while the highest moisture (lowest modulus) was observed during winter. However, in some cases for noncohesive soils, increasing moisture content may result in increasing its modulus if the in situ moisture content is below the optimum moisture content. The data showed also that the moisture content profile through the months of the year could be related to the average monthly rainfall.

In this study, the relationship between subgrade modulus (E) and the gravimetric moisture content was determined for different soil types. A general model relating subgrade modulus to soil moisture and other soil properties was developed and applied for different soil types. Also, a simple model for calculating the modulus shift factor of subgrade soil was developed. The modulus shift factor adjusts subgrade modulus from one reference season (usually summer) to another. This allows determination of subgrade resilient modulus at any season by multiplying the reference value by the SAF for that season. The reference value is the modulus value determined by testing during any selected season (summer). The SAF varies according to subgrade soil type and climatic conditions. The results showed also that pure silty soil is more sensitive to moisture variation than are all other soils.

REFERENCES

Abo-Hashema, M.A., Bayomy, F.M., Smith, R., and Salem H.M. (2002). “Environmental Impacts on Subgrade Resilient Modulus for Idaho Pavements.” The Transportation Research Board, TRB 81st Annual Meeting, January 13-17, 2002, Washington, DC.

Ali, H. and Parker, N. (1996). “Using Time Series to Incorporate Seasonal Variations in Pavement Design.” ”Transportation Research Record,TRR 1539, pp. 33-43.

Al-Kandari, F. (1994). “Mechanistic-Based Overlay Design Procedure for Idaho Flexible Pavements”.S. Thesis, University of Idaho, Moscow, ID.

Bayomy, F., Al-Kandari, F., and Nassar, W. (1997). “Computer-Aided Overlay Design System for Flexible Pavements.”Proceedings of the 8th International Conference on Asphalt Pavements,Seattle, WA.

Bayomy, F., Al-Kandari, F., and Smith, R. (1996). “Mechanistic-Based Overlay Design System for Idaho.”Transportation Research Record, TRR 1543, Transportation Research Board, Washington, DC, pp. 10-19.

Carmichael, R.F. III and Stuart, E. (1985). “Predicting Resilient Modulus: A Study to Determine the Mechanical Properties of Subgrade Soils.”Transportation Research Record TRR 1043, Transportation Research Board, National Research Council, Washington, DC, pp. 145 148.

FHWA (2001). DataPave3.0, CD-ROM. U.S. Department of Transportation, Federal Highway Administration, Washington, DC

Hardcastle, J.H. (1992). Subgrade Resilient Modulus for Idaho Pavements. Final Report of ITD Research Project RP 110-D, Agreement No. 89-47, Department of Civil Engineering, University of Idaho, Moscow, ID, 252 pp.

Hicks, R.G. and Monismith C.L.(1971). “Factors Influencing the Resilient Responses of Granular Materials.”Highway Research Record 345,Highway Research Board, National Research Council, Washington, DC, pp. 15 31.

Jones, M.P. and Witczak, M.W. (1977). “Subgrade Modulus on the San Diego Test Road.”Transportation Research Record, TRR 641, Transportation Research Board, National Research Council, Washington, DC, pp. 1 6.

Lary, J.A. and Mahoney, J.P. (1984). “Seasonal Effects on the Strength of Pavement Structures.”Transportation Research Record, TRR 954, Transportation Research Board, National Research Council, Washington, DC, pp. 88 94.

Rada, G. and Witczak, M.W. (1981). “Comprehensive Evaluation of Laboratory-Resilient Moduli Results for Granular Materials.”Transportation Research Record, TRR 810, Transportation Research Board, National Research Council, Washington, DC, pp. 23 33.

Rada, G., Elkins, G., Henderson, B., Van Sambeek, R., and Lopez, A (1994).“”LTPP Seasonal Monitoring Program: Instrumentation Installation and Data Collection Guidelines.Federal Highway Administration, Report No. FHWA-RD-94-110, Washington, DC.

Scrivner, F.H., Peohl, R., Moore, W.M., and Phillips, M.B. (1969). “Detecting Seasonal Changes in Load-Carrying Capabilities of Flexible Pavements.”NCHRP Report 7Highway Research Board, National Research Council, Washington, DC, 37 pp.

Seed, H.B. Chan, C.K., and Lee, C.E. (1962). “Resilience Characteristics of Subgrade Soils and Their Relation to Fatigue Failures in Asphalt Pavements.”Proceedings, International Conference on the Structural Design of Asphalt Pavements,University of Michigan, Ann Arbor, MI, pp. 611 636.

Sivaneswaran, N., Pierce, L. and Mahoney, J. (1999). “EVERPAVE 5.11: Pavement Overlay Design Program.” Washington State Department of Transportation, Olympia, WA.

Timm, D., Birgisson, B., and Newcomb, D. (2001). “Development of Mechanistic-Empirical Pavement Design in Minnesota.”Transportation Research Board, 80th Annual Meeting, Washington, DC.

Witczak, M.W., Houston, W.N., Zapata, C.E., Richter, C., Larson, G., and Walsh, K. (2000). Improvement of the Integrated Climatic Model for Moisture Content Predictions.Inter- Team Technical Report (Seasonal 4), NCHRP 1-37 A, Arizona State University, Phoenix, AZ.

Previous | Table of Contents | Next

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101