Semisequicentennial Transportation Conference Proceedings
May 1996, Iowa State University, Ames, Iowa

NDT-Evaluation of Seasonal Variation of Subgrade Response in Asphalt Pavements

Mustaque Hossain, Bing Long, and Andrew J. Gisi

M. Hossain and B. Long,
Department of Civil Engineering,
Kansas State University,
Manhattan, Kansas 66506.

A.J. Gisi,
Kansas Department of Transportation,
Bureau of Materials and Research,
2300 Van Buren,
Topeka, Kansas 66611.

Seasonal variations in pavement material properties and behavior due to climatic effects, primarily due to variations in temperature and moisture conditions, have long been known to affect pavement structural performance, or in other words, deflection response. Currently, pavements are tested for deflection response in Kansas by an NDT device, Falling Weight Deflectometer (FWD). Temperature, subgrade moisture content, and FWD-deflection data were collected monthly on four asphalt pavement test sections for a year. The subgrade moduli were backcalculated using the elastic layer theory. It was found that for almost all sites, the variation in subgrade moisture content was not very significant over the seasons. The patterns of subgrade response, in terms of subgrade moduli versus subgrade moisture content, simulated sine-shaped forms signifying a possible temperature effect. Unusually lower values of backcalculated subgrade moduli were obtained for two sections during months when asphalt surface temperatures exceeded 46OC. However, this was found to be attributable to the backcalculation scheme used in this study. The seventh sensor deflection value was found to be independent of temperature and a reasonable indicator of subgrade stiffness.

Several environmental variables affect pavement behavior and performance. These variables include: (i) subgrade moisture content, (ii) temperature, (iii) solar radiation and atmospheric conditions, and (iv) site geological conditions. The first two classes of variables tend to receive primary consideration in many pavement design procedures because temporal variations of these parameters are known to be significant. However, the ability to consider these effects in pavement design has been severely limited primarily due to lack of field data (1). This paper discusses the effects of moisture and temperature on subgrade structural response of asphalt pavements as evaluated by the FWD.


Four 152 m long asphalt pavement sections were selected as test sites in the middle half of the state which somewhat represents the average climatic condition for Kansas. Table 1 lists the locations and characteristics of the sites selected for this study. Two sites (K-18 and K-113) were selected in northeast Kansas and two ( US-160 and US-283) were selected in southwest Kansas. The table also lists the average annual precipitation and average daily maximum temperature for each site. Most of the sites have silty clay soils as subgrade materials.

TABLE 1 Location and Characteristics of the Sites Selected for the Study
Site No. County & Route Asphalt Layer Thickness (mm) Shoulder Type Soil Type (Unified) LL (%) PI (%) % Passing #200 Average Annual Precipitation (mm) Average Daily Max. Temp. (0 C)
1 Clark,
229 unpaved CL-ML 26
533 34.4
2 Clark
305 unpaved CL, CL-
533 34.4
3 Pottawa-tomie,
229 unpaved CL 30
838 33.0
4 Riley,
267 unpaved - - 53
838 33.0

At each site, three time domain reflectometry (TDR) waveguides were installed under the pavement in the subgrade at a spacing of 76 m in 100-mm diameter core holes. An 8-meter coaxial cable was extended from each waveguide to the edge of the shoulder along a saw-cut groove in the paved surface to house the cable end in a secured housing for future moisture measurements. Measurements for soil-moisture content were done for all locations for at least one year with a Tektronix Model 1502 TDR cable tester since October of 1993. Soil-moisture measurements were made twice a month during the thaw-period (February 15 - April 15) for the sites in northeast Kansas. Monthly moisture data collection continued on these sites up to April 1995 yielding 18 months of data.

Deflection data were collected at ten locations at 15-m interval on each test section with a Dynatest 8000 Falling Weight Deflectometer (FWD). The first sensor was located at the center of the loading plate with six others at a uniform radial distance of 0.3 m apart. Three drops of FWD load were used for target loadings of 31, 40, and 67 kN. Tests were done on the outer wheel path of the travel lane.


A linear elastic analysis backcalculation program, MODULUS, was used to backcalculate subgrade moduli from the FWD deflection basins. A number of computer programs, such as BOUSDEF, BKCHEV, MATCH, MODULUS, etc., are available to carry out the backcalculation of the moduli of pavement layers. However, MODULUS was used in this project because this has been approved for use in the SHRP Long Term Pavement Performance (LTPP) study (2).

Effect of Subgrade Moisture Contents on Backcalculated Subgrade Moduli

Figure 1 shows the variation of backcalculated subgrade moduli with the measured subgrade moisture contents for all sites. The results show that the average value of subgrade moduli on K-18 is approximately 117 MPa with a coefficient of variation of about 12 percent over the season despite about 5 percent change in moisture content over the same period. However, majority of the variation is due to the unusually low subgrade modulus of 86 MPa computed during May. The pavement surface temperature during testing on this site was 46OC in that month. The higher deflections measured during test were presumably responsible for lower "backcalculated" subgrade modulus during this month on this site.

As shown in Figure 1, the subgrade moduli on K-113 appear to be quite variable (38 to 51 MPa) in a very narrow range of moisture contents (16 to 20 percent). However, the lower values at moisture contents greater than 20 percent indicate that the other influential factors affect the subgrade response in pavement structure as interpreted by the backcalculation of layer moduli.

The sites on US-160 and US-383 show that the pattern of variation of the subgrade moduli with the average subgrade moisture content displays the sine shape as depicted in Figure 1. This is somewhat a common characteristic for the four asphalt pavement sites studied. Variabilities for both subgrade moisture content and moduli over the seasons are very low on US-283. The coefficients of variation for these parameters are 8.5 and 13.1, respectively. Although the coefficient of variation for the subgrade moisture content on US-160 is 12.1 percent, for subgrade modulus it is 24.1 percent. The majority of the variation is due to the unusually low subgrade modulus, 39 MPa, computed for this site in August. The range in subgrade moduli for other months on this site is from 126 to 1170 MPa. Again, the pavement surface temperature during FWD testing during that month was 54OC.

The above results indicate that a typical assumption regarding moisture influence on subgrade modulus may not always be valid. It is generally assumed that the increase in moisture content should lead to higher subgrade deflections and poorer subgrade response. Based on the results of this study, it appears that this assumption is not valid for the range of moisture content encountered in the in-service pavement subgrades over the seasons. The results also indicate that the temperature variable appears to be a decisive factor of influence on subgrade response in pavement system over the seasons, since subgrade is an unbound layer that may show stress-dependent behavior. Higher temperature lowers the stiffness of the asphalt concrete layer in the pavement structure and in turn, increases the deviator stress in the subgrade. However, for a stress-dependent subgrade, moduli should increase with increase in deviator stress, not decrease, as shown by the backcalculation results of this study. It appears that in order to have a better match, MODULUS was computing lower subgrade moduli whenever the deflection basins were measured by the FWD at higher pavement surface temperature. Thus, it is apparent that in some cases, the backcalculated moduli may become "artifacts" of the calculation scheme. This point needs to be further investigated.

Effect of Variation of Subgrade Moisture Contents on the Seventh Sensor Deflections

The seventh sensor deflection of FWD is assumed to capture the structural response of subgrade due to the FWD load (3). The variations in the seventh sensor deflections with the subgrade moisture contents for the sites generally show sine-shaped relationships. The presence of different seventh sensor deflections values at the same temperature indicates the influence of other variables and most possibly, interaction among the variables. Since during backcalculation, the pavement materials were assumed to be linear and elastic (as in MODULUS), the linear relationship should exist between the seventh sensor deflections and the backcalculated subgrade moduli. The following relationships were developed between these two parameters for different sites:

K-18: Subgrade modulus (MPa) =309.5 - 4.9 * d7 (mm) ( R2 = 0.78)

K-113:Subgrade modulus (MPa) =74.6 + 17.5 * d7 (mm) ( R2 = 0.76)

US-160: Subgrade modulus (MPa) =306.7 - 5.1 * d7 (mm) ( R2 = 0.18)

US-283: Subgrade modulus (MPa) =501.1 - 9.9 * d7 (mm) ( R2 = 0.92)

The values of the coefficient of determination, R2, for the above equations indicate that the linear relationships are good. However, the low value of R2 for US-160 is due to the unusually low subgrade modulus computed in August. The equations for K-18 and US-160 indicate that for each mm increase in the seventh sensor deflection, the subgrade moduli would decrease by approximately 5 MPa. For the site on US-283, the rate of increase would be double. The relationship for K-113 is inexplicable at this point in time. A correlation study among different pavement structural and climatic variables indicate that the seventh sensor deflection values are somewhat correlated with the average temperature apparently signifying the variations in subgrade moduli during different seasons. However, the rates of change were found to be very insignificant due to small slopes of the linear equations between the seventh sensor deflection and the average pavement temperatures. For example, for a 11OC rise in pavement temperature, the seventh sensor deflection on K-18 would increase by 1.83 mm. That would translate into approximately 9 MPa decrease in subgrade modulus for K-18, a small increase per se when the average value of subgrade moduli for K-18 is approximately 117 MPa. Thus the seventh sensor deflection value can safely be assumed to be independent of the influence of the temperature of the surface layer and a good indicator of subgrade stiffness.


Based on this study, the following conclusions can be drawn about the temperature and moisture effects on measured FWD deflections and backcalculated subgrade moduli:

  1. The range of moisture content encountered in the in-service pavement subgrades over the seasons was not wide for most of the pavements and did not appear to affect the subgrade response appreciably.
  2. It was found that for almost all sites, the patterns of subgrade response, in terms of subgrade moduli versus subgrade moisture content, showed sine-shaped forms. This indicates that the temperature may affect the "backcalculated" subgrade moduli, since subgrade is an unbound layer that may show stress-dependent behavior. However, it appeared that in order to have a better match, the backcalculation scheme was consistently computing lower subgrade moduli whenever the deflection basins were measured by the FWD at higher pavement surface temperature. Thus, it is apparent that in some cases, the backcalculated moduli may become "artifacts" of the backcalculation scheme.
  3. The seventh sensor deflection value reasonably indicates the subgrade stiffness. It is hardly influenced by the temperature over the seasons.


This study was supported by the Kansas Department of Transportation.

  1. C.A. Richter. Seasonal Monitoring of Pavements - A Whole Lot More. Proceedings of the Conference on Road and Pavement Response Monitoring Systems, ASCE, New York, 1991, pp. 182–195.
  2. M. Hossain, B. Long, and S.J. Kotdwala. Seasonal and Daily Variations of Deflections Measured with the Falling Weight Deflectometer (FWD). Draft Final Report, Submitted to the Kansas Department of Transportation, December 1995.
  3. P. Ullidtz. Pavement Analysis. Elsevier Science Publishers B.V., New York, 1987.

CTRE is an Iowa State University center, administered by the Institute for Transportation.

Address: 2711 S. Loop Drive, Suite 4700, Ames, IA 50010-8664

Phone: 515-294-8103
FAX: 515-294-0467


Iowa State University--Becoming the Best