Verification testing and alternative test methods for the resilient modulus of fine grained soils

Under the new AASHTO (1986) pavement design method, the resilient modulus, M,, is required to define the elastic modulus for the subgrade soil. This method of designing is mechanistic, which is an improvement over the older empirical method (AASHTO 1972). The cyclic triaxial test has been the standard method of determining the resilient modulus. This method is however, time consuming and does not always produce accurate, reproducible results. Therefore, methods of estimating the resilient response of a subgrade soil are necessary to make the use of resilient modulus in mechanistic design feasible.

A laboratory investigation was conducted on six soils that were obtained from various sites across the state of Tennessee. For each of these six soils, index tests were conducted to determine the properties of each soil. Specimens were compacted at optimum water content and maximum dry density (standard energy), and laboratory measurements of the resilient modulus were performed using the cyclic triaxial method, and an alternative test method (Li 1992). Model parameters were found for two different analytical models. In addition to testing specimens at optimum water content and maximum dry density, three of the soils were compacted at optimum conditions and the degree of saturation was raised to a higher level of saturation before testing. Significant reduction in resilient modulus was observed as the degree of saturation was increased. An alternative test method (Li 1992) based on a falling mass striking a soil specimen compacted in a proctor mold was also used. The ATM detected changes in M, as measured with the cyclic triaxial method. However, the effects of saturation were not distinct as in the triaxial test. Furthermore, the response at low values of deviator stress is not measured accurately with the ATM. It is suggested that other variations of hammer weight and drop height be evaluated. However, from this testing it was concluded that while the cyclic triaxial method produces more consistent data, the alternative test method may represent the in-situ stress state better.