Back-Calculation of Resilient Modulus and Prediction of Permanent Deformation for Fine-Grained Subgrade under Cyclic Loading

Subgrade of roadways is subjected to repeated traffic loading at different loading intensities. The resilient modulus of subgrade is one of the important parameters for the design of pavements, including the estimation of permanent deformation (rutting) in the current Mechanistic-Empirical Pavement Design Guide (MEPDG). A significant portion of total rutting is often contributed by the subgrade, especially soft, fine-grained subgrade. Existing formulas for calculating rut depth of subgrade are only suitable for small deformation. Under certain conditions (e.g., low volume roads), large rut depth may be allowed. In addition, the resilient modulus and the accumulation of the permanent deformation of a fine-grained subgrade are still not well quantified under cyclic loading at different intensities. In this study, cyclic plate loading tests at different loading intensities were performed on six test sections of fine-grained subgrade constructed in a geotechnical testing box [$2\mathrm{m}\text{wide}\times 2.2\mathrm{m}\text{long}\times 2\mathrm{m}\text{high}$]. The California Bearing Ratios (CBRs) of the test sections ranged from 2.9 to 15.8%. The intensities of loading applied on a steel plate 305 mm in diameter increased from 5 to 70 kN, in load increments of 5 kN. In these tests, deformations and vertical/horizontal stresses were measured at varying distances from the center of the loading plate. For each test section, dynamic cone penetrometer (DCP) tests and light weight deflectometer (LWD) tests were performed and a correlation between the CBR and dynamic modulus obtained from LWD tests was developed. Based on the measured surface resilient (elastic or recoverable) deformations, the resilient moduli of test sections at different loading intensities were back-calculated. The MEPDG soil damage model was modified to consider the influences of subgrade stiffness and bearing capacity. The modified MEPDG model was verified by two test sections with the base course over the subgrade, and the permanent deformations predicted by the modified model compared well with the measured results.