The authors present a simple elastostatic approach to estimate a suitable subgrade modulus from falling weight deflectometer (FWD) data. An effective surface modulus is defined on the basis of Boussinesq-Newmark equations. Deviations between the model assumptions and actual in situ conditions are reflected in the effective surface modulus variation with radius. The variation indicates which sensor readings are suitable for backcalculation. The presence of a shallow bedrock, increase in subgrade modulus with depth, or anisotropic properties may lead to nonconservative estimates of subgrade modulus. Assuming that the deflections at the subgrade surface (not directly under the load) may be approximated by the surface deflections, the Boussinesq-Odemark equations are used to define an effective subgrade modulus. In order to determine the subgrade modulus, the equivalent thickness of the pavement structure, which reflects pavement stiffness, is estimated by fitting preselected normalised deflections. Given the equivalent thickness, an effective subgrade modulus profile is established, from which a design subgrade modulus may be determined. An example using data of the Canadian Strategic Highway Research Program (C-SHRP) demonstrates the approach. The uniformity of the subgrade at the C-SHRP site is studied by considering the spectral density functions of the FWD time histories. The group velocity is used to define a weighted average profile modulus. The profile moduli of the two sections studied were higher but consistent with subgrade moduli estimated using elastostatic analysis. (Author/publisher)
Abstract