Determination of effective frequency range excited by falling weight deflectometer loading history for asphalt pavement

Abstract

Falling weight deflectometer (FWD) tests have been widely used to evaluate the layer properties of in-situ asphalt pavement structures. The FWD loading is essentially impulse loading and can excite a wide range of frequency components. This study is to determine the effective frequency range excited by FWD loading through the dynamic back-calculation analysis of FWD data. The FWD deflections were calculated using the spectral element method for asphalt pavements under various field measured FWD loading histories. The calculated deflections were then used as the measured ones to back-calculate the properties of asphalt concrete layers using the dynamic back-calculation procedure. The viscoelastic behaviour of asphalt concrete was characterized using the modified Havriliak-Negami model. For each set of FWD data, several sets of back-calculated model coefficients were obtained by inputting different seed values of model coefficients in the back-calculation procedure. The results show that although all the back-calculated and measured deflection curves are in good agreement, the values of the back-calculated model coefficients can be significantly different and the back-calculated master curves match well only in a specific frequency range that is the effective frequency range excited by FWD loading. The effective frequency ranges for various pavement structures and FWD loading histories were determined. The results show that a higher loading peak level can excite a wider effective frequency range, or more specifically the greater upper limit, regardless of the pavement structures. In general, the effective frequency range excited by a FWD loading history is 5 ~ 65 Hz for asphalt pavement. The existence of effective frequency range indicates that the information contained in FWD data is not sufficient to capture the asphalt concrete properties over the entire frequency domain.