A New Quantitative Procedure To Determine The Location And Embedment Depth Of A Void With Surface Waves

Abstract

The Spectral-Analysis-of-Surface-Waves (SASW) method uses the dispersion of Rayleigh waves to assess elastic properties and layer thicknesses of media, such as soils and pavements. It has also been used for the detection of underground voids. This method can be used in different soil conditions and be performed from the surface without disturbing the medium. The effect of horizontal soil layers is accounted for in the SASW theory; however, the response of a laterally non-homogeneous medium (i.e. a medium with a void) to a transient load is not well understood. In this study, the finite differences method is used to simulate the propagation of Rayleigh waves in a non-homogeneous medium. A model is constructed, which satisfies numerical stability criteria and its applicability is verified through a calibration procedure. Voids with different dimensions and embedment depths are included in a homogeneous medium. Surface waves are measured using a multiple-receiver array; vertical and horizontal surface displacement components are analyzed to locate the void. Time and frequency domain parameters are found to be sensitive to the change in embedment depth and size of the anomaly; however, frequency parameters show higher sensitivity and better resolution. The location of voids is identified from power spectral density functions, dispersion curves, and the relative attenuation of the wave propagation. The attenuation information shows the higher sensitivity to the presence of the void (more than 200% change). Thus a new quantitative procedure based on the relative attenuation of the signals is presented to determine the location and embedment depth of a void. Although attenuated or amplified regions can be seen in the power spectral density functions before or after the void, the pattern of these regions changes with the embedment depth. A critical embedment depth is defined after which, the detection of voids with surface waves is not practical. The new detection procedure is tested with different models (i.e. different material types, and anomalies types). The results are presented in non-dimensional graphs to extend their application to other cases. KEYWORDS: surface waves, void detection, phase velocity, wave propagation, finite differences method, frequency domain analysis, logarithmic-decrement