Experimental investigation of near-field effects on the SASW dispersion curve

When any method of surface wave testing that involves Rayleigh waves is performed, one important assumption is that plane Rayleigh waves are being measured. In the forward modeling or inversion procedure that is used to analyze the field dispersion curve to determine the field V[subscript s] profile, the analysis is based on the wave field consisting of plane Rayleigh waves. Therefore, field dispersion curves that contain near-field data could adversely distort the field V[subscript s] profile. To minimize the influence of near-field effects, several criteria have been recommended in the past. However, most of the criteria were based on empirical equations that implicitly assumed zones of influence, or numerical simulations. There is a lack of experimental investigation, particularly full-scale field investigations. Even, the numerical solutions have been based on simple soil profiles without significant velocity contrasts between soil layers and/or varying thicknesses of soil layers which can significantly influence near-field effects. Data from full-scale field test using the Spectral-Analysis-of-Surface-Waves (SASW) method was used in this thesis research. SASW tests performed at two stages in the construction of a deep, 90-ft thick backfill were studied. The V[subscript s] profiles were normally dispersive, with a substantial increase in the velocity of the layer beneath the backfill. The study shows the adverse distortions that can occur in the field dispersion curve from near-field effects when the spacing of the receiver pair is: (1) above the zone of rapidly increasing V[subscript s] near the surface and (2) less than the depth to the stiffer layer in deeper measurements. Other factors that affect the results are discussed and recommendations are presented to minimize the introduction of near-field effects, at least in these relatively simple V[subscript s] profiles.