Abstract
Spectral Analysis of Surface Waves (SASW) is a nondestructive technique for evaluation of shear wave velocity and thickness of soil and pavement layers. It is based on in situ measurement of the surface wave dispersion curve and inversion of the measured dispersion curve to obtain the site shear wave velocity profile. Most inversion procedures involve a two-step procedure. 1-Forward modeling of the actual SASW test to obtain the theoretical dispersion curve for an assumed profile and 2-some form of optimization to match theoretical and experimental dispersion curves by varying properties of the modeled profile. A common assumption in the existing inversion procedures is that a single Rayleigh mode dominates the wave field. Although this assumption is true for regular profiles, i.e. for profiles where the shear wave velocity increases with depth, it is not correct for irregular profiles, e.g. pavements. Additionally, matching of the theoretical and experimental dispersion curves, which is usually done by minimizing the match error, does not guarantee a match between actual and inverted profiles. This is partly due to the fact that the error space is not a pure concave space and has several local minima, which locations do not match the properties of actual profile. Consequently, identification of one of these minima as the minimum of match error does not necessarily translates to matching of dispersion curves. To overcome these problems, this paper presents a new procedure for the inversion of the SASW test, which considers contribution of all types of waves and benefits from Simulated Annealing (SA), a global optimization algorithm, which discriminates between local and global minimums. The performance of the proposed procedure is then demonstrated through inversion of artificial regular and irregular soil profiles.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call