Multistation Methods for Geotechnical Characterization using Surface Waves

This dissertation deals with soil characterization methods based on surface wave propagation applied to geotechnical engineering purposes. This topic has gained much interest in the last decade because of the appealing possibilities given by non-invasive methods, which are at once very flexible and cost effective. An overview of the properties of Rayleigh waves in layered linear elastic and linear viscoelastic media is presented, together with their applications for site characterization, of whose the SASW (Spectral Analysis of Surface Waves) method is by far the most well-known in geotechnical engineering. The research has been mainly focused on the application of multistation methods, compared with the classical two-station approach typical of the SASW method. Results from both numerical simulations and experimental testing are reported to compare two-station and multistation methods and to clarify the advantages that can be obtained using the latter ones. In particular the research has been developed following two different directions: on the one hand the application of classical geophysical analysis tools (such as domain analysis and slant stack transform) to tests performed with impulsive sources. On the other one the possibility of obtaining from surface wave testing not only a stiffness profile, but also a damping ratio profile for the site. In this respect a new method for simultaneous measurements of Rayleigh dispersion and attenuation curves is proposed. Regarding the first topic, the necessity of a multistation approach to determine the experimental dispersion test is essentially related to the spatial variation of phase velocity. Analyses in the frequency-wavenumber domain and in the frequency-slowness domain are very powerful approaches, still there was a need of studying the effects of the change of scale from geophysical applications to geotechnical ones. Indeed because of the peculiar properties of Rayleigh waves, surface testing is strongly affected by the distance travelled by the analysed wave. The numerical simulations performed in the research show that the phase velocity obtained using multistation methods with a limited number of receivers close to the source is not a modal value as it is for geophysical applications, but an apparent phase velocity arising from modal superposition. The experimental tests showed the good performances of multistation methods when compared to the SASW method. In particular some drawbacks of the latter method, due essentially to its two-station nature, are avoided and the field-testing appears to be very promising for future applications. In particular the application of the frequency-wavenumber domain analysis can lead to much faster and more stable estimates of the experimental dispersion curve and the process is easily automated, with a great saving of time and less requirement for subjective decisions. Another important advantage is given by the stability with respect to a near field effects that lead to a better reconstruction of the dispersion curve for the low frequencies and hence to a deeper characterization. The necessity of a new method for the simultaneous determination of surface wave dispersion and attenuation curves is linked to the strong coupling existing between the two. Such coupling is extremely important for the subsequent inversion process, in a consistent method leading from the field measurements to the stiffness and damping profiles. The proposed method uses a new testing configuration, designed to measure the experimental transfer function. Successively a regression process of the complex quantity with the corresponding expression obtained modelling soil as a linear viscoelastic layered system leads to the experimental dispersion and attenuation curves. Some preliminary results are reported showing very encouraging results, also if a more extensively testing programme is required for the complete validation of the method.