GNGTS 2017 - 36° Convegno Nazionale

358 GNGTS 2017 S essione 2.2 of the most common techniques for site investigation in the context of seismic ground response assessment. To this aim a systematic collection of invasive and non-invasive test results at three reference sites has been organised. Both intra-method and inter-methods comparisons have been implemented. Specifically, for surface wave analysis, a benchmark on a common experimental dataset has involved 14 different teams of well-recognized international experts in the field to assess performances and reliability of interpretation procedures. Three sites in France and Italy have been selected to represent different geological settings: soft soil overlying rock (Mirandola); stiff soil extending to significant depths (Grenoble); rock outcrop (Cadarache). The complete set of results is reported in two companion papers: one for the intra-method comparisons for surface wave analyses (Garofalo et al. , 2016a); the other dealing with the inter- method comparisons between invasive and non-invasive seismic tests (Garofalo et al. , 2016b). On the basis of the evidence from these comparisons and of the substantial experience of the participants in the project, the guidelines for surface wave analyses have been drafted. The guidelines. The main text is written with reference to Rayleigh waves, which are the most commonly exploited surface waves. The same principles apply to the analysis of other kinds of surface waves, such as Love and Scholte waves, which however require specific data acquisition procedures and forward modelling algorithms. The typical mainstream procedure is outlined in the guidelines, considering standard practice for fundamental mode analysis, however the basic concepts to recognize the conditions in which such simplified approach is not applicable are also reported. The guidelines are organized as follows: after a brief introduction on the basic principles of surface wave methods, the typical steps of the test (acquisition, processing and inversion) are discussed and suggestions are provided for their implementation. A series of appendices (provided as additional on line material) cover specific issues and provide selected references for gaining a deeper insight into particular aspects of surface wave methods. Particular attention is devoted in the guidelines to the uncertainties associated to the results of surface wave analyses and to the limitations that have to be taken into account when considering the reliability of the results. Specifically, non- uniqueness of the solution arises from the ill-posedness of the inverse problem that has to be solved to obtain the shear wave velocity from the experimental dispersion curve. In this respect a single best fitting profile is typically not an adequate representation of the results. Other sources of aleatory and epistemic uncertainties (e.g. uncertainties in experimental data, simplification imposed by the initial assumption of the 1D isotropic elastic model, lateral variations, parameterization of the model space) also affect the reliability of the solution. For both active-source and passive (ambient vibrations) surface wave methods, the experimental data acquisition procedure is then treated in details, describing the typical equipment, the acquisition layout and the recording parameters. Procedures to check signal quality are then suggested, considering the peculiarities of the two different kinds of datasets. Several signal processing procedures can be profitably used to obtain the experimental dispersion curve from the field data. In the guidelines the frequency-wavenumber analysis Fig. 1 - Typical shape of the experimental dispersion curve for a site with a soft layer at depth, as indicated by the trough in phase velocity between 2-10 Hz (Grenoble site – InterPACIFIC Project, from combination of active and passive measurements; Foti et al. , 2017).