Seismic site amplification factor (DAF): Geometrical limits of 1D modelling

It is here presented an attempt of quantitative evaluation of the limits of applicability of 1D modelling for evaluation of the site seismic response when dealing with a non horizontal basement. The dynamic evaluation of the numerous situ elastic parameters needed for a correct 2D modelling can represent a considerable part of the economie resources of a project. Therefore it is important to understand the limits of the ID modelling in order to mini mise the expenses without dangerous errors when evaluating the seismic hazard. The response of the most used elastic non-linear 1D and 2D algorithms have been used, in order to estimate the seismic site response. A non-horizontal basement has been modelled for various angles of dip up to 30°. The reference accelerogram is relative of the Umbria earthquake of Colfiorito (ltaly, 26/09/97, M=5.6), recorded at the Department of Geophysics and Volcanology - University of Naples Federico II. The results evidenced that if the dip of the basement is less than 20°, the evaluation of the site response given by the 1D and 2D models is almost similar. For dip values exceeding 20° the DAF spectra starts to diverge, both in terms of amplitude and peak period, therefore the response of the lD models is inaccurate. In these geological conditions the 1D algorithms should not be used for DAF evaluation. The theoretical determinations have also been tested in a real site located in the city of Naples, characterised by a 32° dipping volcanic tuff basement. The results show that the theoretical outcome can be extrapolated to that particular site.