Experimental and numerical modelling of a scaled soil-structure system

Abstract

This paper describes how recent earthquakes (Bucharest, 1997; Mexico City, 1985; and Kobe, 1995) have clearly shown the great influence of soil-structure interaction (SSI) in the seismic response of structures to earthquakes. Unfortunately, man seismic codes do not consider SSI or consider SSI detrimental only for some specific structures, because of the fact that SSI generally leads to higher fundamental periods and then, considering the shape of response and design spectra of seismic codes, to smaller accelerations and stresses in the structure. But, the response spectra of several recent earthquakes were characterized by a very high predominant period. The analysis of SSI could be performed by means of different approaches; the two approaches discussed in the paper are experimental modeling by means of shaking table tests and numerical modeling by means of an FEM code. A 1:6 scaled steel model resting on a sand deposit is subjected to sinusoidal input motions of different amplitudes and constant frequency and duration. In terms of acceleration and displacement, a good agreement is obtained comparing experimental and numerical results.