Implementation of a modified domain reduction technique as an engineering tool for site response analysis: method validation and verification under incident SH waves

Abstract

This work describes a computational framework to conduct site response analysis (SRA) including topographic effects and intended to be used by practicing engineers -- The approach is based upon the domain reduction method (DRM) originally formulated by Bielak et al. (2003), in order to determine the response of small-scale localized topographic features present in large-scale seismic scenarios -- Here we introduce additional assumptions into the reduction method, so it can be used with computational resources typically available at a consulting office -- In particular, as suggested by the original DRM technique, we split the problem into a large-scale regional analysis and into a small-scale or local SRA. However, by contrast with the original approach, in our modified version of the method we include into the free-field the effect of the regional topography after generalizing the concept of the half-space -- Subsequently, this free-field motion is used in the second step of the analysis and conducted to capture the effect of the small-scale (high frequency) topographic irregularities and the mechanical effect -- For the local analysis we use an homogenized version of the half-space, deviating once again from the original Approach -- In this work we conduct a verification study of our modified DRM approach within the context of 2D topographies submitted to incident SH waves -- For that purpose we have implemented the DRM algorithm into a commercial finite element analysis code having standard dynamic analysis capabilities -- In order to test the applicability of the modified technique we selected idealized topographies amenable to be studied by geometrical methods