Development of Acceptance Criteria for Soil-Structure Interaction Solutions Using Linear SSI Techniques

Abstract

Seismic analysis and design of high-hazard nuclear facilities requires evaluation of soil-structure interaction (SSI) effects on structure and soil response due to earthquake ground motions. The industry-wide methodology of computing SSI response of buildings is through linear SSI techniques using the computer code SASSI. Technical issues were identified by users and regulators (Ref. 1) resulting in the U.S. Department of Energy (DOE) commissioning a large scale, multi-year Validation and Verification (V&V) Project for SASSI (Ref. 2). The project was a highly peer reviewed process that included an esteemed Participatory Peer Review Team, DOE oversight, as well as regulatory and stakeholder input. The project goal was to develop benchmark SSI solutions for the range of SSI problems associated with high-hazard facilities within the DOE complex. As per industry software quality control requirements, an acceptance limit for the benchmark solutions must be provided to define the acceptable accuracy of the results produced by SASSI relative to the benchmark. In order to define this limit, variation in the solution of foundation impedance must be related to an expected level of accuracy in structural design quantities of interest (e.g., response spectra, base shear, etc.). Therefore, extensive parametric analyses were performed for coupled soil-structure systems having a broad range of SSI parameters defined by foundation size, soil properties, building stiffness and mass properties, building height, etc. Based on the results of the parametric analyses, ASCE 4 code guidance, and engineering judgment of CJC&A and the peer review team, an acceptable level of accuracy in computed foundation impedance was determined for SASSI solutions. This supported the successful qualification of SASSI for use in two large-scale DOE projects.