SASSI Analytical Methods Compared with SHAKE Results

摘要

The SASSI computer code was developed in the early 1980's to solve Soil-Structure-Interaction (SSI) problems and is commonly used today in the solution of SSI problems for U.S. Department of Energy (DOE) facilities. Two solution methods are available in SASSI, the "direct" method and the "subtraction" method which is computationally more efficient. Recent reviews by regulators have questioned the validity of SASSI results for particular sites, especially when the subtraction method is used. This paper presents a methodology for validating SASSI for use with a particular site profile. Ground motion at the surface is deconvolved in SHAKE to the bottom of the soil column and then brought back to the surface in a SASSI embedded box model. SASSI response spectra at the ground and foundation levels are compared with spectra derived from SHAKE results to validate SASSI properly transmits motion through the soil profile. Two case studies are presented: 1) a deep soil site at the Savannah River Site (SRS) and 2) a shallow stiff soil site at the Hanford Waste Treatment Plant (WTP). Results indicate that for the two cases examined, SASSI properly transmits motion through the soil profile and embedded box structure. Results also indicate that the direct and subtraction methods produce identical results for the soil profiles examined.