High Performance Computing of Dynamic Structural Response Analysis for the Integrated Earthquake Simulation

This paper proposes an application of high performance computing (HPC) to dynamic structural response analysis (DSRA) in order to enhance the capability and increase the efficiency of integrated earthquake simulation (IES). Object Based Structural Analysis (OBASAN) is a candidate DSRA program for IES. With OBASAN, the reliability of structural damage prediction can be increased by means of more complicated structural analysis models, which provide a large number of parameters to generate mathematical equations containing large matrices. In HPC, the general purpose graphic processing unit (GPGPU) is designed to deal with a large amount of data and a large number of simple calculations in mathematical equations. CUDA architecture is a candidate parallel programming model which can extend C/C++ programming code to enable parallel processing in the graphic processing unit (GPU). Since HPC technique can support the solution of mathematical equations in a reasonably short time and since more complicated structural analysis models can be applied to the nonlinear dynamic structural response analysis, IES can be significantly improved to provide more reliable prediction of structural damage in earthquake scenarios.