An efficient platform HSNAS(GPU) for nonlinear static and dynamic analysis of reinforced concrete frames-II. Program verification and analysis

Li Hong-Yu,Teng Jun,Li Zuo-Hua
DOI: https://doi.org/10.13465/j.cnki.jvs.2016.14.009
2016-01-01
Abstract:A simulation platform HSNAS(GPU) for nonlinear static and dynamic analysis based on the parallel computing ability of the graphics processing unit (GPU) was developed. In order to verify the precision and efficiency of the HSNAS(GPU), static cyclic loading tests and a shaking table test on reinforced concrete (RC) columns and frames were simulated using the platform HSNAS(GPU). The results of HSNAS(GPU) show a good agreement with the results of tests. Considering the strength/stiffness degradation and the coupling effects among axial force, shear force, and bending moment, the complicated nonlinear behavior of RC columns can be simulated effectively and precisely. The static and dynamic analyses using the developed platform HSNAS(GPU) achieve a speedup of 12-14 times in computation compared with the traditional serial platform. Based on exploiting the potential of GPU computation, the platform HSNAS(GPU) could provide a computation scenario with high accuracy and efficiency for the large-scale nonlinear analysis of reinforced concrete frames. © 2016, Editorial Office of Journal of Vibration and Shock. All right reserved.
What problem does this paper attempt to address?