%A WANG Jian-Wei, JIN Xian-Long, CAO Lu-Fen, ZHANG Wei-Wei %T Parallel Computation for the Dynamic Response of Connectional Passage in Twin-Tunnel due to Train Load %0 Journal Article %D 2012 %J Journal of Shanghai Jiao Tong University %R %P 591-595 %V 46 %N 04 %U {https://xuebao.sjtu.edu.cn/CN/abstract/article_39975.shtml} %8 2012-04-28 %X Taking a largescale twintunnel as an example, a 3D model of traintunnelconnectional passagessoil system was established. Based on Dawning 5000A in Shanghai Supercomputer Center, a traintunnel interaction balanced domain decomposition algorithm was designed to overcome the computing difficulty induced by large scale model. The dynamic response of connectional passage under train load was analyzed, the computing efficiency of different partition schemes were also compared. The results show that although the additional dynamic stress in connecting passage due to train load is small, two train intersect should be avoided in this location because of the high stress level. The traintunnel interaction balanced domain decomposition algorithm has higher parallel efficiency than the common recursive coordinate bisection method.

map