Parallel analysis method of structural nonlinear seismic response based on Woodbury + OpenMP

被引：0
|
作者：
Yu D. [1 ]
Li G. [1 ]
机构：
[1] State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian
来源：
Zhendong yu Chongji/Journal of Vibration and Shock | 2023年 / 42卷 / 03期
关键词：
OpenMP parallel computation; seismic response analysis; structural nonlinearity; Woodbury formula;
D O I：
10.13465/j.cnki.jvs.2023.03.003
中图分类号：
学科分类号：
摘要：
Nonlinear seismic response analysis becomes an important technical means to evaluate aseismic performance of engineering structures. With increase in structure size, large-scale stiffness matrix iterative updating induced by nonlinearity becomes a key factor to restrict efficiency of structural analysis. Woodbury formula-based nonlinear seismic response analysis method for structures, i. e., Woodbury method, is a new type efficient numerical analysis method. Using local nonlinear characteristics of structure under seismic action, this method can not only ensure higher iterative convergence rate and but also effectively avoid real-time change of structural stiffness matrix and the resulting low computational efficiency. However, the current studies are based on serial computing mode and don' t make full use of parallel computing capabilities of computation hardware. Here, OpenMP mode was introduced to accelerate Woodbury method in parallel, a parallel calculation method for efficient nonlinear seismic response analysis of structures was proposed. Firstly, each iteration calculation step was divided into 3 main calculation parts of updating nonlinear correlation coefficient matrix calculation, solving displacement response based on Woodbury formula, and determining element state. Then, a block calculation method of nonlinear correlation coefficient matrix was established, the calculation process of Woodbury formula was decomposed into 6 parallelizable calculation steps, and each element state was determined separately, OpenMP parallel acceleration strategies suitable for these 3 things were established, respectively to realize parallel optimization for the proposed method's whole process. Finally, the correctness and efficiency of the proposed method were verified with seismic response analysis of a high-rise structure. © 2023 Chinese Vibration Engineering Society. All rights reserved.
引用
收藏
页码：21 / 29and49
页数：2928
相关论文
共 48 条
  • [1] FU Xiaodong, SHENG Qian, ZHANG Yonghui, Parallel computing method for discontinuous deformation analysis using OpenMP, Rock and Soil Mechanics, 35, 8, pp. 2401-2407, (2014)
  • [2] ZHANG Yongbin, TANG Chun'an, JIA Jinghui, Et al., Application of OpenMP in rock dynamic parallel computation, Journal of Wuhan University of Technology, 35, 2, pp. 97-101, (2013)
  • [3] CAO Shengtao, LI Zhishan, YANG Zhiyong, Nonlinear explicit layered shell element and its GPU parallel computing implementation, Journal of Vibration and Shock, 38, 22, pp. 60-69, (2019)
  • [4] HAN Bo, LU Xinzheng, XU Zhen, Et al., Seismic damage simulation of urban buildings based on high performance GPU computing, Journal of Natural Disasters, 21, 5, pp. 16-22, (2012)
  • [5] CAO Shengtao, LU Dechun, DU Xiuli, Et al., Development on software platform for nonlinear dynamic analysis of underground structure based on GPU parallel computing, Engineering Mechanics, 36, 2, pp. 53-65, (2019)
  • [6] LI Hongyu, TENG Jun, LI Zuohua, Analysis method for seismic response of high-risestructure based on CPU-GPU heterogeneous platform, Journal of Vibration and Shock, 33, 13, pp. 86-91, (2014)
  • [7] BOLZ J, FARMER I, GRINSPUN E, Et al., Sparse matrix solvers on the GPU: conjugate gradients and multigrid, ACM Transactions on Graphics, 22, 3, pp. 917-924, (2003)
  • [8] GEORGESCU S, OKUDA H., Conjugate gradients on multiple GPUs, International Journal for Numerical Methods in Fluids, 64, pp. 1254-1273, (2010)
  • [9] LI Hongyu, TENG Jun, LI Zuohua, An efficient platform HSNAS (GPU) for nonlinear static and dynamic analysis of reinforced concrete frames: I program development, Journal of Vibration and Shock, 35, 14, pp. 47-53, (2016)
  • [10] CHEN Xi, WANG Dongyong, REN Jun, Et al., Application of hybrid CPU-GPU computing platform in large-scale geotechnical finite element analysis, China Civil Engineering Journal, 49, 6, pp. 105-112, (2016)
12345