Simulation of near-fault ground motions in complex sites based on CPU-GPU heterogeneous parallelism by spectral element method
-
摘要: 利用基于CUDA编程平台实现的工作站级CPU-GPU异构并行方法开展了实际场地近断层地震动谱元法模拟。通过模拟SECE/USGS提供的自发破裂模型TPV15，测试了工作站级CPU-GPU异构并行方法的计算精度与计算效率，并将该方法应用于1679年三河—平谷M8.0地震的强地面运动模拟，以证实该方法对真实设定地震动模拟的适用性。模拟结果显示：CPU-GPU异构并行计算时间较CPU并行计算时间明显减少，加速比最高值分别为CPU 36核和72核的3.04和2.16倍；1679年三河—平谷M8.0地震的强地面模拟结果清晰地体现出近断层地震动的集中性、破裂的方向性、速度脉冲和永久位移等近断层地震动特征以及真实地形对近断层地震动所产生的影响。结果表明，CPU-GPU异构并行方法有效地提高了谱元法模拟的计算效率，可应用于大尺度复杂场地地震波场模拟。Abstract: Base on CUDA programming platform, the workstation-level CPU-GPU heterogeneous parallel method is implemented, and the spectral element method is used to simulate ground motion near-fault in a real site. In this paper, the computational accuracy and efficiency of the proposed workstation-level CPU-GPU heterogeneous parallelism method are tested by simulating the spontaneous rupture model TPV15 provided by SECE/USGS. Furthermore, the proposed method is applied to the simulation of strong ground motion in 1679 M8.0 Sanhe-Pinggu earthquake, and therefore the applicability of the proposed method to the simulation of real ground motion is verified. The simulation results show that the computing time of CPU-GPU heterogeneous parallelism is significantly reduced than that of CPU parallelism, and the highest acceleration ratio is 3.04 and 2.16 times as long as CPU 36 core and 72 core respectively. The simulation results of M8.0 in Sanhe-Pinggu earthquake in 1679 clearly show the characteristics of near-fault ground motion, such as near-fault ground motion concentration, fault rupture directivity effect, velocity pulse and permanent displacement, and the influence of real terrain on near-fault ground motion. The results show that the CPU-GPU heterogeneous parallelism method can effectively improve the computational efficiency of spectral element method simulation, and it has a good prospect to be applied to seismic wave field simulation of large-scale complex sites.
-
-
![]()
图 6 CPU-GPU异构并行系统计算的观测点位移、速度三分量与SECE/USGS提供结果对比
Figure 6. The results of displacement and velocity in three directions at the observation point compared between CPU-GPU heterogeneous parallelism and SECE/USGS
![]()
图 8 北京地区断层投影位置及整体三维物理模型
Figure 8. Fault projection location and 3-D physical model in Beijing
![]()
图 9 不同周期下三维速度波场快照
（a） 东西方向；（b） 南北方向；（c） 竖直方向
Figure 9. The snapshots of the three-dimensional velocity wave field under different cycles
（a） EW direction；（b） NS direction；（c） Vertical direction
![]()
图 10 各观测点水平 （左） 与竖直 （右）方向的加速度（a）、速度（b）和位移（c）时程
Figure 10. Acceleration （a）， velocity （b） and displacement （c） time histories corresponding to each observation points in horizontal （left） and vertical （right） directions
表 1 CPU和GPU硬件参数
Table 1 CPU and GPU hardware parameters
CPU型号
Xeon Gold 6240主频/GHz 内存容量/GB 核数 2.6 256 36 GPU型号
GeForce RTX2080 TI显存容量/GB 显存带宽/（GB·s−1） 计算能力 流处理器单元 11 616 7.5 4 352 
下载: 导出CSV
表 2 不同观测点对应的震中距以及地震动峰值
Table 2 Different observation points of epicenter distance and the peak value of ground motion
观测点 震中距/km PGA/g PGV/m·s−1 PGD/m 平谷 8.149 0.286 1.535 −1.216 三河 16.566 0.233 1.121 −0.945 通县 19.201 0.151 0.814 −0.497 兴隆 34.051 0.137 1.015 −0.753 北京 60.783 0.098 0.573 −0.413 大兴 70.293 0.043 0.307 0.196 怀柔 71.105 0.031 0.201 0.083 昌平 74.179 0.015 0.151 −0.061
下载: 导出CSV
-
陈曦,王冬勇,任俊,张训维,苗姜龙. 2016. CPU-GPU混合计算构架在岩土工程有限元分析中的应用[J]. 土木工程学报,49(6):105–112. Chen X,Wang D Y,Ren J,Zhang X W,Miao J L. 2016. Application of hybrid CPU-GPU computing platform in large-scale geotechnical finite element analysis[J]. China Civil Engineering Journal,49(6):105–112 (in Chinese).
付长华. 2012. 北京盆地结构对长周期地震动加速度反应谱的影响[D]. 北京: 中国地震局地球物理研究所: 23–33. Fu C H. 2012. A Study on Long-Period Acceleration Response Spectrum of Ground Motion Affected by Basin Structure of Beijing[D]. Beijing: Institute of Geophysics, China Earthquake Administration: 23–33 （in Chinese）.
付长华,高孟潭,俞言祥. 2015. 用数值模拟方法研究北京盆地对3—10 s地震动的放大效应[J]. 地震研究,38(3):448–460. doi: 10.3969/j.issn.1000-0666.2015.03.016 Fu C H,Gao M T,Yu Y X. 2015. Studying on amplification effect of Beijing basin on 3−10 s ground motion by numerical simulation method[J]. Journal of Seismological Research,38(3):448–460 (in Chinese).
高孟潭,俞言祥,张晓梅,吴健,胡平,丁彦慧. 2002. 北京地区地震动的三维有限差分模拟[J]. 中国地震,18(4):356–364. doi: 10.3969/j.issn.1001-4683.2002.04.005 Gao M T,Yu Y X,Zhang X M,Wu J,Hu P,Ding Y H. 2002. Three-dimensional finite-difference simulations of ground motions in the Beijing area[J]. Earthquake Research in China,18(4):356–364 (in Chinese).
郝明辉,张郁山. 2019. 基于DEM数据的地形效应经验预测模型研究[J]. 土木工程学报,52(2):86–96. Hao M H,Zhang Y S. 2019. Research on empirical prediction model of terrain effects based on DEM[J]. China Civil Engineering Journal,52(2):86–96 (in Chinese).
胡元鑫,刘新荣,罗建华,张梁,葛华. 2011. 汶川震区地震动三维地形效应的谱元法模拟[J]. 兰州大学学报(自然科学版),47(4):24–32. Hu Y X,Liu X R,Luo J H,Zhang L,Ge H. 2011. Simulation of three-dimensional topographic effects on seismic ground motion in Wenchuan earthquake region based upon the spectral-element method[J]. Journal of Lanzhou University (Natural Sciences)
,47(4):24–32 (in Chinese). 刘博研,史保平,张健. 2007. 复合地震源模拟强地面运动:以1679年三河—平谷MS8.0地震为例[J]. 地震学报,29(3):302–313. doi: 10.3321/j.issn:0253-3782.2007.03.009 Liu B Y,Shi B P,Zhang J. 2007. Strong motion simulation by the composite source modeling:A case study of 1679 M8.0 Sanhe-Pinggu earthquake[J]. Acta Seismologica Sinica,29(3):302–313 (in Chinese).
刘春成,顾汉明,陈宝书,焦振华,马凯,蔡志成,张立. 2019. 基于GPU和CPU协同并行的三维各向异性介质地震波场正演模拟[J]. 地质科技情报,38(5):240–246. Liu C C,Gu H M,Chen B S,Jiao Z H,Ma K,Cai Z C,Zhang L. 2019. Forward modelling of seismic wavefield in 3D anisotropic media based on GPU and CPU collaboration parallel[J]. Geological Science and Technology Information,38(5):240–246 (in Chinese).
刘培玄,李小军,赵纪生. 2019. 基于断裂两侧应变能积累的地震危险性参数估计:以1679年三河—平谷M8.0地震为例[J]. 地震学报,41(2):259–268. doi: 10.11939/jass.20180110 Liu P X,Li X J,Zhao J S. 2019. Seismic hazard parameters estimation based on strain energy accumulation in both sides of a fault:Taking the 1679 Sanhe-Pinggu M8.0 earthquake as an example[J]. Acta Seismologica Sinica,41(2):259–268 (in Chinese).
刘启方. 2020. 1556年华县大地震地震动场模拟[J]. 自然灾害学报,29(5):1–10. Liu Q F. 2020. Strong motion simulation of the 1556 great Huaxian earthquake[J]. Journal of Natural Disasters,29(5):1–10 (in Chinese).
刘启方,袁一凡,金星,丁海平. 2006. 近断层地震动的基本特征[J]. 地震工程与工程振动,26(1):1–10. doi: 10.3969/j.issn.1000-1301.2006.01.001 Liu Q F,Yuan Y F,Jin X,Ding H P. 2006. Basic characteristics of near-fault ground motion[J]. Earthquake Engineering and Engineering Vibration,26(1):1–10 (in Chinese).
刘少林,杨顶辉,徐锡伟,李小凡,申文豪,刘有山. 2021. 模拟地震波传播的三维逐元并行谱元法[J]. 地球物理学报,64(3):993–1005. doi: 10.6038/cjg2021O0405 Liu S L,Yang D H,Xu X W,Li X F,Shen W H,Liu Y S. 2021. Three-dimensional element-by-element parallel spectral-element method for seismic wave modeling[J]. Chinese Journal of Geophysics,64(3):993–1005 (in Chinese).
潘波,许建东,刘启方. 2009. 1679年三河—平谷8级地震近断层强地震动的有限元模拟[J]. 地震地质,31(1):69–83. doi: 10.3969/j.issn.0253-4967.2009.01.007 Pan B,Xu J D,Liu Q F. 2009. Simulations of the near-fault strong ground motion of the 1679 Sanhe−Pinggu M8 earthquake[J]. Seismology and Geology,31(1):69–83 (in Chinese).
孙吉泽,俞言祥,何金刚,李一琼. 2017. 2013年乌鲁木齐MS5.6和MS5.1地震强地震动模拟研究[J]. 地震学报,39(5):751–763. doi: 10.11939/jass.2017.05.010 Sun J Z,Yu Y X,He J G,Li Y Q. 2017. Ground motion simulation of 2013 Ürümqi MS5.6 and MS5.1 earthquakes[J]. Acta Seismologica Sinica,39(5):751–763 (in Chinese).
熊琛,许镇,陆新征,叶列平. 2016. 城市区域建筑群地震灾害场景仿真的高真实感可视化方法研究[J]. 土木工程学报,49(11):45–51. Xiong C,Xu Z,Lu X Z,Ye L P. 2016. High-fidelity visualization of urban earthquake disaster scenario simulation[J]. China Civil Engineering Journal,49(11):45–51 (in Chinese).
于彦彦. 2016. 三维沉积盆地地震效应研究[D]. 哈尔滨: 中国地震局工程力学研究所: 67–74. Yu Y Y. 2016. Research on Seismic Effects in Three-Dimensional Sedimentary Basins[D]. Harbin: Institute of Engineering Mechanics, China Earthquake Administration: 67–74 （in Chinese）.
张文强. 2020. 破裂动力学的曲线网格有限差分方法研究及高性能计算[D]. 合肥: 中国科学技术大学: 100–111. Zhang W Q. 2020. Study of Curve Grid Finite Difference Method in Rupture Dynamics and High Performance Computing[D]. Hefei: University of Science and Technology of China: 100–111 （in Chinese）.
朱耿尚. 2014. 有限差分方法在强地面运动模拟中的应用[D]. 合肥: 中国科学技术大学: 65–91. Zhu G S. 2014. Strong Ground Motion Simulation by Finite Difference Method[D]. Hefei: University of Science and Technology of China: 65–91 （in Chinese）.
Campa S,Danelutto M,Goli M,González-Vélez H,Popescu A M,Torquati M. 2014. Parallel patterns for heterogeneous CPU/GPU architectures:Structured parallelism from cluster to cloud[J]. Future Gener Comput Syst,37:354–366. doi: 10.1016/j.future.2013.12.038
Chaljub E,Maufroy E,Moczo P,Kristek J,Hollender F,Bard P Y,Priolo E,Klin P,De Martin F,Zhang Z G,Zhang W,Chen X F. 2015. 3-D numerical simulations of earthquake ground motion in sedimentary basins:Testing accuracy through stringent models[J]. Geophys J Int,201(1):90–111. doi: 10.1093/gji/ggu472
Komatitsch D,Liu Q Y,Tromp J,Süss P,Stidham C,Shaw J H. 2004. Simulations of ground motion in the Los Angeles basin based upon the spectral-element method[J]. Bull Seismol Soc Am,94(1):187–206. doi: 10.1785/0120030077
Komatitsch D,Michéa D,Erlebacher G. 2009. Porting a high-order finite-element earthquake modeling application to NVIDIA graphics cards using CUDA[J]. J Parallel Distrib Comput,69(5):451–460. doi: 10.1016/j.jpdc.2009.01.006
Li C G,Maa J P Y,Kang H G. 2012. Solving generalized lattice Boltzmann model for 3-D cavity flows using CUDA-GPU[J]. Sci China Phys Mech Astron,55(10):1894–1904. doi: 10.1007/s11433-012-4856-9
Machidon A L,Machidon O M,Ciobanu C B,Ogrutan P L. 2020. Accelerating a geometrical approximated PCA algorithm using AVX2 and CUDA[J]. Remote Sens,12(12):1918. doi: 10.3390/rs12121918
Shi L,Chen H,Sun J H,Li K L. 2012. vCUDA:GPU-accelerated high-performance computing in virtual machines[J]. IEEE Trans Comput,61(6):804–816. doi: 10.1109/TC.2011.112
Stupazzini M,Paolucci R,Igel H. 2009. Near-fault earthquake ground-motion simulation in the Grenoble valley by a high-performance spectral element code[J]. Bull Seismol Soc Am,99(1):286–301. doi: 10.1785/0120080274
计量
- 文章访问数: 563
- HTML全文浏览量: 214
- PDF下载量: 116
