摘要
为了更有效地模拟和分析地铁站内复杂楼梯上洪水流动的动力学特征,采用光滑粒子流体动力学方法(SPH)并借助GPU并行计算技术对复杂形态(带1、2、3个休息平台的直行、直角转弯L型)阶梯上洪水流动特征和漫延过程进行数值建模和计算分析。基于SPH方法和GPU并行加速的复杂阶梯流模型的数值模拟结果表明:该模型不但可以计算获得更好的阶梯流流态,而且具有较高的计算效率;不同休息平台数、不同形态阶梯上洪水水流呈现差别较大的水流特征;该模型借助GPU并行加速,适合研究大尺度复杂楼梯上洪水漫延问题,可为分析和评估洪水水流对地铁站楼梯上被困人员疏散过程的影响提供有力的工具。
This paper intends to apply the GPU-based smoothed particle hydrodynamics( SPH) method to modulate the dynamic characteristic features of the flooding current over the complex staircases( with 1,2,3 straight rest platform and L-shape turn),in hope to simulate the flood current flow dynamic features of the complex staircases more effectively in the subway stations. For the simulated purpose,we have conducted the GPUbased simulations by following the GPUSPH model with our improvements in the solid-wall boundary treatment,the pressure correction treatment and the complex building modules. What is more,we have established the structure and size of the subway station staircases through simulation according to a real case( a subway station in Beijing City,of the state capital). The results of the complex staircase flow model we have laid out based on SPH method and GPU parallel acceleration indicate that:( a)The proposed model can reproduce fairly nice regimes of the staircase flow and the violent fluid-structure interaction( flooding flow impacts on the wall),and run with significantly high efficiency;( b) The features of the flooding flow on the different rest platforms and different types of staircases are quite different. The impacts of different rest platforms of the straight staircase on the flow features can be stated as below:( a) The number of the rest platforms may likely affect the flow pattern significantly;( b)The greater the number of straight staircase is,the stronger the influence of the flow pattern will be,and,in turn,the lower safety the rest platform will be,and the more violent the flow field distribution on the platform will become;( c) And,consequently,the more unstable the flow pattern is,the greater the flow current velocity at the edge of the staircase will become. To be accurate,the effect of the staircase type on the flow current features can be presented in the following manners:( a) The staircase form can also have great impact on the flooding process and the current flow features,and the flooding flow upstream of the first platform is potential of gliding current on the whole;( b) the flooding flow over the straight staircase with rest platforms will form a free picking flow,which leads to unstable flow feature downstream of the rest platform,and the flow velocity on the step edge increases;( c) With regard to the water flow over the L-shaped staircase,due to the changes in the flow current direction of the staircase water,the flow velocity and the impacting force of the water flow can be reduced,and the pressure distribution on the L-shaped staircase may have dramatically been relieved. All the above mentioned reasonable results demonstrate that the GPU-based SPH model can provide a promising alternative way for numerical investigation of the large-scale flooding cuttent over the complex staircases,which can actually offer a valuable technical support for evaluating the influence of the water current features on the evacuation of the workers trapped in the underground subway station.
作者
吴建松
李娜
刘文雨
周世禄
WU Jian-song;LI Na;LIU Wen-yu;ZHOU Shi-lu(School of Emergency Management and Safety Engineering,ChinaUniversity of Milling and Technology,Beijing 100083,China)
出处
《安全与环境学报》
CAS
CSCD
北大核心
2019年第3期774-779,共6页
Journal of Safety and Environment
基金
国家自然科学基金项目(11502283)
国家重点研发计划项目(2017YFC0805001)
关键词
安全工程
地铁站
洪水
阶梯流
光滑粒子流体动力学
GPU并行计算
safety engineering
underground space
flood
staircase flow
smoothed particle hydrodynamics
GPU parallel computing
