基于SPH方法与GPU并行计算的阶梯流数值模拟

为了更好地模拟分析地下空间阶梯流动力学规律和提升数值计算效率,本文利用光滑粒子流体动力学方法(SPH)并借助GPU并行计算技术对地下空间阶梯流进行数值建模和计算。基于GPU并行化加速的SPH地下空间阶梯流模型的数值计算结果表明：该模型不仅计算获得较好的阶梯流流态,而且计算效率大幅度提高,比串行SPH模型和移动粒子半隐方法(MPS)模型快数百倍,这为研究大尺度城市街区洪水提供了一个比较有前景的研究途径。

Numerical Modeling of Flooding over Underground Staircases Using GPU-based SPH method

In order to better simulate and analyze the dynamic law of flooding flow over underground staircases, this paper applies the GPU-based Smoothed Particle Hydrodynamics (SPH) method to model underground flooding over staircases. The calculated results of the GPU-based SPH model indicate that: the proposed model can reproduce fairly good flow regimes of the staircase flow, and run with significantly high efficiency that is hundreds of times faster than the serial SPH model or the Moving Particle Semi-implicit (MPS) model, which provides a promising alternative tool for numerical modeling of large-scale urban flooding problems.