GPU-acceleration for Moving Particle Semi-Implicit method

Abstract The MPS (Moving Particle Semi-implicit) method has been proven useful in computation free-surface hydrodynamic flows. Despite its applicability, one of its drawbacks in practical application is the high computational load. On the other hand, Graphics Processing Unit (GPU), which was originally developed for acceleration of computer graphics, now provides unprecedented capability for scientific computations. The main objective of this study is to develop a GPU-accelerated MPS code using CUDA (Compute Unified Device Architecture) language. Several techniques have been shown to optimize calculations in CUDA. In order to promote the acceleration by GPU, particular attentions are given to both the search of neighboring particles and the iterative solution of simultaneous linear equations in the Poisson Pressure Equation. In this paper, 2-dimensional calculations of elliptical drop evolution and dam break flow have been carried out by the GPU-accelerated MPS method, and the accuracy and performance of GPU-based code are investigated by comparing the results with those by CPU. It is shown that results of GPU-based calculations can be obtained much faster with the same reliability as the CPU-based ones.

[1]  Diego Rossinelli,et al.  Vortex methods for incompressible flow simulations on the GPU , 2008, The Visual Computer.

[2]  J. Monaghan,et al.  Kernel estimates as a basis for general particle methods in hydrodynamics , 1982 .

[3]  Hitoshi Gotoh,et al.  Development of CMPS Method for Accurate Water-Surface Tracking in Breaking Waves , 2008 .

[4]  Hitoshi Gotoh,et al.  Turbulence particle models for tracking free surfaces , 2005 .

[5]  Hitoshi Gotoh,et al.  Lagrangian Particle Method for Simulation of Wave Overtopping on a Vertical Seawall , 2005 .

[6]  J. Monaghan.  Simulating Free Surface Flows with SPH , 1994 .

[7]  S. Shao,et al.  Corrected Incompressible SPH method for accurate water-surface tracking in breaking waves , 2008 .

[8]  Abbas Khayyer,et al.  Modified Moving Particle Semi-implicit methods for the prediction of 2D wave impact pressure , 2009 .

[9]  Weiguo Liu,et al.  Accelerating molecular dynamics simulations using Graphics Processing Units with CUDA , 2008, Comput. Phys. Commun..

[10]  Hitoshi Gotoh,et al.  Key issues in the particle method for computation of wave breaking , 2006 .

[11]  J. Kulpa,et al.  Time-frequency analysis using NVIDIA compute unified device architecture (CUDA) , 2009, Symposium on Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments (WILGA).

[12]  Hitoshi Gotoh,et al.  Enhancement of stability and accuracy of the moving particle semi-implicit method , 2011, J. Comput. Phys..

[13]  Diego Rossinelli,et al.  GPU accelerated simulations of bluff body flows using vortex particle methods , 2010, J. Comput. Phys..

[14]  Michael Garland,et al.  Efficient Sparse Matrix-Vector Multiplication on CUDA , 2008 .

[15]  Satoshi Matsuoka,et al.  Software-Based ECC for GPUs , 2011 .

[16]  P. J. Narayanan,et al.  Accelerating Large Graph Algorithms on the GPU Using CUDA , 2007, HiPC.

[17]  Ester M. Garzón,et al.  The sparse matrix vector product on GPUs , 2011 .

[18]  S. Shao,et al.  INCOMPRESSIBLE SPH METHOD FOR SIMULATING NEWTONIAN AND NON-NEWTONIAN FLOWS WITH A FREE SURFACE , 2003 .

[19]  Hitoshi Gotoh.  Lagrangian Particle Method As Advanced Technology For Numerical Wave Flume , 2009 .

[20]  Abbas Khayyer,et al.  A higher order Laplacian model for enhancement and stabilization of pressure calculation by the MPS method , 2010 .

[21]  S. Koshizuka,et al.  Moving-Particle Semi-Implicit Method for Fragmentation of Incompressible Fluid , 1996 .

[22]  Masashi Kashiwagi,et al.  A CIP-based method for numerical simulations of violent free-surface flows , 2004 .