Parallel Performance of a Discontinuous Galerkin Spectral Element Method Based PIC-DSMC Solver

Ortwein, Philip; Binder, Tilman; Copplestone, Stephen M.; Mirza, A.; Nizenkov, P.; Pfeiffer, Marcel; Stindl, T.; Fasoulas, Stefanos; Munz, Claus‐Dieter

doi:10.1007/978-3-319-10810-0_44

Cited by 9 publications

(3 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on Aleph, researchers have performed large-scale DSMC/PIC simulations with up to billions of simulation particles and thousands of cores [15], [16]. Other works [17] use the parallel-coupled PIC/DSMC algorithm to simulate the active plasma flow and chemical reactions in a rarefied condition. The work presented in [18] implemented a parallel high-order DSMC/PIC solver using MPI to simulate the 250ps diffusion process of a laser-driven plasma plume.…”

Section: Related Workmentioning

confidence: 99%

Parallelizing and Balancing Coupled DSMC/PIC for Large-scale Particle Simulations

Qiu

et al. 2022

2022 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

View full text Add to dashboard Cite

Particle simulations are important workloads in high performance and parallel computing. Due to massive particle migration between parallel processes during simulations, efficient and balanced parallel computing is a practical challenge in large-scale realistic particle applications. This paper proposes a novel approach to enable highly-efficient dynamic load balance in a coupled DSMC/PIC solver for large-scale numerical simulations of the plasma plume. We employ dual unstructured grids of different granularity, with a coarse grid for DSMC simulations of flow fields and an embedded fine grid for PIC simulations of electric fields, to facilitate coupled DSMC/PIC calculation and grid partition for parallel computing. We then design and implement a centralized as well as a distributed communication strategies to dynamically migrate particles among arbitrary parallel processes. During the timestep iterations, we present a lightweight dynamic load balancer, composed of a load imbalance factor, a weighted load model and an efficient grid remapping mechanism, to adaptively rebalance the simulation among parallel processes with little extra overheads. We perform 3D unsteady simulations of the plasma plume induced by the pulsed vacuum arc in a cylindrical nozzle with hydrogen atoms and ions to validate the coupled solver. Parallel performance results scaling up to thousands of processes with billions of particles demonstrate the efficiency and effectiveness of our dynamic load balancer and parallel implementation.

show abstract

Section: Related Workmentioning

confidence: 99%

Parallelizing and Balancing Coupled DSMC/PIC for Large-scale Particle Simulations

Qiu

et al. 2022

2022 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

View full text Add to dashboard Cite

show abstract

“…with κ(ρ(k, n), l) = κ( k, l) given in (6). This does not introduce any further terms into the equation since | β kn (s)| = 1.…”

Section: Preprocess and Initializementioning

confidence: 99%

“…The reasons for this success are many [4,5]: an extremely flexible framework easily lending itself to many different types of equations, stability and conservation properties comparable to those of the finite volume method, natural support for high order discretizations and various types of adaptivity (h-, p-, r-) as well as for complex domain geometries, etc. Due to a favorable computation-to-communication ratio, the method also fits extremely well into the popular parallel and hybrid computational paradigms [6].…”

Section: Introductionmentioning

confidence: 99%

FESTUNG: A MATLAB/GNU Octave toolbox for the discontinuous Galerkin method. Part III: Hybridized discontinuous Galerkin (HDG) formulation

Jaust

Reuter

Aizinger

et al. 2018

Computers & Mathematics with Applications

View full text Add to dashboard Cite

The third paper in our series on open source MATLAB / GNU Octave implementation of the discontinuous Galerkin (DG) method(s) focuses on a hybridized formulation. The main aim of this ongoing work is to develop rapid prototyping techniques covering a range of standard DG methodologies and suitable for small to medium sized applications. Our FESTUNG package relies on fully vectorized matrix / vector operations throughout, and all details of the implementation are fully documented. Once again, great care is taken to maintain a direct mapping between discretization terms and code routines as well as to ensure full compatibility to GNU Octave. The current work formulates a hybridized DG scheme for a linear advection problem, describes hybrid approximation spaces on the mesh skeleton, and compares the performance of this discretization to the standard (element-based) DG method for different polynomial orders.

show abstract

Coupled PIC-DSMC Simulations of a Laser-Driven Plasma Expansion

Copplestone

Ortwein

Munz

et al. 2016

High Performance Computing in Science and Engineering ´15

View full text Add to dashboard Cite

Parallel Performance of a Discontinuous Galerkin Spectral Element Method Based PIC-DSMC Solver

Cited by 9 publications

References 8 publications

Parallelizing and Balancing Coupled DSMC/PIC for Large-scale Particle Simulations

Parallelizing and Balancing Coupled DSMC/PIC for Large-scale Particle Simulations

FESTUNG: A MATLAB/GNU Octave toolbox for the discontinuous Galerkin method. Part III: Hybridized discontinuous Galerkin (HDG) formulation

Coupled PIC-DSMC Simulations of a Laser-Driven Plasma Expansion

Contact Info

Product

Resources

About