Development of Parallel CFD Applications with the Chapel Programming Language

Parenteau, Matthieu; Bourgault-Cote, Simon; Plante, Frédéric; Kayraklioglu, Engin; Laurendeau, Éric

doi:10.2514/6.2021-0749

Cited by 6 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It involves computing the airflow around an aircraft, the impingement of the supercooled water droplets on the geometry, the thermodynamic exchanges and the ice growth. Numerical simulations are performed using an unstructured RANS software named CHApel Multi-Physics Simulation (CHAMPS-ICE) [4]. The latter is programmed using the Chapel programming language [5] and uses a cell-centered finite volume approach.…”

Section: Introductionmentioning

confidence: 99%

Recent developments of the multi-physics solver champs-ice

Zayni,

Blanchet,

Laurendeau

2023

Progress in Canadian Mechanical Engineering. Volume 6

View full text Add to dashboard Cite

This paper presents recent improvements in CHAMPS (Chapel Multi-Physics Software) developed at Polytechnique Montréal with emphasis on the aero-icing capabilities. This software, written in the Chapel language, is designed to simulate two-dimensional and three-dimensional multi-physics phenomena involving aerodynamics such as ice accretion and fluid-structure interactions through an Unstructured Finite-Volume Unsteady Reynolds-Averaged Navier-Stokes (URANS) realm. It is programmed using the open-source Chapel language, which facilitates parallel computing on laptops, desktops and High-Performance Computing (HPC) platforms. The basic approach to model ice shapes involves several modules with a segregated strategy: aerodynamic, droplet, thermodynamic and geometric. Four newly implemented state-of-the-art features expand the aero-icing suite, CHAMPS-ICE. A first feature is the implementation of a transitional turbulence model to enhance the physical representation of the boundary layer state near the airfoil (1). Furthermore, a local ice roughness model is added (2) to properly quantify the position of the laminar-turbulent transition which influences the surface convective heat transfer. As for the droplet module, it considers the first and second-order effects of Supercooled Large Droplets (SLD) such as droplet splashing and deformation (3). This extension brings the collection efficiency amplitude and impingement limits closer to experimental data. Another development addresses the stochasticity of the ice accretion process using an advancing front technique (4). Those newly implemented features are discussed and validated on 2D and 3D rime and glaze ice cases taken from the American Institute of Aeronautics and Astronautics (AIAA) Ice Prediction Workshops and the European Ice Genesis project.

show abstract

Section: Introductionmentioning

confidence: 99%

Recent developments of the multi-physics solver champs-ice

Zayni,

Blanchet,

Laurendeau

2023

Progress in Canadian Mechanical Engineering. Volume 6

View full text Add to dashboard Cite

show abstract

“…In various domains, empirical studies compare different parallel programming environments for a typical problem in the respective field of research. In [13], Parenteau et al develop Chapel-based alternatives for two CFD applications. These were compared to conventional MPI-based algorithms, and the competitiveness of Chapel is highlighted.…”

Section: Introductionmentioning

confidence: 99%

A performance-oriented comparative study of the Chapel high-productivity language to conventional programming environments

Helbecque

Gmys

Carneiro

et al. 2022

Proceedings of the Thirteenth International Workshop on Programming Models and Applications for Multicores and Manycores

View full text Add to dashboard Cite

The increase in complexity, diversity and scale of high performance computing environments, as well as the increasing sophistication of parallel applications and algorithms call for productivity-aware programming languages for high-performance computing. Among them, the Chapel programming language stands out as one of the more successful approaches based on the Partitioned Global Address Space programming model. Although Chapel is designed for productive parallel computing at scale, the question of its competitiveness with well-established conventional parallel programming environments arises. To this end, this work compares the performance of Chapel-based fractal generation on shared-and distributed-memory platforms with corresponding OpenMP and MPI+X implementations. The parallel computation of the Mandelbrot set is chosen as a test-case for its high degree of parallelism and its irregular workload. Experiments are performed on a cluster composed of 192 cores using the French national testbed Grid'5000. Chapel as well as its default tasking layer demonstrate high performance in shared-memory context, while Chapel competes with hybrid MPI+OpenMP in distributed-memory environment.

show abstract