Strong Scaling of Numerical Solver for Supersonic Jet Flow Configuration

Carlos Junqueira-Junior,João Luiz F. Azevedo,Jairo Panetta,William R. Wolf,Sami Yamouni

DOI: https://doi.org/10.1007/s40430-019-2055-6

2020-03-19

Abstract:Acoustics loads are rocket design constraints which push researches and engineers to invest efforts in the aeroacoustics phenomena which is present on launch vehicles. Therefore, an in-house computational fluid dynamics tool is developed in order to reproduce high-fidelity results of supersonic jet flows for aeroacoustic analogy applications. The solver is written using the large eddy simulation formulation that is discretized using a finite-difference approach and an explicit time integration. Numerical simulations of supersonic jet flows are very expensive and demand efficient high-performance computing. Therefore, non-blocking message passage interface protocols and parallel input/output features are implemented into the code in order to perform simulations which demand up to one billion degrees of freedom. The present work evaluates the parallel efficiency of the solver when running on a supercomputer with a maximum theoretical peak of 127.4 TFLOPS. Speedup curves are generated using nine different workloads. Moreover, the validation results of a realistic flow condition are also presented in the current work.

Computational Engineering, Finance, and Science,Distributed, Parallel, and Cluster Computing

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to simulate supersonic jet configurations by developing and optimizing a high - performance computational fluid dynamics (CFD) tool in order to study its unsteady property fields, especially acoustic phenomena. These phenomena are crucial to rocket design because the high noise levels generated during rocket launches may damage the launch structures or be reflected onto the aircraft itself and the equipment on its top, resulting in structural vibrations and acoustic stress damage. Therefore, understanding and predicting these acoustic loads are very important for the design of large - scale launch vehicles.

To achieve this goal, the author has developed a numerical tool - JAZzY based on the large - eddy simulation (LES) method for performing time - dependent supersonic jet simulations. The LES method can provide high - fidelity numerical data and is suitable for aero - acoustic applications. However, LES simulations require a large amount of computing resources. Therefore, the paper also evaluates the parallel efficiency of this solver on supercomputers, especially its performance when dealing with problems with up to one billion degrees of freedom. Specifically, the paper evaluates the strong scalability of the solver by generating speed - up ratio curves under different workloads and also shows the verification results under actual flow conditions.

**The main research contents include: **
1. **Numerical tool development**: Developed a CFD tool JAZzY based on LES, using the finite difference method for discretization and explicit time integration.
2. **Parallel computing optimization**: Implemented the non - blocking message passing interface protocol and parallel input / output functions to improve the efficiency of large - scale simulations.
3. **Performance evaluation**: On a supercomputer with a theoretical peak performance of 127.4 TFLOPS, the parallel efficiency of the solver was evaluated through nine different grid configurations with the number of grid points ranging from 5.8 million to 1 billion.
4. **Verification and analysis**: Showed the verification results under actual flow conditions and discussed the strong scalability of the solver.

Through these studies, the paper aims to provide an efficient and reliable numerical simulation tool for aero - acoustic problems in rocket design.