For complex engineering and scientific applications, Computational Fluid Dynamics (CFD) simulations require a huge amount of computational power. As such, it is of paramount importance to carefully assess the performance of CFD codes and to study them in depth for enabling optimisation and portability. In this paper, we study three complex CFD codes, OpenFOAM, Alya and CHORUS representing two numerical methods, namely the finite volume and finite-element methods, on both structured and unstructured meshes. To all codes, we apply a generic performance analysis method based on a set of metrics helping the code developer in spotting the critical points that can potentially limit the scalability of a parallel application. We show the root cause of the performance bottlenecks studying the three applications on the MareNostrum4 supercomputer. We conclude providing hints for improving the performance and the scalability of each application.

No abstract

No abstract

The large disparity between the time and length scales of an acoustically active flow field, and the ones of the resulting generated acoustic field, is an issue in computational aeroacoustics (CAA). Numerical schemes used to calculate the time and space derivatives in CAA should exhibit a low dispersion and dissipation error. This paper presents the implementation of a high-order finite difference scheme to model CAA problems. The numerical scheme consists of a sixth-order prefactored compact scheme coupled with the 4-6 Low Dispersion and Dissipation Runge-Kutta (LDDRK) time marching scheme. Onesided explicit boundary stencils are implemented and a buffer zone is used to eliminate spurious numerical waves. Results are in good agreement with a 1-D advection equation benchmark problem. This constitutes a preliminary validation for the scheme to explore and investigate the acoustic propagation of sound generated aerodynamically.

A new family of prefactored cost-optimized schemes is developed to minimize the computational cost for a given level of error in linear wave propagation applications, such as aerodynamic sound propagation. This work extends the theory of Pirozzoli 1 to the prefactored compact high-order schemes of Hixon, 2 which are MacCormack type schemes that use discrete Padé approximations. An explicit multi-step Runge-Kutta scheme advances the states in time. Theoretical predictions for spatial and temporal error bounds are used to drive the optimization process. Theoretical comparisons of the cost-optimized schemes with a classical benchmark scheme are made. Then, two numerical experiments assess the computational efficiency of the costoptimised schemes for computational aeroacoustic applications. A polychromatic sinusoidal test-case verifies that the cost-optimized schemes perform according to the design highorder accuracy characteristics for this class of problems. For this test case, upwards of a 50% computational cost-saving at the design level of error is recorded. The final test case shows that the cost-optimized schemes can give substantial cost savings for problems where a fully broadband signal needs to be resolved.