Large Scale Simulations of Realistic Flow Problems

Iglberger K, Götz J, Donath S, Feichtinger C, Rüde U (2009)


Publication Type: Journal article

Publication year: 2009

Journal

Book Volume: 7

Pages Range: 40-45

Journal Issue: 2

URI: http://inside.hlrs.de/assets/pdfs/inside_autumn09.pdf

Abstract

The simulation of many realistic engineering problems requires enormous computing power to cope with the complexity of the simulated scenarios. Especially fluid simulations are amongst the most compute intense engineering applications. In this article we present the state of the art of two currently studied engineering problems in the field of fluid simulation: the simulation of particulate flows and the simulation of free surface flows for the optimization of low temperature fuel cells. For the simulation of fluids there exist several methods. Our simulations are based on the Lattice Boltzmann Method (LBM) [5], a mesoscopic simulation approach that is very well suited for the simulation of complex geometries (as required for both problems) and that can be efficiently implemented for the purpose of large scale simulations. The basic approach of LBM is the discretization of the computational domain by a finite number of fixed lattice nodes that contain information about the fluid particles at a specific position at a certain time step. Our C++ library for LBM simulations is the waLBerla framework [2], a software package specifically designed for large scale LBM simulations for various CFD applications.

Authors with CRIS profile

Additional Organisation(s)

How to cite

APA:

Iglberger, K., Götz, J., Donath, S., Feichtinger, C., & Rüde, U. (2009). Large Scale Simulations of Realistic Flow Problems. Innovatives Supercomputing in Deutschland : inSiDE, 7(2), 40-45.

MLA:

Iglberger, Klaus, et al. "Large Scale Simulations of Realistic Flow Problems." Innovatives Supercomputing in Deutschland : inSiDE 7.2 (2009): 40-45.

BibTeX: Download