Personal tools
You are here: Home Science & Projects Deisa Extreme Computing Initiative Projects 2009 - 2010 Simulation of transport phenomena in open-cell foams with a scalable lattice Boltzmann flow solver

Simulation of transport phenomena in open-cell foams with a scalable lattice Boltzmann flow solver

Project ScalLB
Research Area Engineering
Principal Investigator(s) Dr. Thomas Zeiser
Institution(s)
  • FAU Erlangen-Nuernberg, Germany
  • Max-Planck Institute for Dynamics of Complex Technical Systems, Magdeburg, Germany

Abstract

The aim of the present project is twofold: (1) Starting point is a challenging problem from chemical reaction engineering, i.e. the application of computational flow simulations to support the development of new chemical reactors. (2) These simulations are complemented by research from the area of computational science to investigate the characteristics and suitability of different high performance architectures for the applied highly parallel lattice Boltzmann flow solver.
In the chemical process industries, the majority of heterogeneously catalyzed gasphase reactions are carried out in tubular fixed-bed reactors that contain a so-called ’random packing’. However, owing to this inhomogeneous and incoherent geometrical structure, such conventional fixed-bed reactors exhibit several drawbacks that result from the non-uniform distribution of the velocity field and the concentration field, respectively. A promising alternative is the use of consolidated structures (e.g. ceramic foams or monoliths) as catalyst support. Despite their clear advantages, the application of consolidated catalytic supports is still limited to very few examples, because as of today, there is no quantitative understanding of the structural influence of e.g. the local foam structure on the fluid dynamics and the heat or mass transport (and thus on the reactor performance). The lattice Boltzmann flow simulations carried out within the present project will shed some light on local transport phenomena and their correlation to the local structure.
Parallel lattice Boltzmann codes have been used in e.g. the procurements of LRZ and HLRS to quantify performance characteristics. The computational science aspect of this project therefore is to continue with architecture specific optimizations of the applied lattice Boltzmann code and to compare architectural characteristics of the wide range of different systems available within DEISA.

Document Actions