Validation of LBM based on BGK on Poiseuille s Flow and Vortex Street in a Channel Flow

Abstract

The aim of research is to simulate a large scale bioreactor using Lattice Boltzmann method (LBM). Design and scale up a bioreactor is a highly complex process since it involves various physical phenomena occurring together. Computational Fluid Dynamics (CFD) has been developed over the years to simulate complex phenomenon occurring in a bioreactor and can make up for the shortcomings associated with traditional scale-up methods. It allows the simulation of all the relevant physics like species and thermal transport along with millions of bubbles. Due to various recent advances in numerical techniques and computer hardware (particularly the Graphical Processing Units-GPUs, aimed at number crunching applications), the simulation of an industrial scale reactor is now possible in reasonable time. While conventional pressure based solvers, such as the SIMPLE algorithm and its variants, do not scale well with large number of processors and may not utilize the GPUs cores effectively, Lattice Boltzmann Methods are highly parallelizable and scale almost linearly with number of cores. In this paper we present the development and validation of a basic LBM code which will be used for the bioreactor simulation. The Bhatnagar-Gross-Krook (BGK) model is used and is validated by theoretically computed parameters for Poiseuille's flow and with experimentally determined parameters for Vortex Street in a channel flow. Velocity profile, Strouhal number and vortex distance are compared. Reynolds numbers (Re) of 30, 60, 120, and 140 have been considered.