A ghost fluid Lattice Boltzmann method for complex geometries

2011 ◽  
Vol 69 (2) ◽  
pp. 481-498 ◽  
Author(s):  
A. Tiwari ◽  
S. P. Vanka
2020 ◽  
Vol 31 (05) ◽  
pp. 2050070 ◽  
Author(s):  
Jean-Michel Tucny ◽  
David Vidal ◽  
Sébastien Leclaire ◽  
François Bertrand

Accurate imposition of boundary conditions (BCs) is of critical importance in fluid flow computation. This is especially true for the Lattice Boltzmann method (LBM), where BC imposition is done through operations on populations rather than directly on macroscopic variables. While the regular Cartesian structure of the lattices is an advantage for flow simulation through complex geometries such as porous media, imposition of correct BCs remains a topic of investigation for rarefied flows, where slip BCs need to be imposed. In this work, current kinetic BCs from the literature are reviewed for rarefied flows and an extended version of a technique that combines bounce-back and diffusive reflection (DBB BC) is proposed to solve such flows that exhibit effective viscosity gradients. The extended DBB BC is completely local and addresses ambiguities as regards to the definition of boundary populations in complex geometries. Numerical tests of a rarefied flow through a slit were performed, confirming the intrinsic second-order convergence of the proposed extended DBB BC. It settles a long-standing debate regarding the convergence of BCs in rarefied flows. Good agreement was also found with existing numerical schemes and experimental data.


2016 ◽  
Author(s):  
Kartik Jain

The present contribution reports direct numerical simulations of pulsatile flow through a 75% eccentric stenosis using the Lattice Boltzmann Method (LBM). The stenosis was previously studied by Varghese, Frankel, and Fischer in a benchmark computation, and the goal of this work is to evaluate the LBM and the solver Musubi for transitional flows in anatomically realistic geometries. A part of the study compares the LBM simulation results against the benchmark and evaluates the efficacy of most basic LBM scheme for simulation of such flows. The novelty lies in the computation of Kolmogorov micro-scales by performing simulations that consist of up to ∼ 700 million cells. Recommendations on the choice of spatial and temporal resolutions for simulation of transitional flows in complex geometries naturally arise from the results. The LBM results show an excellent agreement with the previously published results thereby validating the method and the solver Musubi for the simulation of transitional flows. The study suggests that with a prudent calibration of the parameters, the LB method, due to its simplicity and compute efficiency has advantages for the simulation of such flows.


Sign in / Sign up

Export Citation Format

Share Document