Title: Application of a parallel solver to the LES modelling of turbulent buoyant flows with heat transfer

Authors: Ilyas Yilmaz; Hasan Saygin; Lars Davidson

Addresses: Department of Mechanical Engineering, Faculty of Engineering and Natural Sciences, Istanbul Bilgi University, Eyup, 34060, Istanbul, Turkey ' Department of Mechanical Engineering, Faculty of Engineering, Istanbul Aydin University, Florya, 34295, Istanbul, Turkey ' Division of Fluid Dynamics, Department of Applied Mechanics, Chalmers University of Technology, SE-412 96 Gothenburg, Sweden

Abstract: An existing fully implicit, non-dissipative direct numerical simulation (DNS) algorithm is reformulated to utilise the sub-grid scale (SGS) models in large eddy simulation (LES). The Favre-filtered equations with low-Mach number scaling are derived. The wall-adapting local eddy-viscosity (WALE) is used as SGS model. A fully parallel, finite volume solver is developed based on the resulting LES algorithm using PETSc library and applied to buoyancy- and thermally-driven transitional/turbulent flows in Rayleigh-Taylor instability and turbulent Rayleigh-Bénard convection. Results verify that the proposed low-Mach number LES approach, which is physically more accurate than pure incompressible methods for flows with variable properties, perfectly captures the evolution and complex physics of turbulent buoyant flows with or without heat transfer by taking the effects of density and viscosity changes into account without the Oberbeck-Boussinesq (OB) assumption even at large temperature differences with uniform accuracy and efficiency.

Keywords: large eddy simulation; LES; low-Mach; variable-density; wall-adapting local eddy-viscosity; WALE; Rayleigh-Bénard convection; RBC; Rayleigh-Taylor instability; RTI.

DOI: 10.1504/PCFD.2018.090338

Progress in Computational Fluid Dynamics, An International Journal, 2018 Vol.18 No.2, pp.89 - 107

Received: 21 Oct 2015
Accepted: 01 Sep 2016
Published online: 13 Mar 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article