Methodologies for RANS-LES interfaces in turbulence-resolving simulations
Doctoral thesis, 2017

Hybrid Reynolds-Average Navier-Stokes (RANS)-Large-Eddy Simulations (LES) techniques are now seen by the aeronautical industry as the most promising turbulence-resolving approaches for complex high-Reynolds-number turbulent flows with regard to cost and accuracy. However, connecting RANS and LES simulated flows is a challenging task and needs special attention in order to increase simulation robustness and accuracy. This thesis presents a new low-Reynolds-number k − ω based zonal hybrid RANS-LES turbulence model with which novel interface methodologies for connecting RANS and LES regions are explored. The emphasis of the work reported in the thesis is on RANS-LES interface techniques for grey-area mitigation and reduction of log-layer mismatch since these are the two most important issues to resolve in hybrid RANS-LES modeling in order to meet industry requirements for robustness and accuracy. The proposed hybrid RANS-LES model was applied to Decaying Homogeneous Isotropic Turbulence (DHIT) and channel flow for calibration purposes. It was concluded from the simulations of fully developed channel flow that a wall distance based LES length scale was superior in reducing the log-layer mismatch compared to other LES length scales discussed in the literature. It was shown from simulations of spatially developing boundary layer flow and flow over a wall-mounted hump that a RANS-LES interface technique combining commutation terms, introduced in the k and ω equations to reduce the turbulent viscosity, and synthetic turbulent fluctuations substantially mitigates the grey area as compared to commonly used RANS-LES interface methods and gives results that are in good agreement with experimental data. In simulations of a plane mixing layer flow, commutation terms were introduced in the k, ω and momentum equations in order to represent the transfer of energy between modeled and resolved turbulent scales at the  wall-normal RANS-LES interface, located at the trailing edge of the flat plate. It was found that the commutation terms in the momentum equations are able to trigger the equations to resolve turbulence, thus mitigating the RANS-to-LES transition region and improving the prediction of the resolved turbulent stresses. These were in good agreement with experimental data. Moreover, the proposed hybrid RANS-LES model has been successfully used to simulate a transonic flow in a rectangular duct with shock-induced corner flow separations.

zonal RANS-LES

aeronautics

RANS-LES interfaces

Hybrid RANS-LES

grey-area mitigation

log-layer mismatch

embedded LES

LES length scale

LES

HA2, Hörsalsvägen 4
Opponent: Apl. Prof. Dr.-Ing. habil. Suad Jakirlic, Department of Mechanical Engineering, Darmstadt University of Technology, Germany

Author

Sebastian Arvidson

Mechanics and Maritime Sciences (M2)

Hybrid RANS-LES Modeling Using a Low-Reynolds-Number k−ω Based Model

52nd AIAA Aerospace Sciences Meeting - AIAA Science and Technology Forum and Exposition, SciTech 2014, National Harbor, United States, 13-17 January 2014,; (2014)

Paper in proceeding

Feasibility of Hybrid RANS-LES Modeling of Shock/Boundary-Layer Interaction in a Duct

Notes on Numerical Fluid Mechanics and Multidisciplinary Design,; Vol. 117(2012)p. 345-356

Journal article

Prediction of Transonic Duct Flow Using a Zonal Hybrid RANS-LES Modeling Approach

Notes on Numerical Fluid Mechanics and Multidisciplinary Design,; Vol. 130(2015)p. 229-242

Journal article

Arvidson, S., Davidson, L., Peng, S-H., Grey-area mitigation using commutation terms at the interfaces in hybrid RANS-LES modeling

Arvidson, S. Davidson, L., Peng, S-H., Hybrid RANS-LES interface methods for grey-area mitigation in turbulence-resolving simulations

Flödessimuleringar är idag ett viktigt verktyg för design och analys av flygande farkoster. Trots att flödessimuleringar är vida använt inom flygindustrin behövs förbättrade metoder för att prediktera komplicerade turbulenta flöden för att erhålla högre noggrannhet och förbättrad robusthet.

Denna avhandling belyser användningen av avancerade turbulensmodelleringstekniker för industrirelevanta applikationer. Avhandlingen presenterar en ny turbulensmodell för prediktering av komplexa turbulenta flöden samt metoder för att kombinera olika turblensmodelleringstekniker med syfte att göra flödessimuleringar mindre beräkningsintensiva. De turbulensmodelleringstekniker som kombineras i denna avhandling är instationär Reynolds-Average Navier-Stokes (URANS) samt Large-Eddy simulation (LES).

Den föreslagna turbulensmodellen tillsammans med de föreslagna metodikerna för att kombinera URANS och LES ger resultat i bra överensstämmelse med data från experiment för olika typer av flöden. Vidare är de föreslagna metoderna av generel art vilket gör dem robusta och användbara i olika typer av applikationer.

Flow simulations are today the main tool for design and analysis of aerial vehicles. However, there is a need for improved prediction of turbulence in complex flow fields, for increased accuracy and robustness in aeronautical applications.

This thesis highlights the use of advanced turbulence modeling techniques for industrial relevant applications. The thesis, presents a new turbulence model for prediction of turbulence in complex flow fields and methodologies for combining different turbulence modeling techniques in order to reduce the computational cost. The turbulence modeling techniques combined in this thesis are unsteady Reynolds-Average Navier-Stokes (URANS) and Large-Eddy simulation (LES).

The proposed turbulence model with the methodologies for combining URANS and LES, give results in good agreement with experimental reference data for various kinds of flows. Moreover, the methodologies presented are of general type which makes them robust and applicable to various kinds of applications.

Subject Categories

Mechanical Engineering

Fluid Mechanics and Acoustics

Driving Forces

Sustainable development

Areas of Advance

Transport

Infrastructure

C3SE (Chalmers Centre for Computational Science and Engineering)

ISBN

978-91-7597-583-2

Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 4264

Publisher

Chalmers

HA2, Hörsalsvägen 4

Opponent: Apl. Prof. Dr.-Ing. habil. Suad Jakirlic, Department of Mechanical Engineering, Darmstadt University of Technology, Germany

More information

Latest update

10/19/2018