A new paradigm for wall-modeled large eddy simulations using the volume-filtering framework

TL;DR

This paper introduces a novel volume-filtered wall-modeled LES framework that avoids traditional assumptions, enabling more accurate and flexible simulations of wall-bounded turbulent flows with explicit volume-filtering and coupling with complex geometries.

Contribution

The paper proposes a new volume-filtered WMLES framework that eliminates the need for temporal averaging and assumptions on pressure gradients, improving wall flow modeling accuracy.

Findings

Accurately predicts mean velocity profiles in turbulent channel flow.

Validates the framework with flow over periodic hills.

Demonstrates compatibility with coarse Cartesian meshes.

Abstract

In the present paper, we apply the framework of volume-filtering for particle-laden flows, to large eddy simulations (LES) of wall-bounded flows leading to a new perspective on wall-modeled LES (WMLES) that we refer to as volume-filtered WMLES (VF-WMLES). In contrast to existing wall-models, the VF-WMLES framework does not rely on temporal averaging, does not make a priori assumptions on the pressure gradient, and can be used with a uniform spatial filter, which avoids the appearance of commutation closures in spatial derivatives of the filtered momentum and continuity equation. Volume-filtering is well-defined, even close to the wall, and it is shown that a non-zero slip and penetration velocity at the wall is a direct consequence of volume-filtering the flow. With the VF-WMLES concept, new wall models can be directly assessed in a priori and a posteriori studies by comparing the predicted slip and penetration velocities at the wall with the velocities from explicitly volume-filtered direct numerical simulations (DNS). Based on the VF-WMLES concept, we derive an LES modeling strategy that is based on the recently proposed PC-IBM, a modeling framework based on volume-filtering allowing to couple the flow with arbitrarily shaped solid boundaries using relatively coarse Cartesian fluid meshes. The proposed VF-WMLES is validated with two cases, a turbulent channel flow and a turbulent flow over periodic hills, and shown to accurately predict the mean velocity profiles for both cases.