An Initial Attempt of Converged Machine-Learning Assisted Turbulence Modeling in RANS Simulations with Eddy-Viscosity Hypothesis

TL;DR

This paper introduces a converged machine-learning framework for turbulence modeling in RANS simulations, leveraging high-fidelity data and ensuring convergence through a closed-loop system, improving accuracy in turbulent flow predictions.

Contribution

It presents a novel, converged ML-assisted turbulence modeling framework that integrates traditional models and ensures solver convergence in RANS simulations.

Findings

The MLATM converges well to DNS data for mean velocity profiles.

The model accurately predicts turbulence viscosity profiles.

The framework ensures stable convergence in turbulent flow simulations.

Abstract

This work presents a converged framework of Machine-Learning Assisted Turbulence Modeling (MLATM). Our objective is to develop a turbulence model directly learning from high fidelity data (DNS/LES) with eddy-viscosity hypothesis induced. First, the target machine-learning quantity is discussed in order to avoid the ill-conditioning problem of RANS equations. Then, the novel framework to build the turbulence model using the prior estimation of traditional models is demonstrated. A close-loop computational chain is designed to ensure the convergence of result. Besides, reasonable non-dimensional variables are selected to predict the target learning variables and make the solver re-converge to DNS mean flow field. The MLATM is tested in incompressible turbulent channel flows, and it proved that the result converges well to DNS training data for both mean velocity and turbulence viscosity profiles.