Progressive augmentation of Reynolds stress tensor models for secondary flow prediction by computational fluid dynamics driven surrogate optimisation

Abstract

Generalisability and the consistency of the a posteriori results are the most critical points of view regarding data-driven turbulence models. This study presents a progressive improvement of turbulence models using simulation-driven surrogate optimisation based on Kriging. We aim for the augmentation of secondary-flow reconstruction capability in a linear eddy-viscosity model without violating its original performance on canonical cases e.g. channel flow. Explicit algebraic Reynolds stress correction models (EARSCMs) for $k-\omega$ SST turbulence model are obtained to predict the secondary flow which the standard model fails to capture. The optimisation of the models is achieved by a multi-objective approach based on duct flow quantities, and numerical verification of the developed models is performed for various test cases. The results of testing new models on channel flow cases guarantee that new models preserve the performance of the original $k-\omega$ SST model. Regarding the generalisability of the new models, results of unseen test cases demonstrate a significant improvement in the prediction of secondary flows and streamwise velocity. These results highlight the potential of the progressive approach to enhance the performance of data-driven turbulence models for fluid flow simulation while preserving the robustness and stability of the solver.

arXiv comment25 pages, 24 figures

References in corpus (14)

• Turbulence Modeling in the Age of Data
• Direct numerical simulation of turbulent channel flow up to $R{e}_{\tau }\approx 5200$
• A Physics Informed Machine Learning Approach for Reconstructing Reynolds Stress Modeling Discrepancies Based on DNS Data
• A Comprehensive Physics-Informed Machine Learning Framework for Predictive Turbulence Modeling
• Physics-Informed Machine Learning Approach for Augmenting Turbulence Models: A Comprehensive Framework
• RANS Turbulence Model Development using CFD-Driven Machine Learning
• Perspectives on Machine Learning-augmented Reynolds-averaged and Large Eddy Simulation Models of Turbulence
• Discovery of Algebraic Reynolds-Stress Models Using Sparse Symbolic Regression
• Numerical studies of space filling designs: optimization of Latin Hypercube Samples and subprojection properties
• Data-Driven Modelling of the Reynolds Stress Tensor using Random Forests with Invariance
• Multi-Objective CFD-Driven Development of Coupled Turbulence Closure Models
• CFD-driven Symbolic Identification of Algebraic Reynolds-Stress Models
• Deep Structured Neural Networks for Turbulence Closure Modelling
• Frozen propagation of Reynolds force vector from high-fidelity data into Reynolds-averaged simulations of secondary flows

Cited by in corpus (1)

• A progressive data-augmented RANS model for enhanced wind-farm simulations