Improving secondary flow prediction in turbulence models

Paper Title:

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

Published on:

24 August 2023

Primary Category:

Fluid Dynamics

Paper Authors:

M. J. Rincón,

A. Amarloo,

M. Reclari,

X. I. A. Yang,

M. Abkar

Bullets

Key Details

•

Uses CFD-driven optimization to improve turbulence model

•

Focuses on adding secondary flow prediction capabilities

•

Verifies models preserve performance on channel flows

•

Tests models on diverse unseen duct flow cases

•

Models show improved generalization and physics prediction

Explore the topics in this paper

cfd validation

computational fluid dynamics

model optimization

secondary flows

turbulence modeling

AI generated summary

Improving secondary flow prediction in turbulence models

This paper develops enhanced turbulence models that improve prediction of secondary flows, without compromising performance for simple canonical flows. Using computational fluid dynamics (CFD) simulations and optimization, the models augment a standard two-equation model to capture more complex physics. The progressive approach first focuses on adding secondary flow capabilities, then verifies on channel flows. Two models are developed and tested on various duct flows, showing improved generalization.