Data-driven prediction of complex flow field over an axisymmetric body of revolution using Machine Learning

TL;DR

This paper introduces machine learning models trained on CFD data to rapidly and accurately predict flow fields over axisymmetric bodies of revolution, significantly reducing computational time while maintaining high accuracy.

Contribution

The study develops ML surrogate models for flow prediction over ABRs, demonstrating their efficiency and accuracy compared to traditional CFD methods.

Findings

ML models predict pressure, velocity, and turbulence energy accurately

Speed up of orders of magnitude over CFD simulations

Validated hyper-parameters ensure model reliability

Abstract

Computationally efficient and accurate simulations of the flow over axisymmetric bodies of revolution (ABR) has been an important desideratum for engineering design. In this article the flow field over an ABR is predicted using machine learning (ML) algorithms, using trained ML models as surrogates for classical computational fluid dynamics (CFD) approaches. The flow field is approximated as functions of x and y coordinates of locations in the flow field and the velocity at the inlet of the computational domain. The data required for the development of the ML models were obtained from high fidelity Reynolds stress transport model (RSTM) based simulations. The optimal hyper-parameters of the trained ML models are determined using validation. The trained ML models can predict the flow field rapidly and exhibits orders of magnitude speed up over conventional CFD approaches. The predicted results of pressure, velocity and turbulence kinetic energy are compared with the baseline CFD data, it is found that the ML based surrogate model predictions are as accurate as CFD results. This investigation offers a framework for fast and accurate predictions for a flow scenario that is critically important in engineering design.