Sensitivity Analysis of Lift and Drag Coefficients for Flow over Elliptical Cylinders of Arbitrary Aspect Ratio and Angle of Attack using Neural Network

TL;DR

This paper employs neural networks trained on CFD data to efficiently analyze how lift and drag coefficients of elliptical cylinders vary with geometric and flow parameters, enabling rapid sensitivity analysis for engineering design.

Contribution

It introduces a neural network-based method for sensitivity analysis of lift and drag on elliptical cylinders, reducing computational costs compared to traditional CFD simulations.

Findings

Neural networks accurately predict lift and drag coefficients.

Sensitivity of coefficients to aspect ratio, angle of attack, and Reynolds number quantified.

Method applicable to various bluff body configurations.

Abstract

Flow over bluff bodies has multiple engineering applications and thus, has been studied for decades. The lift and drag coefficients are practically important in the design of many components such as automobiles, aircrafts, buildings etc. These coefficients vary significantly with Reynolds number and geometric parameters of the bluff body. In this study, we have analyzed the sensitivity of lift and drag coefficients on single and tandem elliptic cylinders to cylinder aspect ratios, angles of attack, cylinder separation, and flow Reynolds number. Sensitivity analysis with Monte-Carlo algorithm requires several function evaluations, which is infeasible with high-fidelity computational simulations. We have therefore trained multilayer perceptron neural networks (MLPNN) using computational fluid dynamics data to estimate the lift and drag coefficients efficiently. Line plots of the variations in lift and drag as functions of the governing parameters are also presented. The present approach is applicable to study of various other bluff body configurations.