MACHINE LEARNING APPROACH TO AERODYNAMIC ANALYSIS OF NACA0005 AIRFOIL: ANN AND CFD INTEGRATION

Abstract

This study presents a machine learning approach to predict the unsteady aerodynamic performance of a NACA0005 airfoil. Data generated by computational fluid dynamics (CFD) is used to train the model for Reynolds numbers Re ∈ [1000 − 5000] and angles of attack ranging from 9◦ to 11◦. A robust Scaled Conjugate Gradient (SCG) algorithm is employed for efficient training of data. The ANN has a two-layer architecture, 9 fixed neurons in the first hidden layer and a varying number of neurons in the second layer to achieve optimal performance. The model yielded coefficients of determination (R 2 ) of 0.994 (Coefficient of lift (Cl)) and 0.9615 (Coefficient of drag (Cd )) for training, and 0.9563 (Cl) and 0.9085 (Cd) for testing. Overall mean errors are found to be less than 1%. It offers a powerful surrogate modelling approach for aerodynamic studies at ultra-low Reynolds numbers. Moreover, it provides rapid and reliable alternatives to traditional CFD simulations in aerodynamic analysis for unseen cases.