Eventos Anais de eventos
EPTT 2024
14th Spring School on Transition and Turbulence
Prediction of Transition to Turbulence in Airfoils using Artificial Neural Networks
Submission Author:
Leandra Abreu , SP , Brazil
Co-Authors:
Leandra Abreu, Ivan Aldaya, Gabriel Pereira Gouveia da Silva
Presenter: Leandra Abreu
doi://10.26678/ABCM.EPTT2024.EPT24-0051
Abstract
Transition to turbulence in airfoils plays a critical role in aerodynamic performance and efficiency. This study explores the application of Artificial Neural Networks (ANNs) to predict laminar to turbulent transition in airfoils, using data generated by XFOIL software. XFOIL provides a robust framework for simulating aerodynamic behavior, generating a comprehensive database of airfoil characteristics under various flow conditions. Using this database, ANNs are trained to accurately predict the location of transition to turbulence based on input parameters such as airfoil geometry, Reynolds number, angle of attack and Mach number. The hyper-parameters used to training the ANNs were optimized using grid search. The trained ANNs demonstrate promising performance, achieving high accuracy in predicting transition points across a range of airfoil configurations. This research not only show the efficacy of ANNs in turbulence prediction but also underscores the potential for leveraging computational tools like XFOIL to enhance aerodynamic modeling and analysis.
Keywords
Artificial neural networks, Multilayer Perceptron, Turbulence transition, artifitial intelligence
