In this project, an innovative approach blends Artificial Neural Networks (ANN) with Finite Element Method (FEM)-based CFD simulations to predict drag and lift forces. By training ANN models on FEM datasets, we create a system that achieves near real-time predictions with high accuracy. This fusion, named FEM-NET, aims to transform traditional CFD workflows by cutting down computational time without compromising precision, paving the way for smarter and faster engineering designs.
Keywords: Artificial Neural Networks (ANN) ,Computational Fluid Dynamics (CFD) ,Hybrid Modeling (FEM-NET) ,Fluid Flow Simulation ,Time-Dependent Fluid Forces ,Numerical Simulation ,Physics-Informed Machine Learning ,Reduced Computational Cost
Tools: COMSOL Multiphysics (for FEM-based CFD simulations), MATLAB (optional: for preprocessing data handling or additional simulations),• Finite Element Method (FEM) for CFD simulation, Artificial Neural Networks (ANN) for force prediction,Data Generation from FEM simulations,Supervised Machine Learning (training ANN models),Regression Analysis (prediction of drag and lift forces),Error Metrics Calculation (MSE
Department: Department of Mathematics
Tools: COMSOL Multiphysics (for FEM-based CFD simulations), MATLAB (optional: for preprocessing data handling or additional simulations),• Finite Element Method (FEM) for CFD simulation, Artificial Neural Networks (ANN) for force prediction,Data Generation from FEM simulations,Supervised Machine Learning (training ANN models),Regression Analysis (prediction of drag and lift forces),Error Metrics Calculation (MSE
Department: Department of Mathematics
Project Poster
