A Novel CFD-ANN Framework for the prediction of Real Time Flow Features.

This project introduces a hybrid Computational Fluid Dynamics (CFD) and Artificial Neural Network (ANN) model for predicting drag and lift coefficients in real-time for non-Newtonian fluid flow over a backward-facing step. Time-dependent CFD simulations were conducted by varying Reynolds numbers and flow behavior indices. The generated data was used to train an ANN, which achieved near-CFD accuracy while drastically reducing prediction time. This approach has strong potential for real-time decision-making in fluid-based systems.

Project Poster