Prediction of Non-Uniform Distorted Flows, Effects on Transonic Compressor Using CFD, Regression Analysis and Artificial Neural Networks

Non-uniform inlet flows frequently occur in aircrafts and result chronological distortions of total temperature pressure at the engine inlet. Distorted flow operation axial compressor deteriorates aerodynamic performance, which reduces stall margin increases blade stress levels, turn causes failure. Deep learning is an efficient approach to predict catastrophic failure, its stability for better performance minimum computational cost time. The current research focuses on development a transonic instability prediction tool comprehensive modeling dynamics. A novel predictive founded by extensive CFD-based dataset supervised has been implemented behavior different ambient temperatures conditions. Artificial Neural Network-based results accurately parameters minimizing Root Mean Square Error (RMSE) loss function. Computational show that, as compared tip radial distortion, hub distortion improved range compressor. Furthermore, combined effect with bulk qualitative deteriorator