The Flow of Jeffrey Nanofluid through Cone-Disk Gap for Thermal Applications using Artificial Neural Networks

This study investigates the flow of Jeffrey nanofluid through the gap between a disk and a cone, incorporating the influences of thermophoresis and Brownian motion within the flow system. Suitable variables have used to convert the modeled equations to dimension-free notations. This set of dimensionless equations has then solved by using Levenberg Marquardt Scheme through Neural Network Algorithm (LMS-NNA). In this study, it has been observed that the absolute error (AE) between the reference and target data consistently falls in the range 10-4 to 10-5 demonstrating the exceptional accuracy performance of LMS-NNA. In all four scenarios it has noticed that transverse velocity distribution has declined with augmentation in magnetic and Jeffery fluid factors by keeping all the other parameters as fixed. It is evident that the optimal validation performance 2.8227×10-9 has been achieved at epoch 1000 for the transverse velocity when cone and disk gyrating in opposite directions.