AI-Neural Networking Analysis (NNA) of Thermally Slip Magnetized Williamson (TSMW) fluid flow with heat source

Abstract

The thermal case study is conducted by using artificial intelligence to examine the heat transfer traits in Williamson fluid flow with heat source and thermal slip effects. The thermal flow field interacts with the externally applied magnetic field and velocity slip is additionally considered at the surface. The flow is formulated in terms of energy and momentum equations. All inputs (Prandtl number, magnetic field, thermal slip, and Weissenberg number) are represented in a 4 × 72 matrix and the 72 samples of Nusselt number are represented by a 1 × 72 matrix. The samples are randomly divided into three stages: 70%(50) for training, and 15%(11) each for validation and testing. The number of neurons is set to ten. To train the neural networking model, the Levenberg-Marquardt algorithm is employed. The best validation performance is noticed 5.9676e-08. This indicates that the neural network was successfully trained to predict NN at a magnetized surface. For heat generation and thermal slip parameters, the magnitude of Williamson fluid temperature is higher for the non-magnetic flow field. Further, the Nusselt number admits a declining trend for temperature slip, magnetic parameters, and Weissenberg number while it shows inciting values for the Prandtl number.

Keywords

• Williamson fluid
• Heat transfer
• Heat generation
• Nusselt number
• Neural networking