Stochastic numerical computing for hydro-magnetic flow of Carreau-nanofluid model

Abstract

This study focused on mass and heat transport in a time-dependent hydro-magnetic Carreau-nanofluid flow model (HF-CNM) over a radially stretched surface. The impacts of Brownian motion and thermophoresis on the fluid are analyzed through Buongiorno’s model. Using similarity transformation, the model’s controlling nonlinear PDEs are transformed into ODEs. These ODEs are utilized to get datasets using the Homotopy analysis methodology (HAM). By implementing Intelligent Backpropagation Networks with the Levenberg Marquardt scheme (IBPNs-LMS), the reference datasets are divided into testing, training, and validation to calculate and analyze the solution of HF-CNM. The performance analysis of the proposed IBPNs-LMS approach is validated by the MSE representation, the histogram of error, regression analysis, and transition measurements. Moreover, the effects of various variables such as magnetic parameter (M), local Weisenberg number (We), thermophoresis parameter (Nt), Brownian number (Nb), Eckert number (Ec), unsteadiness parameter (A), Schmidt number (Sc), chemical reaction parameter () and Power-law index (n) on velocity, skin friction, temperature, Nusslet number and concentration field are also investigated. Results reveal the velocity field declined as the values of We and M decreased, and the thickness of the velocity boundary layer is decreasing as well. The temperature field rises by increasing the values of Nb and Nt.