Simulation and prediction of MHD dissipative nanofluid flow on a permeable stretching surface using artificial neural network

• A similarity solution is implemented for a complicated problem of nanofluid flow.
• Simultaneous effects of surface stretching and magnetic field are considered.
• Injection or suction from the wall and viscous dissipation are taken into account.
• A multilayer artificial neural network is used to predict the desired parameters.
• The ANN predictions are well coincided with the numerical results.

This paper concerns with modeling nanofluid boundary layer flow in the presence of a magneto hydrodynamic field over a horizontal permeable stretching flat plate using artificial neural network. The flow is generated due to the linear stretch of the sheet. The governing PDEs are transformed into ODEs and numerically solved using a double precision Euler's procedure. We studied numerically the effects of injection or suction from the surface, the volume fraction of nanoparticles, the viscous dissipation, and the magnetic parameter on the skin friction factor, Nusselt number and hydraulic and thermal boundary layer thicknesses. The results show that both the friction factor and Nusselt number increase as the volume fraction of nanoparticles increases. Moreover, the magnetic field increases the friction factor while reduces the Nusselt number. By a multilayer neural network model, we also calculated the skin friction factor and Nusselt number with respect to the aforementioned effective parameters. The model is able to compute the test data set with mean relative errors of 0.19% and 0.36% for the friction factor and Nusselt number, respectively. This means the applied neural network model can accurately predict the output results.