Abstract
Abstract
To characterise complex non-Newtonian motion, the Williamson-Casson (WC) nanofluid flow past a curved, elongating sheet is considered, which is important for promoting heat transfer, nanotechnology, and controlling flow. It is used in advanced cooling systems, extrusion of plastics, bio-medical (blood flow simulation), and in the manufacturing industry. The present paper provides a comprehensive numerical exploration of the unsteady MHD flow of a WC nanofluid past a curved stretching sheet, using the Cattaneo-Christov heat flux model. The impact of ferro-hydrodynamics and the use of complex boundary conditions, such as melting, second order slip, and chemical reactions are also examined. The flow equations are reformed into a system of non-dimensional ODEs. The numerical solution is obtained through the parametric continuation method (PCM) to generate the date for artificial neural networking based (ANN) on the Levenberg-Marquardt back-propagation algorithm (LMBPA). To ensure the legitimacy of the numerical outputs, the results are compared with published findings, and reveals the percent error up to 0.00059480 % by taking
Rd
= 0.3 (thermal radiation parameter). From the graphical results, it can be observed that the skin friction drops up to 9.89189 % and 48.0315 %, whereas the energy transfer rate enhances by up to 21.6389 % and 41.4158 % by varying the 1st order and 2nd order slip parameters from 0.4 to 0.7, respectively. The skin friction increases up to 27.311 %, while the Nusselt number drops by up to 8.2858 % by varying the ferrohydrodynamic interaction factor from 1.5 to 3.5. Furthermore, the ANN-LMBPA show a very steady training procedure as the magnitude of the gradient and adaptive learning rate essentially converged towards near-zero values (e.g., 9.9 × 10
−9
and 1 × 10
−9
) at the end of up to 597 epochs.