This study aims to develop high-order computational framework to model the steady flow and heat transfer characteristics of a Williamson nanofluid over an electrically driven Riga plate. Two nanofluid configurations, namely, Ag-based and ZrO2-based nanofluids with C2H6O2 as the base fluid, are considered. This study also assesses the capability of artificial neural networks (ANNs) in predicting the thermal performance of nanofluid systems.
The governing nonlinear partial differential equations describing the Williamson nanofluid flow are transformed into dimensionless ordinary differential equations using appropriate similarity variables. The resulting boundary value problem is numerically solved using the MATLAB bvp5c solver. An ANN model is coupled with the numerical results to predict the obtained data for both nanofluid scenarios.
The results of this study indicate that increasing radiative heat intensifies internal heating, making thermal radiation the dominant mode of heat transfer. The momentum diffusion associated with increased viscosity because of nanoparticle inclusion slows fluid motion and promotes heat retention. Furthermore, the Ag-based nanofluid exhibits more efficient heat transfer performance compared to the ZrO2-based nanofluid. The ANN demonstrates high accuracy in predicting the numerical results for both cases.
This study presents an original numerical and data-driven investigation of Williamson nanofluid flow over a Riga plate by combining a similarity-based numerical solution with ANN prediction. The comparative analysis of Ag and ZrO2-based nanofluids and validation of ANN accuracy provide valuable insights for improving thermal system performance and support the application of intelligent prediction techniques in advanced mining heat transfer systems.
