PINN-driven heat transfer in tangent hyperbolic fluid with Cattaneo-Christov model over variable thick surface

Flow over stretching surface is of utmost importance, owing to its applications in solar energy, geothermal energy, polymer extrusion. Most of the fluids used in these sectors are non-Newtonian. The present study investigates the flow of a tangent hyperbolic fluid (a non-Newtonian fluid) over a stretching plate with variable thickness, with heat transfer modelled using Cattaneo-Christov (CC) heat flux model. The heating process is supported by a space and temperature-dependent heat source/sink. Modeling of the fluid flow over such a surface with stagnation point is achieved utilizing the boundary layer approximation, and solution is obtained reducing the governing PDEs to non-dimensional ODEs using similarity transformation. And, solution is obtained by implementing data free unsupervised physics informed neural networks (PINNs). The PINN is trained to minimize the loss function composed of physical loss (due to the differential equations) and the loss due to boundary conditions. The effectiveness of this approach is validated by comparing the results with previously published data for a few special cases. Computational results demonstrate a significant increase in the Local Nusselt number, with a 28.71 % rise as the power law index ( n ) increases from 0 to 1, and an additional 20.51 % increase as n further rises from 1 to 2. Furthermore, in the presence of a heat source, increasing τ enhances the heat transfer rate, indicating that CC model captures the non-Fourier heat transfer effects more effectively.