Numerical investigation of thermal transport in magnetized Casson tri‐hybrid nanofluid with Cattaneo‐Christov heat flux model: Application of cascade neural network

Abstract This study conducts a comprehensive numerical analysis of thermal transport phenomena in a class of magnetohydrodynamic Casson tri‐hybrid nanofluid flows. The analysis considers the effects of nonlinear thermal radiation, heat generation, and convective boundary conditions. The Casson tri‐hybrid nanofluid (CTHNF) consists of CuO, TiO 2 , and SiO 2 nanoparticles dispersed in a blood‐based fluid. The Casson fluid model captures the shear‐dependent rheological behaviour of blood, while the Cattaneo–Christov heat‐flux (CCHF) theory addresses non‐Fourier thermal transport effects. Under suitable physical assumptions, the nonlinear partial differential equations for momentum and energy are transformed into a coupled system of ordinary differential equations using similarity transformations. A cascade neural network utilizing the Levenberg–Marquardt backpropagation algorithm (CaNN‐LMBA) is employed to predict fluid velocity and temperature. Training data sets are generated using the Lobatto IIIA method, which is known for its accuracy and stability with stiff, nonlinear systems. The findings show that increasing the Casson parameter enhances fluid flow while a higher Hartmann number reduces flow due to magnetic forces. The addition of nanoparticles increases both velocity and temperature. The radiation parameter, Biot number, and surface heating all contribute to higher temperatures. The CaNN‐LMBA model delivered accurate, consistent results with minimal errors and regression values close to one, demonstrating its robustness and reliability. Overall, this study demonstrates that the proposed work can be valuable in medical treatments, including cancer therapy, where controlled heating is essential. It can also help deliver medicine precisely to the targeted location in the body, as accurate temperature regulation is crucial in these processes.