Neuro-computing prediction of thermal transfer rate in micropolar hybrid nanofluid flow past a stretching/shrinking surface: a Keller box numerical scheme

Purpose Micropolar fluids are used to simulate complex fluids that contain micro-structure, where the rotation of particles and micro-inertia effects play a significant role in the fluid flow. The purpose of this study is to illustrate the significance of micropolar fluid flow considering hybrid nanofluid mixture past a permeable stretching/shrinking surface incorporating heat source/sink and radiation. Moreover, the introduction of convective heating evidently improves the thermal transfer characteristics. In this model, two different nanoparticles, zinc oxide (ZnO) and copper (II) oxide (CuO), with the base fluid sodium alginate (SA) are considered. Design/methodology/approach The behaviour of the flow through stretching/shrinking surfaces is investigated by applying appropriate similarity transformations to convert the system of equations into non-dimensional form. Furthermore, robust numerical method, i.e. the Keller box method (KBM) is employed to solve these transformed equations. In addition, the artificial neural network (ANN) model was constructed using the radiation parameter, heat source parameter and Biot number as the input variables while the Nusselt number as the output variable and demonstrated high prediction accuracy with an R2 value of 0.99849. Findings The important findings indicate that nanoparticle concentration reduces fluid velocity profiles and notably enhances temperature distribution. However, the current research has numerous practical implications for advanced cooling technologies, where the development of nanofluids has substantially improved the heat transfer process across various applications in biological, physical sciences and engineering sector. Originality/value The novelty of this study is based on the investigation of micropolar hybrid nanofluid flow, which considers the effect of micro-rotation and micro-structure for accurate simulation of complex fluids. The non-linear boundary layer equations have been solved by employing an efficient KBM. In addition, ANN have been employed for accurate prediction of heat transfer performance, which improves computational efficiency.