Neural network – assisted analysis of magnetohydrodynamic Jeffrey nanofluid flow over a stretching angled wedge with couple stress and chemical reaction

This investigation explores the steady movement of magnetohydrodynamic Jeffrey nanofluid streaming past an angled wedge, incorporating couple stress, thermal radiation, and first-order chemical reaction effects within the Buongiorno two-phase framework. The governing partial differential system was reduced to ordinary differential equations via similarity transformations and solved numerically using the bvp4c routine in MATLAB. To authenticate the computational predictions, a feedforward neural network trained with the Levenberg-Marquardt backpropagation algorithm was deployed as an independent validation tool, where the numerical solutions served as target datasets for training, testing and regression analysis. Key parametric outcomes revealed that increasing the couple stress parameter from 0.1 to 2.1, attenuated the velocity field while substantially amplifying the skin friction coefficient. The Deborah number exhibited analogous decelerating behaviour, attributed to the elastic memory effects that generate internal resistive forces opposing the primary motion. Thermal field analysis indicated that thermophoresis and thermal radiation intensified the temperature distribution, whereas the thermal relaxation parameter demonstrated a mitigating influence. Notably, the Nusselt number experienced a pronounced enhancement of approximately 50 % when the radiation parameter increased from 0.1 to 2.5. The concentration profiles declined monotonically with increasing Brownian motion, chemical reaction, and Lewis number effects. The neural network validation achieved exceptional correlation coefficients exceeding 0.99999 across all datasets, with mean squared errors on the order of 10 −10 , confirming the numerical solver’s accuracy. These findings offer practical insights for thermal management systems, polymer processing operations, and microfluidic device design, where precise control of flow and heat transfer characteristics remains paramount for optimizing industrial performance.