Modeling nanomaterial transport with chemical reaction and thermal radiation effects using intelligent learning techniques

Some of the key nanofluid properties that govern transport phenomena, including energy and mass transfer within the boundary layer and throughout the flow domain, are the effects of thermo-migration and the random motion of nanoparticles suspended in the base fluid. In this work, a trihybrid model comprising alumina, multi-walled carbon nanotubes, and graphene is used to explore the effects of magnetohydrodynamic flow, with C₃H₈O₂ as the base fluid in three distinct cases. This work presents the computational outcomes for the basic expression that models the dynamics of a colloidal suspension of C₃H₈O₂ with spherical Al₂O₃ particles, cylindrical MWCNTs, and platelet graphene nanoparticles under magnetic effects, considering Darcy–Forchheimer, activation energy, thermal radiation, viscous dissipation, Joule heating, and heat source scenarios. The BVP4C solver is used to generate the reference dataset for training the Bayesian regularization backpropagation neural network (BR-BANN) model, taking into account the variations in the model parameters. The results show that fluid velocity decreases as the magnetic field strength and porosity parameter rise, whereas the temperature field increases as the thermal radiation and heat source parameters grow. The model's excellent convergence and high accuracy are indicated by its low mean-squared error (MSE). Additionally, the regression coefficient's closeness to unity indicates a high degree of agreement between the numerical and predicted outcomes. The accuracy of the model is further confirmed by the absolute error analysis, which demonstrates the resilience of the suggested BR-ANN framework, with errors ranging from 10⁻¹¹ to 10⁻⁴ across all case studies.