Data-driven intelligent framework for energy optimization viscoelastic hybrid nanofluid over a permeable surface with heterogeneous-homogeneous reactions

Viscoplastic hybrid nanofluids (HNFs) have demonstrated significant potential as working fluids in technological and industrial applications, such as energy-effective heat exchangers, polymer extrusion, catalytic chemical reactors, and biomedical membrane transfer. Convective thermal conditions, reactive species movement, and yield stress rheology all concurrently control the overall process efficiency and product quality in these systems. The boundary layer flow literature has not adequately addressed the precise characterization of heat and mass transport behavior in such multi-physics arrangements over stretched surfaces embedded in porous media, despite their practical significance. For the first time, the two-dimensional Darcy-Forchheimer dissipative flow of a Papanastasiou-Bingham HNF with shape-specific CNT (platelet, shape factor 5.7) and Al₂O₃ (spherical, shape factor 3.0) nanoparticles distributed in dihydrogen as base fluid over a convectively heated stretching surface embedded in a porous medium was investigated in this work. Forchheimer inertia dissipation was directly incorporated into the energy equation to physically represent frictional heating in the porous matrix. The regulating flow equations as a set of partial differential equations (DEs) were transformed into ordinary DEs. An artificial neural network (ANN) was developed using eight different training functions for predicting flow, temperature, and concentration. I found that the mixed convection due to a H-H reaction amplified the flow and energy profile, while conflicting behavior was seen for concentration. The viscoplastic fluid containing hybrid nanoparticles uplifted the wall stress, heat, and mass fluxes. The temperature of Papanastasiou-Bingham HNF was much higher than Papanastasiou-Bingham fluid. Moreover, the wall shear stress increased with Darcy and Bingham factors.