Surface transport characteristics of EMHD stratified Oldroyd-B flow in porous stretching processes with arrhenius chemical kinetics: A hybrid computational and machine learning framework

-Diverse applications of fluids contained in stratified environments are observed in engineering processes, such as reducing pressure and enhancing pumping efficiency in oil recovery processes, preparation of films used in food preservation, coating protection, plastic sheet extrusion, and targeted drug delivery and so forth. In view of the exemplary utilization of stratified fluids, the primary intent of the present study is to explore the attributes of viscoelastic material modelled through the Oldroyd-B formulation across a stretching surface immersed in a permeable medium encompassed by the Darcy model. Subsequently, the physical aspect of electric and magnetic fields is also accounted. Linear stratification is considered by producing the associated convective potential while considering different wall and ambient temperatures and concentration constraints. Enhancement of stratification is envisioned by including the propagation of gyrotactic microorganisms with activation energy aspects. The mathematical framework for the aforementioned physical factors is manifested through a differential setup by employing conservative formulations. Subsequently, similar solutions are obtained by implementing appropriate variables using the BVP4C approach. The results are computed in a comparative manner for stratified and non-stratified situations for associated distributions against sundry parameters. Fluxes noticing the surface drag, thermal and solutal gradients, and movement of microorganisms across the boundaries are also extracted. A machine learning approach based on intelligent neural networking was also used to validate and predict the numerical data extracted through numerical simulations. A comparative analysis of numerically computed results with ANN outcomes is presented in tabular manner. The thermal and solutal distributions were reduced against the associated stratification parameters. Furthermore, the Sherwood number declines when stratification is considered in the domain, whereas the opposite is observed in the case of the motile microorganism density flux coefficient. Learning of data for stratified environment is more efficient which is evidenced from magnitude of mean square 2.78 x 10 -9 in comparison to value of MSE (9.41 x 10 -10 ) attained for . A marginal elevation of up to 0.7% in the heat flux coefficient is observed against the radiation parameter when it is changed from 0 to 0.2, whereas an increment in Prandtl raises the heat transfer rate up to 40%. The consideration of Brownian movement and the thermophoretic phenomenon increased the Nusselt number by approximately 18 %.