Data-driven neurocomputational analysis of couple-stress bioconvective nanofluid transport with Arrhenius activation over a magnetically actuated Riga surface

The present study develops a computational framework for the analysis of multi-physical transport phenomena, thermal characteristics and bioconvection dynamics of a conducting fluid in contact with an elongating stretching surface coupled with a Riga plate electromagnetic actuator. The transport model also explains the localized MHD control effect through the Lorentz force, which is parallel to the wall and described by the exponential Grinberg term, while also taking into account the stable bioconvection field, which is affected by the microscopic floating of the motile microbes. In order to address the natural convergence and stability, the Parametric Continuation Method (PCM) is used to generate the dataset. The dataset is further solved through the Artificial Neural Network (ANN) based on the Levenberg-Marquardt backpropagation algorithm (LMBPO). The network is trained using the baseline dataset and is able to model the complex cross coupling behavior over a wide parametric range, including changes in the Hartmann and bioconvection Péclet numbers. The surrogate framework provides excellent predictive accuracy by having mean squared errors (MSE) close to zero and a correlation coefficient ( R ) close to on1. Ultimate validation shows that the ANN-LMBPO approach is capable of justifying the numerical instabilities that are typically seen in PCM during aggressive parametric sweep and is also able to instantaneously evaluate the critical engineering quantities like the skin friction coefficient, local Nusselt number and motile microbe’s density number. The energy transfer rate enhances up to 34.6496%, and 44.0358% by varying the mixed convection parameter from 0.5 to 1.5, and heat radiation factor from 0.1 to 0.5, whereas it drops up to 47.5205% with the effect of thermophoresis factor. The mass propagation increases up to 38.196%, 13.1664% and 19.1389% with the variation in mixed convection parameter from 0.5 to 1.5, thermal radiation from 0.1 to 0.5 and thermophoresis from 0.2 to 0.6, respectively.

Hali tarjima qilinmagan