Slip driven and Darcy–Forchheimer flow of Maxwell nanofluid with thermal radiation subjected to porous cylinder via machine learning approach

This analysis accentuates the importance of nanomaterial suspensions and their flow characteristics, particularly in the context of their real‐world appliances. Especially, the consumption of these materials in biomedical rheological models has gained substantial distinction. Further, nanofluids play a crucial role in cooling small electronic devices, such as heat exchangers, thermal energy storage, microchips, and similar components. Keeping the applications, the steady flow of magnetized Maxwell nanofluid inside Darcy porous cylinder has been investigated. Further, the influence of nonlinear thermal radiation, thermal conductivity, and chemical reaction in the presence of motile bacteria is taken into account. Similarity functions are implemented to reduce the nonlinear PDEs of the cylindrical model into ODEs. The machine learning feed forward neural network (MLFFNN) technique is employed to improve the accuracy of simulations that effectively optimize the fluid flow models. Employing a newly developed methodology, MLFFNN can be successfully effective in fluid flows. Our suggested technique addresses many facets of fluid issues by combining training, testing, and validation based on datasets obtained through the traditional bvp4c technique. The performance of the least mean square (LMS) neural network method (MLFFNN) is assessed using the regression plots, training state, performance plots, and mean square error. For every scenario of this study, the absolute error is around for skin friction, for Nusselt number, and for Sherwood number. In order to lower mean square error, the numerical solutions for the nonlinear fluid dynamics system were taken into consideration. The outcomes of involving parameters on physical quantities of interest and involved profiles are deliberated in form of graphs, tables, and literature. A higher Peclet number declines the mobile microbes concentration by approximately 10%–18% reflecting the dominance of advective transport over diffusion in microbial motility. This study demonstrates how a neural network (MLFFNN) may be used to treat complex problems of fluid dynamics issues more accurately and efficiently.