Artificial neural network based predictive modeling for entropy optimized flow of tri-hybrid nanomaterials invoking Darcy Forchheimer law

This research explores the entropy-optimized magnetohydrodynamic slip flow of a tri-hybrid nanofluid on a curved stretching surface with non-similarity conditions via an artificial neural network (ANN) predictive model with Levenberg–Marquardt algorithm. The governing model accounts for Darcy–Forchheimer effects, Joule heating, heat generation and radiation, and is numerically solved using the Bvp4c method in MATLAB. This data set is used to train the ANN model to predict the nonlinear transport properties. It is found that the magnetic parameter markedly decreases velocity (Lorentz force) and promotes the temperature field. The curvature parameter inhibits fluid flow but enhances heat transfer in the boundary layer. Moreover, increasing Brinkman number and radiation parameter leads to higher temperature and entropy generation, implying an increase in the degree of irreversibility. The results show good correlation between the ANN model and numerical results, validating the accuracy of the model. The ANN model demonstrates high predictive accuracy, with a mean squared error (MSE) of the order of 10 −5 and regression coefficients (R) approaching unity for training, validation, and testing datasets. The present study has significant applications in advanced materials processing and manufacturing technologies where precise thermal and flow control are crucial. The findings are also advantageous for the design of cooling systems in electronic manufacturing, where tri-hybrid nanofluids can enhance thermal conductivity and efficiency.