Peristaltic flow of electromagnetic tri-hybrid Carreau nanofluid using backpropagated Levenberg-Marquardt technique: an entropy generation analysis in blood cells.

The present research concentrates on examining entropy generation during the flow phenomenon of a three-dimensional peristaltic motion of a magnetized tri-hybrid nanofluid within a curved rectangular duct using a machine learning technique called backpropagated Levenberg-Marquardt (BLMT). The Carreau constitutive model is used for base liquid (blood). To obtain the most accurate solutions for the governing equations, an analytical tool called the Homotopy Perturbation Method (HPM) is utilized along with a machine learning methodology ANN-BLMT method on MatLab. The data of HPM and machine learning are also compared to assess how the framework of partial differential equations (PDEs) occurring in the problem can be improved. It shows the highest correlations between output and prediction of ANN-BLMT method. The convergence analysis reveals that for two scenarios, velocity exhibits the best validation performance values around $7.3117\times {10}^{-11}$ and $1.0082\times {10}^{-10}$. A detailed comparison between blood and nanofluid has been presented graphically to enhance the benefits of ternary hybrid nanoparticles in a simple base fluid. It is also found that the velocity of the blood can be slowed by the curvature increase and because of the increment of tri-hybrid nanoparticles in pure blood. It is also noted that the rate of heat transfer for ternary hybrid nanofluids is greater than that of a simple blood. Research findings have obvious implications for comprehending and enhancing peristaltic dynamics in biological processes such as the intestinal tract.