Chemical reaction impact on Buongiorno model for trihybrid nanofluid blood flow in a squeezed porous channel using the Levenberg–Marquardt neural network algorithm

This study explores Darcy Forchheimer tri-hybrid nanofluid (CuO+Au+Fe3O4) in blood for drug delivery function in a time dependent squeezed channel with the impacts of MHD, heat generation/absorption, non-linear thermal radiation. The novelty of the current study is to examine chemical reaction, Brownian and thermophoretic diffusions using the Buongiorno model. The range of CuOfrom 30 % to 50 %, range of Aufrom 20 % to 40 % and 20 % to 35 % for Fe3O4 respectively, have been selected to obtain optimal particles concentrations in ternary hybrid nanofluid. The inclusion of nanoparticlesCuO, Auand Fe3O4 in a base fluid blood predicts a potential reduction in blood viscosity by 10 % to 25 %, and a 15 % to 30 % enhancement in thermal conductivity compared to untreated blood. Also predicts a sustained drug release profile with an estimated delivery efficiency of 60 % to 80 % over a specified period. The tri-hybrid nanofluid increases the localized temperature from 1 °C to 3 °C within the targeted regions, demonstrating its potential for hyperthermia treatment in theoretical scenarios. The study employs a multilayer perceptron (MLP) neural network with Levenberg-Marquardt backpropagation, utilizing bvp4c numerical data, and incorporating physical constraints like chemical reactions, Prandtl number, and porosity. Furthermore, these parameters are for the range of different scenarios about neural network mapping and the solution. The benchmark datasets are allocated as 70 % for training, 15 % for testing, and 15 % for validation. The results demonstrated the model's effectiveness across all phases of evaluation, achieving highly accurate predictions with minimal error values. This research highlights the potential of tri-hybrid nanofluids for enhancing drug delivery and hyperthermia treatments, offering novel insights into fluid dynamics and biomedical applications.