Thermal heat transfer analysis of trihybrid nanofluids using an intelligent Levenberg-Marquardt neural network (ANN-LMA) approach, with a focus on entropy generation, has been conducted. The flow equations were modeled in Cartesian coordinates and simplified using dimensionless variables. Partial differential equations were converted into ordinary differential equations through appropriate similarity transformations. These ordinary differential equations were then solved using the finite element method applied to a data set evaluated from (ANN-LMA) approach. This dataset can be input into MATLAB to generate predicted solutions for flow patterns. The ANN-LMA technique was employed to evaluate the efficiency of heat transfer characteristics for nanofluids in various scenarios. Incorporating carbon nanotubes (both single-wall (SWCNT) and multi-wall (MWCNT)) along with iron oxide in water, the study demonstrates their effectiveness in enhancing heat transfer. These nanofluids have broad industrial applications, such as in coolant enhancement, cancer therapy, and solar radiation management, and show promising results. This study specifically examines the flow properties of water-based CNT cross-trihybrid nanofluids over a convectively heated surface, leveraging their unique characteristics. Improvements in heat transfer are achieved through the introduction of dissipative heat, thermal radiation, and external heat sources or sinks. The performance of the computational solver was assessed using error histograms, regression analyses, and Mean Squared Error (MSE) results. The physical significance of the designed factors is graphically depicted and discussed in detail. It was found that radiative heat increases surface heat energy through substantial accumulation, thereby enhancing heat transfer properties, while dissipative heat, due to Joule dissipation and other external sources, significantly raises the fluid temperature.
Read full abstract