Exploring nanoparticle dynamics in binary chemical reactions within magnetized porous media: a computational analysis

Area of Science:

• Computational fluid dynamics
• Heat transfer
• Nanofluidics

Background:

• Artificial neural networks (ANNs) excel at complex mathematical problems.
• Hybrid nanofluids offer enhanced thermal properties for various applications.
• Accurate modeling of heat transfer is crucial for engineering applications.

Purpose of the Study:

• To investigate the use of back-propagation ANNs with the Levenberg-Marquardt algorithm for heat transfer analysis.
• To computationally analyze the mixed convection flow of a MgO+GO/EG hybrid nanofluid over an exponentially stretching sheet.
• To evaluate the impact of slip conditions, heat generation, and thermal radiation on heat transfer.

Main Methods:

• Developed a computational model using ANNs and the Levenberg-Marquardt algorithm.
• Transformed partial differential equations into ordinary differential equations using similarity transformations.
• Generated benchmark datasets using the bvp4c method for training, testing, and validation of the ANN.
• Validated the ANN model using mean squared error, error histograms, and regression analysis.

Main Results:

• The ANN model demonstrated outstanding agreement with numerical results.
• The methodology effectively handles nonlinear problems in heat transfer.
• Flow properties (temperature, velocity, concentration) were accurately predicted.
• Graphical and numerical representations of flow characteristics were provided.

Conclusions:

• The ANN approach is a reliable tool for analyzing heat transfer in hybrid nanofluids.
• Understanding heat transfer in hybrid nanofluids is vital for applications in drug delivery, microelectronics, and nuclear cooling.
• The study highlights the practical importance of analyzing hybrid nanofluid behavior under various conditions.