Deep learning-based thermal modeling of slipping velocity in oblique magnetohydrodynamic non-Newtonian nanofluid flows
In recent years, artificial neural networks (ANNs) have emerged as reliable and efficient computational tools for solving highly nonlinear and complex mathematical models in fluid dynamics. Motivated by the limitations in existing studies on nanofluid flows under magnetic influence, this work invest...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | en |
| Published: |
Elsevier
2025
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/120652/1/120652.pdf http://psasir.upm.edu.my/id/eprint/120652/ https://linkinghub.elsevier.com/retrieve/pii/S2590123025026672 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In recent years, artificial neural networks (ANNs) have emerged as reliable and efficient computational tools for solving highly nonlinear and complex mathematical models in fluid dynamics. Motivated by the limitations in existing studies on nanofluid flows under magnetic influence, this work investigates the inclined magnetohydrodynamic (MHD) flow of a Carreau nanofluid incorporating cubic autocatalysis and homogeneous heterogeneous chemical reactions. The physical model considers temperature-dependent thermal conductivity and a magnetized environment to better capture real-world transport phenomena. A system of partial differential equations (PDEs) governing the flow, heat, and mass transfer is formulated and reduced to a set of ordinary differential equations (ODEs) using similarity transformations, with boundary conditions imposed over the interval [0, ∞). The resulting ODEs are solved using the Artificial Neural Network (ANN) approach, with the MATLAB-based bvp4c solver employed for validation. Statistical comparisons confirm the accuracy of ANN with a mean square error below 10^-5. The results reveal that an increase in the homogeneous reaction parameter enhances nanoparticle dispersion, thereby increasing the nanofluid temperature by approximately 18.6 % for the shear-thinning case and 22.3 % for the shear-thickening case. Moreover, the application of a stronger magnetic field intensifies the Lorentz force, leading to a 27 % reduction in velocity. The findings demonstrate the robust influence of chemical reactions and magnetic forces on heat and mass transport in non-Newtonian nanofluids, offering valuable insights for industrial and biomedical applications. |
|---|