EUVIMEDEUVIMED entdecken
Crossref

Hybrid Numerical–AI Framework for Three‐Dimensional Magnetohydrodynamic Nanofluid Flow Over a Biaxially Stretching Surface in a Darcy–Forchheimer Porous Medium With Activation Energy and Joule Heating

Mumtaz Khan, M. S. Anwar, Zakir Hussain, Taseer Muhammad, M. Irfan, Taoufik Saidani

ZAMM - Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik · 2026-11 · 106 (11)

DOI: 10.1002/zamm.70641 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ZAMM - Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik, online erschienen am 18.11.2026. Verfasst von Mumtaz Khan und 5 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reactive transport in electrically conducting hybrid nanofluids underpins a growing class of thermal‐management technologies, yet existing boundary‐layer models rarely capture the full interplay of magnetic braking, non‐Darcian porous resistance, finite‐rate surface chemistry, and convective wall transfer within a single self‐consistent framework. The present work addresses this gap by combining a high‐fidelity numerical scheme with an AI‐based surrogate to study the steady, three‐dimensional MHD mixed‐convection flow of an –/engine‐oil hybrid nanofluid over a biaxially stretching surface in a Darcy–Forchheimer porous medium. The formulation simultaneously retains nonlinear thermal radiation under the Rosseland approximation, temperature‐dependent thermal conductivity, viscous and Joule dissipation, modified‐Arrhenius chemical kinetics with finite activation energy, Brownian and thermophoretic species transport, frame rotation, and partial velocity slip. A Robin‐type convective condition is imposed on both the temperature and the species concentration at the wall, characterised by independent thermal and mass‐transfer Biot numbers, and , that permit the heat‐ and mass‐transfer interfaces to be modelled with comparable physical fidelity. The reduced boundary‐value problem is solved with the bvp4c collocation algorithm to six‐decimal‐place accuracy. Benchmark validation against published limiting cases agrees within . A scaled conjugate gradient artificial neural network () is then trained on the high‐fidelity dataset and achieves across all data splits, accelerating engineering‐quantity evaluation by roughly four orders of magnitude relative to the reference solver. The parametric analysis reveals several quantitatively significant trends. Increasing the magnetic parameter from to suppresses the primary and secondary wall velocities by approximately and , respectively. The disproportionate retardation of the secondary flow reflects the absence of a direct wall‐driving term in the cross‐stream momentum equation. The Forchheimer inertia correction elevates the wall skin friction by about over the linear‐Darcy prediction, demonstrating that quadratic drag cannot be neglected at moderate interstitial Reynolds numbers. Activation energy and reaction rate exert opposing, monotone influences on the Sherwood number, while the mass‐transfer Biot number governs the wall concentration and dominates the reactive transport response. Contour analysis over the plane delineates a diffusion‐dominated regime in which the Lewis number controls almost independently of activation energy, and provides a compact design chart for selecting the parameter combination that yields a target mass‐transfer rate. The combined numerical–AI framework therefore furnishes both physical insight and a fast, accurate surrogate suitable for inverse design, uncertainty quantification, and real‐time control of magneto‐convective thermal‐management devices.

Zitieren

APA

Khan, M., Anwar, M. S., Hussain, Z., Muhammad, T., Irfan, M., & Saidani, T. (2026). Hybrid Numerical–AI Framework for Three‐Dimensional Magnetohydrodynamic Nanofluid Flow Over a Biaxially Stretching Surface in a Darcy–Forchheimer Porous Medium With Activation Energy and Joule Heating. ZAMM - Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik, 106(11). https://doi.org/10.1002/zamm.70641

Vancouver

Khan M, Anwar MS, Hussain Z, Muhammad T, Irfan M, Saidani T. Hybrid Numerical–AI Framework for Three‐Dimensional Magnetohydrodynamic Nanofluid Flow Over a Biaxially Stretching Surface in a Darcy–Forchheimer Porous Medium With Activation Energy and Joule Heating. ZAMM - Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik. 2026;106(11). doi:10.1002/zamm.70641

BibTeX
@article{Khan2026,
  author = {Khan, Mumtaz and Anwar, M. S. and Hussain, Zakir and Muhammad, Taseer and Irfan, M. and Saidani, Taoufik},
  title = {Hybrid Numerical–AI Framework for Three‐Dimensional Magnetohydrodynamic Nanofluid Flow Over a Biaxially Stretching Surface in a Darcy–Forchheimer Porous Medium With Activation Energy and Joule Heating},
  journal = {ZAMM - Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik},
  year = {2026},
  volume = {106},
  number = {11},
  doi = {10.1002/zamm.70641}
}
Kostenlos starten →