Skip to main navigation Skip to search Skip to main content

AI-optimized thermodynamic boundary layer analysis of nanofluid coatings for hypersonic space plasma sheath mitigation

  • Umar Farooq*
  • , Chao Shen
  • , M. Mahtab Alam
  • *Corresponding author for this work
  • Harbin Institute of Technology
  • King Khalid University

Research output: Contribution to journalArticlepeer-review

Abstract

Hypersonic vehicles face significant challenges due to space plasma sheath effects, which cause communication blackout and thermal stresses on curved surfaces. This study addresses these issues by developing a computational model to optimize the thermodynamic and electromagnetic performance of a hybrid nanofluid coating, composed of graphene oxide, titanium dioxide, and 1-butyl-3-methylimidazolium tetrafluoroborate [BMIM][BF₄], on curved surfaces of hypersonic vehicles to mitigate space plasma sheath effects. The main goal is to enhance heat transfer and reduce electromagnetic wave. The novelty lies in coupling the Reiner-Philippoff hybrid nanofluid model with an artificial neural network using Levenberg–Marquardt backpropagation to analyze surface-plasma interactions. Governing equations in cylindrical coordinates model the flow, capturing mass, momentum, and energy transfer with viscous heating and radiative effects. These equations, solved numerically, describe the coating's interaction with the plasma sheath under magnetic fields and shear stress. The neural network analyses four scenarios and nine cases, using 300 grid points for velocity and temperature profiles and 120 for heat transfer rate analysis, with a dataset split of 80 % training, 10 % testing, and 10 % validation. Results show that increasing the curvature parameter (Λ = 2, 4, 6) raises the skin friction coefficient by 3.14 % to 15.92 %, while adjusting the radiation parameter (Rd = 0.5, 1, 1.5) enhances the Nusselt number by 25.23 % to 26.78 %. The hybrid nanofluid outperforms mono-nanofluid in velocity, and temperature at nanoparticle volume fractions of 0.01 and 0.02. The neural network achieves a coefficient of determination of 1.00, with mean squared error values of 2.0913e−10 for heat transfer and −1.6e−6 for heat transfer rate. These findings demonstrate the coating's potential to improve thermodynamic efficiency and mitigate plasma sheath-induced debonding.

Original languageEnglish
Article number109626
JournalInternational Communications in Heat and Mass Transfer
Volume169
DOIs
StatePublished - Dec 2025
Externally publishedYes

Keywords

  • Artificial neural network
  • Hypersonic vehicle surface
  • Nanoparticles
  • Reiner–Philippoff fluid
  • Space plasma
  • Thermodynamic analysis

Fingerprint

Dive into the research topics of 'AI-optimized thermodynamic boundary layer analysis of nanofluid coatings for hypersonic space plasma sheath mitigation'. Together they form a unique fingerprint.

Cite this