Skip to main content

A Physics-Informed Neural Networks Framework for Simulating Natural Convection of Cu-Water Micropolar Nanofluid in a Porous Enclosure with Internal Heat Generation

Submitted by vrh59ir on

“A Physics-Informed Neural Networks framework for simulating natural convection of Cu–water micropolar nanofluid in a porous enclosure with internal heat generation”

The paper has been published in the European Journal of Mechanics / B Fluids.

In this work, we develop a Physics-Informed Neural Network (PINN) framework for a strongly coupled heat-transfer and fluid-flow problem involving micropolar nanofluids in porous media. The approach combines Fourier feature encoding with adaptive loss weighting and is validated against meshless RBF-FD solutions. The results also provide insights into the effects of internal heat generation, porosity, vortex viscosity, and nanoparticle concentration on heat-transfer performance.

I am very happy to see this work published and grateful for the collaboration and effort behind it.

Authors: Marzieh Biglari and Vahid Reza Hosseini
DOI: https://doi.org/10.1016/j.euromechflu.2026.204633

Link: https://www.researchgate.net/publication/413763098_A_Physics-Informed_N…

#PhysicsInformedNeuralNetworks #PINNs #ComputationalMechanics #HeatTransfer #FluidMechanics #ScientificMachineLearning #PorousMedia