Multi-physics study on double-diffusion convective flow in an inverted T-shaped porous enclosure: ANN-based parametric estimation

• Published in: Physics letters. A Vol. 518; p. 129681
• Main Authors: Kumar, Sumant, Murthy, S.V.S.S.N.V.G. Krishna, Kumar, B.V. Rathish, Parmar, Deepika
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 05.09.2024
• Subjects: Chemical reaction; Deep learning; Double-diffusion; Parameter-estimation; Porous media; Thermal radiation; Deep learning; Porous media; Parameter-estimation; Thermal radiation; Chemical reaction; Double-diffusion
• ISSN: 0375-9601; 1873-2429
• DOI: 10.1016/j.physleta.2024.129681

This study explored the multi-physics effects (chemical reaction, magnetic field, etc.) on double-diffusion natural convection within an inverted T-shaped porous enclosure saturated with hybrid nanofluid. A penalty FEM is employed for numerical simulation, whereas an ANN-based approach was proposed for efficient parameter estimation. Results show a significant reduction in computational cost using the ANN method. Streamlines, isotherms, and heat/mass transfer rates are analyzed for various flow parameters, which reveals that the higher buoyancy forces enhance convection while stronger magnetic fields suppress it. The influence of other flow parameters is complex and depends on specific parameter values. •A nonlinear mathematical model is proposed for investigating the multiphysics interactions in double-diffusive flow.•Developed nonlinear mathematical model is numerically simulated through penalty finite element approach.•Set of flow parameters identified for better convective fluid flow, heat, and mass transport characteristics.•The proposed ANN-based model efficiently predicts and saves computational time for parametric estimation.