Scientific Machine Learning and Data-Driven Modeling for Multiphysics Transport Phenomena and Heat Transfer

Multiphysics transport phenomena involving fluid flow, heat transfer, mass transfer, porous media, reactive transport, and complex materials play a vital role in numerous engineering and scientific applications, including energy systems, advanced manufacturing, aerospace, biomedical engineering, environmental processes, polymer processing, and thermal management technologies. These systems are governed by highly coupled nonlinear physical mechanisms involving magnetic fields, porous-medium resistance, viscous and Joule dissipation, Brownian motion, thermophoresis, radiation, phase change, and chemical reactions, presenting significant mathematical and computational challenges.Recent advances in machine learning, scientific machine learning (SciML), data-driven modeling, deep learning, neural operators, surrogate modeling, and physics-informed artificial intelligence have created new opportunities for modelling, predicting, and optimizing complex engineering systems. These emerging approaches complement conventional computational techniques by improving predictive capability, computational efficiency, parameter estimation, uncertainty quantification, and inverse modeling while maintaining consistency with governing physical principles. Among these methods, Physics-Informed Neural Networks (PINNs) have emerged as a promising framework for solving forward and inverse partial differential equations; however, they represent one of several rapidly developing approaches within the broader landscape of data-driven computational modeling.This Research Topic aims to provide a multidisciplinary platform for recent advances in machine learning, scientific machine learning, data-driven computational modeling, computational mathematics, and intelligent numerical methods for multiphysics transport phenomena and engineering systems. Contributions are encouraged on the mathematical modelling, numerical simulation, algorithm development, and experimental validation of complex transport processes involving Newtonian and non-Newtonian fluids, nanofluids, hybrid nanofluids, porous media, multiphase systems, magnetohydrodynamics, thermal sciences, reactive transport, and energy systems. Studies integrating artificial intelligence with traditional computational methods and comparing data-driven approaches with established numerical techniques are particularly encouraged.The Research Topic welcomes high-quality original research and review articles addressing theoretical developments, computational innovations, and practical engineering applications. Topics of interest include, but are not limited to:- Machine learning and data-driven modeling for engineering and physical systems- Scientific Machine Learning (SciML) for computational mechanics and transport phenomena- Physics-Informed Neural Networks (PINNs) for forward and inverse problems- Deep learning, neural operators, surrogate models, and hybrid AI approaches- Mathematical modelling of fluid flow, heat transfer, and mass transfer- Non-Newtonian fluids, nanofluids, hybrid nanofluids, and biofluids- Magnetohydrodynamics (MHD), porous media, and reactive transport- Thermal and solutal stratification, radiation, phase change, and multiphase transport- Brownian motion, thermophoresis, chemical reactions, and coupled transport phenomena- Boundary-layer theory, similarity transformations, and nonlinear mathematical models- Numerical methods including finite difference, finite element, finite volume, spectral, mesh-free, and lattice Boltzmann methods- Comparative assessment of machine learning, PINNs, neural operators, and conventional numerical techniques- Model accuracy, convergence, stability, robustness, uncertainty quantification, and explainable AI- Optimization, parameter estimation, reduced-order modeling, digital twins, and intelligent computational frameworks- Applications in energy systems, renewable energy, advanced manufacturing, aerospace, biomedical engineering, environmental sciences, and industrial thermal systemsThis Research Topic seeks to advance the integration of machine learning, data-driven modeling, scientific computing, computational mathematics, and engineering sciences by promoting innovative methodologies for modelling, simulation, prediction, optimization, and decision-making across a broad spectrum of multiphysics transport and thermal engineering applications. Within this broader framework, Physics-Informed Neural Networks (PINNs) are welcomed as an important research direction alongside other emerging machine learning and data-driven computational approaches. Topic editors: Dr. Amit Kumar Gupta, Carlos M. Travieso-Gonzalez, Priya Mathur, Ibrahim Alade Source: https://www.frontiersin.org/research-topics/84408/scientific-machine-learning-and-data-driven-modeling-for-multiphysics-transport-phenomena-and-heat-transfer