Scientific Machine Learning for Coating Degradation and Corrosion Prediction: A Critical Review and Thermodynamic Research Perspective
A Thermo-Fractional Graph Phase-Field Neural Operator (TFG–PFNO) is proposed as a research framework that combines fractional memory, graph-based geometry, phase-field interface evolution, metriplectic dynamics, and calibrated uncertainty to preserve physical admissibility and propagate uncertainty to structural risk,...