Very happy to share that our paper "Physics-based Machine Learning for Computational Fracture Mechanics" is now open-access online at Machine Learning for Computational Science and Engineering: https://doi.org/10.1007/s44379-025-00019-x
💡 What makes this different?
✅ A physics-based machine learning (ϕML) framework captures brittle and ductile fracture behavior, enhancing predictive accuracy and generalization.
✅ Thermodynamic consistency is ensured by embedding fundamental physical principles directly into the neural network architecture.
✅ Problem-specific retraining is eliminated, enabling adaptability to a wide range of boundary value problems while improving computational efficiency.
✅ Bridges machine learning and computational fracture mechanics by minimizing data dependency and ensuring physically consistent predictions.
Fadi Aldakheel, Elsayed Saber Elsayed, Yousef Heider, Oliver Weeger
#MachineLearning #FractureMechanics
#ComputationalMechanics #PhysicsBasedML
#AI #Research #Engineering
Institut für Baumechanik und Numerische Mechanik (IBNM)
