Physics-Informed Surrogate Modelling of Concrete Self-Healing via Coupled FEM-ML with Active Learning

Key Points

• The research aims to develop a predictive framework for assessing the self-healing capabilities of microbial concrete.
• Combined finite element method (FEM) simulations with supervised machine learning (ML)
• Utilized a physics-informed approach to enhance model accuracy
• Implemented active learning to optimize the training of the predictive model
• The model successfully predicts self-healing performance with high accuracy
• Demonstrated enhanced reliability in simulations compared to traditional methods
• Highlighted the potential for improved durability in concrete applications