Implementation of Deep Learning and Machine Learning Algorithms for Predicting the Stress Intensity Factor (KI) in Cracked Mechanical Components Deep Learning for Stress Intensity Factor Prediction

In Fracture Mechanics, predicting the stress intensity factor (SIF, Mode I, KI) is crucial, as it enables the assessment of a component’s resistance in the presence of cracks and, consequently, the prevention of catastrophic failures. The correlational study between machine learning (ML) and deep learning (DL) methodologies allowed us to demonstrate that crack length is closely related to SIF. However, the primary focus was to evaluate both methodologies and achieve a more precise prediction of KI. To this end, a numerical analysis was conducted using the finite element method (FEM)