In this study, the research investigates the prediction of fatigue life for Functionally Graded Materials (FGM) specimens comprising Polylactic acid (PLA) and Thermoplastic Polyurethane (TPU). For this, Machine learning (ML) techniques, including Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN) are utilized. A predictive in-house code is developed for each technique, thereby facilitating the fatigue performance of layered deposited specimens subjected to varying cyclic loadings. In order to verify the effectiveness of the ML technique, a comparative analysis among all is reported based on empirically determined fatigue life obtained values. RF is proven to be the most suitable technique with minimal error percentage in obtained results with optimally synchronized data sets in a minimum time frame. Subsequently, the application of ML in those predictions is reported for future aspects in augmenting the operational efficiency associated with fatigue life prediction.
Read full abstract