| Abstract Scope |
This study investigates the vulnerability of composite structures to low-velocity impacts, a prevalent source of internal damage often undetected by traditional analyses. Incidents such as tool drops and encounters with debris pose threats to material strength, necessitating a novel approach. Leveraging machine learning and data science tools, the research utilizes 150 experimental data points encompassing material configuration, temperature, and impact energy to predict crucial parameters, including dent depth, delamination length, and various mechanical properties. Building upon experimental insights, the research pivots towards predicting flexural failure modes using a random forest machine learning model. The employed model showcases exceptional efficacy, boasting an impressive 95% accuracy in foreseeing post-impact strength. The study discerns influential contributors to flexural failure, notably identifying bending type, temperature, and facesheet elastic modulus as key parameters. The model not only accurately predicts the residual bending after impact strength of specimens but also provides valuable insights into the flexural failure mode. Furthermore, this study opens new horizons for the improvement of composite structures. The predictive capabilities offer opportunities to optimize material selection and tailor structural performance based on specific application requirements. The integration of machine learning insights into composite engineering holds the potential to revolutionize the field, ushering in a new era of resilient, efficient, and cost-effective composite structures. |