Defect detection on RTM composite parts via robotic contact measurement system

Abstract

In this work a new approach to detect and to evaluate defects typical of carbon fiber parts manufactured via resin transfer molding (RTM) is presented. The approach is based on a robotic contact measurement system, as an alternative or combined solution to vision methods. The research is meant to provide more accurate local geometry information to the global information acquired using vision systems. The system consists of a collaborative robot equipped with a force-torque sensor that scans the surface of the workpiece with a defined motion strategy. To evaluate the system, tests were conducted on a 3D printed part presenting a feature representative of the defects found on carbon fiber components. The results are compared with the CAD model and data acquired using a structured light scanner, showing that robotic contact measurement can be a viable method to enhance the reconstruction of a part with a higher accuracy. The system is then applied and validated on a real part made with RTM.