A novel monitoring index for mechanical properties prediction in composite manufacturing based on electromechanical behavior and statistical approach

Abstract

Carbon-fiber-reinforced plastics (CFRPs) are commonly utilized in the aerospace and automotive industries owing to their high strength-to-weight ratios. Given the complexity of manufacturing CFRPs, several process monitoring techniques have been developed. This study investigated the correlation between monitoring data and mechanical properties of CFRP panels, aiming at a real-time prediction of the mechanical properties without post-evaluation testing. Electrical resistance changes were analyzed on the basis of electromechanical behaviors between the resin and carbon fibers. Process monitoring was conducted during the infusion and curing of 400 mm x 600 mm panels, similar size to those of industrial panels. The cured panels were cut into 26 specimens and tested for flexural strength using three-point bending. Monitoring index was introduced to establish a quantitative correlation between the monitoring data and mechanical properties of the CFRP panels. This index provides an integrated indicator for diagnosing quality at the section level within the panel. The mechanical properties were distributed based on the monitoring index, offering a probability-based evaluation. For robustness, the process monitoring results were incorporated into the index, imposing a penalty on the monitoring index for defects, particularly partial impregnation, based on their severity. The adjusted index predicted flexural strength with low root-mean-square error and mean absolute error values, indicating high reliability. Further, optical microscopy confirmed the effect of the void content on the mechanical properties. The proposed system can enhance the efficiency and reliability of composite manufacturing by reducing the reliance on post-processing in nondestructive evaluations.