Vision-Based Defect Measurement of Drilled CFRP Composites Using Double-Light Imaging

Abstract

Drilling-induced defects in carbon fiber-reinforced polymers (CFRP) composite laminate can lead to the risk of aircraft structural failure and even aviation disaster. Measuring these defects is, then, inevitable for aerospace vehicles assembly. Tear is a representative drilling-induced delamination defect. But the most widely used measurement method is still based on human vision and vernier calipers. With low measurement efficiency and large measurement error, this is no longer suitable for industrialized large-scale production. Machine vision provides a new technology to address this problem. However, the semi-specular and textured surface around the drilled holes in CFRP laminates makes machine vision-based tear measurement difficult. To address these issues, we propose a novel automatic tear measuring system to obtain the tear point with maximum radial length (TPMRL). The method consists of three parts: (1) a double-light imaging frame is designed to illuminate the object hole under two different lighting conditions and acquire images; (2) a series of image processing procedures are proposed to accurately obtain inner hole information and tear edge information; (3) combined processing of inner hole information and tear edge information is performed to obtain the maximum radial length of the tear area (MRLTA). The experimental results show that the proposed system is able to measure the MRLTA and is superior to some existing relative approaches.