Abstract
Ultrasonic inspection technique has been widely used in defect detection of carbon fiber reinforced polymer (CFRP) materials. Although a variety of signal processing methods have been applied to highlight the defect features contained in ultrasonic signals, the most usual way to identify the defective regions is still not automatic, which is not only time-consuming but also critically dependent upon operator's performance and experience. In order to solve the problem, a feature-selective unsupervised clustering method which adaptively chooses the subset of the features is applied to ultrasonic data processing for the defect detection of CFRP. Compared with the conventional clustering without feature selection, it shows more accurate in size and location identification of the defects and, furthermore, provides the depth estimation of defects at same time. In order to demonstrate the feasibility of the proposed method, the experiment of the defect detection for the CFRP specimen with known artificial defects is conducted, resulting in better performances than other methods.