Quantitative characterization of out-of-plane fiber wrinkling in thick CFRP with Double-side inverse-variance weight-synthetic ultrasonic imaging

Fiber wrinkling-induced beam deviation and wave attenuation in thick carbon fiber reinforced polymer (CFRP) pose a great challenge to accurate and robust ultrasonic imaging of fiber-wrinkling. Addressing this challenge, based on finite element and experimental studies, a double-side inverse-variance weight-synthetic ultrasonic imaging method is developed featuring i) circular variance-based weight estimation for synthetic imaging and ii) circular-circular correlation-based evaluation of imaging quality. The phase-based B-scan images with upper and lower side scanning are segmented into multiples patches, whose circular variances of ply angles extracted with the structure tensor are reciprocal of the weights to synthesize the phases. The imaging quality is quantitatively characterized with the circular-circular correlation of ply angles between the imaging and the baseline for CFRP up to 10 mm thick and ply angles around 15°. Results prove the developed approach can achieve a circular-circular correlation coefficient ≥ 0.7631, compared to the single-side imaging only ≥ 0.3783.