Hits:

Indexed by:Journal Papers

Date of Publication:2019-12-01

Journal:COMPOSITE STRUCTURES

Included Journals:EI、SCIE

Volume:229

ISSN No.:0263-8223

Key Words:Carbon Fibre Reinforced Plastics (CFRP); Porosity; Ultrasonic backscattered signal; Recurrence quantitative analysis (RQA); Real Morphology Void Model (RMVM); Void morphology

Abstract:The porosity characterization of Carbon Fibre Reinforced Plastics (CFRP) by ultrasonic attenuation measurement is restricted when the back-wall echoes are weak or absent at high porosity, and the estimation accuracy is decreased with increment of porosity due to complex void morphology. In this paper, Real Morphology Void Models (RMVMs) with different porosity were established to simulate the interaction between ultrasonic waves and voids with complex morphology. Subsequently, Recurrence Quantification Analysis (RQA) was introduced to process ultrasonic backscattered signals generated by voids so as to present Recurrence Plots (RPs) for porosity characterization. Recurrence Rate (RR) was selected as RQA variable to quantitatively analyze RPs, avoiding the determination of additional parameters. The correlation between porosity P and RR was revealed by simulation for the first time, and verified by experiments with CFRP laminates having porosity of 0.8%-4.2%. The complex and random void morphology induced the non-mapping relationship between P and RR, influencing estimation accuracy of porosity. The normalized results indicated that porosity characterization with RR was less affected by void morphology for P > 1.97% in comparison with ultrasonic attenuation measurement. It is concluded that RQA is valuable for porosity characterization of CFRP with high porosity and strong attenuation.