Nondestructive evaluation of delamination cracks in textile composites by using 3-dimensional reconstruction from acoustical imaging - Université Polytechnique des Hauts-de-France Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Nondestructive evaluation of delamination cracks in textile composites by using 3-dimensional reconstruction from acoustical imaging

Résumé

An acoustical imaging technique performing 3-dimentional (3D) reconstruction and visualization of materials was developed and used to evaluate delamination cracks in textile composites. The method is based on 2-dimentional (2D) mechanical scan combined with ultrasonic A-scan to obtain 3D volumetric data. The use of focus transducer working at high frequency (20MHz) allows the imaging of curved interfaces and woven structure of the composite with high contrast and resolution. An efficient and fast algorithm was proposed and implemented for the reconstruction of surface objects from 3D volumetric data, determined by the time peak value of A-scan pulses. Such representation makes clear distinction and observation of crack defects, especially delamination, from regular pattern of the composite. 2D textile composite plate samples (20x10x2mm 3 in size) were used for the imaging before and after impact loading test. The results show in detail an internal structure of [(0°/90°)12K ,(+45°,-45°) NCF ,(-45°,+45°) NCF , (0°/90°)3K ]. The delamination fractures for samples after impact test can be precisely recognized and evaluated in their localization, size and shape.

Dates et versions

hal-03432285 , version 1 (17-11-2021)

Identifiants

Citer

Wei-Jiang Xu, Yao Xu, Sylvain Treutenaere, Franck Lauro. Nondestructive evaluation of delamination cracks in textile composites by using 3-dimensional reconstruction from acoustical imaging. 7th International Congress on Ultrasonics, ICU 2019, Sep 2019, Bruges, Belgium. pp.030001, ⟨10.1121/2.0001045⟩. ⟨hal-03432285⟩
38 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More