Open Access Open Access  Restricted Access Subscription Access

Internal Damage Prediction in CFRP Laminates Using BVID Profiles and Machine Learning

SAKI HASEBE, RYO HIGUCHI, TOMOHIRO YOKOZEKI, SHIN-ICHI TAKEDA

Abstract


The purpose of this study is to investigate the possibility of estimating impact information only from the surface profile of impacted CFRP laminates using machine learning. The low-velocity impact test was conducted to collect the data about barely visible impact damage (BVID) under various impact conditions: stacking sequence, impactor shape, and impact energy. A 3D optical profiler and an ultrasonic C-scanning system were utilized to measure external and internal damages. After feature engineering using only the specimen surface data, the machine learning models were able to estimate impactor shape with over 80% accuracy by a classification problem and over 90% by a regression problem, and delamination area and its length with over 70% R2, without significant outliers. It demonstrated that considering both local and total deformation that is unique to the low-velocity impact test is prominent in this estimation.


DOI
10.12783/asc36/35818

Full Text:

PDF

References


Dubary, N., Bouvet, C, Rivallant, S. and Ratifandrihana, L. 2018. “Damage tolerance of an impacted composite laminate,” Composite Structures, 206: 261–271.

Davies, G.A.O. and Olsson, R. 2004. “Impact on composite structures,” The Aeronautical Journal, 108(1089): 541–563.

Figure

. Importance after Random Forest: (a) Delamination area and (b) Delamination

(a)

(b)

Delaney, M.P., Fung, S.Y. and Kim, H. 2018. “Dent depth visibility versus delamination damage for impact of composite panels by tips of varying radius,” Journal of Composite Materials, 52(19): 2691–2705.

Baaran, J. July 2009. “Study on visual inspection of Composite Structures,” European Aviation Safety Agency, EASA research project/2007/3 Final report, DLR.

Chhabra, A., Bottcher, P. and Klaas, E. 2015. “‘GO/No-Go’ 3D digital dent-mapping,” presented at the 58th annual A4A NDT Forum, September 23-25.

Chen, C.-T. and Gu, G.X. 2019. “Machine learning for composite materials,” MRS Communications, 9(2): 556–566.

Takeda, S., Aoki, Y. and Nakamura, T. 2019. “Detailed evaluation of visible deformation in CFRP laminates subjected to out-of-plane impact loading,” presented at the 22nd International Conferences on Composites Materials, August 11-16.

Maugis, D. 2000. Contact, Adhesion and Rupture of Elastic Solids, Springer, pp. 203-343.


Refbacks

  • There are currently no refbacks.