Artificial intelligence for non-destructive testing of CFRP prepreg materials
- 93 Downloads
This paper presents a concept of the quality assurance for CFRP prepreg materials and focusses on the classification of thermographic images using convolution neural networks (CNNs). The method for non-destructive testing of CFRP prepreg materials combines a laser-triangulation sensor and an infrared camera to monitor both, the geometry and the impregnation of the prepreg material. The aim is to ensure a high material quality excluding any defective material in an early stage of the process chain of the production of CFRP components. As a result, the reliability of Automated-Fiber-Placement processes utilizing this previously tested material increases. Therefore, an artificial intelligence is set up to classify the thermal images of the CFRP material. Two different architectures of CNN are trained and validated with data sets consisting of thermal images of several prepreg materials and different material defects, such as geometric deviations and varying fiber-matrix-ratios caused by an incorrect impregnation. The CNNs are able to differentiate prepreg materials and to detect and classify certain material-independent defects for known and trained prepreg materials.
KeywordsArtificial Intelligence Automated-Fiber-Placement Defects Prepreg Quality assurance
The authors would like to thank the Federal state of Lower Saxony and the Volkswagen Foundation for funding the research project “Multi-Matrix-Prepreg”. For further information, visit the website www.hpcfk.de (Grand No. ZN3063).
- 1.Rudberg T, Neilson J, Henschied M, Cemenska J (2014) Improving AFP cell performance. In: SAE International Journal of Aerospace Manufacturing and Automated Fastening ConferenceGoogle Scholar
- 2.Alexandra K, Linb S, Brabandta D, Böhlkeb T, Lanzaa, G (2014) Quality control in the production process of SMC lightweight material. In: Proceedings of the 47th CIRP conference on manufacturing, 47: 772–777, https://doi.org/10.1016/j.procir.2014.01.138
- 3.Lukaszewicz D (2011) Optimization of high-speed automated layup of thermoset carbon fiber preimpregnates. Dissertation, University of BristolGoogle Scholar
- 4.Schulz M, Goldbach S, Heuer H, Meyendorf N (2011) Ein Methodenvergleich—ZfP an Kohlefaserverbundwerkstoffen mittels wirbelstrom- und ultraschallbasierender Prüfverfahren. DGZfP—JahrestagungGoogle Scholar
- 8.Zhang H et al (2018) A novel optical air-coupled ultrasound NDE sensing technique compared with infrared thermographic NDT on impacted composite materials. In: proc. SPIE 10661, thermosense: thermal infrared applications XL, 106610X, https://doi.org/10.1117/12.2303494
- 9.Schumacher D, Meyendorf N, Hakim I, Ewert U (2018) Defect recognition in CFRP components using various NDT methods within a smart manufacturing process. In: AIP conference proceedings. 44th annual review of progress in quantitative nondestructive evaluation, Vol. 37, https://doi.org/10.1063/1.5031521
- 14.Schmidt C, Denkena D, Hocke T, Völtze K (2017) Thermal imaging as a solution for reliable monitoring AFP processes. In: 3rd international symposium on automated composites manufacturing (ACM)Google Scholar
- 15.Weimer D, Scholz-Reiter B, Shpitalni M (2016) Design of deep convolutional neural network architectures for automated feature extraction in industrial inspection. In: CIRP Annals—Manufacturing Technology, Vol. 65, pp. 417–420. https://doi.org/10.1016/j.cirp.2016.04.072
- 16.Marani R, Palumbo D, Galietti U, Stella E, D’Orazio T (2016) Automatic detection of subsurface defects in composite materials using thermograpgy and unsurpervised machine learning. In: 8th IS. https://doi.org/10.1109/is.2016.7737471
- 17.He K, Zhang X, Ren S, Sun J (2016) Deep Residual Learning for Image Recognition. In: IEEE conference on computer vision and pattern recognition, pp. 770–778. https://doi.org/10.1109/CVPR.2016.90