Damage Identification in CFRP/plywood Laminate Drilling Based on Digital Image Processing, Artificial Neural Network and Fuzzy Modelling
摘要
Hybrid structures made of carbon fiber-reinforced plastics (CFRP) find many applications due to their favorable properties (high strength with low weight). In contrast to hybrid metal-based CFRP laminates, there is a lack of research on hybrid laminates consisting of CFRP and plywood. Spallings and delamination of holes are the main disadvantages when drilling CFRP and wood-based composites. The CFRP/plywood hybrid laminates are widely used in areas that benefit from the obtained properties of these materials, that is, high strength and low density, in particular as a structural material for the production of finishing components of lightweight structures. In this study, a vision-based method for recognizing and evaluating burrs and delamination generated during drilling was developed based on threshold segmentation methods and using fuzzy Takagi–Sugeno–Kang (TSK) subtractive clustering. Digital images and selected time-domain measures of cutting torque and thrust force and they were adopted as input parameters in the TSK model. The proposed method was verified by full drilling experiments on CFRP/wood-based composite samples and the results were compared based on images of the holes obtained with a stroboscopic microscope. A good agreement was found between the results of manual measurement method and developed TSK-based model.