Abstract

Carbon Fiber Reinforced Plastic (CFRP) has been increasingly utilized to repair damaged steel structures, and inspection of the resulted CFRP-reinforced steel structures is crucial but still challenging. Different physical characteristics of steel, CFRP and epoxy resin bring difficulties to accurate detection and clear visualization of internal defects in the hybrid structures. In this study, eddy current pulsed thermography (ECPT) was exploited to inductively heat the inspected structures and a stacked autoencoder (SAE) model was constructed to extract pixel-wise thermal features and generate images for visualization of internal defects. Experimental results on specimens with prefabricated debondings, delaminations and cracks validated efficiency of the proposed method. Clear visualization of defects provided by the proposed method is greatly beneficial for accurate detection and efficient information retrieval of internal defects in structures. In future studies, performance of the proposed method for other kinds of defects in CFRP-reinforced steel structures should be investigated.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call