Identification of Fatigue Crack Propagation in Open-Hole Plates Based on Ultrasonic Guided Waves and Hybrid Deep Learning

The damage identification method based on ultrasonic guided waves can be used for structural health monitoring, but it is affected by complex environmental and structural forms, and there is still insufficiency in the accuracy of damage diagnosis. For the common fatigue crack extension problem in open-hole plate structures in aviation, this paper proposes a real-time crack length prediction method based on ultrasonic guided waves and a hybrid deep learning model. First, the damage features are automatically extracted from the guided wave signals by a one-dimensional convolutional neural network, then a bidirectional long short-term memory network is used to establish a mapping relationship between the damage features and the crack length, and finally, the hybrid deep learning model is trained to predict the crack length. The effectiveness of the method was verified by the fatigue crack extension and monitoring test on an open-hole aluminum alloy plate. The results show that the proposed method can accurately predict the fatigue crack extension length, with an average prediction error of only 0.31 mm, and has higher identification accuracy than other single deep learning prediction models.