Artificial intelligence‐assisted fatigue fracture recognition based on morphing and fully convolutional networks

Abstract

Fatigue fracture is one of the most common metallic component failure cases in manufacturing industries. The observation on fractography can provide direct evidence for failure analysis. In this study, an image semantic segmentation method based on fully convolutional networks (FCNs) was proposed to figure out the boundary between fatigue crack propagation and fast fracture regions from optical microscope (OM) fractography images. Furthermore, a novel morphing‐based data augmentation method was adopted to enable few‐shot learning of sample images. The proposed framework can successfully segment two categories, namely, the crack propagation and fast fracture regions, thus differentiating the boundary of two regions in one image. This artificial intelligence (AI)‐assisted fatigue analysis architecture can complete the failure analysis procedure in 0.5 s and prove the feasibility of fatigue failure analysis. The segmentation accuracy of self‐developed network achieves 95.4% for the fatigue crack propagation region, as well as 97.2% for the fast fracture region. Not only for semantic segmentation DNN, we also prove that our novel data augmentation method can applied at the instance segmentation DNN, such as mask regional convolutional neural network (mask R‐CNN), one state‐of‐the‐art deep learning network for instance segmentation, to achieve similar accuracy.

Deep learning networks for microbiology images recognition

“…However, we can draw a conclusion that each of them has its own strength and weakness under different circumstances, which is inevitable. A new method is supposed to be introduced which can combine strengths and avoid weakness [14][15][16].…”

Exploration of continual learning based on the comparison and analysis of LWF and EWC

“…Furthermore, the LWF needs additional networks to assist in the construction of new models, it will occupy more space when the quantity of tasks is large. The above shortcomings may be addressed in future research using CNN based networks [11][12][13][14][15] with a pyramidal structure for feature extraction.…”

Fatigue evolution prediction for fiber‐reinforced plastics based on frequency‐wavenumber wavefield of guided wave using deep‐learning model

“…However, in recent years, the rapid development of artificial intelligence (AI) technology has provided a promising avenue for handling these fatigue-related problems, such as structural fatigue-resistance design, 21 structural property evaluation, 22 and fatigue fracture recognition. 23 In order to characterize the fatigue state of composites, many feature extraction methods based on deep learning models can be considered, such as the deep neural network (DNN), 22 the convolutional neural network (CNN), 24 and the autoencoder (AE). 25 The notable advantage of the aforementioned methods is the ability to learn the intrinsic features through multi-layered neural networks.…”

“…Due to the complex fatigue behavior exhibited by FRPs, 20 addressing numerous engineering issues related to structural fatigue poses inherent challenges. However, in recent years, the rapid development of artificial intelligence (AI) technology has provided a promising avenue for handling these fatigue‐related problems, such as structural fatigue‐resistance design, 21 structural property evaluation, 22 and fatigue fracture recognition 23 . In order to characterize the fatigue state of composites, many feature extraction methods based on deep learning models can be considered, such as the deep neural network (DNN), 22 the convolutional neural network (CNN), 24 and the autoencoder (AE) 25 .…”