Wissenschaftliche Artikel der BAM
Filtern
Erscheinungsjahr
- 2023 (3)
Dokumenttyp
Sprache
- Englisch (3)
Schlagworte
- Artificial Intelligence (3) (entfernen)
Organisationseinheit der BAM
AbstractLaser beam welding has become widely applied in many industrial fields in recent years. Solidification cracks remain one of the most common welding faults that can prevent a safe welded joint. In civil engineering, convolutional neural networks (CNNs) have been successfully used to detect cracks in roads and buildings by analysing images of the constructed objects. These cracks are found in static objects, whereas the generation of a welding crack is a dynamic process. Detecting the formation of cracks as early as possible is greatly important to ensure high welding quality. In this study, two end-to-end models based on long short-term memory and three-dimensional convolutional networks (3D-CNN) are proposed for automatic crack formation detection. To achieve maximum accuracy with minimal computational complexity, we progressively modify the model to find the optimal structure. The controlled tensile weldability test is conducted to generate long videos used for training and testing. The performance of the proposed models is compared with the classical neural network ResNet-18, which has been proven to be a good transfer learning model for crack detection. The results show that our models can detect the start time of crack formation earlier, while ResNet-18 only detects cracks during the propagation stage.
It is shown that preconditioning of experimental X‐ray computed tomography (XCT) data is critical to achieve high‐precision segmentation scores. The challenging experimental XCT datasets and deep convolutional neural networks (DCNNs) are used that are trained with low‐resemblance synthetic XCT data. The material used is a 6‐phase Al–Si metal matrix composite‐reinforced with ceramic fibers and particles. To achieve generalization, in our past studies, specific data augmentation techniques were proposed for the synthetic XCT training data. In addition, two toolsets are devised: (1) special 3D DCNN architecture (3D Triple_UNet), slicing the experimental XCT data from multiple views (MultiView Forwarding), the i.S.Sy.Da.T.A. iterative segmentation algorithm, and (2) nonlocal means (NLM) conditioning (filtering) for the experimental XCT data. This results in good segmentation Dice scores across all phases compared to more standard approaches (i.e., standard UNet architecture, single view slicing, standard single training, and NLM conditioning). Herein, the NLM filter is replaced with the deep conditioning framework BAM SynthCOND introduced in a previous publication, which can be trained with synthetic XCT data. This leads to a significant segmentation precision increase for all phases. The proposed methods are potentially applicable to other materials and imaging techniques.
The present work is part of the AIFRI project (Artificial Intelligence For Rail Inspection), where we and our project partners train a neural network for defect detection and classification. Our goal at BAM is to generate artificial ultrasound and eddy current training data for the A.I. This paper has an exploratory nature, where we focus on the simulation of eddy current signals for head check cracks, one of the most important rail surface defects. The goal of this paper is twofold. On the one hand, we present our general simulation setup. This includes geometric models for head check cracks with features like branching and direction change, a model for the HC10 rail testing probe, and the configuration of the Faraday simulation software.
On the other hand, we use the Faraday software to simulate eddy current testing signals with a strong focus on the influence of the damage depth on the signal, while differentiating between different crack geometries. Here, we observe an early saturation effect of the test signal at a damage depth of 2 mm (at a crack angle of 25◦ to the surface). That is about 2 mm earlier than we would expect from measurements at a crack angle of 90◦. This behavior will be investigated further in a future paper. Finally, we interpolate the simulated signals in a two-step curve fitting process. With these interpolations we may generate eddy current test signals for any damage depth within the simulated range.