Ingenieurwissenschaften und zugeordnete Tätigkeiten
Filtern
Erscheinungsjahr
- 2021 (2) (entfernen)
Dokumenttyp
Sprache
- Englisch (2)
Referierte Publikation
- ja (2) (entfernen)
Schlagworte
- Artificial intelligence (1)
- Automated image analysis (1)
- Cracking (1)
- Electron microscopy (1)
- Hydrogen (1)
- Image segmentation (1)
- Neural networks (1)
- Offshore (1)
- Steel (1)
- Welding (1)
Organisationseinheit der BAM
Paper des Monats
- ja (2) (entfernen)
Offshore wind turbines (OWT) are a major goal of the energy strategy of Germany encompassing the increase of the installed wind power. OWT components are manufactured from welded steel plates with thicknesses up to 200 mm. The underlying standards and technical recommendations for construction of OWTs encompass specifications of so-called minimum waiting time (MWT) before non-destructive testing of the weld joints is allowed. Reason is the increased risk of time-delayed hydrogen assisted cold cracking as hydrogen diffusion is very slow due to the very thick plates. The strict consideration of those long MWT up to 48 h during the construction of OWTs leads to significant financial burden (like disproportionately high costs for installer ships as well as storage problems (onshore)). In this study, weld joints made of S355 ML were examined in comparison with the offshore steel grade S460 G2+M. The aim was to optimize, i.e., reduce, the MWT before NDT considering varied heat input, hydrogen concentration and using self-restraint weld tests. This would significantly reduce the manufacturing time and costs of OWT construction. To quantify the necessary delay time until hydrogen-assisted cold cracks appear, acoustic emission analysis was applied directly after welding for at least 48 h.
We present a workflow for obtaining fully trained artificial neural networks that can perform automatic particle segmentations of agglomerated, non-spherical nanoparticles from scanning electron microscopy images “from scratch”, without the need for large training data sets of manually annotated images. The whole process only requires about 15 minutes of hands-on time by a user and can typically be finished within less than 12 hours when training on a single graphics card (GPU). After training, SEM image analysis can be carried out by the artificial neural network within seconds. This is achieved by using unsupervised learning for most of the training dataset generation, making heavy use of generative adversarial networks and especially unpaired image-to-image translation via cycle-consistent adversarial networks. We compare the segmentation masks obtained with our suggested workflow qualitatively and quantitatively to state-of-the-art methods using various metrics. Finally, we used the segmentation masks for automatically extracting particle size distributions from the SEM images of TiO2 particles, which were in excellent agreement with particle size distributions obtained manually but could be obtained in a fraction of the time.