Lab scale additive manufacturing of Fe-Nd-B based powders was performed to realize bulk nanocrystalline Fe-Nd-B based permanent magnets. For fabrication a special inert gas process chamber for laser powder bed fusion was used. Inspired by the nanocrystalline ribbon structures, well-known from melt-spinning, the concept was successfully transferred to the additive manufactured parts. For example, for Nd16.5-Pr1.5-Zr2.6-Ti2.5-Co2.2-Fe65.9-B8.8 (excess rare earth (RE) = Nd, Pr; the amount of additives was chosen following Magnequench (MQ) powder composition) a maximum coercivity of µ₀H_c = 1.16 T, remanence J_r = 0.58 T and maximum energy density of (BH)_max = 62.3 kJ/m³ have been achieved. The most important prerequisite to develop nanocrystalline printed parts with good magnetic properties is to enable rapid solidification during selective laser melting. This is made possible by a shallow melt pool during laser melting. Melt pool depths as low as 20 to 40 µm have been achieved. The printed bulk nanocrystalline Fe-Nd-B based permanent magnets have the potential to realize magnets known so far as polymer bonded magnets without polymer.

Numerous studies show that vat photopolymerization enables near-net-shape printing of ceramics and plastics with complex geometries. In this study, vat photopolymerization was investigated for cemented carbide specimens. Custom-developed photosensitive WC-12 Co (wt%) slurries were used for printing green bodies. The samples were examined for defects using quantitative microstructure analysis. A thermogravimetric analysis was performed to develop a debinding program for the green bodies. After sintering, the microstructure and surface roughness were evaluated. As mechanical parameters, Vickers hardness and Palmqvist fracture toughness were considered. A linear shrinkage of 26–27% was determined. The remaining porosity fraction was 9.0%. No free graphite formation, and almost no η-phase formation occurred. WC grain growth was observed. 76% of the WC grains measured were in the suitable size range for metal cutting tool applications. A hardness of 1157 HV10 and a Palmqvist fracture toughness of 12 $\mathrm{MPa}\sqrt{\mathrm{m}}$ was achieved. The achieved microstructure exhibits a high porosity fraction and local cracks. As a result, vat photopolymerization can become an alternative forming method for cemented carbide components if the amount of residual porosity and defects can be reduced.

As a result of their outstanding properties hard metals offer great potential in the tool industry. They are however expensive and geometrically limited when manufactured using conventional machining techniques. Additive manufacturing has the possibility to revolutionise the manufacture of such tools. The composition of hard metals consisting of a hard component within a binder phase in combination with laser based additive manufacturing can produce particular effects in the microstructural development of such materials. This article describes those particular microstructural developments which can be achieved in the powder metallurgical additive manufacturing of hard metals.

We examine the impact of water (160 ± 41 ppm of reference) on the anode, cathode, separator and electrolyte in two aging scenarios: calendric aging (60 °C, 80 days, charged state), resulting in a triggered current interrupt device (CID), and cycling 1680 times (charge/discharge with 1C, 2.75–4.2 V, 20 ± 2 °C), resulting in 24.5% residual capacity. We applied computer tomography (CT), scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS) and secondary ion mass spectrometry (SIMS) to understand the capacity loss. The aged NMC/LCO–graphite cells were compared to a reference cell in pristine state. Both aging scenarios showed (a) thick depositions on the anode and cathode consisting mainly of oxygen, fluorine and phosphorous, (b) reduced separator pore sizes, (c) the deposition of Mn, Co and Ni on top of the anode and (d) the decomposition of the conductive salt LiPF₆ accompanied by HF formation and a loss of active lithium. Calendric aging consumes the water content and additionally leads to (e) the decomposition of the organic solvent followed by CO₂ gas formation. Cyclic aging increases the water content and additionally results in (f) the consumption of the additive fluoroethylene carbonate (FEC). These findings show how water affects aging phenomena and results in a capacity decrease in the cell.

Training of semantic segmentation models for material analysis requires micrographs as the inputs and their corresponding masks. In this scenario, it is quite unlikely that perfect masks will be drawn, especially at the edges of objects, and sometimes the amount of data that can be obtained is small, since only a few samples are available. These aspects make it very problematic to train a robust model. Therefore, we demonstrate in this work an easy-to-apply workflow for the improvement of semantic segmentation models of micrographs through the generation of synthetic microstructural images in conjunction with masks. The workflow only requires joining a few micrographs with their respective masks to create the input for a Vector Quantised-Variational AutoEncoder (VQ-VAE) model that includes an embedding space, which is trained such that a generative model (PixelCNN) learns the distribution of each input, transformed into discrete codes, and can be used to sample new codes. The latter will eventually be decoded by VQ-VAE to generate images alongside corresponding masks for semantic segmentation. To evaluate the quality of the generated synthetic data, we have trained U-Net models with different amounts of these synthetic data in conjunction with real data. These models were then evaluated using real microscopic images only. Additionally, we introduce a customized metric derived from the mean Intersection over Union (mIoU) that excludes the classes that are not part of the ground-truth mask when calculating the mIoU of all the classes. The proposed metric prevents a few falsely predicted pixels from greatly reducing the value of the mIoU. With the implemented workflow, we were able to achieve a time reduction in sample preparation and acquisition, as well as in image processing and labeling tasks. The idea behind this work is that the approach could be generalized to various types of image data such that it serves as a user-friendly solution for training models with a smaller number of real images.

FastCAR is a novel task consolidation approach in Multi-Task Learning (MTL) for a classification and a regression task, despite the non-triviality of task heterogeneity with only a subtle correlation. The approach addresses the classification of a detected object (occupying the entire image frame) and regression for modeling a continuous property variable (for instances of an object class), a crucial use case in science and engineering. FastCAR involves a label transformation approach that is amenable for use with only a single-task regression network architecture. FastCAR outperforms traditional MTL model families, parametrized in the landscape of architecture and loss weighting schemes, when learning both tasks are collectively considered (classification accuracy of 99.54%, regression mean absolute percentage error of 2.4%). The experiments performed used "Advanced Steel Property Dataset" contributed by us https://github.com/fastcandr/AdvancedSteel-Property-Dataset. The dataset comprises 4536 images of 224x224 pixels, annotated with discrete object classes and its hardness property that can take continuous values. Our proposed FastCAR approach for task consolidation achieves training time efficiency (2.52x quicker) and reduced inference latency (55% faster) than benchmark MTL networks.