Ingenieurwissenschaften und zugeordnete Tätigkeiten
Filtern
Dokumenttyp
- Zeitschriftenartikel (3)
- Posterpräsentation (2)
- Vortrag (1)
Sprache
- Englisch (6)
Schlagworte
- Image processing (6) (entfernen)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (1)
This article presents a method to use the dispersive behavior of ultrasonic guided waves and neural networks to determine the isotropic elastic constants of plate-like structures through dispersion images. Therefore, two different architectures are compared: one using convolutions and transfer learning based on the EfficientNetB7 and a Vision Transformer-like approach. To accomplish this, simulated and measured dispersion images are generated, where the first is applied to design, train, and validate and the second to test the neural networks. During the training of the neural networks, distinct data augmentation layers are employed to introduce artifacts appearing in measurement data into the simulated data. The neural networks can extrapolate from simulated to measured data using these layers. The trained neural networks are assessed using dispersion images from seven known material samples. Multiple variations of the measured dispersion images are tested to guarantee the prediction stability. The study demonstrates that neural networks can learn to predict the isotropic elastic constants from measured dispersion images using only simulated dispersion images for training and validation without needing an initial guess or manual feature extraction, independent of the measurement setup. Furthermore, the suitability of the different architectures for generating information from dispersion images in general is discussed.
Parameter studies are a common step in selecting process parameters for powder bed fusion of metals with laser beam (PBF-LB/M). Density cubes manufactured with varied process parameters exhibit distinguishable surface structures visible to the human eye. Industrial visual in-situ monitoring systems for PBF-LB/M currently have limited resolution and are incapable of reliably capturing differences in the surface structures. For this work, a 65 MPixel high resolution monochrome camera is integrated in an industrial PBF-LB/M machine together with a high intensity led bar. Post-exposure images are taken to analyze differences in light reflection on the specimen’s surface. The grey level co-occurrence matrix is used to quantify the in-situ measured visual surface structure of nickel-based super alloy IN939 density cubes. The properties of the grey level co-occurrence matrix correlate to the energy input and resulting porosity of specimens. Low energy samples with lack of fusion flaws show an increased contrast in the grey level co-occurrence matrix compared to specimens with an optimal energy input. The potential of high-resolution images as reference data in in-situ process monitoring in PBF-LB/M is discussed.
Additive manufacturing (AM) in general and laser powder bed fusion (PBF-LB/M) in particular are becoming increasingly important in the field of production technologies. Especially the high achievable accuracies and the great freedom in design make PBF-LB/M interesting for the manufacturing and repair of gas turbine blades. Part repair involves building AM-geometries onto an existing component. To minimise the offset between component and AM-geometry, a precise knowledge of the position of the component in the PBF-LB/M machine is mandatory. However, components cannot be inserted into the PBF-LB/M machine with repeatable accuracy, so the actual position will differ for each part. For an offset-free build-up, the actual position of the component in the PBF-LB/M machine has to be determined. In this paper, a camera-based position detection system is developed considering PBF-LB/M constraints and system requirements. This includes finding an optimal camera position considering the spatial limitations of the PBF-LB/M machine and analysing the resulting process coordinate systems. In addition, a workflow is developed to align different coordinate systems and simultaneously correct the perspective distortion in the acquired camera images. Thus, position characteristics can be determined from images by image moments. For this purpose, different image segmentation algorithms are compared. The precision of the system developed is evaluated in tests with 2D objects. A precision of up to 30μm in translational direction and an angular precision of 0.021∘ is achieved. Finally, a 3D demonstrator was built using this proposed hybrid strategy. The offset between base component and AM-geometry is determined by 3D scanning and is 69μm.
During Rutherford cable production the wires are plastically deformed and their initially round shape is distorted. Using X-ray absorption tomography we have determined the 3D shape of an unreacted Nb3Sn 11 T dipole Rutherford cable, and of a reacted and impregnated Nb3Sn cable double stack. State-of-theart image processing was applied to correct for tomographic artefacts caused by the large cable aspect ratio, for the segmentation of the individual wires and subelement bundles inside the wires, and for the calculation of the wire cross sectional area and shape variations. The 11 T dipole cable cross section oscillates by 2% with a frequency of 1.24 mm (1/80 of the transposition pitch length of the 40 wire cable). A comparatively stronger cross sectional area variation is observed in the individual wires at the thin edge of the keystoned cable where the wire aspect ratio is largest.
During Rutherford cable production the wires are plastically deformed and their initially round shape is distorted. Using X-ray absorption tomography we have determined the 3D shape of an unreacted Nb3Sn 11 T dipole Rutherford cable, and of a reacted and impregnated Nb3Sn cable double stack. State-of-theart image processing was applied to correct for tomographic artefacts caused by the large cable aspect ratio, for the segmentation of the individual wires and subelement bundles inside the wires, and for the calculation of the wire cross sectional
area and shape variations. The 11 T dipole cable cross section oscillates by 2% with a frequency of 1.24 mm (1/80 of the transposition pitch length of the 40 wire cable). A comparatively stronger cross sectional area variation is observed in the individual wires at the thin edge of the keystoned cable where the wire aspect ratio is largest.