Filtern
Dokumenttyp
- Vortrag (8)
- Zeitschriftenartikel (3)
- Beitrag zu einem Tagungsband (2)
- Dissertation (1)
- Posterpräsentation (1)
Schlagworte
- Crystal plasticity (7)
- Additive manufacturing (6)
- Anisotropy (6)
- Scale-bridging (4)
- Virtual experiments (4)
- Structural Health Monitoring (3)
- Structural health monitoring (3)
- COPV (2)
- Damage localization (2)
- Geführte Ultraschallwellen (2)
Organisationseinheit der BAM
- 5 Werkstofftechnik (8)
- 5.2 Metallische Hochtemperaturwerkstoffe (8)
- 8 Zerstörungsfreie Prüfung (8)
- 8.4 Akustische und elektromagnetische Verfahren (7)
- 5.5 Materialmodellierung (6)
- 3 Gefahrgutumschließungen; Energiespeicher (1)
- 3.5 Sicherheit von Gasspeichern (1)
- 5.1 Mikrostruktur Design und Degradation (1)
- 8.5 Röntgenbildgebung (1)
- 8.6 Faseroptische Sensorik (1)
Eingeladener Vortrag
- nein (8)
The underlying cause of mechanical anisotropy in additively manufactured (AM) parts is not yet fully understood and has been attributed to several different factors like microstructural defects, residual stresses, melt pool boundaries, crystallographic and morphological textures.
To better understand the main contributing factor to the mechanical anisotropy of AM stainless steel 316L, bulk specimens were fabricated via laser powder bed fusion (LPBF).
Tensile specimens were machined from these AM bulk materials for three different inclinations relative to the build plate. Dynamic Young's modulus measurements and tensile tests were used to determine the mechanical anisotropy. Some tensile specimens were also subjected to residual stress measurement via neutron diffraction, porosity determination with X-ray micro-computed tomography, and texture analysis with electron backscatter diffraction (EBSD). A crystal plasticity model was used to analyze the elastic anisotropy and the anisotropic yield behavior of the AM specimens, and it was able to capture and predict the experimental behavior accurately. Overall, it was shown that the mechanical anisotropy of the tested specimens was mainly influenced by the crystallographic texture.
Virtual-lab-based determination of a macroscopic yield function for additively manufactured parts
(2019)
This work aims for a yield function description of additively manufactured parts of S316L steel at the continuum-mechanical macro-scale by means of so-called virtual experiments using a crystal plasticity (CP) model at meso-scale. Additively manufactured parts require the consideration of the specific process-related microstructure, which prevents this material to be macroscopically treated as isotropic, because of crystallographic as well as topological textures. From virtual experiments, yield loci under various loading conditions are simulated. The scale bridging from meso- to macro-scale is realised by the identification of the simulated yield loci as a modified anisotropic Barlat-type yield model representation.
Virtual-lab-based determination of a macroscopic yield function for additively manufactured parts
(2019)
This work aims for an yield function description of additively manufactured (AM) parts of S316L steel at the continuum-mechanical macro-scale by means of so-called virtual experiments using a crystal plasticity (CP) model at meso-scale. Additively manufactured parts require the consideration of the specific process-related microstructure, which prevents this material to be macroscopically treated as isotropic, because of crystallographic as well as topological textures.
EBSD/CT-Scans from in-house additively manufactured specimen extract the unique microstructural topology which is converted to a representative volume element (RVE) with grain structure and crystal orientations.
Crystal plasticity model parameters on this RVE are calibrated and validated by means of mechanical testing under different texture angles. From virtual experiments on this RVE, yield loci under various loading conditions are simulated. The scale bridging from meso- to macro-scale is realised by the identification of the simulated yield loci as a modified anisotropic Barlat-type yield model representation.
Structural Health Monitoring (SHM) mittels geführter Ultraschallwellen ermöglicht eine kontinuierliche Überwachung komplexer Bauteile und liefert umfangreiche Informationen über deren strukturelle Integrität. Die aufgezeichneten Signale sind jedoch komplex und werden daher zunehmend mittels KI verarbeitet, um Anomalien in der Systemintegrität automatisch zu erkennen. Eine Herausforderung dabei ist jedoch das Trainieren solcher KI-Modelle, da dafür große Mengen an Trainingsdaten inklusive diverser Schadensfälle benötigt werden, um eine hohe Genauigkeit und Zuverlässigkeit zu gewährleisten. Experimentelle Generierung dieser Daten ist wegen des enormen Zeit- und Kostenaufwands nicht möglich.
Eine Alternative hierfür bieten die numerischen Simulationen, da dadurch Trainingsdaten für beliebige Anzahl an Schadenskonfigurationen erzeugt werden können. Das Hauptproblem dabei ist jedoch die Abweichung zwischen den experimentellen und simulierten Daten (durch z.B. Abweichungen bei der Geometrie, Randbedingungen oder Materialparameter). Um diese Abweichungen zu reduzieren bzw. zu eliminieren, wird ein KI-Modell mit einer Autoencoder-Decoder Architektur vorgestellt, die mittels eines Experiments kalibriert werden kann und anschließend in der Lage ist, die Genauigkeit der Simulationsergebnisse nachträglich zu erhöhen.
The applications of composite overwrapped pressure vessels (COPVs) in extreme conditions, such as storing hydrogen gases at very high pressure, impose new requirements related to the system's integrity and safety. The development of a structural health monitoring (SHM) system that allows for continuous monitoring of the COPVs provides rich information about the structural integrity of the component. Furthermore, the collected data can be used for different purposes such as increasing the periodic inspection intervals, providing a remaining lifetime prognosis, and also ensuring optimal operating conditions. Ultimately this information can be complementary to the development of the envisioned digital twin of the monitored COPVs. Guided waves (GWs) are preferred to be used in continuous SHM given their ability to travel in complex structures for long distances. However, obtained GW signals are complex and require advanced processing techniques. Machine learning (ML) is increasingly utilized as the main part of the processing pipeline to automatically detect anomalies in the system's integrity. Hence, in this study, we are scrutinizing the potential of using ML to provide continuous monitoring of COPVs based on ultrasonic GW data. Data is collected from a network of sensors consisting of fifteen Piezoelectric (PZT) wafers that were surface mounted on the COPV. Two ML algorithms are used in the automated evaluation procedure (i) a long short-term memory (LSTM) autoencoder for anomaly detection (defects/impact), and (ii) a convolutional neural network (CNN) model for feature extraction and classification of the artificial damage sizes and locations. Additional data augmentation steps are introduced such as modification and addition of random noise to original signals to enhance the model's robustness to uncertainties. Overall, it was shown that the ML algorithms used were able to detect and classify the simulated damage with high accuracy.
Additive manufacturing (AM) offers significantly greater freedom of design compared to conventional manufacturing processes since the final parts are built layer by layer. This enables metal AM, also known as metal 3D printing, to be utilized for improving efficiency and functionality, for the production of parts with very complex geometries, and rapid prototyping. However, despite many technological advancements made in recent years, several challenges hinder the mass adoption of metal AM. One of these challenges is mechanical anisotropy which describes the dependency of material properties on the material orientation. Therefore, in this work, stainless steel 316L parts produced by laser-based powder bed fusion are used to isolate and understand the root cause of anisotropy in AM parts. Furthermore, an efficient and accurate multiscale numerical framework is presented for predicting the deformation behavior of actual AM parts on the macroscale undergoing large plastic deformations. Finally, a novel constitutive model for the plastic spin is formulated to capture the influence of the microstructure evolution on the material behavior on the macroscale.
The underlying cause of mechanical anisotropy in additively manufactured (AM) parts is not yet fully understood and has been attributed to several different factors like microstructural defects, residual stresses, melt pool boundaries, crystallographic and morphological textures. To better understand the main contributing factor to the mechanical anisotropy of AM stainless steel 316L, bulk specimens were fabricated via laser powder bed fusion (LPBF). Tensile specimens were machined from these AM bulk materials for three different inclinations: 0◦, 45◦, and 90◦ relative to the build plate. Dynamic Young’s modulus measurements and tensile tests were used to determine the mechanical anisotropy. Some tensile specimens were also subjected to residual stress measurement via neutron diffraction, porosity determination with X-ray micro-computed tomography (μCT), and texture analysis with electron backscatter diffraction (EBSD). These investigations revealed that the specimens exhibited near full density and the detected defects were spherical. Furthermore, the residual stresses in the loading direction were between −74 ± 24 MPa and 137 ± 20 MPa, and the EBSD measurements showed a preferential ⟨110⟩ orientation parallel to the build direction. A crystal plasticity model was used to analyze the elastic anisotropy and the anisotropic yield behavior of the AM specimens, and it was able to capture and predict the experimental behavior accurately. Overall, it was shown that the mechanical anisotropy of the tested specimens was mainly influenced by the crystallographic texture.
The applications of composite overwrapped pressure vessels (COPVs) in extreme conditions, such as storing hydrogen gases at very high pressure, impose new requirements related to the system's integrity and safety. The development of a structural health monitoring (SHM) system that allows for continuous monitoring of the COPVs provides rich information about the structural integrity of the component. Furthermore, the collected data can be used for different purposes such as increasing the periodic inspection intervals, providing a remaining lifetime prognosis, and also ensuring optimal operating conditions. Ultimately this information can be complementary to the development of the envisioned digital twin of the monitored COPVs. Guided waves (GWs) are preferred to be used in continuous SHM given their ability to travel in complex structures for long distances. However, obtained GW signals are complex and require advanced processing techniques. Machine learning (ML) is increasingly utilized as the main part of the processing pipeline to automatically detect anomalies in the system's integrity. Hence, in this study, we are scrutinizing the potential of using ML to provide continuous monitoring of COPVs based on ultrasonic GW data. Data is collected from a network of sensors consisting of fifteen Piezoelectric (PZT) wafers that were surface mounted on the COPV. Two ML algorithms are used in the automated evaluation procedure (i) a long short-term memory (LSTM) autoencoder for anomaly detection (defects/impact), and (ii) a convolutional neural network (CNN) model for feature extraction and classification of the artificial damage sizes and locations. Additional data augmentation steps are introduced such as modification and addition of random noise to original signals to enhance the model's robustness to uncertainties. Overall, it was shown that the ML algorithms used were able to detect and classify the simulated damage with high accuracy.
Virtual-lab-based determination of a macroscopic yield function for additively manufactured parts
(2018)
This work presents a method for the yield function determination of additively manufactured parts of S316L steel. A crystal plasticity model is calibrated with test results and used afterwards to perform so-called virtual experiments, that account for the specific process-related microstructure including crystallographic and morphological textures. These simulations are undertaken on a representative volume element (RVE), that is generated from EBSD/CT-Scans on in-house additively manufactured specimen, considering grain structure and crystal orientations. The results of the virtual experiments are used to determine an anisotropic Barlat yield function, that can be used in a macroscopical continuum-sense afterwards. This scale-bridging approach enables the calculation of large-scale parts, that would be numerically too expensive to be simulated by a crystal plasticity model.