Browsing by Author "Jarwitz, Michael"

Now showing 1 - 8 of 8

Open Access
Application of a physics-informed hybrid model with additional output constraints for the prediction of the threshold of deep-penetration laser welding
(2025) Jarwitz, Michael; Michalowski, Andreas
The quantitative prediction of process constraints, such as the threshold of deep-penetration laser welding, plays a crucial role for the fast and reliable development of robust process windows for laser manufacturing processes. A physics-informed hybrid model with additional output constraints for the prediction of the threshold of deep-penetration laser welding is presented. A “residual model” approach is used, where a machine learning model, employing Gaussian processes, is used to model and compensate for the deviations between experiments and a physical model, and output warping is used to incorporate additional output constraints into the model. The main benefits that result from applying such a model are found to be (1) an increased prediction accuracy compared to only using the physical model, leading to a reduction of the mean relative error of about 76%; (2) a reduction of the number of required training data compared to only using a black-box machine learning model; (3) an increased prediction accuracy compared to only using a black-box machine learning model; (4) and an increased compliance with physical boundary conditions by applying the additional output constraints.
Open Access
Application of output constraints to a physics-informed hybrid model for the prediction of the threshold of deep-penetration laser welding
(2024) Jarwitz, Michael; Michalowski, Andreas
Physics-informed hybrid models, the combination of physics and machine learning, have already shown considerable benefits for quantitative predictions of process constraints, such as the threshold of deep-penetration laser welding. However, despite the improved prediction accuracy and extrapolation capability of such models, there can still be cases where the predictions of the model, including the confidence region, result in values that are not consistent with physical boundary conditions. Therefore, this paper presents the application of additional output constraints to a physics-informed hybrid model to further improve the compliance of the model with physics. Gaussian processes are used for the machine learning model and output warping is used to incorporate the output constraints directly into the model. The approach is demonstrated at the example of a hybrid model for the prediction of the threshold of deep-penetration laser welding.
Open Access
Influence of geometry variations during pyrometric temperature measurement in laser material processing
(2024) Traunecker, David; Jarwitz, Michael; Michalowski, Andreas
Open Access
Laserstrahlschweißen von Metallen mit unterschiedlichen thermophysikalischen Eigenschaften
(München : utzverlag, 2020) Jarwitz, Michael; Graf, Thomas (Prof. Dr. phil. nat.)
Im Rahmen dieser Arbeit wurden prozessangepasste Strategien für die gezielte Energieeinbringung beim Laserstrahlschweißen von Metallen mit unterschiedlichen thermophysikalischen Eigenschaften untersucht und entwickelt. Die Umsetzung wurde anhand der Beispiele des Laserstrahlschweißens von Nickelschaum an Nickelblech und des Laserstrahlschweißens von Aluminium-Kupfer-Mischverbindungen demosntriert.
Open Access
Local laser heat treatment of AlSi10Mg as-built parts produced by Laser Powder Bed Fusion
(2024) Kramer, Steffen; Jarwitz, Michael; Schulze, Volker; Zanger, Frederik
Today, complex structural components for lightweight applications are frequently manufactured by laser powder bed fusion (PBF-LB), often using aluminum alloys such as AlSi10Mg. However, the application of cyclic load cases can be challenging as PBF-LB produced AlSi10Mg parts typically have low ductility and corresponding brittle failure behavior in the as-built condition. Therefore, this paper presents investigations on the feasibility of a laser heat treatment of PBF-LB produced AlSi10Mg parts to locally increase the ductility and decrease the hardness in critical areas. Potential heat treatment process parameters were derived theoretically based on the temperature fields in the material calculated assuming three-dimensional heat conduction and a moving heat source. PBF-LB produced specimens were then laser heat treated at varying laser power and scan speed. Hardness measurements on metallographic cross sections showed hardness reductions of over 35 % without inducing hydrogen pore growth.
Open Access
Potential of contactless support structures for improving the part quality of AlSi10Mg PBF-LB parts
(2023) Kramer, Steffen; Drechsel, Kai; Jarwitz, Michael; Schulze, Volker; Zanger, Frederik
Support structures are usually required to prevent thermally induced deformation of parts during the PBF-LB process, but also cause an increased effort for post-processing and an insufficient surface quality of the produced parts. An approach to overcome these limitations, suggested in the literature, is the use of contactless support structures (CSS), which should act as a heat sink to avoid thermally induced deformation. This approach was already successfully tested for electron beam melting (PBF-EB). Therefore, the aim of the present study is the experimental investigation of the application of CSS for the PBF-LB process. Different part and support geometries were built to investigate the potential of CSS to reduce thermally induced deformation during the PBF-LB process. Several use cases were covered, including increased overhang angles for thin features and maximized overhang thickness. The literature results for PBF-EB and Ti6Al4V could not be confirmed for PBF-LB using AlSi10Mg. The thermally induced deformation could not be reduced and the experimental results indicate that the use of CSS is accompanied by counterproductive process mechanisms.
Open Access
Towards a universal manufacturing node: requirements for a versatile, laser-based machine tool for highly adaptable manufacturing
(2022) Jarwitz, Michael; Traunecker, David; Arnim, Christian von; Müller, Niklas; Kramer, Steffen
The current trend in the context of Industry 4.0 towards small batch sizes and increasing product variety results in ever-changing requirements for both, the products and the production. This requires highly versatile, fully and easily adaptable, and efficient manufacturing environments that can meet these demands, ideally already on the level of the machine tool. Because of its versatility, the laser is a promising tool for such a machine tool, but there is still a considerable need for research in the field of system technology. We consider the requirements for a versatile, laser-based machine tool for highly adaptable manufacturing, that utilizes the combination of laser-based manufacturing processes on one machine. The focus of the considerations lies on remote processes and the processing of metals. Five key research topics for the development of such a universal laser manufacturing node are identified: highly dynamic and precise kinematics (1); ‘on-the-fly’ reconfigurable, distributed control architectures (2); adaptable process diagnostics for online quality monitoring (3); technological interactions in laser-based process chains (4); and models for a fast estimation of the process parameters for each production step (5). The relevance and current needs for research for each topic are discussed and corresponding solution concepts are proposed.
Open Access
Updating the Linux TAPRIO scheduler in deterministic time
(2022) Arnim, Christian von; Gessner, Gernot; Jarwitz, Michael; Lechler, Armin; Riedel, Oliver
In flexible production systems with distributed control systems which communicate with each other via a real time network, changes in the requirements for real time communication during operation must not lead to a temporary failure of the deterministic communication. Rather, the changes of the network configuration must become active at exactly predictable times in order to guarantee the functionality of the deterministic communication all the time. This paper shows a realization of how the configuration of a real time network schedule under Linux can be adjusted at predetermined times without interrupting the communication. For this purpose, the real time scheduler TAPRIO is integrated into the library libnl, and the performance of this extension is evaluated in a test case. It is shown that the modification of the network configuration is reliably possible in several successive communication cycles.