FSP3: Produktion
Refine
Year of publication
Document Type
- conference proceeding (article) (72)
- Article (66)
- conference proceeding (presentation) (20)
- Part of a Book (6)
- conference proceeding (summary) (3)
- conference proceeding (volume) (3)
- Doctoral Thesis (3)
- Report (3)
- Other (2)
- Bachelor´s thesis (1)
Language
- English (94)
- German (86)
- Multiple languages (2)
Publication reviewed
- begutachtet (157)
- nicht begutachtet (23)
Keywords
- KLEVERTEC / Vorlaufforschung (7)
- LILIPUT (7)
- Automatisierungstechnik (6)
- MEDAL (6)
- Produktionstechnik (6)
- ARCHIVE (5)
- DrGaN (5)
- Künstliche Intelligenz (5)
- Digitaler Zwilling (4)
- Lehre (4)
Institute
- Fakultät Maschinenbau (69)
- Fakultät Elektrotechnik (56)
- EPT – Institut für Effiziente Produktionstechnik (25)
- IMS - Institut für Maschinelles Sehen (23)
- KLEVERTEC (14)
- IPI – Institut für Produktion und Informatik (12)
- Fakultät Informatik (8)
- IEAT - Institut für Energie- und Antriebstechnik (6)
- ECC-ProBell - Europäisches Kompetenzzentrum für Glocken (4)
- IFA – Institut für Innovative Fahrzeugantriebe (4)
The increasing use of simulation technologies, especially virtual commissioning, in the context of modern plant development for manufacturing discrete parts is driven by the pressure to shorten time-to-market cycles and overcome supply bottlenecks. The need for robust technologies to seamlessly integrate the digital and physical world is growing as machine data becomes more readily available. A challenge to this integration is presented by the need to continuously adjust the movement parameters, especially for event-discrete actuators based on live data, taking wear, ageing and process-time fluctuations into account. A lack of synchronization leads to discrepancies between the simulation and reality renders them useless. Related works in this field are discussed, which highlight the complexities of achieving synchronization between simulation and reality, particularly in event-discrete signals and systems. The aim of this article is to present a method for reusing virtual commissioning models for operation-synchronized simulations at actuator level. This approach includes introducing of a methodology called prescheduling in order to compensate process uncertainties and also defines the necessary requirements for the simulation tool and model. The method is validated using an industrial test system and a commercial virtual commissioning tool to confirm ist suitability for real-life implementation in industrial plants, which suggests its suitability for improving production efficiency and reducing costs by means of machine monitoring and proactive control interventions.
Using a Machine Learning Regression Approach to Predict the Aroma Partitioning in Dairy Matrices
(2024)
Aroma partitioning in food is a challenging area of research due to the contribution of several physical and chemical factors that affect the binding and release of aroma in food matrices. The partition coefficient measured by the Kmg value refers to the partition coefficient that describes how aroma compounds distribute themselves between matrices and a gas phase, such as between different components of a food matrix and air. This study introduces a regression approach to predict the Kmg value of aroma compounds of a wide range of physicochemical properties in dairy matrices representing products of different compositions and/or processing. The approach consists of data cleaning, grouping based on the temperature of Kmg analysis, pre-processing (log transformation and normalization), and, finally, the development and evaluation of prediction models with regression methods. We compared regression analysis with linear regression (LR) to five machine-learning-based regression algorithms: Random Forest Regressor (RFR), Gradient Boosting Regression (GBR), Extreme Gradient Boosting (XGBoost, XGB), Support Vector Regression (SVR), and Artificial Neural Network Regression (NNR). Explainable AI (XAI) was used to calculate feature importance and therefore identify the features that mainly contribute to the prediction. The top three features that were identified are log P, specific gravity, and molecular weight. For the prediction of the Kmg in dairy matrices, R2 scores of up to 0.99 were reached. For 37.0 °C, which resembles the temperature of the mouth, RFR delivered the best results, and, at lower temperatures of 7.0 °C, typical for a household fridge, XGB performed best. The results from the models work as a proof of concept and show the applicability of a data-driven approach with machine learning to predict the Kmg value of aroma compounds in different dairy matrices.
Effiziente Kühlung und Schmierung für Fräsprozesse: Frästechnologie und Hochdruckkühlschmierstoff
(2024)
Insbesondere bei der 5-achsigen Fräsbearbeitung existiert noch kein System, das die optimale bzw. minimale Hochdruckkühlschmierstoffmenge bei jeweils gegebenen Bearbeitungszuständen beschreibt.
Die Entwicklung, Umsetzung und Potentiale eines seriennah einsetzbaren, externen und vom Werkzeugmaschinenhersteller unabhängigen Systems zur optimierten Bereitstellung von Hochdruckkühlschmierstoff werden gezeigt.
Bisher existiert noch kein System, das die optimale bzw. minimale Kühlschmierstoffmenge bei jeweils gegebenen Bearbeitungssituationen unter besonderer Berücksichtigung der zu verarbeitenden Werkstoffe und deren signifikantem Verschleißverhalten beschreibt.
Die Entwicklung eines seriennahen unabhängigen Systems zur optimierten Bereitstellung von Kühlschmierstoffen mit angepassten Drücken für die jeweilige Werkstoffapplikationen wird gezeigt.
Diese Veröffentlichung beleuchtet die Herausforderungen und Lösungsansätze zur Bewältigung des Fachkräftemangels in der deutschen produzierenden Industrie. Der Mangel an qualifizierten Arbeitskräften verursacht erhebliche Kosten und verringert das Produktionspotenzial. Hauptursachen sind demografische Veränderungen und veränderte Wertvorstellungen der Beschäftigten. Die Attraktivität des Arbeitsumfelds soll durch die Sichtbarmachung individueller Beiträge und die Stärkung des Gemeinschaftsgefühls erhöht werden. Maßnahmen aus dem Umfeld von Large-Language-Models werden vorgestellt. Abschließend wird die Vision einer modernen, vollvernetzten und arbeitnehmerfreundlichen Produktionsstätte skizziert.
In an industrial context, AI-based methods are becoming increasingly important in the optical systems used for identification, inspection and classification. The reasons for this are that AI-based image processing algorithms are easy to use on the operator side and often achieve superior results. E.g. in complex classification tasks. In the sand cast industry, the complexity in optical inspection of cast parts is connected with strong variations in the local surface topography and in the global object geometry change. Despite the great potential of AI-based methods, application is often hindered by the immense effort involved in acquiring a suitable training dataset. This refers not only to the acquisition of the required number of images but also to the tedious labelling. In this work, we investigate the capabilities and limits of synthetic training data on an AI-based optical scanner used to identify and track cast parts. The optical scanner is capable of detecting and classifying a codification specifically designed for the casting industry. By reading the code, the scanner can deduce the specific number of the cast part. For synthetic image generation, we use physically based rendering, which has advantage of full control over all rendering parameters. This allows for both a systematic investigation of the importance of the parameters and, an automatic labelling process of the training datasets. Our results show that, in particular, a detailed geometric modelling of the local surface topography and global object geometry of the pins have a positive influence on the recognition rate of the neural network. With that accuracy rates up to 56 % are achieved using synthetic training datasets, only.
Anomalie-Detektion im Kontext der Industrie 4.0 hat gerade seit der wachsenden Popularität von maschinellen Lernverfahren und aufkommenden großen Datenmengen an Relevanz gewonnen. So kann in Herstellungsprozessen über die Zeit unerwartetes Verhalten auftreten, das für den Ausschuss von Teilen verantwortlich ist. Es besteht ein Interesse an den Ursachen dieses Verhaltens, sodass ggf. einem nochmaligen Auftreten proaktiv entgegengewirkt werden kann. Aus den aufgenommenen Daten der Sensoren für den Prozess lassen sich Ursachen für dieses Verhalten ableiten. Es besteht bereits Forschung für die Detektion von Anomalien in Sensordaten, jedoch meist im univariaten Fall, d.h unter Beobachtung einer Zielgröße. Zudem besteht ein Forschungsbedarf bei der Erkennung von Anomalien, die sich nicht punktweise manifestiert, sondern über einer Menge von beobachteten Daten. Dafür verfügbare sind allerdings ebenfalls nur spärlich vorhanden und verfügen über keine Nähe zu aus Sensoren gewonnen Daten. Diese Arbeit widmet sich diesen Herausforderungen, indem sie die Problemstellung ausformuliert, den aktuellen Forschungsstand wiedergibt und mit einem praktischen Teil Lösungsansätze zur Verfügung stellt. So wird eine Komponente zur Generierung von synthetischen hoch-dimensionalen Daten entwickelt, die Prozesskurven nachempfunden sind, und damit über eine Ähnlichkeit zu Sensordaten verfügen. Zudem enthalten die erzeugten Daten Informationen über eingepflegte Anomalien, was eine Evaluation von Algorithmen und Modellen ermöglicht. Es werden Verfahren basierend auf dem Stand der Forschung entwickelt und auf synthetischen Daten aus dieser Datengenerierungskomponente evaluiert. Abschließend wird ein Fazit über die Ergebnisse dieses Benchmarks getroffen und ein Ausblick auf die weitere Forschung in diesem Themenfeld gegeben.
Clubfoot is a common congenital foot deformity that leads to constant pain and significant limitations if left untreated or not treated adequately. The most used method for treating clubfoot is the Ponseti method. It involves a correction phase where about five plaster casts are applied and changed weekly. This treatment lasting about 2 to 3 months, is the most chosen method due to its high success rate. However, treated babies often experience skin complications caused by stiff and tight casts. Previous research showed that viable solutions already exist including orthoses. In this research, a developed method known as VDI 2221 was applied and the printable orthosis using 3D printer was selected as an alternative to Ponseti method. Calculations and finite element method (FEM) analysis demonstrated that the orthosis made of PA6-CF provides sufficient stiffness and strength, assuming the weight force of the foot is 10 N. The selected design was developed based on requirements and functional analysis, effectively mitigating the disadvantages of the Ponseti method. The developed orthosis can be manufactured globally using the 3D printing process, with a manufacturing cost of around €150, excluding assembly costs. In summary, a new solution was proposed within the same treatment method, effectively eliminating skin complications, and enabling cost-effective manufacturability on a global scale.
Algorithms for causal discovery have recently undergone rapid advances and increasingly draw on flexible nonparametric methods to process complex data. With these advances comes a need for adequate empirical validation of the causal relationships learned by different algorithms. However, for most real and complex data sources true causal relations remain unknown. This issue is further compounded by privacy concerns surrounding the release of suitable high-quality data. To tackle these challenges, we introduce causalAssembly, a semisynthetic data generator designed to facilitate the benchmarking of causal discovery methods. The tool is built using a complex real-world dataset comprised of measurements collected along an assembly line in a manufacturing setting. For these measurements, we establish a partial set of ground truth causal relationships through a detailed study of the physics underlying the processes carried out in the assembly line. The partial ground truth is sufficiently informative to allow for estimation of a full causal graph by mere nonparametric regression. To overcome potential confounding and privacy concerns, we use distributional random forests to estimate and represent conditional distributions implied by the ground truth causal graph. These conditionals are combined into a joint distribution that strictly adheres to a causal model over the observed variables. Sampling from this distribution, causalAssembly generates data that are guaranteed to be Markovian with respect to the ground truth. Using our tool, we showcase how to benchmark several well-known causal discovery algorithms.
Wide bandgap semiconductors, SiC and GaN-based power devices represent key candidates in the development of more efficient devices due to their superior electrical and thermal properties compared to silicon. To achieve maximal performance from WBG semiconductors, new packaging technologies and thermo-electric designs must be developed to ensure efficient and fast switching of devices while minimizing losses. The paper aims to investigate the thermal and mechanical behavior of new prepackage embedding technologies by finite element simulation. The focus is on insulated substrates including direct bonded copper (DBC) with various dielectrics such as AlN, Al 2O 3, Si3N 4 and new insulated metal substrates (IMS) with emphasis on commercially available materials and thicknesses. This study proposes a thermo-mechanical pareto-optimization methodology able to identify the best substrate configuration. The sintered silver layer (in both sides of the chip), which is the most prone to failure due to delamination, has been modelled with a temperature-dependent bilinear hardening model to account for plasticity. Pareto-optimization accounts for the module thermal resistance and the plastic strain or Von Mises Stress in the sintered layer. Results demonstrate that the best candidate from the thermo-mechanical point of view is the DBC with AlN showing a thermal resistance of 0.34 K/W, accumulative plastic strain of 0.18 % and Von Mises stress of 274 MPa. Finally, the parasitic inductance of multiple pre-packages is evaluated to scale the power of the module. Proper design allows to achieve a stray inductance as small as 1.23 nH for two prepackages and 2.85 nH for four prepackages.
Horizontal chip cracks have been reported in various scientific publications on PCB embedded power semiconductor devices. This study investigates in detail the root cause of the cracks. Experimental evidence indicates that the chip fractures in the mechanical grinding process during preparation of the cross-sections. Here, two different factors are relevant: First, the mechanical fracture strength of the semiconductor die decreases when grinding its edge. The use of P320 sand paper reduces the characteristic fracture strength from 719 MPa to 211 MPa. Second, the tensile stresses in the chip edge increase considerably when, part of the die and package is removed by grinding. Both effects together result in a failure probability of 100%. The use of finer grain sandpaper for target preparation helps to reduce the probability of generating horizontal chip cracks during cross-sectioning.
Die Veröffentlichung beleuchtet die Herausforderungen und Lösungsansätze zur Bekämpfung des Fachkräftemangels innerhalb der produzierenden Industrie am Standort Deutschland. Der Mangel an qualifizierten Arbeitskräften führt zu erheblichen Kosten und reduziert das Produktionspotenzial. Hauptursachen sind der demographische Wandel und veränderte Wertvorstellungen der Beschäftigten. Es gillt, das Arbeitsumfeld Produktion attraktiver zu gestalten, indem individuelle Beiträge sichtbarer gemacht und das Gemeinschaftsgefühl gestärkt werden. Zudem soll die wahrgenommene Komplexität reduziert und die Autonomie des Fertigungspersonals erhöht werden. Beispiele wie ein digitales Ampelsystem und die Visualisierung individueller Beiträge verdeutlichen diese Ansätze. Die Zukunft der Produktion wird durch sieben Thesen skizziert, die die Bedeutung einer partizipativen Planung, die Veränderung der Anforderungen und die Notwendigkeit eines Kommunikationsraums betonen. Die Präsentation endet mit der Vision einer modernen, vollvernetzten und arbeitnehmerfreundlichen Produktionsstätte.
Many cities in Europe and around the world are concerned with reducing their CO2-emissions. One step on this agenda is the introduction of electric buses to replace combustion engines. The electrification of urban buses requires an accurate prediction of the energy demand. In this pa per, an energy model and the underlying calibration process is presented. This approach leverages substantial tracking data from 10 electric buses operated in Göttingen, Germany. It was shown that, with the use of additional information from the directly measured tracking data, like auxiliary power, charging power and vehicle weight, it is possible to precisely calibrate models based on physical equations with regard to generally poorly identifiable parameters like rolling friction coefficient or efficiency of the electric machine. With a multilayered approach for simulating the energy demand, it is possible to validate the results on the mechanical layer and on the electrical layer separately. This enables a far better parametrization and elimination of uncertainties from the different parameters. Furthermore, we compare the results to other publications for sections with 1 km, 100 km and 230 km, respectively. The relative errors between the simulated and measured electrical power consumption are below 0.3%, 3% and 6.5%, respectively. Hence, the yielded model is appropriate for electric urban bus network planning applications. And the found parameters should be a good starting point for other energy prediction models. To further enable comparability with other approaches the dataset used for calibration is made publicly available.
Many cities in Europe and around the world are concerned with reducing their CO2-emissions. One step on this agenda is the introduction of electric buses to replace combustion engines. The electrification of urban buses requires an accurate prediction of the energy demand. In this pa per, an energy model and the underlying calibration process is presented. This approach leverages substantial tracking data from 10 electric buses operated in Göttingen, Germany. It was shown that, with the use of additional information from the directly measured tracking data, like auxiliary power, charging power and vehicle weight, it is possible to precisely calibrate models based on physical equations with regard to generally poorly identifiable parameters like rolling friction coefficient or efficiency of the electric machine. With a multilayered approach for simulating the energy demand, it is possible to validate the results on the mechanical layer and on the electrical layer separately. This enables a far better parametrization and elimination of uncertainties from the different parameters. Furthermore, we compare the results to other publications for sections with 1 km, 100 km and 230 km, respectively. The relative errors between the simulated and measured electrical power consumption are below 0.3%, 3% and 6.5%, respectively. Hence, the yielded model is appropriate for electric urban bus network planning applications. And the found parameters should be a good starting point for other energy prediction models. To further enable comparability with other approaches the dataset used for calibration is made publicly available.
Produkte werden heutzutage immer variantenreicher und individueller. Für die industrielle Fertigung wachsen die Anforderungen, die Produkte effizient zu fertigen und schnell auf sich verändernde Marktbedingungen zu reagieren. Entsprechend rapide steigt die Nachfrage nach flexiblen Produktionslösungen, die sich möglichst autonom an die neuesten Marktanforderungen anpassen. Wandlungsfähige und modulare Produktionssysteme sollen es ermöglichen, auch bei großer Variantenvielfalt effizient zu produzieren - sogar bis zur Losgröße 1. Unterstützt wird die Produktion durch digitale Lösungen, die bereits an vielen Stellen zum Einsatz kommen.
Der Arbeits- und Fachkräftemangel wird seitens der produzierenden Industrie im DACH-Raum als zunehmend wachstumshemmend und geschäftsmodellbedrohend empfunden. Als zentrale Ursachen können gleichermaßen der demographische Wandel wie auch ein entsprechender Wertewandel ausgemacht werden. Es erscheint daher empfehlenswert, die "Ressource Mensch" in dieser Gemengelage neu zu denken. Die Frage, wie ein modernes Produktionssystem beschaffen sein muss, damit dieses aus Arbeitnehmersicht langfristig als attraktiv wahrgenommen wird, rückt hierbei in den Mittelpunkt wissenschaftlicher Überlegungen.
„Hybride Montage“ als Antwort auf Modell Mix und Variantenvielfalt im produzierenden Mittelstand
(2024)
Die "Hybride Montage" als Kombination aus konventioneller Fließfertigung und innovativer Matrixproduktion offeriert vielfältige Potentiale, dem zunehmenden Maß an Variantenvielfalt in der produzierenden Industrie Rechnung zu tragen. Im Zeitalter des Industrial Metsverse und der damit verbundenen "Servitisierung" und des "Manufacturing as a Service" - d.h. der Güterproduktion als Dienstleistung - ermöglicht die Hybride Montage außerdem eine flexiblere Arbeitsplanung.
Durch globale Unsicherheiten, steigende Variantenvielfalt und den Kundenbedarf nach ökologisch nachhaltig produzierten Produkten nehmen die Anforderungen an heutige Produktionssysteme stetig zu. Dies stellt bisherige Produktionsformen wie die Linienproduktion zunehmend vor Herausforderungen. Als hochflexible Produktionsform bietet sich die Matrixproduktion als alleinige oder in Kombination mit der Linie als hybride Produktionsform an.
Die produzierende Industrie des DACH-Raumes erlebt im Jahr 2023 vielfältige Herausforderungen. Gerade scheint die Covid-Pandemie überwunden, welche die Vulnerabilität globaler Lieferketten schonungslos offenbart hat, treten neue Herausforderungen zutage. Veränderungen der gesetzlichen Anforderungen (wie EU Data Act), steigende Ansprüche an ökologische Nachhaltigkeit (z.B. Kreislaufwirtschaft) oder veränderte Kundenbedürfnisse (wie insbesondere Servitisierung) führen – bei einem konstant hohen Niveau an Variantenvielfalt – zu erheblichen technologischen Herausforderungen. Diese werden darüber vielerorts flankiert und im negativen Sinne überlagert durch einen erheblichen Mangel an Arbeits- und Fachkräften in der Produktion. Das Produktionssystem der Zukunft wird unserer Einschätzung nach daher nicht nur unternehmensübergreifenden Datenaustausch ermöglichen, Kreislaufwirtschaft befähigen und verstärkt kundenzentriert ausgerichtet sein. Es wird Arbeits- und Fachkräfte in der Produktion auf vielfältige Art und Weise „begeistern“. Auf diese Weise wird Fluktuation reduziert, generisches Wissen im Unternehmen gehalten und so die Grundlage für nachhaltigen Geschäftserfolg und technologischen Fortschritt sowie die langfristige Sicherung von attraktiven Arbeitsplätzen in der produzierenden Industrie geschaffen.
Nach unserer Überzeugung setzen sich in der produzierenden Industrie des DACH-Raumes diesbezüglich zwei Erkenntnisse durch:
1. Der „kritische Wettbewerb“ stammt selten aus Europa
2. Für die Bewerkstelligung dieser Herausforderungen sind authentische Impulse von außen – insbesondere von anderen, vergleichbaren Unternehmen – ein zentraler Erfolgsfaktor.
In dieser Gemengelage haben wir mit dem SUMMIT ALLGÄU eine Managementkonferenz „von der Industrie für die Industrie“ ins Leben gerufen. Im Zentrum des Veranstaltungskonzeptes stehen hierbei keine wissenschaftlichen Fachvorträge, sondern authentische Erfahrungsberichte hochkarätiger Referenten aus der industriellen Praxis. Beim ersten SUMMIT ALLGÄU Produktion am 23. und 24. Oktober 2023 in Marktoberdorf standen inhaltlich insbesondere die Themenkomplexe „Transformation & Nachhaltigkeit“, „Faktor Mensch in der Produktion“ sowie „Variantenvielfalt“ im Fokus. Auf überfachlicher Ebene wurden vor allem der persönliche Austausch und das Networking zwischen den zahlreichen Teilnehmern, Referenten und Ausstellern fokussiert.
The spectrometer-free chromatic confocal measurement technique enables 3D surface measurements with just one exposure and without scanning. To reduce the need for a spectrometer for the spectral analysis of the reflected light composition and thus the extraction of the local surface height, an optical spectral analysis unit is used. This unit determines the first momentum of the spectral composition reflected from the surface under probe for a large number of lateral measurement points simultaneously. This work investigates the impact of the spectral composition and light power of the light source on the sensitivity and accuracy of this method. A thorough optimization of the light source will be conducted, demonstrating the impact of various spectral compositions and light source power on the system performance, taking into account the system-related etendue. In addition, the optimization of the spectral transmission filter used in the optical spectral analysis unit and its influence on the accuracy and sensitivity of the system over the entire measurement range is shown.
The spectrometer free areal chromatic confocal metrology (ChromaCAM) is an optical 3D surface measurement technology, which allows a simultaneous measurement of a large array of measuring points within a single exposure. In this work, we investigate the accuracy of a first prototype sensor system utilizing this new singleshot 3D measurement technique. It is found that surface height measurement errors smaller 1μm within a total measurement range of about 1000 μm are achievable. Furthermore, several influential factors are investigated showing the advantages and limits of the presented system. Investigating different surface materials it is found that frame rates up to approximately 800 fps for highly reflecting surfaces and up to 30 fps for ceramics, aluminum, and plastics are achievable.
Portable Emission Measurement Technology and RDE on Motorcycles as Instruments for Future Challenges
(2020)
A considerable amount of enabling technologies are being explored in the era of fifth generation (5G) mobile system. The dream is to build a wireless network that substantially improves the existing mobile networks in all performance metrics. To address this 5G design targets, massive MIMO (multiple input multiple output) and mmWave (millimeter wave) communication are also candidate technologies. Luckily, in many respects these two technologies share a symbiotic integration. Accordingly, a logical step is to integrate mmWave communications and massive MIMO to form mmWave-massive MIMO which substantially increases user throughput, improve spectral and energy efficiencies, increase the capacity of mobile networks and achieve high multiplexing gains. Thus, this work analyses the concepts, performances, comparison and discussion of these technologies called: massive MIMO, mmWave Communications and mmWave-massive MIMO systems jointly. Besides, outcomes of extensive researches, emerging trends together with their respective benefits, challenges, proposed solutions and their comparative analysis is addressed. The performance of hybrid analog-digital beamforming architecture with a fully digital and analog beamforming techniques are also analyzed. Analytical and simulation results show that the low-complexity hybrid analog-digital precoding achieves all round comparable precoding gains for mmWave-Massive MIMO technology.
Das Industrial Metaverse scheint der nächste große Game Changer in der modernen Industrie zu werden und wird diese signifikant verändern. Die Vision ist nicht nur die Simulation einer einzelnen Anlage, wie sie Unternehmen heute beispielsweise bei der virtuellen Inbetriebnahme einsetzen, sondern ganzheitlicher Anlagenverbunde aus heterogenen Maschinen samt Fabrikperipherie und Infrastruktur – einer Digitalen Fabrik. Um das Industrial Metaverse zum Leben zu erwecken, wird eine Vielzahl zusätzlicher Technologien und Innovationen benötigt, als lediglich Digitale Zwillinge. Dieser Beitrag schlägt eine ganzheitliche Referenzarchitektur eines Industrial Metaverse vor und fokussiert auf dessen Nutzen.
The packaging of fresh meat has been studied for decades, leading to improved packaging types and conditions such as modified atmosphere packaging (MAP). While commonly used meat packaging uses fossil fuel-based materials, the use of biodegradable packaging materials for this application has not been studied widely. This study aimed at evaluating the sustainability of biodegradable packaging materials compared to established conventional packaging materials through analyses of the quality of freshly packaged pork. The quality was assessed by evaluating sensory aspects, meat color and microbiological attributes of the pork products. The results show no significant differences (p > 0.05) in ground pork and pork loin stored in biodegradable MAP (BioMAP) and conventional MAP for the evaluated sensory attributes, meat color or total bacterial count (TBC) over extended storage times. The data suggest that BioMAP could be a viable alternative to MAP using conventional, fossil fuel-based materials for the storage of fresh meats, while simultaneously fulfilling the customers’ wishes for a more environmentally friendly packaging alternative.
The dissertation presents two approaches for novel and improved measurement techniques, in the area of optical confocal surface metrology.
Firstly, a highly miniaturized endoscopic point distance sensor based on a spatial confocal measurement principle is presented. The sensor utilizes a new technique called spatial confocal point distance measurement. A special feature of the proposed sensor design is the high degree of miniaturization through femto-second direct laser writing and the use of optical fiber bundles, which enable an endoscopic application.
The second part of the dissertation will present a novel method for chromatic confocal metrology, that enables high-speed and high-resolution one-shot areal surface metrology without the need for a spectrometer. After deriving the theoretical basis of the approach, an optical design is conducted and verified using a preliminary optical simulation, After the design of a suitable housing and implementation of a calibration and processing algorithm, the experimental validation and the acquired results will be presented.
In summary, this work adds two novel measurement techniques to the field of fast in-line optical confocal metrology.
This study investigated the thermal performance of a packaging solution designed to manage the electrical isolation and cooling of high voltage ( ) SiC power semiconductor > 3300 V devices. The proposed packaging merges the ceramic substrate and the heat exchanger into a single component, streamlining the overall design. Specifically, a novel heat exchanger is developed for a multi-chip module (20 kV), utilizing a combination of jet impingement and channel- flow cooling techniques. Computational fluid dynamics (CFD) simulations and experimental validation are conducted on a multi-chip module to assess the thermal resistance of this new cooling solution. The results demonstrate a
low thermal resistivity of 0.118 cm2K/W, indicating the potential for improved cooling performance in high voltage and power density semiconductor applications.
Battery electric buses (BEBs) are gaining prominence in public transportation systems. In this paper, we investi-gate the impact of road grade, passenger load, and recuperation power limits on the energy consumption of BEBs using a physics-based model with tuned parameters. The model was employed to conduct a sensitivity analysis taking into account different altitude data sources, passenger load assumptions, and maximum recuperation power limits. The results highlight the importance of considering the route topology and its interaction with dynamic passenger loading for energy consumption predictions. Further-more, the results indicate that various altitude data sources are feasible to estimate the road grade for this purpose. Apart from that, the sensitivity for recuperation power limitations is shown and put into context. Within a broader framework, the findings suggest that physics-based energy consumption models with optimised parameters can serve as a powerful tool for enhanced operations and planning of BEBs.
Various software environments have been developed in the past to create digital twins of single cells or a digital twin of a factory. Each environment has its own strengths and weaknesses and has been designed with a specific focus. The environments that are able to holistically simulate complete factories are limited in terms of the modelling details required for the analysis of single manufacturing cells (e.g., manufacturer-independence of the individual digital twins) and their ability for virtual commissioning. This paper presents three options for realising a virtual commissioning of linked cells using a 3D integration platform with NVIDIA Omniverse, consisting of two different digital models fused into a combined model, also representing material flow. First, with a source/sink solution and unidirectional connector controlled by OPC UA; secondly, with a bidirectional connector, developed in the course of this elaboration, and an extension of the 3D integration platform controlled by Apache Kafka; thirdly, with a bidirectional connector and using only an extension of the 3D integration platform. The research demonstrates that virtually commissioning multiple linked digital twins from different manufacturers in a 3D platform with material flow makes a significant contribution to the industrial metaverse.
In der vorliegenden Arbeit werden Methoden zur datengetriebenen Qualitäts- und Verschleißbestimmung beim Scherschneiden entwickelt. Zunächst wird ein optisches Überwachungssystem in einen Scherschneidprozess integriert und eine Inline-Überwachung in Bezug auf die Schnittflächenkenngrößen ermöglicht. Für die automatische Qualitätsbestimmung wird am Beispiel der Glattschnitthöhe ein Algorithmus zur automatisierten Bildverarbeitung gezeigt. Die neuen Möglichkeiten, die durch diese Entwicklung entstehen, werden anhand einer Fallstudie analysiert. Um die Schwächen des entwickelten Algorithmus mit Blick auf die Verarbeitungszeit auszugleichen, werden ebenfalls neuronale Netzwerke zur Bildverarbeitung untersucht. Zunächst wird hierzu eine Datenmenge aus dem Fertigungsprozess aufgezeichnet und ein manuelles Labelling sowie eine künstliche Datenerweiterung durchgeführt. Beruhend auf Normen und Expertenwissen, wird eine Methode zur Bewertung der Prognosegüte entwickelt. Diese beinhaltet eine Bewertung anhand von verschiedenen Koeffizienten, die sowohl konturbasierte als auch regionenbasierte Ansätze kombinieren. Durch einen Vergleich unterschiedlicher Netzwerkarchitekturen zur Bildsegmentierung und deren Optimierung wird eine Netzwerkarchitektur ermittelt, die die Qualitätsbewertung im Hinblick auf die Glattschnitthöhe innerhalb des Fertigungsprozesses ermöglicht.
Abschließend wird ein Sensornetzwerk bestehend aus unterschiedlichen Sensoren und dem entwickelten Überwachungssystem an einem Scherschneidprozess installiert. In einer Versuchsreihe werden Daten der unterschiedlichen Datenquellen über die Lebensdauer eines Schneidstempels aufgezeichnet. Zusätzlich wird zu definierten Intervallen der vorliegende Schneidstempelverschleiß ermittelt. Die Daten werden in Verschleißklassen eingeteilt und durch unterschiedliche Datenvorverarbeitung entstehen mehrere Merkmalsräume. Diese unterscheiden sich zum einen durch die Kombination der Datenquellen als auch durch die angewendete Datenreduzierung. Die unterschiedlichen Merkmalsräume dienen zum Training von sowohl klassischen Algorithmen des maschinellen Lernens als auch neuronalen Netzwerken. Eine Bewertung und Analyse erfolgt anhand der jeweiligen Konfusionsmatrix und mittels SHAP-Werten. Es zeigte sich, dass durch eine Vorhersage der Verschleißklassen beruhend auf den Bildaufnahmen der Glattschnittfläche nur minimal schlechtere Ergebnisse erzielt werden als durch ein Sensornetzwerk beruhend auf einer Kombination aller Datenquellen.
Intensive research in the field over the past decades highlighted the complexity of aroma partition. Still, no general model for predicting aroma matrix interactions could be described. The vision outlined here is to discover the blueprint for the prediction of aroma partitioning behavior in complex foods by using machine learning techniques. Therefore, known physical relationships governing aroma release are combined with machine learning to predict the 𝐾𝑚𝑔 value of aroma compounds in foods of different compositions. The approach will be optimized on a data set of a specific food product. Afterward, the model should be transferred using explainable artificial intelligence (XAI) to a different food category to validate its applicability. Furthermore, we can transfer our approach to other relevant questions in the food field such as aroma quantification, extraction processes, or food spoilage.
Das Forschungsprojekt TwinMaP hat zum Ziel, eine effiziente Komplettbearbeitung von Bauteilen variabler Losgrößen in einem heterogenen Maschinenpark zu ermöglichen. Der wirtschaftlichste Produktionsablauf soll mittels digitalen Zwillingen, die mit den realen Anlagen und einer Integrationsplattform vernetzt sind, in einer Vorsimulation ermittelt werden.
Die Fertigungsindustrie orientiert sich derzeit in Richtung von Technologien, die Sprungbrettinnovationen in der Produktion ermöglichen. In diesem Zusammenhang werden häufig das Industrial Metaverse und die Künstliche Intelligenz (KI) genannt. In diesem Artikel werden erste Implementierungen dieser Technologien und aktuelle Forschungsaktivitäten im Produktionsumfeld vorgestellt.
The Industrial Metaverse (IM) is an upcoming topic for companies and offers new possibilities to digitalize and optimize their business processes together with AI capabilities. In the production domain, the Industrial Metaverse is a step towards the vision of predicting factory behavior for optimization purposes. A central challenge is a complete factory model necessary as the base to predict its behavior. Therefore, the IM approach is promising to build and contain this model out of available single Digital Twins of factory parts. Consequently, an IT target landscape is required to build an Industrial Metaverse for Digital Twins. This paper evaluates different design pattern options for an industrial IT architecture reference implementation of an IM that companies can use in current IT landscapes. It also proposes a high-level roadmap towards the proposed target IT architecture of an IM.
Tests are carried out on two different series-connected switches made of six SiC MOSFETs capable of blocking 10 kV and 20 kV respectively. High voltage capacitors are connected to the drain terminals of the MOSFETs to emulate the common mode currents’ paths due to parasitic capacitances of the packaging. Due to high dV/dt, common mode currents become dangerous to the switch as they unsettle the voltage sharing during turn-off and unbalance switching losses. This could lead some of the MOSFETs to experience avalanche breakdown. A novel packaging approach that suppresses or balances the common mode currents is introduced.
Gallium nitride transistors have a smaller die area compared to silicon-based devices, which can lead to thermal challenges in high current density applications. Therefore, thermally optimized packages with a high heat spreading capability in combination with small parasitic are necessary. This work investigates the thermal performance a 7 mΩ, 100 V GaN HEMT in a thermally optimized single chip package with integrated RTD and compares it to the commercial counterpart. The thermal optimized package shows a significantly better transient thermal impedance resulting in a static thermal resistance of 3.1 K/W, which is a 20 % reduction in comparison to the COTS package. The integrated RTD trace has a relative reaction time of 590 ms, which is 30-fold slower in comparison to the junction temperature. To show the identical electrical behavior, although the single chip package is larger, it is compared with the commercial off-the-shelf package and a 5 mΩ, 100 V GaN single chip package in a 300 kHz, 48 V buck converter. Both 7 mΩ versions have identical efficiencies of ≈97.5 % up to 50 A output current, slightly outperforming the 5 mΩ GaN transistor. With its combination of improved thermal characteristics and lowinductance, the thermally optimized package of the GaN device offers more degrees of freedom in the design of power converter to exploit trade-offs between longer lifetime, higher temperature operation and power density. Index Terms—gallium nitride, high electron mobility transistor, (thermal) performance evaluation, thermally optimized package, DC/DC converter, high current application, parallelization
In diesem Beitrag wird ein Manipulatormodell zur Verwendung in einem Entwurfsprozess für die Manipulatormechanik vorgestellt. Das Ziel ist Schwingungen zu minimieren und die Genauigkeit applikationsspezifisch zu entwerfen. Aus geeigneten Modellansätzen der einzelnen Roboterkomponenten, wie der Tragstruktur, den Wälzlagern, der Verkabelung und der Fügestellen wird ein Gesamtmodell einer seriellen Sechs-Achs-Kinematik in Modelica erstellt. Für den Antriebsstrang werden Modellansätze für Kompaktgetriebe, Riementriebe, Wellen und Zahnräder abgeleitet. Eine Validierung erfolgt durch eine Modalanalyse durch Selbstanregung.
使工厂生产更灵活的创新
(2023)
Die wichtigsten Elemente von Industrie 4.0 auf der Feldebene sind Aktoren und Sensoren. Durch den Einsatz von Sensoren und Aktoren in Verbindung mit einer bestehenden Infrastruktur und basierend auf der Internettechnologie wird eine Automatisierung und Überwachung der Prozesse ermöglicht. Sensoren dienen dabei als Datenlieferant, Aktoren sind die ausführende Hardware. Die Zielsetzung „Losgröße 1“ von Industrie 4.0 erfordert schnellste Datengenerierung und Umsetzung in Aktionen. Als „Smart Field Devices“ erhalten sie mehr lokale Verarbeitungskapazität. Es geht aber nicht nur um die dezentrale Datenverarbeitung, sondern Innovationen sind auch im Bereich der Basisfunktionen relevant. Bei den Aktoren sind Innovationen auf der Basis der Piezo-Keramik aufzuführen, speziell für Industrie-4.0-Anforderungen sind es Innovationen wie z.B. flexible Bewegungssysteme für Roboter. Auch auf der Sensor-Seite sind neue Basisfunktionen zu entwickeln, im Zusammenhang mit einer zunehmenden Miniaturisierung und Multi-Sensor-Systemen. Unabdingbar für Sensoren wie Aktoren ist die Einbindung in die Daten- und Informationsvernetzung, die über verschiedene Ebenen hergestellt werden muss. Für diese „Field Device Integration (FDI)“ stehen anerkannte Protokolle zur Verfügung, die einen Datentransfer der verschiedensten Geräte ermöglichen. Hier besteht noch ein Entwicklungsbedarf, bestehende Aktoren und Sensoren mit entsprechenden Schnittstellen hierfür auszurüsten.
Die Verfügbarkeit von Produktionsdaten ist eine zentrale Herausforderung der zerspanenden Industrie. Herstellern von Maschinen, Maschinenbauteilen und Sensoren fehlt es an Informationen über das Verhalten ihrer Produkte im Einsatz, um Optimierungspotenziale zu erkennen. Im Fokus des Forschungsprojektes Dione-X stehen daher der sichere und souveräne Datenaustausch entlang der Wertschöpfungskette und die Entwicklung von datengetriebenen Geschäftsmodellen.
In general, current systems for the Digital Factory implement a product-process-resource (PPR) data model in a monolithic rich-client/server architecture with a single database persistence layer. Common data objects are the product bills of material, descriptions of the production processes, or the resource structure, e.g. bill of equipment. The main drawback of the current monolithic architecture is the slow rate of development, which prevents fast adoption of the software to the new production planning process (i.e., due to new technologies for the transformation of the automotive industry with the goal of electrification) is not possible. Furthermore, time-consuming and error-prone export-import operations characterize the collaboration of the engineering supply chain. Mercedes-Benz has created a new IT system architecture for their Digital Factory. The core idea of this architecture is a module-based approach. Each planning step has its own module, e.g. product analysis, layout planning or cost calculation. One single module consists of a server-based business logic, a web-based user interface and its own database. Each module is the source of master data objects that originate from the corresponding planning step and refers to data objects from predecessor planning steps. The single modules communicate mostly via KAFKA. The usage of a model based application engine allows the fast creation of different modules. Best-of-breed third-party systems for specific planning steps can be integrated into the system architecture. Web technologies allow suppliers to access the Mercedes-Benz systems directly for a fully integrated supplier collaboration. Roll-out has started and has already led to significant efficiencies.
Die mittelständisch geprägte deutsche Gießereibranche steht vor akuten
Herausforderungen. Steigende Energie- und Rohstoffpreise wirken sich als Standortnachteil aus und erhöhen den Druck, die Effizienz der Produktion weiter zu erhöhen. Probleme bei der Rekrutierung qualifizierter Fachkräfte erfordern neue Ansätze zur Sicherung des Prozesswissens erfahrener Mitarbeiter. Zudem fehlt aufgrund der mangelnden analytischen Beschreibbarkeit und der Komplexität der Prozesse häufig eine adäquate Modellierung der Kausalketten, was die Ursachenforschung bei Qualitätsproblemen erschwert.
Datenbasierte multimodale Assistenzsysteme können gewinnbringend in
Gießprozessen eingesetzt werden und einen Teil dieser Probleme lösen. Sie nutzen fortgeschrittene Methoden der Datenanalyse (Data Analytics), ein Prozess der Datenuntersuchung, des Datenverständnisses und des Erkenntnisgewinns, in dem unbekannten Muster und Zusammenhänge zwischen den Variablen durch die Analyse von Rohdaten identifiziert werden. So können komplexe Prozesse modelliert, Probleme frühzeitig erkannt und Lösungen in Echtzeit generiert werden. Viele Softwaretools und Datenbanken unterstützen bereits heute die sogenannte „deskriptive Analyse“ und die „diagnostische Analyse“. Ein weiterer Schritt ist jedoch, dieselben Daten zu nutzen, um Modelle zu speisen, die Vorhersagen mit ähnlicher oder besserer Genauigkeit als ein menschlicher Experte treffen können (prädiktive Analyse).
Punching is a process that is sensitive to a multitude of parameters. The estimation of part and punch quality is often based on expert knowledge and trial-and-error methods, mostly carried out as a separate offline process analysis. In a previous study, we developed an optical inline monitoring system with subsequent image processing which showed promising results in terms of capturing every manufactured part, but was limited by slow image processing. Here, we present a more efficient image processing technique based on neural networks. For our approach, we manually identify the burnish parts in images based on criteria established via an expert survey in order to generate a training dataset. We then employ a combination of region-based and boundary-based losses to optimize the neural network towards a segmentation of the burnish surface which allows for an accurate measurement of the burnish height. The hyperparameter optimization is based on custom evaluation metrics that reflect the requirements of the burnish surface identification problem as well. After comparing different neural network architectures, we focus on optimizing the backbone of the UNet++ structure for our task. The promising results demonstrate that neural networks are indeed capable of an inline segmentation that can be used for measuring the burnish surface of punching parts.
Kompaktes Lehrbuch für Studium und Praxis.
Dieses Lehrbuch für Studium und Praxis ist speziell abgestimmt auf das Ingenieurwesen, das Wirtschaftsingenieurwesen und die Betriebswirtschaft. Es wird ein breites Wissen über die wesentlichen Elemente des Qualitätsmanagements dargestellt: vom Qualitätsmanagement in der Beschaffung, der Statistik, den Messsystemen, der ISO 9001, den Qualitätsmethoden, bis hin zum Projekt- und Risikomanagement und der Servicequalität. Zudem wird ein Schwerpunkt auf das internationale Qualitätsmanagement gelegt. Zentral sind kommunikative und interkulturelle Kompetenzen – sei es beim Umgang mit Fehlern, bei der Lösungssuche oder beim Aufbau eines Qualitätsmanagementsystems.
In order to enable the packaging-free purchase of highly viscous foodstuffs requiring refrigeration (such as dairy products), the Kempten University of Applied Sciences developed a dosing dispenser for this area of application and built it with the aid of additive manufacturing processes. The main requirements were to ensure functionality, i.e. sufficient cooling capacity and reliable, hygienic dosing of the products to be filled, as well as easy handling by the customers. Furthermore, a modular system was to be developed that would allow different product categories from quark to yogurt to be filled. With the equipment concept developed for this purpose, different configurations can be flexibly combined and easily exchanged via a modular system. Scalability of the product chamber was also demonstrated by developing a series.
Following a proof of concept, a virtual prototype of the food fill system was created and optimized. Based on this, a physical prototype was successfully realized and presented to potential target groups in field tests. In these user tests, both the required functionality under real test conditions and the acceptance by the respective target group were successfully confirmed.
The energy absorption capacity of braided composite frames under bending loads was studied by conducting quasi-static four-point-bending tests. As specimen geometry C-shaped frame segments where chosen which show the typical failure behaviour of frames with open cross section, such as local buckling and crippling. The braiding manufacturing process offers the possibility to influence the fracture mechanics by a local hybridization of the braider yarns. Different hybridization concepts were investigated to identify design principles for braided frame structures with enhanced energy absorption capacity. The test results show that the post-failure energy absorption of braided frame segments can be significantly increased by a local modification of the braid architecture.
A successful application of braiding technology for the manufacturing of an industrial product is shown on the example of a carbon fiber-reinforced plastic (CFRP) manipulator. Due to its lightweight design, the CFRP system offers highly ergonomic advantages for the worker compared to the previously available steel version. In the course of the braided manipulator design, braid-specific manufacturing options were used to improve bending and torsion stiffness by combining design aspects and fiber architecture variations. It is also shown how braid-related manufacturing constraints can be overcome by additional subsystems, eg, to realize strongly curved part geometries. Finally, a process of the mechanical testing procedure for the whole system is presented which follows the classical building-block approach used for the validation of aeronautical structures.
Breaking down the cost structure of state-of-the-art CFRP part shows that a major share of the costs is caused by labor and equipment as well as process energy consumption.
Therefore, the main goal of the EU funded FP7 project LOWFLIP (Low Cost Flexible Integrated Composite Process) has been the reduction of these costs by introducing new technologies into CFRP production processes.
The LOWFLIP concept focuses on three main aspects:
• Development of a new out-of-autoclave (OOA) prepreg system with snap cure capabilities.
• Development of a direct 3D placement technology for plies and tapes.
• Development of energy efficient and fast heating toolings.
The main content of this paper is detailed information on a novel direct 3D prepreg layup process for automated production of large-scale fiber reinforced parts of small and medium lot sizes. The advanced ply placement process, which is able to drape and compact unidirectional prepreg tapes with currently up to 300 mm ply width directly into a double curved tooling, is being introduced. Two large-scale demonstrator parts from the transport and aerospace sector will be presented. Experiences gained during prototype manufacturing will be reflected and benchmarks of the equipment are presented.
The effect of the cover factor on the in-plane mechanical properties of biaxial and triaxial braided carbon fiber composites is investigated. Low braid coverage occurs if the mandrel circumference exceeds a specific length, so that gaps occur in between the different yarn systems.
Braids are manufactured on different mandrel sizes with a braid angle of ±45°. The cover factor is calculated analytically and measured by means of gray scale analysis of scanned samples. A new analytical model is presented which allows for calculating the cover factor of triaxial braids in good approximation with real results, taking into account the off-centered alignment of the axial yarns. Tensile and compression test results are presented for the thermoset resin systems EPIKOTE MGS® RIM 235 and HexFlow® RTM6. The results show that there is a significant influence of decreased cover factors on the mechanical properties, especially in case of triaxial braids tested in axial direction. Micro computed tomography images based on synchrotron radiation reveal that low coverage leads to an increased fiber undulation of the axial yarns, which mainly determine the mechanical properties of triaxial braids. An influence of the different resin systems is observed for matrix-dominated properties of biaxial braids, where the composite failure strain is in the range of neat resin failure strain.
Scope of the presented work is a detailed comparison of a macroscopic draping model with real fibre architecture on a complex non-crimp-fabric preform using a new robot-based optical measurement system. By means of a preliminary analytical process design approach, a preforming test centre is set up to manufacture dry non-crimp-fabric preforms. A variable blank holder setup is used to investigate the effect of different process parameters on the fibre architecture. The real fibre architecture of those preforms is captured by the optical measurement system, which generates a three-dimensional model containing information about the fibre orientation along the entire surface of the preform. The measured and calculated fiber orientations are then compared with the simulation results in a three-dimensional overlay file. The results show that the analytical approach is able to predict local hot spots with high shear angles on the preform. Macroscopic simulations show a higher sensitivity towards changes in blank holder pressure than reality and limit the approach to precisely predict fibre architecture parameters on complex geometries.
Carbon fiber reinforced plastics (CFRP) allow to redesign well established metallic structures like manipulators in order to decrease their weight and to increase their level of performance. This task can only be successful, when the CFRP-design process does not copy the metallic structure, but solves the underlying functions in a new way. The present study shows, how an I-beam steel manipulator was transferred to a braided design with hollow cross sections and locally adapted material properties. Step by step we show, how to define the manipulator’s functions and mechanics and how they can be considered with a braided CFRP-structure. This includes analytical pre-design as well as detailed finite element method (FEM) design. A mechanical testing program was performed in order to measure input-values for the numerical calculations on coupon-level and validate the simulation on system level. Static and dynamic tests of the CFRP-manipulator showed that it fulfils all requirements to be certified for sale. Therefore, the braiding process for commercial production was investigated for two machine setups and the state of serial production was reached. The braided CFRP-manipulator is 70% lighter than the steel reference and has increased handling properties because of locally adapted material properties.
Ziel der vorliegenden Arbeit ist die Entwicklung eines messtechnischen Verfahrens, um prozessbedingte Variationen der Faserarchitektur textiler Preforms im Entwicklungsprozess von Bauteilen aus Faser-Kunststoff-Verbunden berücksichtigen zu können. Anhand mechanischer Prüfreihen von geflecht- und gelegeverstärkten Verbundwerkstoffen können klare Wirkzusammenhänge zwischen visuellen Textilmerkmalen und den mechanischen Eigenschaften identifiziert werden. Schwerpunkt der Untersuchungsreihen sind Auswirkungen von Lücken, Ondulationen und Streuungen des Faserwinkels auf die Zug- und Druckeigenschaften kohlenstofffaserverstärkter Kunststoffe. Die Erkenntnisse fließen in die Entwicklung eines optischen Messsystems zur 3D-Preformanalyse ein, welches mittels grauwertbasierter Bildverarbeitung die vollflächige Texturanalyse komplexer, dreidimensionaler Oberflächen ermöglicht. In der Arbeit werden zuletzt Anwendungspotenziale der detaillierten Faserarchitekturanalyse zur Optimierung von Fertigungsprozessen und zur Bauteildimensionierung vorgestellt.
This work describes why additive manufacturing is a key technology for efficient design iterations and rapid production ramp-up with large-scale manufacturing technologies. Laser cutting, injection moulding and folding were used to increase the production capacity of face shields for health care workers during the COVID-19 pandemic. We applied systematic learnings from the iterative processes used for additive manufacturing to these large-scale manufacturing technologies and the respective face shield designs. In cooperation with medical experts, structural and functional design requirements of face shields were identified and are described in detail in this work. The regulatory design requirements according to EN 166 are introduced, which were considered to receive a CE certification for three of the presented designs. The employed manufacturing techniques are specified and the respective implications on the design solutions are discussed. The paper concludes with a summary of the production initiative at the research campus ARENA2036 with a total output of over 13 000 face shields from April to June 2020, which were distributed internationally.
In case of electric vehicles (EV) powered by lithium ion traction batteries (LIB), remanufacturing processes nbecome increasingly important due to their rising market share and valuable raw materials. LIB can account for up to 40% of the total EV cost. Often, only a small portion of the cells are significantly degraded when the usable battery capacity falls below 80%, which is currently considered the standard end-of-life criterion. However, in order to enable efficient remanufacturing, novel battery design principles are required. This paper discusses the requirements, opportunities and challenges of future remanufacturing processes of LIBs down to the cell level using a battery system of a commercial vehicle as an example. It gives an overview of the current state-of-the-art manufacturing processes of battery systems and shows the developed overall remanufacturing process including condition assessment, disassembly and reassembly. Subsequently, requirements on future designs are discussed. The state-of-the-art of EV batteries is evaluated based on these requirements to determine where incompatible connections such as welded contacts or adhesive joints conflict with remanufacturing design principles.
Two large groups of materials, namely metals and ceramics, are used in mass quantities in today’s industry because of their outstanding properties. To achieve higher product performance dissimilar materials need to be combined in assemblies, but their joining is challenging. Using friction surfacing technology Al2O3 ceramic substrates were coated with an aluminium alloy (AlMg4.5Mn0.7). Earlier research by the authors suggested that two major bonding mechanisms, namely mechanical interlocking and van der Waals forces, are responsible for the bonding strengths achieved between the coating and the substrate. Further scanning electron microscopy, scanning transmission electron microscopy, high-resolution transmission electron microscopy and energy dispersive X-ray spectroscopy analysis at a sub nanometre resolution were conducted and are presented in this article. These analytical methods revealed that the aluminium coating and the Al2O3 grains form a sharp boundary without evidence of either a chemical reaction or diffusion at the interface and suggest that the main bonding mechanisms for the Al/Al2O3 system are van der Waals forces. In addition, mechanical interlocking may serve to hold in position the interface surfaces, to preserve their close proximity, allowing the van der Waals forces to persist.
The data represents tensile tests of 2D braided biaxial and triaxial composite laminates. Braids were produced on a Herzog radial braiding machine RF 1-176-100 with carbon fiber yarns (Toho Tenax F13 HTS40 12K 800 tex). Test laminates are infused with a VARI process with the 2-component epoxy resin system Hexion RIM235/RIMH237. The infusion is performed at vacuum pressure leves of 20-30 mbar and the resin is cured at room temperature for 24 hours at a pressure level of 250 mbar. A subsequent post-cure step maximizes matrix properties at 70°C for 10 hours.
The braid setups have been chosen to achieve four different braids with a fiber angle of 45° on braid mandrels with a diameter of 120 mm, two biaxial ones and two triaxial ones. By using different sizes of braiding ring diameters and yarn forces, homogeneous and inhomogeneous fiber architectures are generated.
Braiding setups 1 (biaxial) and 2 (triaxial) with 130 mm braiding ring diameter and 350 g yarn force result in braids with homogeneous fiber architecture. 3 (biaxial, 400 mm braiding ring, 600 g yarn force) and 4 (triaxial, 250 mm braiding ring, 600 g yarn force) in braids with inhomogeneous scattering of visual characteristics. Single layers are braided on mandrels with a length of 1500 mm and laminates with 6 layers of biaxial and 4 layers of triaxial braids are subsequently infused.
Tensile tests are performed on a Schenck-Trebel RM250 universal testing machine according to AITM 1-0007. For each specimen, the individual fiber orientation in each layer is measured prior to infusion by image analysis performed on scans of the braided layers. The results show a strong dependence of the individual fiber orientation of single test specimens on their mechanical properties in case of the presented tensile tests
Chromatic confocal metrology suffers from a limitation in the number of measurement points that can be measured simultaneously in a single frame acquisition. We propose chromatic confocal areal metrology (ChromaCAM), in which the surface height for each point in a 2D grid of measurement spots, generated by a rectangular micro-lens array, is parallely analyzed through the utilization of a pinhole multiplexer unit, an analog optical analysis unit, and postprocessing algorithms. An experiment shows the viability of the simultaneous acquisition of multiple measurement points and the advantages over exisiting areal chromatic confocal approaches. Compared with conventional chromatic confocal metrology, the increase in the acquisition rate is significant and enables one-shot measurements.
In manufacturing industry, product failure is costly, as it results in financial and time losses. Understanding the causes of product failure is critical for reducing the occurrence of failure and optimising the manufacturing process. As a result, a number of studies utilising data-driven approaches such as machine learning have been conducted to reduce the occurrence of this failure and to improve the manufacturing process. While these data-driven approaches enable pattern recognition, they lack the advantages associated with knowledge-driven approaches, such as knowledge representation and deductive reasoning. Similarly, knowledge-driven approaches lack the pattern-learning capabilities inherent in data-driven approaches such as machine learning. Therefore, in this paper, leveraging the advantages of both data-driven and knowledge-driven approaches, we present a strategy with a prototype implementation to reduce manufacturing product failure. The proposed strategy combines a data-driven technique, Bayesian structural learning, with a knowledge-based technique, knowledge graphs.
Artificial intelligence (AI) is a crucial technology of industrial digitalization. Especially in the production industry, a great potential is present in optimizing existing processes, e.g., concerning resource consumption, emission reduction, process and product quality improvements, predictive maintenance, and so on. Some of this potential is addressed by methods of industrial analytics beyond specific production technology. Furthermore, particular technological aspects in production systems address another part of this potential, e.g., mechatronics, robotics and motion control, automation systems, and so on. The problem is that the field of AI includes many research areas and methods, and many companies are losing the overview of the necessary and appropriate methods for solving the company problems. The reasons for this are, on the one hand, a lack of expertise in AI and, on the other hand, high complexity and risks of use for the companies (especially for SMEs). As a result, many potentials cannot yet be exploited. The KI-NET project aims to fill this gap, whereby a project overview is presented in this contribution.
In this work an approach for a direct experimental comparison of the application-oriented performance between two high current GaN DC/DC converters based on 7mΩ, 100V GaN HEMTs in a commercial off-the-shelf top-cooled package and 5mΩ, 100V GaN HEMTs embedded in a thermally-optimized single-chip package is given. The two packaging versions are compared by a maximally identical implementation of the power and gate loops in a 48V, 300kHz buck converter with two parallel GaN HEMTs. In the single-chip package the die is directly mounted on a 12×6mm 2 copper-heat spreader offering a significantly lower thermal resistance to heatsink (1.69KW −1 ) in comparison to the commercial off-the-shelf version (2.45KW −1 ). For a direct benchmark, the COTS-based converter has an identical power-loop to the SCP-version with a novel gate-drive concept. Both versions have an efficiency of 96.8% at 65A output current (output power: 1.5kW), while the commercial off-the-shelf version has a better efficiency for lower currents, due to its better hard-switching Figure-of-Merits and therefore lower switching losses and reaches an output current of up to 80A (output power: 1.75kW). A detailed analytical loss breakdown for the different transistors in dependence of the temperature and output current is given to proof the measured current point, where the efficiencies of both converters are identical, since the higher switching losses of the SCP version are compensated by lower conduction losses at higher currents and temperatures compared to the smaller COTS transistor. Finally, an outlook on further improvements for reaching higher output currents and potential converters for a more fair comparison of different thermally optimized SMD packages are given.
Das Einbetten von Halbleiterbauelementen stellt einen innovativen Ansatz für die Aufbau- und Verbindungstechnik (AVT) leistungselektronischer Systeme dar. Im Rahmen von zahlreichen Forschungsvorhaben wurde gezeigt, dass im Vergleich zu einer konventionellen AVT geringere parasitäre Induktivitäten und Widerstände [1, 2]und damit geringere Leitungs- und Schaltverluste ermöglicht werden. Auch kann der thermische Widerstand durch den Einsatz von Dickkupfersubstraten und wärmeleitfähigen organischen Isolatoren erheblich reduziert werden [2, 3, 4]. Dadurch lässt sich auf Systemebene eine höher Leistungsdichte erreichen [5], ohne dass dadurch die Lastwechselfestigkeit leidet [6, 3].
Dennoch findet die Technologie bis heute nur im Low-Power Bereich oder zum Chip-Scale Packaging von einzelnen Leistungsschaltern kommerziellen Einsatz.
Dieser Vortrag diskutiert mögliche Hemmnisse und Grenzen der Technologie.
As a part of the overarching project KI-Net (AB292) a robot cell including a delta kin-ematic system is developed, which will act as a testbed for different software proto-types. The board game “nine men´s morris” serves as an application, which can be implemented with simple pick-and-place movements of the robot. This application of-fers different interfaces to further instances and thus forms a modularity to the outside.
In the following extended abstract, an overview including motivation, objectives and expected results of the Interreg Austria-Bavaria project "KI-Net" (AB292; www.kinet.eu/) is summarized. The project develops a cross-border competence network that investigates, researches, and develops fundamental methods for AI-based optimizations in industrial manufacturing.
Zunächst wird die aktuelle Situation der Schnittflächenkenngrößenermittlung innerhalb des Stanzprozesses in der Industrie beleuchtet und den steigenden Qualitätsanforderungen sowie anwendungsindividualisierbaren Produkten/Prozessen gegenübergestellt.
Auf dieser Basis wird ein optisches Inline-System entwickelt. Dieses System ermöglicht die Aufzeichnung der Schnittflächenqualität in einer Prozessgeschwindigkeit von 250 Hub/min mittels Bildaufnahme. Zusätzlich werden über einen integrierten Triangulationssensor dreidimensionale Schnitte der Schnittkante aufgenommen. Eine entwickelte Bildverarbeitung ermittelt aus den Daten der Bildaufnahme automatisiert und reproduzierbar die Glattschnitthöhe. Die Ermittlung weiterer Schnittflächenkenngrößen ist denkbar, wobei dreidimensionale Daten aus dem Triangulationssensor entstehen. Zusätzlich wird die erreichte Genauigkeit der Systems validiert. Abschließend wird ein Ausblick gegeben, welcher im Besonderen dieses System als Grundlage für die Öffnung der Steuerung des Stanzprozesses mittels intelligenter Methoden sieht. Dies wird an einem Praxisbeispiel unterstrichten.
In komplexen Fertigungsprozessen wie dem horizontalen Schleudergießen stoßen klassische Modellierungsmethoden und analytische Betrachtungen der Zusammenhänge an ihre Grenzen. Dies ist einerseits der Vielzahl von Einflussgrößen und Wechselwirkungen geschuldet, andererseits lassen sich manche Umgebungsbedingungen und prozessspezifische Merkmale durch simulationsgestützte Verbesserungsansätze nicht adäquat abbilden. Datengetriebene Modellierung bietet die Möglichkeit, Prozesse und Produktqualitäten auf empirischer Grundlage zu optimieren. In diesem Beitrag wird gezeigt, wie unüberwachtes und überwachtes maschinelles Lernen sowie mehrkriterielle datengetriebene Optimierung genutzt werden können, um das Wandstärkenprofil duktiler Gusseisenrohre vorherzusagen, effizient anzupassen und zu optimieren. Sie gewähren eine höhere Prozesssicherheit und –transparenz.
Von unüberwachtem Lernen in Form einer robusten Hauptkomponentenanalyse wird für eine verteilungsunabhängige Störgrößenkompensation Gebrauch gemacht, um latente Muster in den Prozessdaten offenzulegen und eine solide Grundlage für die Modellierung zu schaffen. Ensemblelernverfahren wie Random Forests und CatBoost werden zur Vorhersage der Wandstärkenverläufe herangezogen. Empirische Methoden der BlackBox-Modellevaluation fungieren als Sensitivitätsanalyse und ermöglichen eine datengetriebene virtuelle DoE, die im Gegensatz zur analytischen Form eine signifikant höhere Variantenzahl simulieren kann. Die trainierten Prognosefunktionen werden schließlich in einer gradientenfreien Optimierung mit Hilfe von genetischen Algorithmen und dem Downhill-Simplex-Verfahren eingesetzt, um Wandstärkenprofile zu homogenisieren und Ausschuss zu reduzieren.
The lack of suitable techniques for joining Si3N4 ceramics with metals has limited the usage of this otherwise outstanding material for composite applications. In this study, aluminum AlMgSi0.5 (EN AW-6060) was coated onto silicon nitride Si3N4 ceramic substrates using friction surfacing technology. Experimental work revealed that the harmful effects of thermal shock (e.g., substrate cracking, coating delamination) observed with other material combinations can be avoided by selecting materials with a low coefficient of thermal expansion, low Young’s modulus and high thermal conductivity. Design of experiments derived models for coating thickness and bonding strength fit the data well (i.e., the regression model accounts for most of the variation in the response variable). Whereas the coating thickness is predominately dependent on the rotational speed used, the bonding strength is also affected by the traverse speed. Coating thicknesses upto 2.03 mm and bonding strengths of 42.5 MPa were achieved. Deposition rates exceed those of physical vapor deposition by a magnitude of ×1000 and bonding strength is on-par with thin-film metallization. Scanning transmission electron microscope analysis revealed formation of a glassy phase at the interface. Using energy-dispersive X-ray spectroscopy analysis high silicon and oxygen content with smaller percentages of aluminum and nitrogen were detected. High-resolution transmission electron microscope imaging revealed no distinct lattice structure leading to the assumption that the composition is predominantly amorphous and consists of SiAlON.
This work investigates a packaging solution for high power density semiconductors (> 200 W/cm 2), allowing for a dramatic reduction in size and complexity of power electronics modules. The multiple layers in standard packaging structures degrade the cooling efficiency due as they lengthen the path between dies and heatsinks. Here, we reduce the layer count by merging the ceramic substrate and the heat exchanger in a single part. CFD simulations and experimental validation are performed on a single-chip cooling packaging, and demonstrate a 10-20 % reduction in thermal resistance over more traditional cooling solutions.
The ever-increasing switching speed of semiconductor devices requires a precise measurement of steep current transients. The M-shunt concept offers high signal fidelity, good cooling, and simple manufacturing. Depending on the resistive material used, temperature as well as skin and proximity effects impede static and dynamic measurements to a different degree. A step forward has been derived from the ideal coaxial shunt, so far, a purely theoretical concept, which is hardly producible due to its sophisticated structure. By transferring this concept to the M-shunt structure with its improved PCB manufacturing technologies it can now be realised in practice. Nevertheless, the calibration and the correct degree of delay compensation remain challenging and are investigated more closely within this paper. Furthermore, it will be discussed why the conventional method of bandwidth determination doesn't work for the M-shunt structure. In addition to the low inductance introduced into the load circuit, the high bandwidth of the shunts could be demonstrated, as well as the possibility to extend this by design rules. Supplemented by the advantages of the lower load inductance, the M-shunt will become the tool of choice for characterising switching transients at least up to 200 MHz required bandwidth eventually. Although it is obviously difficult to improve the 3 dB bandwidth with suitable design rules, the range of nearly entirely unaffected measurement frequencies (e.g. < 1 dB) can be significantly extended by limited coupling. For even higher frequencies, measurements of the current M-shunt models, as well as for the coaxial shunts used as reference, should be corrected by post processing to get precise measurement results.
Previously, a novel concept for PCB embedding of power semiconductors with reinforced top contacts has been published by our group. In this work, a thermomechanical study is performed to understand the fracture probability of the chip during fabrication. The ball-on-ring (BOR) and ball-on-edge (BOE) tests are employed to characterize the chip strength. A probabilistic model is used to evaluate the failure probability. The contribution of different fabrication steps to the overall failure probability is investigated. A parametric analysis is performed to analyze the impact of material choice and the thickness of the substrate and the interposer. A single-step sintered package with a 200 μm thick Cu interposer and an 800 μm thick Cu substrate with hard Cu leads to a 12% failure probability compared to 52% for an identical stack with a soft Cu substrate. The top surface failure probability for a 150 μm thick CIC interposer is 0.016% compared to 8.6% for a 100 μm Cu interposer., for an identical 800μm thick Cu substrate. A sinterlamination process, where the die-attach and curing of the prepreg material simultaneously take place, results in a 0.03% failure probability.
This work investigates a packaging solution for high voltage semiconductors (20 kV), allowing for a dramatic reduction in size and complexity of power electronics modules. The standard packaging structures typically introduce a competition between electrical insulation (which requires thick insulating layers) and thermal performance (where thin, high thermal conductivity layers are preferred). Here, we introduce a concept which addresses this competition and is based on direct cooling using dielectric liquid. Single-chip heatsinks are designed, optimized using computational fluid dynamics (CFD), built and tested.
PCB embedding is an attractive packaging technology for highly integrated future power systems. Electrical and thermal benefits and limitations are well understood, but only few publications focus on reliability. Here we are addressing high-humidity high temperature reverse bias (H3TRB) tests, which are performed to gain deeper understanding of the limitations of this technology. Cu filament growth is observed on both, PCB embedded samples and silicone potted reference samples. These filaments may form a conductive path and lead to an increased leakage current in blocking mode. Electrochemical corrosion of the Cu-plated guard ring is identified as root cause.
Embedding power semiconductor devices into printed circuit boards (PCB) provides several benefits compared to conventional packaging technologies. Integrating the semiconductor dies into the circuit board reduces the converter size. This results in short current loops, enabling low interconnection resistances and parasitic inductances. Both contribute to a higher system-level efficiency, as conduction and switching losses are reduced. Moreover, the use of thick Cu substrates allows efficient heat removal, due to a low thermal resistance. Therefore, PCB embedding has received a lot of attention in the power electronics community for more than a decade. This article aims to provide a comprehensive review of the scientific literature on the topic ranging from basic fabrication technology over module or system-level demonstrators for electrical and thermal testing to reliability studies. Performance indicators, such as the commutation loop inductance Lσ, the chip area independent thermal resistance Rth × Achip, allow a comparison of different approaches and benchmarking with conventional power modules. Several publications report stray inductances below 1 nH and chip area independent thermal resistances in the range of 20…30 mm²K/W.
Most current measurement techniques suffer from temperature impacts, though to a different degree. This is also true for shunt resistors, which are usually improved significantly by using temperature compensated alloy materials like Manganin(r) or by proper calibration and subsequent temperature compensation. The concept of the copper-based PCB M-shunt has already been reported to yield excellent current measurement results at room temperature. However, due to the high-temperature coefficient of copper (3920 ppm/K), the current measurement with those M-shunts showed considerable measurement errors for high temperatures and due to self-heating. This can be improved by the use of Manganin. However, still the connection of the resistor with the copper of the PCB and their respective thermal behavior needs to be considered as further possible sources of error. This paper presents challenges associated with the improvement of PCB-based shunts by temperature-compensated material, while not compromising the critical issue of bandwidth and manufacturing capability. It explicitly addresses the problems and difficulties as reported for previous development stages. In its new version, the M-shunt, represents a low inductance measurement system that can be optimised for much higher energies than the coax shunt, while not increasing the inductance introduced into the circuit. Alternatively, it can be configured for a significantly increased bandwidth and lower inductance compared to its coaxial reference.
Embedding power semiconductor devices in pre-packages may enable low-inductive power semiconductor module designs with superior thermal performance and reliability. However, it is crucial to understand mechanical stress formation due to the differences in thermal expansion of the materials used. This paper presents a systematic study of thermomechanical stresses in Si and SiC power semiconductor pre-packages based on PCB embedding or compression molding. The analysis is based on thermomechanical FEM simulations and complemented by passive thermal cycling of different test carrier designs. Patterning and CTE-matching of contacts are proven as strategies to minimize thermomechanical stress.
A novel tracking system for the iron foundry field based on deep convolutional neural networks
(2022)
In modern manufacturing the ability of retracing produced components is crucial for quality management and process optimization. Tracking is essential, especially for analyzing the influence of the production parameters on the final quality of the castings. In the iron foundry industry, common marking methods, such as a datamatrix code, cannot be used due to harsh environmental conditions and the rough surface of the cast parts. This work presents a new coding and reading system that guarantees unique marking in the casting process.The coding is built up over several beveled pins and is read out using an optical 2D handheld scanner. With a deep convolutional neural network approach of object detection and classification, a stable image processing algorithm is presented. With a first prototype a reading accuracy of 99.86% for each pin was achieved with an average scanning time of 0.43 s. The presented code is compatible with existing foundry processes, while the handheld scanner is intuitive and reliable. This allows immediate benefits for process optimization.
Development of Data-Driven Machine Learning Models for the Prediction of Casting Surface Defects
(2022)
This paper presents an approach for the application of machine learning in the prediction and understanding of casting surface related defects. The manner by which production data from a steel and cast iron foundry can be used to create models for predicting casting surface related defect is demonstrated. The data used for the model creation were collected from a medium-sized steel and cast iron foundry in which components ranging from 1 to 100 kg in weight are produced from wear and heat resistant cast iron and steel materials. This includes all production-relevant data from the melting and casting process, as well as from the mold production, the sand preparation and component quality related data from the quality management department. The data are tethered together with each other, the information regarding the identity and number of components that resulted in scrap due to the casting surface defect metal penetrations was added to the dataset. Six different machine learning algorithms were trained and an interpretation of the models outputs was accomplished with the application of the SHAP framework.
Gießereien müssen sich in der Zukunft einer Vielzahl von technologischen, wirtschaftlichen und nachhaltigkeitsbezogenen Herausforderungen sowie einem zunehmenden Wettbewerbsdruck stellen. Dies wird nicht ohne die konsequente Weiterentwicklung der Prozesstechnik und -steuerung durch den Einsatz intelligenter, adaptiver Produktionssysteme möglich sein. Ziel des hier vorgestellten Projekts war es, den Energieverbrauch und dessen Optimierungspotenzial beim Schmelzen von Stählen im Mittelfrequenz-Induktionsofen zu modellieren.
Purposefully induced axial chromatic aberration is the core of the chromatic confocal metrology technique.Through the resulting generation of separated focal planes for each wavelength of a broadband light sourcea measurement volume is created and a three-dimensional reconstruction of the topography of technical and biomedical surfaces and layers can be performed. Based on the chromatic confocal metrology technique various metrology sensors and measurement systems have been developed, with high axial and lateral resolution, accuracy and precision. For a significant increase in measurement points, that can be measured simultaneously and the resulting reduction in measurement time, a chromatic confocal method utilizing a micro-lens array in combination with a improved spectral peak detection, has been developed. Through a single image acquisition, the object topography can be measured for multiple points simultaneously and therefore mitigating the need for axial aswell as lateral scanning of the object. For this reason in-situ applications have become a viable domain. First preliminary results of testing a laboratory setup of the proposed system design are presented.
Thermal impedance evaluation of optimized PCB-based GaN HEMT Single-ChipPrepackage using VGS Method
(2022)
This study evaluates the measurement of the thermal resistance and impedance of GaN HEMTs packages with gatesource voltage (VGS) for junction temperature measurement. First, calibration measurement of VGS vs temperature is performed in an oven. Large deviations from chip to chip require that each device is calibrated separately. Zth is measured for an experimental single-chip prepackage (SCP), developed in our lab, and a commercial reference package (GaNpx). The measured junction to ambient thermal resistance of the SCP is 35% lower compared to the reference. This is caused by a copper heat spreader integrated in the package. The thermal resistance of multiple packages attached on one PCB is also performed. This measurement indicates a maximum difference of 35%. This difference remained the same in several experiments with two thermal interface materials and pressure levels.
This paper presents design, fabrication, and analysis of a PCB embedded half-bridge. Here, a pre-package that contains a 100 V / 100 A MOSFET half bridge is fabricated and embedded into a converter level power PCB, allowing for two routing layers on top of the semiconductors. This enables to minimize the commutation loop inductance by placing a decoupling capacitor directly on top of the switches. Experimentally a parasitic inductance of 1.4 nH is deducted from the ringing frequency in switching experiments. The second design goal is an optimized thermal
performance. Both, simulation with ANSYS Icepak and experiments indicate a value of 1.77 K/W. Based on the results, a 12 V / 48 V bidirectional converter was implemented and operated with a maximum power of 620 W.
Thermisch und elektrisch optimiertes Leiterplatten-Prepackage für 100V/100A GaN Leistungshalbleiter
(2021)
In dieser Arbeit wird ein Integrationskonzept für GaN Leistungshalbleiter in Leiterplattentechnologie vorgestellt. Das
Konzept beruht auf thermisch optimierten Prepackages mit integrierten Temperatursensoren, die in Leiterplattenembeddingtechnologie hergestellt werden. Durch ein asymmetrisches Layout wird ein niederinduktiver Kommutierungskreis ermöglicht. Funktionsmuster mit einem 100 V / 120 A GaN HEMT wurden aufgebaut und erprobt. Die Bauteile sind elektrisch voll funktionsfähig und weisen einen thermischen Widerstand von 3,34 K/W auf. Dieser Wert liegt 35% unter dem eines kommerziellen Referenzbauteils.
Datengetriebene Modellierung, insbesondere maschinelles Lernen, lässt sich nutzen, um Prozesse und Produktqualitäten zu optimieren. Dies gilt insbesondere für komplexe Fertigungsprozesse wie das horizontale Schleudergießen, bei dem eine analytische Betrachtung der zusammenhänge schwerfällt. Unüberwachtes Lernen in Form der Hauptkomponentenanalyse wird herangezogen, um latente Strukturen in den Daten freizulegen und eine solide Datenbasis zu schaffen. Auf diese Weise werden die Modellkomplexität reduziert und prozessbedingte Störungen korrigiert. Durch Gradient Boosting werden datengetriebene Prozessmodelle erzeugt, um den Wanddickenverlauf duktiler Gusseisenrohre auf Grundlage von Prozessparametern mit hoher Genauigkeit vorherzusagen. Mithilfe von Methoden der empirischen Modellevaluation werden schließlich die treibenden Einflussgrößen des Prozesses identifiziert, wodurch eine erhöhte Prozesstransparenz gewährleistet wird. Die Haupt-Performanceindikatoren können anschließend verwendet werden, um Wandstärkenverteilungen effizient anzupassen und zu optimieren.
Die Digitalisierung der Industriezweige macht auch vor der Gießerei-Industrie keinen Halt. Die Potenziale dieses Wandels sind groß, doch die praktische Umsetzung erweist sich oft als schwierig. Die Institute der Akademischen Interessensgemeinschaft Gießereitechnik (AKAGuss) haben hier eine gemeinsame Expertise entwickelt, die sie im Sinne einer erfolgreichen Chancennutzung interessierten Firmen zur Verfügung stellen möchten. In diesem ersten von zwei Teilbeiträgen berichten das Gießerei-Institut der RWTH Aachen, die Hochschule Kempten und das Österreichische Gießerei-Institut sowie die Montanuniversität Leoben über ihre Digitalisierungsaktivitäten.
Im Sommersemester 2021 wurden unterschiedliche Personengruppen an den Hochschulen für angewandte Wissenschaften befragt. Fast 12.000 Studierende und über 2.000 Lehrende haben daran teilgenommen. Ergebnisse aus Gruppendiskussionen zur Hochschulentwicklung und Daten zu Studienverläufen und Prüfungen ergänzen die Studie.
Chromatic confocal metrology is a widely established optical metrology technique, that allows for non-contact high-speed three-dimensional surface profiling without the need of mechanical depth scanning. However current methods are limited by the use of some sort of surface scanning method with mechanically moving parts. Furthermore the setups involve a spectrometer setup, either through prisms, gratings or multi-spectral cameras. This drastically limits the simultaneously measureable positions in lateral direction, as the spectrometer setup
will utilize one spatial dimension for the wavelength domain. We present a novel method for chromatic confocal metrology, that enables high-speed and high resolution one-shot aerial surface metrology. This method is scalable with respect to measurement range in axial as well as in lateral direction and in the number of measurement points that can be measured simultaneously. After deriving the theoretical basis of the approach a virtual optical design with a FOV of 10mm by 10mm and a depth range of 1.5mm with roughly 1000 measurement points, based mainly on off-the-shelf components will be presented. This virtual system design was used to perform various simulations and explain the design process and considerations as well as the expected system response of the proposed system.
The embedding of power electronics components into the PCB increases their thermal and electrical performance. Due to lower parasitic inductances and a lower thermal resistance, the overall power density of the system can be increased. However, there is a lack on studies about the reliability of PCB embedded power electronics, especially under humidity stress. This study uses test vehicles with functional diodes embedded into the PCB. The reliability of the test vehicles is studied under HV-H3TRB conditions. A critical leakage current is observed after 200h with a benzoxazine based prepreg and after 600h for an epoxy based prepreg. A rootcause analysis revealed lift-off of the polyimide layer and subsequent electrochemical corrosion as failure mechanism. The delamination is most likely triggered by the Au surface on the chips top metal pad. A second set of test samples was fabricated using an improved surface treatment before lamination. 75% of these samples passed the 1000h H3TRB test successfully.
PCB embedding in combination with direct-bonded copper (DBC) substrates is an attractive approach for packaging of power semiconductors facilitating low-inductive designs while relying on a proven insulating material. However, the CTE mismatch of these materials could cause reliability issues. This study presents an initial reliability screening using simple IGBT prepackages with alumina-based DBC as test vehicles. After -40/150 °C temperature cycles, fracture of the substrate and the chip is observed, resulting in an increased on-state resistance. Literature data suggest that the substrate failure is independent from the embedding. To gain a deeper understanding of the limitations of the technology, further research with optimized DBC substrates is required.
To make the full performance of the intrinsic 100 V, 5 mΩ gallium nitride transistors available on system level, in this work an asymmetrical & thermally optimized PCB embedded single chip package with integrated resistance thermometer, high temperature capability and a thermal resistance Rth,j−hs of 3.3 K W−1 is characterized in a 48 V to 24 V 300 kHz mild-hybrid DC/DC operation with two paralleled chips in each low- (LS) and high-side (HS). The transistors are mounted on a 4-layer multilayer PCB with 1 mm copper inlays to achieve a high current capability, while allowing narrow logic traces on the same PCB. The designed converter is achieving a light load efficiency of ≥99 % and an efficiency of 97 % at 60 A output current and ≈1.3 kW output power in a 48 V to 24 V 300 kHz buck-converter operation. The on-board temperature readout circuit and the phase output current sensor offer the possibility to extend the GaN transistors to an intelligent power module by the compact and simple sensors.