@techreport{SternerThemaEckertetal., author = {Sterner, Michael and Thema, Martin and Eckert, Fabian and Moser, Albert and Sch{\"a}fer, Andreas and Drees, Tim and Christian Rehtanz, and Ulf H{\"a}ger, and Kays, Jan and Seack, Andr{\´e} and Dirk Uwe Sauer, and Matthias Leuthold, and Philipp St{\"o}cker,}, title = {Stromspeicher in der Energiewende - Untersuchung zum Bedarf an neuen Stromspeichern in Deutschland f{\"u}r den Erzeugungsausgleich, Systemdienstleistungen und im Verteilnetz}, publisher = {Agora Energiewende}, address = {Berlin}, doi = {10.13140/RG.2.2.31804.56964}, abstract = {Wie groß ist der Speicherbedarf in Deutschland in der weiteren Umsetzung der Energiewende? Welche Rolle spielen Batteriespeicher, Pumpspeicher, Power-to-Gas etc. im Kontext anderer Flexibilit{\"a}tsoptionen auf den verschiedenen Netzebenen? Wie entwickelt sich der Markt f{\"u}r Batterien und Wasserstoff? In unserer Agora-Speicherstudie haben wir auch erstmalig den Begriff Power-to-X definiert und damit die bis dato entstandenen Begriffe Power-to-Gas, Power-to-Liquids, Power-to-Products, Power-to-Chemicals etc. zusammengefasst.}, language = {de} } @article{LehrerStockerDuddecketal., author = {Lehrer, Tobias and Stocker, Philipp and Duddeck, Fabian and Wagner, Marcus}, title = {UCSM: Dataset of U-shaped parametric CAD geometries and real-world sheet metal meshes for deep drawing}, series = {Computer-aided design}, volume = {188}, journal = {Computer-aided design}, publisher = {Elsevier}, doi = {10.1016/j.cad.2025.103924}, pages = {17}, abstract = {The development of machine learning (ML) applications in deep drawing is hindered by limited data availability and the absence of open-access benchmarks for validating novel approaches, including domain generalization over distinct geometries. This paper addresses these challenges by introducing a comprehensive U-shaped dataset tailored to this manufacturing process. Our U-Channel sheet metal (UCSM) dataset combines 90 real-world meshes with an infinite number of synthetic geometry samples generated from four parametric Computer-Aided Design (CAD) models, ensuring extensive geometry variety and data quantity. Additionally, a ready-to-use dataset for drawability assessment and segmentation is provided. Leveraging CAD and mesh data sources bridges the gap between sparse data availability and ML requirements. Our analysis demonstrates that the proposed parametric models are geometrically valid, and real-world and synthetic data complement each other effectively, providing robust support for ML model development. While the dataset is confined to U-shaped, thin-walled, deep drawing scenarios, it considerably aids in overcoming data scarcity. Thereby, it facilitates the validation and comparison of new geometry-generalizing ML methodologies in this domain. By providing this benchmark dataset, we enhance the comparability and validation of emerging methods for ML advancements in sheet metal forming.}, language = {en} } @misc{LehrerStockerDuddecketal., author = {Lehrer, Tobias and Stocker, Philipp and Duddeck, Fabian and Wagner, Marcus}, title = {Comparison of Low- vs. High-Dimensional Machine Learning Approaches for Sheet Metal Drawability Assessment}, series = {Third International Conference on Computational Science and AI in Industry (CSAI 2023), Trondheim, Norway, 28-30 August 2023}, journal = {Third International Conference on Computational Science and AI in Industry (CSAI 2023), Trondheim, Norway, 28-30 August 2023}, publisher = {International Centre for Numerical Methods in Engineering}, doi = {10.35096/othr/pub-6477}, url = {http://nbn-resolving.de/urn:nbn:de:bvb:898-opus4-64772}, abstract = {Developing new deep-drawn sheet metal parts is a challenging task due to conflicting demands for low costs, durability, and crash properties. Ensuring manufacturability throughout geometrical changes adds to the complexity, leading engineers to rely on experience-driven iterative design changes that compromise requirements and lack reproducibility. Finite Element (FE) simulation models are employed to ensure manufacturability, albeit at the expense of high computational costs and delays in part development. To improve efficiency, a Machine learning (ML)-centered approach was proposed to ensure manufacturability. However, the limited availability of data raises uncertainty about whether a low- or high-dimensional ML approach is most suitable for drawability assessment. This work compares the accuracy of a low-dimensional, feature-based Linear Support Vector surrogate and an adapted high-dimensional PointNet model under different dataset sizes. The dataset is composed of parametrically generated, U-shaped structural sheet metal parts. We use a one-step simulation scheme and evaluate results with a Forming Limit Diagram (FLD) to label drawability. Results show the point of transition to be at about 500 training samples, from which onwards Deep learning is advantageous. Moreover, the generalizability of these models is tested on a second dataset with topologically similar components. This is to assess the potential for a geometrically more comprehensive evaluation. We discuss several influences on model performances and outline future potentials.}, language = {en} } @inproceedings{StockerLehrerDuddecketal., author = {Stocker, Philipp and Lehrer, Tobias and Duddeck, Fabian and Wagner, Marcus}, title = {Face-wise prediction of sheet-metal drawability using graph neural networks}, series = {Journal of Physics: Conference Series}, volume = {3104}, booktitle = {Journal of Physics: Conference Series}, number = {1}, publisher = {IOP Publishing}, issn = {1742-6588}, doi = {10.1088/1742-6596/3104/1/012053}, pages = {11}, abstract = {The early design phase of deep-drawn structural components involves time-consuming iterative development. Traditional drawability assessments rely on finite element simulations, which are computationally expensive and slow the design process. Alternative machine learning (ML) approaches show promise in accelerating this process but face challenges with existing methods. Existing low-dimensional ML models only provide global predictions without identifying specific geometric regions prone to failure. High-dimensional models provide local predictions but require significant amounts of training data. We propose a data-driven approach leveraging graph neural networks (GNNs) for face-wise drawability prediction of sheet metal components in their computer-aided design (CAD) representation. Our method aims to bridge the gap between the computational efficiency of ML and the spatial resolution of simulation by providing face-wise insight into potential failure regions. This study utilises a dataset of parametric U-channel geometries with variability in both geometry and topology. Ground-truth labels are generated using inverse analysis simulations. Geometric entities are represented through the use of UV parameterisations, whereby 3D surfaces are mapped into 2D space to facilitate geometric encoding. Concurrently, the topological relationships are captured using a face adjacency graph. To address data scarcity, we evaluate how different amounts of training data affect model performance and perform ablation studies to analyse the impact of different CAD representation features. Our results show that the proposed approach achieves high accuracy even with limited training data. In addition, the ablation studies provide insights into the most critical CAD features, guiding future research. These results highlight the potential of our GNN to predict face-wise drawability in the early design phase.}, language = {en} } @unpublished{LehrerStockerDuddecketal., author = {Lehrer, Tobias and Stocker, Philipp and Duddeck, Fabian and Wagner, Marcus}, title = {UCSM: Dataset of U-Shaped Parametric CAD Geometries and Real-World Sheet Metal Meshes for Deep Drawing}, publisher = {SSRN}, doi = {10.2139/ssrn.5268323}, pages = {19}, abstract = {The development of machine learning (ML) applications in deep drawing is hindered by limited data availability and the absence of open-access benchmarks for validating novel approaches, including domain generalization over distinct geometries. This paper addresses these challenges by introducing a comprehensive U-shaped dataset tailored to this manufacturing process. Our U-Channel sheet metal (UCSM) dataset combines 90 real-world meshes with an infinite number of synthetic geometry samples generated from four parametric Computer-Aided Design (CAD) models, ensuring extensive geometry variety and data quantity. Additionally, a ready-to-use dataset for drawability assessment and segmentation is provided. Leveraging CAD and mesh data sources bridges the gap between sparse data availability and ML requirements. Our analysis demonstrates that the proposed parametric models are geometrically valid, and real-world and synthetic data complement each other effectively, providing robust support for ML model development. While the dataset is confined to U-shaped, thin-walled, deep drawing scenarios, it considerably aids in overcoming data scarcity. Thereby, it facilitates the validation and comparison of new geometry-generalizing ML methodologies in this domain. By providing this benchmark dataset, we enhance the comparability and validation of emerging methods for ML advancements in sheet metal forming.}, language = {en} } @misc{StockerLehrer, author = {Stocker, Philipp and Lehrer, Tobias}, title = {Machine Learning Dataset of U-Channel Sheet Metal Geometry Representations with Supervision Information for Drawability Assessment and Part Segmentation [Data set]}, doi = {10.5281/zenodo.15327950}, abstract = {The dataset contains 2533 geometries from four different parametric CAD models sampled from the U-Channel python package. The representations include the original CAD geometries (.step), graph binaries (.bin), meshes (.off), and point clouds (.xyz). Additionally, we provide labels for supervised learning use cases of local and global drawability assessment and part segmentation (.json) for all representations. Custom label computation for drawability assessment is enabled by the provided strains.zip file, which contains minor and major true strains for the given meshes. For more details, refer to the associated publication.}, language = {en} }