@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} } @article{HollweckLeidingerHartmannetal., author = {Hollweck, Christoph and Leidinger, Lukas and Hartmann, Stefan and Wagner, Marcus and W{\"u}chner, Roland}, title = {An analysis of the critical time step size for explicit dynamics using trimmed B-splines, LR-splines, and THB-splines}, series = {Computer Methods in Applied Mechanics and Engineering}, volume = {450}, journal = {Computer Methods in Applied Mechanics and Engineering}, publisher = {Elsevier}, issn = {0045-7825}, doi = {10.1016/j.cma.2025.118656}, pages = {36}, abstract = {Isogeometric analysis (IGA) combined with explicit dynamics is increasingly used in academia and has already been successfully applied in industrial simulations, including crash and sheet-metal forming. Since explicit schemes are only conditionally stable, accurate estimation of the critical time step is essential for both stability and efficiency. Adaptive mesh refinement is widely used to balance accuracy and computational cost. In IGA, THB- and LR-splines break the tensor-product structure of standard B-splines and enable local refinement, but their effect on the critical time step under trimming has not been systematically studied - a key requirement for reliable time step estimation. We investigate the critical time step in explicit dynamic simulations using trimmed B-splines, LR-splines, and THB-splines, based on a lumped mass matrix obtained by simple row summation. One-dimensional bar, two-dimensional membrane, and trimmed shell models are analyzed to determine how trimming and local refinement influence element and system eigenfrequencies, which directly control the stable time step. Refined boundary elements in open knot vectors are identified as the main bottleneck. Trimming these elements can increase the stable time step, though certain trimming configurations introduce new restrictions. Results show that LR- and THB-splines impose time step constraints similar to B-splines, making them equally suitable for explicit simulations. We also present a general method for computing element-wise B{\´e}zier extraction operators for LR- and THB-splines, enabling straightforward integration into standard finite element solvers. The findings are validated through nonlinear sheet-metal forming simulations in LS-DYNA using shells discretized with trimmed B-, LR-, and THB-splines. This represents the first such application and demonstrates their practical feasibility for industrial use.}, 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} } @article{SittlMarburgDeckersetal., author = {Sittl, Christopher and Marburg, Steffen and Deckers, Elke and Wagner, Marcus}, title = {Model order reduction for unbounded second-order vibroacoustic systems using infinite elements and Dirichlet-to-Neumann map}, series = {Computer Methods in Applied Mechanics and Engineering}, volume = {453}, journal = {Computer Methods in Applied Mechanics and Engineering}, publisher = {Elsevier}, doi = {10.1016/j.cma.2026.118846}, pages = {22}, abstract = {This work addresses the efficient numerical simulation of time-harmonic vibroacoustic problems in unbounded domains, with a focus on fluid-structure interaction. The underlying mathematical model is a second-order dynamical system arising from the coupling of structural and acoustic domains, incorporating material damping effects, relevant in structural acoustics and noise control applications. A central novelty of the proposed method is its unified computational framework that supports two distinct strategies for treating unbounded fluid domains: (1) non-local absorbing boundary conditions based on Dirichlet-to-Neumann map, and (2) infinite elements, which extend the computational domain rather than truncate it. Both approaches are integrated into a consistent formulation that enables flexible and accurate modeling of exterior wave propagation. To efficiently evaluate frequency-domain transfer functions, the method employs model order reduction using the Pad{\´e}-via-Lanczos technique. While this algorithm typically targets first-order systems, the present approach uses a Schur complement strategy to reduce the second-order system in a way that maintains computational efficiency and storage requirements comparable to first-order formulations. Importantly, the framework seamlessly embeds both interior structural damping and the additional dissipation introduced by the acoustic-domain truncation into the model-order reduction process. The exterior acoustic field is represented via spherical harmonic expansions, with expansion coefficients computed from the reduced system. Numerical results demonstrate the method's accuracy, efficiency, and scalability, making it well-suited for high-fidelity vibroacoustic analysis in unbounded domains.}, language = {en} } @misc{LehrerStockerDuddecketal., author = {Lehrer, Tobias and Stocker, Philipp and Duddeck, Fabian and Wagner, Marcus}, title = {Parametric CAD Models to Create an Infinite Number of Structural U-Shaped Sheet Metal Geometries [Data set]}, doi = {10.5281/ZENODO.14191374}, abstract = {The U-Channel Software is a python package containing four parametric models of U-shaped Sheet metals. The parameters refer to geometric dimensions, angles, fillet radii. Using the models, one can generate an infinite number of U-shaped sheet metal geometries, e.g. as STEP files. The parametric models are inspired by real-world shapes. A feasibility check is provided to ensure plausiblity of the shapes. Further features of the code are creating images of the geometries and generating geometries and images in parallel. For more details, refer to the repository and associated publications.}, language = {en} }