TY - JOUR A1 - Franke, Markus A1 - Wagner, Marcus T1 - Transient surrogate modeling of modally reduced structures with discontinuous loads and damping JF - Archive of Applied Mechanics N2 - In this work, a surrogate model for structural, transient and discontinuously excited finite element method simulations is developed. This allows to reduce the computational effort of repeated calculations of identical models under different load cases. The architecture of the surrogate combines fully connected neural network layers with long short-term memory layers. For the reproduction of different damping ratios, a categorical variable is added to the continuous input data. Based on a recursive flow of the predicted data back to the input layer, long-term dependencies do not vanish due to short-input sequences. The system dimension is reduced by applying the model-order reduction technique for modal decomposition. The high accuracy of the surrogate and the reduction of computational costs are shown on an academic example of a cantilever beam and a real-world example of a robot. The advantages of our approach are illustrated in comparison with state-of-the-art surrogates for transient finite element analysis. By using the surrogate proposed in this study, oscillations due to discontinuous excitation of mechanical structures can be reproduced. For this purpose, only short-input sequences are necessary since the excitation of the oscillations does not have to be part of the input sequence during the whole duration of the oscillations. Due to the categorical variable for the damping ratio, the surrogate can account for the influence of different damping in parameter studies. KW - Discontinuous loads KW - Modal decomposition KW - Surrogate modeling KW - Long short-term memory KW - Finite element method Y1 - 2024 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-72834 N1 - Corresponding author der OTH Regensburg: Markus Franke VL - 94 IS - 5 PB - Springer Nature ER - TY - CHAP A1 - Buchner, Stefan A1 - Wagner, Marcus A1 - Höller, Christoph T1 - Inverse Acoustic Characterization of Rigid Porous Media using Artificial Neural Networks T2 - Proceedings of DAS|DAGA 2025, 51st Annual Meeting on Acoustics, March 17-20, 2025, Copenhagen N2 - Porous sound absorbers can be described by the isotropic Biot model, and its fluid phase can be represented by the Champoux-Allard model. To apply these models to a given absorber material, the five acoustical and the four mechanical material parameters must be known. The direct measurement of these parameters is complex and requires cost-intensive measurement equipment. Current inverse methods to obtain these material parameters solve an optimization problem, trying to fit the absorption or impedance curve of the material model to the impedance tube measurement data. Solving this optimization problem, i.e. finding the global minimum, is not guaranteed in an acceptable amount of time, as the optimization problem possesses a multitude of local minima. This work proposes an alternative, data driven approach using artificial neural networks to obtain the material parameters necessary for the characterization of open porous materials. The approach only requires the results of standard impedance tube measurements. The characterization of rigid- and elastic frame materials has been investigated. The datasets were generated using the rigid and elastic frame models for porous absorbers. The approach shows good results for impedance curves generated by the analytical models, the validation with real-world impedance tube measurement data is currently under investigation. Y1 - 2025 SN - 978-3-939296-23-2 U6 - https://doi.org/10.71568/dasdaga2025.142 SP - 847 EP - 850 CY - Berlin ER - TY - JOUR A1 - Scharfenberg, Georg A1 - Mottok, Jürgen A1 - Artmann, Christina A1 - Hobelsberger, Martin A1 - Paric, Ivan A1 - Großmann, Benjamin A1 - Pohlt, Clemens A1 - Wackerbarth, Alena A1 - Pausch, Uli A1 - Heidrich, Christiane A1 - Fadanelli, Martin A1 - Elsner, Michael A1 - Pöcher, Daniel A1 - Pittroff, Lenz A1 - Beer, Stefan A1 - Brückl, Oliver A1 - Haslbeck, Matthias A1 - Sterner, Michael A1 - Thema, Martin A1 - Muggenthaler, Nicole A1 - Lenck, Thorsten A1 - Götz, Philipp A1 - Eckert, Fabian A1 - Deubzer, Michael A1 - Stingl, Armin A1 - Simsek, Erol A1 - Krämer, Stefan A1 - Großmann, Benjamin A1 - Schlegl, Thomas A1 - Niedersteiner, Sascha A1 - Berlehner, Thomas A1 - Joblin, Mitchell A1 - Mauerer, Wolfgang A1 - Apel, Sven A1 - Siegmund, Janet A1 - Riehle, Dirk A1 - Weber, Joachim A1 - Palm, Christoph A1 - Zobel, Martin A1 - Al-Falouji, Ghassan A1 - Prestel, Dietmar A1 - Scharfenberg, Georg A1 - Mandl, Roland A1 - Deinzer, Arnulf A1 - Halang, W. A1 - Margraf-Stiksrud, Jutta A1 - Sick, Bernhard A1 - Deinzer, Renate A1 - Scherzinger, Stefanie A1 - Klettke, Meike A1 - Störl, Uta A1 - Wiech, Katharina A1 - Kubata, Christoph A1 - Sindersberger, Dirk A1 - Monkman, Gareth J. A1 - Dollinger, Markus A1 - Dembianny, Sven A1 - Kölbl, Andreas A1 - Welker, Franz A1 - Meier, Matthias A1 - Thumann, Philipp A1 - Swidergal, Krzysztof A1 - Wagner, Marcus A1 - Haug, Sonja A1 - Vernim, Matthias A1 - Seidenstücker, Barbara A1 - Weber, Karsten A1 - Arsan, Christian A1 - Schone, Reinhold A1 - Münder, Johannes A1 - Schroll-Decker, Irmgard A1 - Dillinger, Andrea Elisabeth A1 - Fuchshofer, Rudolf A1 - Monkman, Gareth J. A1 - Shamonin (Chamonine), Mikhail A1 - Geith, Markus A. A1 - Koch, Fabian A1 - Ühlin, Christian A1 - Schratzenstaller, Thomas A1 - Saßmannshausen, Sean Patrick A1 - Auchter, Eberhard A1 - Kriz, Willy A1 - Springer, Othmar A1 - Thumann, Maria A1 - Kusterle, Wolfgang A1 - Obermeier, Andreas A1 - Udalzow, Anton A1 - Schmailzl, Anton A1 - Hierl, Stefan A1 - Langer, Christoph A1 - Schreiner, Rupert ED - Baier, Wolfgang T1 - Forschungsbericht 2015 T3 - Forschungsberichte der OTH Regensburg - 2015 Y1 - 2015 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-13867 SN - 978-3-00-048589-3 CY - Regensburg ER - TY - JOUR A1 - Kaps, Arne A1 - Lehrer, Tobias A1 - Lepenies, Ingolf A1 - Wagner, Marcus A1 - Duddeck, Fabian T1 - Multi-fidelity optimization of metal sheets concerning manufacturability in deep-drawing processes JF - Structural and Multidisciplinary Optimization N2 - Multi-fidelity optimization, which complements an expensive high-fidelity function with cheaper low-fidelity functions, has been successfully applied in many fields of structural optimization. In the present work, an exemplary cross-die deep-drawing optimization problem is investigated to compare different objective functions and to assess the performance of a multi-fidelity efficient global optimization technique. To that end, hierarchical kriging is combined with an infill criterion called variable-fidelity expected improvement. Findings depend significantly on the choice of objective function, highlighting the importance of careful consideration when defining an objective function. We show that one function based on the share of bad elements in a forming limit diagram is not well suited to optimize the example problem. In contrast, two other definitions of objective functions, the average sheet thickness reduction and an averaged limit violation in the forming limit diagram, confirm the potential of a multi-fidelity approach. They significantly reduce computational cost at comparable result quality or even improve result quality compared to a single-fidelity optimization. KW - Multi-fidelity optimization KW - Efficient global optimization KW - Sheet metal forming KW - Deep drawing Y1 - 2023 U6 - https://doi.org/10.1007/s00158-023-03631-8 VL - 66 PB - Springer Nature ER - TY - GEN A1 - Lehrer, Tobias A1 - Kaps, Arne A1 - Daiminger, Tobias A1 - Lepenies, Ingolf A1 - Duddeck, Fabian A1 - Wagner, Marcus T1 - Analysis of Global Sensitivities for One-Step and Multi-step Deep-Drawing Simulations T2 - 16th German LS-DYNA Forum : October 11-13, 2022, Bamberg, Germany and online : Book of Abstracts N2 - In practical use cases, simulation engineers are confronted with uncertainties in the simulation parameters. Normally, trust in a model is built from experience, practical assumptions, and parameter studies. This approach, though, is based on the assumption that few parameter combinations are sufficient to represent the whole design space. This lacks an appraisable mathematical basis. To get insights into which parameters most strongly affect the results, a global sensitivity analysis can be conducted [2, 3]. The results are utilized to rank the most influential parameters and to filter less relevant ones. This gives feedback which improved set of input data will lead to more certainty in the simulation results. To enable this in the framework of multi-fidelity analysis and optimization, we compare here global sensitivities and uncertainties of the implicit One-Step approach (low-fidelity) with those of the explicit multi-step deep drawing approach (high-fidelity). Y1 - 2022 UR - https://www.dynamore.de/de/fortbildung/konferenzen/vergangene/16-ls-dyna-forum-2022/2022-proceedings.pdf SN - 978-3-9816215-8-7 SP - 117 EP - 118 PB - DYNAmore GmbH CY - Stuttgart ER - TY - JOUR A1 - Mauerer, Wolfgang A1 - Rexhepaj, Tanja A1 - Monkman, Gareth J. A1 - Sindersberger, Dirk A1 - Diermeier, Andreas A1 - Neidhart, Thomas A1 - Wolfrum, Dominik A1 - Sterner, Michael A1 - Heberl, Michael A1 - Nusko, Robert A1 - Maier, Georg A1 - Nagl, Klaus A1 - Reuter, Monika A1 - Hofrichter, Andreas A1 - Lex, Thomas A1 - Lesch, Florian A1 - Kieninger, Bärbel A1 - Szalo, Alexander Eduard A1 - Zehner, Alexander A1 - Palm, Christoph A1 - Joblin, Mitchell A1 - Apel, Sven A1 - Ramsauer, Ralf A1 - Lohmann, Daniel A1 - Westner, Markus A1 - Strasser, Artur A1 - Munndi, Maximilian A1 - Ebner, Lena A1 - Elsner, Michael A1 - Weiß, Nils A1 - Segerer, Matthias A1 - Hackenberg, Rudolf A1 - Steger, Sebastian A1 - Schmailzl, Anton A1 - Dostalek, Michael A1 - Armbruster, Dominik A1 - Koch, Fabian A1 - Hierl, Stefan A1 - Thumann, Philipp A1 - Swidergal, Krzysztof A1 - Wagner, Marcus A1 - Briem, Ulrich A1 - Diermeier, Andreas A1 - Spreitzer, Stefan A1 - Beiderbeck, Sabrina A1 - Hook, Christian A1 - Zobel, Martin A1 - Weber, Tim A1 - Groß, Simon A1 - Penzkofer, Rainer A1 - Dendorfer, Sebastian A1 - Schillitz, Ingo A1 - Bauer, Thomas A1 - Rudolph, Clarissa A1 - Schmidt, Katja A1 - Liebetruth, Thomas A1 - Hamer, Markus A1 - Haug, Sonja A1 - Vernim, Matthias A1 - Weber, Karsten A1 - Saßmannshausen, Sean Patrick A1 - Books, Sebastian A1 - Neuleitner, Nikolaus A1 - Rechenauer, Christian A1 - Steffens, Oliver A1 - Kusterle, Wolfgang A1 - Gömmel, Roland A1 - Wellnitz, Felix A1 - Stierstorfer, Johannes A1 - Stadler, Dominik A1 - Hofmann, Matthias J. A1 - Motschmann, Hubert A1 - Shamonin (Chamonine), Mikhail A1 - Bleicher, Veronika A1 - Fischer, Sebastian A1 - Hackenberg, Rudolf A1 - Horn, Anton A1 - Kawasch, Raphael A1 - Petzenhauser, Michael A1 - Probst, Tobias A1 - Udalzow, Anton A1 - Dams, Florian A1 - Schreiner, Rupert A1 - Langer, Christoph A1 - Prommesberger, Christian A1 - Ławrowski, Robert Damian ED - Baier, Wolfgang T1 - Forschungsbericht 2016 T3 - Forschungsberichte der OTH Regensburg - 2016 Y1 - 2016 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-13840 CY - Regensburg ER - TY - JOUR A1 - Lehrer, Tobias A1 - Stocker, Philipp A1 - Duddeck, Fabian A1 - Wagner, Marcus T1 - UCSM: Dataset of U-shaped parametric CAD geometries and real-world sheet metal meshes for deep drawing JF - Computer-aided design N2 - 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. Y1 - 2025 U6 - https://doi.org/10.1016/j.cad.2025.103924 VL - 188 PB - Elsevier ER - TY - JOUR A1 - Lehrer, Tobias A1 - Kaps, Arne A1 - Lepenies, Ingolf A1 - Raponi, Elena A1 - Wagner, Marcus A1 - Duddeck, Fabian T1 - Complementing Drawability Assessment of Deep-Drawn Components with Surrogate-Based Global Sensitivity Analysis JF - ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part B: Mechanical Engineering N2 - In the early-stage development of sheet metal parts, key design properties of new structures must be specified. As these decisions are made under significant uncertainty regarding drawing configuration changes, they sometimes result in the development of new parts that, at a later design stage, will not be drawable. As a result, there is a need to increase the certainty of experience-driven drawing configuration decisions. Complementing this process with a global sensitivity analysis can provide insight into the impact of various changes in drawing configurations on drawability, unveiling cost-effective strategies to ensure the drawability of new parts. However, when quantitative global sensitivity approaches, such as Sobol's method, are utilized, the computational requirements for obtaining Sobol indices can become prohibitive even for small application problems. To circumvent computational limitations, we evaluate the applicability of different surrogate models engaged in computing global design variable sensitivities for the drawability assessment of a deep-drawn component. Here, we show in an exemplary application problem, that both a standard kriging model and an ensemble model can provide commendable results at a fraction of the computational cost. Moreover, we compare our surrogate models to existing approaches in the field. Furthermore, we show that the error introduced by the surrogate models is of the same order of magnitude as that from the choice of drawability measure. In consequence, our surrogate models can improve the cost-effective development of a component in the early design phase. KW - sheet metal forming KW - deep drawing KW - global sensitivity analysis KW - variance-based sensitivity analysis KW - metamodeling Y1 - 2024 U6 - https://doi.org/10.1115/1.4065143 SN - 2332-9025 SP - 1 EP - 10 PB - ASME ER - TY - JOUR A1 - Lehrer, Tobias A1 - Kaps, Arne A1 - Lepenies, Ingolf A1 - Duddeck, Fabian A1 - Wagner, Marcus T1 - 2S-ML: A simulation-based classification and regression approach for drawability assessment in deep drawing JF - International Journal of Material Forming N2 - New structural sheet metal parts are developed in an iterative, time-consuming manner. To improve the reproducibility and speed up the iterative drawability assessment, we propose a novel low-dimensional multi-fidelity inspired machine learning architecture. The approach utilizes the results of low-fidelity and high-fidelity finite element deep drawing simulation schemes. It hereby relies not only on parameters, but also on additional features to improve the generalization ability and applicability of the drawability assessment compared to classical approaches. Using the machine learning approach on a generated data set for a wide range of different cross-die drawing configurations, a classifier is trained to distinguish between drawable and non-drawable setups. Furthermore, two regression models, one for drawable and one for non-drawable designs are developed that rank designs by drawability. At instantaneous evaluation time, classification scores of high accuracy as well as regression scores of high quality for both regressors are achieved. The presented models can substitute low-fidelity finite element models due to their low evaluation times while at the same time, their predictive quality is close to high-fidelity models. This approach may enable fast and efficient assessments of designs in early development phases at the accuracy of a later design phase in the future. KW - One-step KW - Deep drawing KW - Drawability KW - Meta-modeling KW - Machine learning Y1 - 2023 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-64664 N1 - Corresponding author: Tobias Lehrer VL - 16 SP - 1 EP - 17 PB - Springer ER - TY - GEN A1 - Lehrer, Tobias A1 - Stocker, Philipp A1 - Duddeck, Fabian A1 - Wagner, Marcus T1 - Comparison of Low- vs. High-Dimensional Machine Learning Approaches for Sheet Metal Drawability Assessment T2 - Third International Conference on Computational Science and AI in Industry (CSAI 2023), Trondheim, Norway, 28-30 August 2023 N2 - 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. Y1 - 2023 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-64772 PB - International Centre for Numerical Methods in Engineering ER -