TY - JOUR A1 - Ledentsov, Dmitry A1 - Düster, Alexander A1 - Volk, Wolfram A1 - Wagner, Marcus A1 - Heinle, Ingo A1 - Rank, Ernst T1 - Model adaptivity for industrial application of sheet metal forming simulation JF - Finite Elements in Analysis and Design N2 - finite element simulation of sheet metal forming, shell elements are widely used. The limits of applicability of the shell elements are sometimes disregarded, which leads to an error in predictions of important values such as springback geometry. The underlying kinematic assumptions of the shell elements do not hold where the thickness of In the metal sheet approaches the value of the radius of curvature. Complex three-dimensional material behavior effects cannot be represented precisely as the result of the simplified kinematics. Here we present a model adaptivity scheme based on a model error indicator. The model-adaptive technique presented in this paper aides to resolve only the critical areas of the structure with a three-dimensional discretization while keeping reasonable computational cost by utilizing shell elements for the rest of the structure. The model error indicator serves as a guide for subsequent automatic adaptive re-meshing of the work-piece followed by a model-adaptive finite element analysis. The accuracy of the approximation obtained by the model-adaptive technique coincides well with that of a more expensive solution obtained with solid elements only. KW - Finite elements KW - Model adaptivity KW - Coupled analysis KW - Model error KW - Sheet metal forming Y1 - 2010 U6 - https://doi.org/10.1016/j.finel.2010.02.006 VL - 46 IS - 7 SP - 585 EP - 600 ER - TY - JOUR A1 - Wagner, Marcus A1 - Franke, Markus A1 - Krause, Tom Janis A1 - Heinle, Ingo T1 - Data augmentation of material properties for machine learning in industrial production - a case study in an automotive press shop JF - Engineering Computations : International journal for computer-aided engineering and software N2 - In the context of industrial production, the utilisation of data recording and processing techniques is becoming increasingly prevalent across the manufacturing sector. The solutions integrate sensors, facilitate the transmission of data, and enable data-driven decision-making, thereby reducing downtime and optimising quality. However, challenges emerge due to the limited non-transferable data or models between processes. Alterations to the production process can render collected data invalid, resulting in restricted datasets and potential overfitting. To address these issues, techniques such as data augmentation are employed. This study aims to develop a data augmentation methodology applicable in dynamic, data-scarce production environments, enhancing the robustness of regressor predictions. KW - Data augmentation KW - Machine learning KW - Manufacturing KW - Deep drawing Y1 - 2025 U6 - https://doi.org/10.1108/EC-08-2024-0787 PB - Emerald Publishing CY - Leeds ER -