DETECTION OF SINGULARITIES IN FINITE ELEMENT SIMULATIONS: A 3D DEEP LEARNING APPROACH
Bickel S, Götz S, Wartzack S (2023)
Publication Type: Conference contribution
Publication year: 2023
Publisher: American Society of Mechanical Engineers (ASME)
Book Volume: 2
Conference Proceedings Title: Proceedings of the ASME Design Engineering Technical Conference
Event location: Boston, MA
ISBN: 9780791887295
Abstract
The growing digitalization trend has a great influence and impact on the entire product development domain. The new algorithms and methods transform and improve many established processes and workflows. In addition to converting existing know-how into the new processes, available data can also be used to implement various models to enhance the development process. At the same time, the lack of experienced engineers leads to an increasing number of novice engineers from the field of 3D design, who have to perform tasks from the simulation department. Therefore, support for these less experienced users of product development software through newly available methods from the field of data-driven methods is reasonable. Correspondingly, the goal of this paper is to present a method for singularity detection in Finite Element (FE) simulations that detects the presence of a singularity in a computed simulation with the help of Deep Learning. For this purpose, a dataset of calculated simulations with several components with and without singularities is first generated. Then, a new approach based on PointNet and labeled datasets of FE simulations is proposed. Afterwards, the new procedure is compared against a classic Machine Learning approach, whereby different parameter settings were investigated for both approaches.
Authors with CRIS profile
Sebastian Bickel
Chair of Engineering Design
Stefan Götz
Chair of Engineering Design
Sandro Wartzack
Chair of Engineering Design
How to cite
APA:
Bickel, S., Götz, S., & Wartzack, S. (2023). DETECTION OF SINGULARITIES IN FINITE ELEMENT SIMULATIONS: A 3D DEEP LEARNING APPROACH. In Proceedings of the ASME Design Engineering Technical Conference. Boston, MA, US: American Society of Mechanical Engineers (ASME).
MLA:
Bickel, Sebastian, Stefan Götz, and Sandro Wartzack. "DETECTION OF SINGULARITIES IN FINITE ELEMENT SIMULATIONS: A 3D DEEP LEARNING APPROACH." Proceedings of the ASME 2023 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2023, Boston, MA American Society of Mechanical Engineers (ASME), 2023.
BibTeX: Download