- Organisation
- Minisymposia
- Participation
- Contact
- Archive
Keywords: Computational homogenization; Microstructure modeling; Multi-scale simulation; Multi-field coupling; Neural networks in material modeling
Organizers:
M. Vorwerk (1) – maximilian.vorwerk@uni-due.de
E. Donval (2) – elodie.donval@uni-due.de
B.H. Nguyen (2) – binh.nguyen@uni-due.de
M. Krause (2) – maximilian.krause@kit.edu
Affiliations:
(1) Institute of Mechanics, University of Duisburg-Essen, Germany
(2) Institute of Engineering Mathematics, University of Duisburg-Essen, Germany
Abstract:
Composite materials are ubiquitous in various engineering applications, such as ultra-lightweight fiber composites that are exceptionally robust, multi-field coupled functional materials for transducing smart devices or high-performance magnetic materials that achieve peak performance through particularly pronounced grain boundary layers. These composites are characterized by particularly pronounced interactions of different physical fields across various length-scales. Their behavior also relies heavily on the precise description of their complex microstructure. Therefore, classical analytical homogenization methods often fail to describe the field at the microscopic scale precisely and to provide an accurate macroscopic response.
Several computational methods, including Finite-Element and FFT-based homogenization or FE² approaches for instance, have been introduced to bridge the scales more accurately. In recent times, Neural-Network-based approaches also aim at providing a precise description of the micro and macro fields while reducing the computational costs associated with the previous approaches. This mini-symposium aims to cultivate scientific exchange among researchers across various fields and striving to accurately represent the different physical fields at different length scales. The primary emphasis remains on bridging various length scales to interact and effectively coupling their corresponding physical fields.
Topics of interest include, but are not limited to: