GRK 2379:  Hierarchical and Hybrid Approaches in Modern Inverse Problems

• Subject Area: Mechanics and Constructive Mechanical Engineering ; Construction Engineering and Architecture ; Computer Science ; Mathematics
• Term: since 2018
• Website: Homepage
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 333849990

Computational methods permeate every aspect of engineering and science, from analysis to discovery and optimization. Their evolution continues at a rapid pace, driven not only by ever faster computing hardware, but also by our growing understanding of the true potential of computer-aided methods. Practical simulations transition from single numerical experiments towards robust predictive tools; models of isolated phenomena evolve into model hierarchies representing complex systems; numerical methods expand to deal with sensitivities with respect to parameters, uncertainties in those parameters and in models themselves. The educational system on all levels must keep up with and foster these advances; International Research Training Group 2379 has established, and continues to evolve, the required framework for doctoral training. The IRTG builds on a unique and complementary consortium, at RWTH Aachen University, with its School for Simulation and Data Science as an integral part of the Center for Simulation and Data Science (JARA-CSD), and at the University of Texas at Austin (UT), with its Oden Institute for Computational Engineering and Sciences. The projects are embedded in the field of modern inverse problems and introduce an innovative perspective into the education of future scientists and engineers. They are focusing on the challenges that arise in the interaction of the four specific themes: geometry, data, models, and applications. Within each theme, the expertise at one institution is significantly augmented by that of the other partner. As one example, in the area of geometry, we name the isogeometric concept championed by Oden Institute which is influencing many engineering analysis and design methods, but requires new approaches to mesh generation and sensitivity propagation. Furthermore, in modeling, hybrids of physics-driven and data-driven models and methods are starting to play a crucial role; UT's strength in data science, and the breadth of application domains at RWTH, are equally important to achieve the full potential of such hybrid approach.This training group unites experts in Aachen and Austin, providing unparalleled critical mass in computational engineering. Extensive experience with doctoral training on both sides of the partnership is being perfected. A joint research training with the optimal combination of structure and individuality, a tailored academic program of courses and colloquia, and a common supervision concept makes this transatlantic cooperation a success and a blueprint for future collaborations.

• DFG Programme: International Research Training Groups
• International Connection: USA
• Applicant Institution: Rheinisch-Westfälische Technische Hochschule Aachen
• IRTG-Partner Institution: The University of Texas at Austin
• Spokesperson: Professor Marek Behr, Ph.D.
• Participating Researchers: Professor Dr. Benjamin Berkels; Professorin Dr.-Ing. Stefanie Elgeti; Professor Dr. Michael Herty; Professor Dr. Leif Kobbelt; Professorin Dr. Julia Kowalski; Professor Alexander Mitsos, Ph.D.; Professorin Dr. Abigail Morrison; Professor Dr.-Ing. Raul Tempone, Ph.D.; Professor Dr. Manuel Torrilhon; Professor Florian Wellmann, Ph.D.
• IRTG-Partner: Spokesperson: Professor Leszek F. Demkowicz, Ph.D.