Relation of the RQs in PN3
Particle- and continuum-based simulations are important tools for understanding and designing materials, biological matter, and technical systems. PN 3 investigated how their predictive capabilities could be improved by integrating data from experiments, simulations, and growing scientific databases across different scales.
The research focused on combining data-assisted model reduction with physically informed machine learning. The aim was to bridge scales, reduce the complexity of continuum models and develop data-driven models that retained essential physical properties, thereby enabling more efficient and predictive simulations for materials design and other scientific applications.
Research Focus
The research addressed three closely connected questions:
- Scale bridging for particle models: How could data-assisted model reduction connect detailed particle-level descriptions with models on larger scales?
- Complexity reduction in continuum models: How could data be used to reduce the complexity of continuum-based models while retaining their essential behaviour?
- Machine-learned models: How could experimental and simulation data be used to develop machine-learned particle and continuum models that respected physical constraints?
Projects
| Data-integrated scale bridging for all-solid state batteries: micro- to mesoscale | Felix Fritzen (Blazej Grabowski) |
| Reduced model reanalysis for path-dependent multifield problems | Manfred Bischoff (Jörg Fehr) |
| Electrokinetic characterization of conducting two-phase flows in model porous media | Christian Holm (Holger Steeb) |
| Simulation of surface processes using machine-learned potentials | Johannes Kästner (Germán Molpeceres, Kristyna Pluhackova) |
| Data-driven multi-scale stability analysis of multi-stimuli-responsive hydrogels | Marc-André Keip (Tim Ricken, Felix Fritzen) |
| Pore topology and surface design for energy storage applications | Niels Hansen, Joachim Groß |
| Data-integrated scale bridging for all-solid state batteries: From electrons to atoms | Blazej Grabowski (Felix Fritzen, Yuji Ikeda) |
| Biological Molecular Dynamics Simulations 2.0 | Kristyna Pluhackova |
| Data-integrated simulation of magnetic gels | Rudolf Weeber (Dirk Pflüger) |
| A data-driven approach to viscous fluid dynamics | Christina Lienstromberg (Marc-André Keip) |
| Approximation and learning density matrices | Benjamin Stamm (Johannes Kästner) |
| Bottom-up modeling of conducting porous materials via molecular simulation | Alexander Schlaich (Holger Steeb) |
| Soft deformation sensor with high spatial resolution through impedance spectroscopy | Philipp Rothemund (Felix Fritzen) |
| Processing uncertain microstructural data | Felix Fritzen (Andrea Barth) |
| Data-based model reduction and reanalysis | Manfred Bischoff (Jörg Fehr) |
| Electrokinetic characterization of conducting two-phase flows in model porous media | Christian Holm (Holger Steeb) |
| Characterization of potential energy surfaces using machinge-learning techniques | Johannes Kästner (Bernard Haasdonk) |
| Data-driven surrogate modeling of structural instabilities in electroactive polymers | Marc-André Keip (Tim Ricken) |
| Data-integrated simulation of enzymes | Jürgen Pleiss (Niels Hansen) |
| Strengthening mechanisms of Cu-Ni-Si-Cr alloys through simulations and machine learning potentials | Maria Fyta (Siegfried Schmauder) |
| Transferable force fields and transport properties | Niels Hansen, Joachim Groß |
| Simulations of hydrogen embrittlement in Ni-based super alloys | Siegfried Schmauder (Blazej Grabowski, Maria Fyta) |
Associated projects
| Data-supported simulation and surrogate modeling of mechanical systems | Felix Fritzen |
| Materials 4.0 | Blazej Grabowski |
| Molecular transport in nanoporous materials | Kristyna Pluhackova |
| Advanced learning strategies for machine learned interatomic potentials | Christian Holm |
| Certified coupled model order reduction (CCMOR) | Jörg Fehr |
| Bottom-up modelling of COF/electrode systems for CO2 reduction in confinement | Alexander Schlaich |
| Data-driven investigation of three-dimensional instabilities in magneto-active thin films heterogeneously patterned by design | Marc-André Keip, Felix Fritzen |
| Machine learning methods for the simulation of physical systems | Christian Holm |
| Simulation and experimental validation of residual stresses in laser-generated composite materials | Felix Fritzen |
| AI software tools for materials development | Johannes Kästner, Dirk Pflüger |
Project Network Coordinators
Felix Fritzen
Prof. Dr.-Ing. Dipl.-Math. techn.Data Analytics in Engineering
Blazej Grabowski
Prof. Dr. rer. nat.Computational Materials Science