Materials Informatics and Data Science (MIDS)
The research group „Materials informatics and data sciences“ is currently being established. Key elements of the group are the application of statistical and machine learning methods as well as data science approaches to materials science problems mostly based on simulated data. With this focus the new research group strengthens and extends the “informatics” activities at ICAMS. Applications of these methods focus on bridging between simulation methods like atom-continuum or mesoscale-continuum and connecting knowledge from simulations to experiments. Current materials science research questions are related to plasticity of crystalline materials. In the past and present this includes mechanical aspects of tribological contacts, strain hardening and dislocation multiplication, dislocation grain boundary interaction, development of interatomic potentials to assess energetics and dynamics of crystalline defects on the atomic level, and data fusion from simulation and experiment to understand and quantify observed material behavior from simulations and experiments. Research questions on the methodical side revolve around appropriate data formats, standardization and representation of information.
Snapshot from a nudged elastic band (NEB) simulation of the basal cross-slip process in pure Magnesium calculated with a Behler-Parrinello neural network interatomic potential. Colors indicate crystallographic arrangement with green: face centred cubic; blue: body centred cubic; grey: others; hexagonally close packed atoms are not shown.
- Plasticity modeling with mesoscale simulations and scale bridging
- Validation of interatomic potentials based on machine-learning methods
- Simulation of atomistic processes
- Data fusion from simulations and experiments
Prof. Dr. Markus Stricker
Tel: +49 234 32 29377