Invited Lecture
From experiment to simulation and back – data science approaches for understanding plasticity

Experimental observations and simulation data should – in principle – help to shed light on the „inner workings“ of a physical system, say, a material or specimen. There, the „inner workings“ would be the interaction of microstructural features (such as dislocations) among themselves, with the surfaces of the specimen, or with phase boundaries, to name but a few. Both experiment and simulation, however, suffer from particular problems which in many situations makes it difficult to directly compare them or to use results from one as input or support for the other.

In this presentation, we will start by giving an overview over current attempts for integrating experiment and simulation in the context of dislocation plasticity. We will then demonstrate, on the one hand, how data science approaches might be used to access data from experiments that would be otherwise inaccessible and, on the other hand, how data science also might help to reduce the high level of abstraction inherent to most simulations. With those methods, experiment and simulation might get a little closer to each, thereby helping to understand relevant mechanisms in plasticity and fracture from a new point of view.