Skip to main content

Time-scale bridging potentials for realistic molecular dynamics simulations

Periodic Reporting for period 4 - TIME-BRIDGE (Time-scale bridging potentials for realistic molecular dynamics simulations)

Reporting period: 2019-03-01 to 2020-12-31

The main aim of TIME-BRIDGE is to design novel methodologies that will enable a more realistic simulation of materials. These simulations will open the pathway to a further improvement of materials properties such as, e.g. strength or deformability. Present-day atomistic simulations face the enormous challenge of bridging the time scale from the atomic motion—i.e. thermal vibrations which happen on the time scale of femtoseconds—to the time scale of seconds where processes relevant for materials properties take place, the so-called rare events. The main idea of TIME-BRIDGE is to develop a special type of potentials that treat the troublesome thermal vibrations in an effective manner. Such potentials are known from the field of electronic structure calculations where they replace effectively the electronic motion. Within TIME-BRIDGE this concept is transferred to the atomic motion. In addition to the time scale problem, also the length scale problem present in atomistic simulations is addressed within the TIME-BRIDGE project. Finally, the third part of TIME-BRIDGE is a complementary experimental branch. In this part, we perform idealized experiments that will serve as a well-defined reference to our new atomistic approaches, allowing to scrutinize critically their performance.

Solving the time and length scale problems is a main objective of the project as it will enable better, more realistic atomistic computer simulations of materials. These simulations will allow materials scientists to improve the properties of materials. The benefit for society can be expected to be substantial in many respects. For example, improved materials properties can lead to safer cars or airplanes. They can also lead to lighter cars or airplanes and thus decrease energy consumption.
Multiple research directions have been initiated in the first funding period of TIME-BRIDGE in order to reach our ambitious goal. A main route of development was centered around the optimization of the so-called order parameter. The order parameter is crucial in the development of the potentials because it enables to distinguish the thermal motion of atoms from the rare events. Presently employed order parameters have problems with determining certain important rare events. We are currently working on a better definition of the order parameter and on a new, automated procedure for detecting the rare events. For that purpose, we make for example use of techniques from the field of machine learning that are able to process quickly large datasets.

Progress has been achieved also in the research direction focusing on the length scale problem, for example, by developing new and improved atomic interactions or by coupling the atomistic domain to larger length scale continuum approaches. As for the experiments, miniaturized, or so-called small scaled compression experiments have been performed on copper single crystalline micro pillars oriented in two distinct crystallographic orientations, which is a well-studied system and consequently serves as an idealized model system to study the processes simulated with our atomistic developments.
Within all three branches of TIME-BRIDGE we have gone beyond state of the art with our developments. The order parameter investigated within the time scale bridging part is based on a novel approach that has not been applied in this community before. The general concept has been known in a distinct research community and, in collaboration with colleagues from Imperial College, we have transferred this concept to the definition of the order parameter used for describing rare events. The new coupling scheme implemented in the length scale part of TIME-BRIDGE is novel with respect to its adaptive coupling that can follow travelling defects inside the material. The experiments go beyond state of the art with respect to their idealization and with respect to the size of the investigated pillars.