MedeA: An environment for ICME starting with atoms, and what we can learn from it

P. Saxe
Materials Design, Inc.,
United States

Keywords: ICME, ab initio, DFT, forcefields, modeling, simulation

Summary:

ICME is a grand vision, but there is an enormous amount of work needed to bring together experiment and simulation across the entire spectrum of length- and time-scales to go from atoms to cars, airplanes, computer chips or any other manufactured product. The challenge is not only to pick the most adequate from the multitude of theoretical methods and computing codes to complement the experiments, but also to integrate them in a coherent framework delivering results of reliable accuracy – and effectively harnessing growing computing power. For the last two decades we have been advancing the field by systematically developing an integrated simulation environment that lives up to that goal. It embraces the best of the quantum mechanical as well as forcefield methods, linking the two through tools creating forcefield parameters from quantum results – thus bridging the scale from electrons to atoms to molecules. Such integration boosts the spectrum of the few chosen methods and, combined with the experimental databases, results in an unprecedented range of applications, all within a single environment. With this foundation solidly in place, we are reaching up to the next scales with TCAD in electronics and CALPHAD methods for structural materials. In this talk we will briefly cover what has been accomplished to date, focusing mainly on the common themes that emerge over time. Examples of materials research in a variety of areas such as metallurgy, electronics, catalysis and polymers will be used to illustrate the points. The attention will then turn to the challenges ahead – what they are, and what the technical, societal and financial hurdles that will have to be overcome to sustain this effort, as well as the more general ICME effort, over the long time it will require to reach the goals laid out by e.g. the Materials Genome Initiative.