The Computational Materials Design Facility (CMDF): A powerful framework for multi-paradigm multi-scale simulations
- PDF / 163,401 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 78 Downloads / 153 Views
0894-LL03-03.1
The Computational Materials Design Facility (CMDF): A powerful framework for multiparadigm multi-scale simulations Markus J. Buehler1,2, Jef Dodson2, Adri C.T. van Duin2, Peter Meulbroek2, William A. Goddard III2 1 2
Massachusetts Institute of Technology, Cambridge, MA 02139, USA Materials and Process Simulation Center, California Institute of Technology, Pasadena, CA, 91125, CA
ABSTRACT Predicting the properties and behavior of materials by computer simulation from a fundamental, ab initio perspective has long been a vision of computational material scientists. The key to achieving this goal is utilizing hierarchies of paradigms and scales that connect macrosystems to first principles quantum mechanics (QM). Here we describe a new software environment, the “Computational Materials Design Facility” (CMDF), capable of simulations of complex materials studies using a variety of simulation paradigms. The CMDF utilizes a Python scripting layer to integrate different computational tools to develop multi-scale simulation applications. We have integrated DFT QM methods, the first principles ReaxFF reactive force field, empirical all atom force fields (FFs), mesoscale and continuum methods. The central data structure Extended OpenBabel (XOB) plays a critical role as glue between applications. We demonstrate the usefulness of CMDF in examples that couple complex chemistry and mechanical properties during dynamical failure processes, as for example in a study of cracking of Ni under presence of O2. Figure 1: Schematic representation of the Computational Materials Design Facility (CMDF). To calculate materials properties, fundamental information is obtained from the QM level, and then used to train the ReaxFF level, which in turn is used to train ordinary FF and mesoscale levels to inform the macroscale simulations needed for engineering design. In principle, this scheme allows the process to be inverted. CMDF allows re-usage of software for multiscale applications. The concepts developed within the CMDF framework will help with seamlessly integrated multiscale modeling.
INTRODUCTION Calculating macroscopic properties of materials from first principles or ab-initio computations is one of the foremost goals of computational materials science. There have been enormous advances in efficient and accurate quantum mechanical (QM) methods [1], new accurate force fields for molecular dynamics (MD) [2-5], coarse grain descriptions for treating very large systems, and techniques to couple scales and paradigms (e.g. [6-8]). Furthermore, coupling these advances in methods with the enormous growth of computing power has enabled studies with millions to billions of particles [9]. However, remains cumbersome to apply all these scales of simulation to the same problem in order to transcend completely from the electrons of QM to continuum plasticity, within a single model. We report development of the Computational Materials Design Facility (CMDF), a new multi-scale multi-paradigm simulation framework. The CMDF is developed
Data Loading...
