Multiscale modeling of homo- and heterogeneous system

Abstract

As a way of solving physical problems with important features at multiple scales, multiscale modeling is widely used in engineering, physics, meteorology, computer science and so forth. Multiscale simulation is to be attempted by using a combination of Molecular Dynamics (MD) and Direct Simulation Monte Carlo (DSMC) and which represent micro- and meso-scale in dimension, respectively. In this thesis, we introduce two different projects in different scales. In the first project, we study the physical and thermodynamic properties of crystals with defects. In particular, vacancies and their effects to the materials are our interest. Etomica, which is a Java-based open source package, is used to simulate atom-based crystals of interest. A statistical ensemble method, which was developed by Pronk and Frenkel for monovacancy, has been modified to calculate different degree of vacancies. In the second project, we try to solve the problems in a fluid flow with the DSMC algorithm which is widely used in a rarefied gas. A modified DSMC method is attempted to obtain correct transport properties in a gas flow to create a new equation of state for a real gas. In this regard, this work contains two creative points. The first is that one can control to study the degree of vacancy from mono-to-poly so that the extreme case of void in crystal can be studied properly. The second is to break the traditional inconsistency that DSMC yields transport properties for a real gas yet has an ideal gas equation of state, so that it recovers the exact hard sphere (HS) equation of state.