Modeling of Glasses: An Overview

Glass modeling includes unique advantages and challenges with respect to other fields of materials modeling owing to lack of long-range order, strong dependence on temperature and pressure history, statistical nature of glassforming liquid, and the availability of almost the entire periodic table for constituents in glass. In this chapter, we introduce a range of methods used for glass modeling and overcoming these challenges. We first briefly compare how glass modeling is different from crystalline materials. Next, we briefly outline some of the techniques used for modeling glass and finally present the outstanding challenges in glass modeling and design. As glass modeling merges empirical techniques (i.e., data-driven machine learning, finite element models for mechanical and acoustic properties, composition/property/processing relationships) with fundamental physical methods (i.e., statistical physics, diffusion, first principles quantum mechanical theories, energy landscapes), many orders of magnitude in time- and length scales may be simultaneously modeled across vast composition spaces whose experimental exploration would be prohibitively expensive.