Glass modeling has unique advantages and challenges over other areas of material modeling. This is due to the absence of a long-range or a strong dependence on temperature and pressure history, statistical nature of the glass-forming fluid, and availability. The glass periodic table contains almost all the constituents.
There are many ways to shape glass. Modeling glass is very different from modeling crystalline materials. These are some of the methods used to form glass.
Modeling glass uses empirical techniques to fuse them (i.e. Data-driven machine learning, finite elements models for mechanical and/or acoustic properties. Composition / property/ processing relationships. statistical physics, diffusion, first principles, theories and energy landscapes of quantum mechanics.
It is possible to model many orders of magnitudes on both time and length scales simultaneously in large compositional spaces, which would be prohibitively costly for experimental exploration.
Computational codes are essential tools for the geochemical modeling and analysis of mineral alteration. They can handle key mechanisms like dissolution, precipitation and diffusion at many spatial and time resolutions.
A description of the amorphous layer on the glass surface that forms during glass corrosion modeling is essential. Its effect on the alteration kinetics and the glass's behavior is also required.
GRAAL (glass reactivity in allowance of the alteration layer) is designed to allow for a simple implementation of passivation in a reactive transport code. It also provides data about the composition and solubility of an amorphous layer.
Glass alteration rates are determined by the size and properties of the protective layer. In terms of passivation, the higher the quantity of protective layer, the lower is the primary mineral's dissolution rate. To apply the model and measure parameters, simple glass alteration experiments can be used.