A Potential Short-Cut to Quantitative Mineralogy

Date

2015-08

Journal Title

Journal ISSN

Volume Title

Publisher

New York State College of Ceramics at Alfred University. Kazuo Inamori School of Engineering.

Abstract

Quantitative mineralogy is a critical aspect of material performance and is typically obtained via quantitative X-ray diffraction (QXRD). This thesis proposes that quantitative mineralogy can be reasonably predicted from readily measured material properties (such and density and coefficient of thermal expansion) using a volume based rule of mixtures (ROM) approach. Obtaining the mineralogy of a two-component system is straightforward from a single property, but to obtain the mineralogy of a three-component system requires two property measurements. It should also be possible to use the starting chemistry as a constraint if the chemical reactions and resulting reaction products are simple. The ROM approach has two initial restrictions: (1) the difference of component property values should be significantly different, (2) the properties of different phases should not increase consistently. Other restrictions are possible, with the most obvious being the potential contribution of anisotropy. Three systems were evaluated to determine feasibility of this approach: porcelain, mullite, and spinel. The predicted mineralogy was compared to results obtained using QXRD via an internal standard method. Results show that the predictions based on ROM approach match the QXRD results with approximately a 95% confidence. It is proposed that this approach may predict the mineralogy of systems composed of more than three components (n=3) using n-1 measured properties or constraints (such as chemistry).

Description

Advisory committee members: Doreen Edwards, Dawei Liu. Dissertation completed in partial fulfillment of the requirements for the degree of Masters of Science in Ceramic Engineering at the Kazuo Inamori School of Engineering, New York State College of Ceramics at Alfred University

Keywords

Citation

DOI