Date

2014

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Physics

First Adviser

Kim, Yong W.

Other advisers/committee members

Gunton, James D.; Stavola, Michael; Huang, Xiaolei; Vavylonis, Dimitrios

Abstract

The morphology of disordered binary metallic alloys is investigated. The structure of disordered binary metallic alloys is modeled as a randomly close packed (RCP) assembly of atoms. It was observed through a 2-D binary hard sphere experiment that RCP structure can be modeled as a mixture of nano-crystallites and glassy matter. We define the degree of crystallinity as the fraction of atoms contained in nano-crystallites in an RCP medium. Nano-crystallites by size in a crystallite size distribution were determined experimentally to define the morphology of the RCP medium. Both the degree of crystallinity and the crystallite size distribution have been found to be determined by the composition of a given binary mixture. A 2-D Monte Carlo simulation was developed in order to replicate the RCP structure observed in the experiment which is then extended to cases of arbitrary composition. Crystallites were assumed to be spherical with isotropic cross sections. The number of atoms in an individual crystallite in 2-D is simply transformed into the number of atoms in 3-D; we then obtain the crystallite size distribution in 3-D. This experiment accounts for the contribution from the repulsive core of the inter-atomic potential. The attractive part of the potential is recovered by constructing spherical nano-crystallites of a given radius from a crystalline specimen of each given alloy. A structural model of a disordered alloy is thus obtained.With the basic structure of the RCP medium defined, the response to heating would be in the form of changes to the crystallite size distribution. This was first investigated in a hard sphere mechanical oven experiment. The experimental setup consists of a 2-D cell which is driven by two independent stepper motors. The motors drive a binary RCP bed of spheres on a slightly tilted plane according to a chaotic algorithmm. The motors are driven at four different speed settings. The RCP medium was analyzed using a sequence of digital images taken of the beds. The bursts of images provide a Gaussian distribution of particle speeds in x and y directions thus giving rise to the notion of "temperature." This temperature scales with the motor speed settings. The measured average degree of crystallinity is found to decrease as the effective temperature was raised suggesting that nano-crystallites dissociate under thermal forcing. The evolution of a specimen's structure is calculated rigorously by means of the law of mass action formalism. A system of thermal dissociation reaction equations is written out for the set of nano-crystallites according to the 3-D crystallite size distribution. The equilibrium treatment is justified because the energy differences between metastable RCP structures fall within kT. Thermal dissociation of one surface atom at a time is assumed because the energy cost in dissociation of a surface atom on a nano-crystallite is significantly less than that of a multi atom cluster. The full set of reaction equations cover all possible dissociation steps, which may amount to several thousand for a disordered alloy specimen. The primary determining factor in each of these dissociation equations is the dissociation potential or the amount of attractive energy needed to remove a surface atom on a nano-crystallite of a given size. The attractive potential between atoms is calculated using a Lennard-Jones potential between a pair of atoms for which quantum chemistry calculations exist in the literature. All interactions impinged on the surface atom by all other atoms in a crystallite are summed. As the nano-crystallites dissociate due to heating, the structure of the alloy changes, and this leads to modifications of alloy's transport properties. The model is found to predict the melting temperature of various disordered binary alloys as well as refractory metals in good agreement with known data. The structure model for disordered binary alloys gives an excellent characterization of the alloy morphology. It therefore provides fruitful avenues for making predictions about how thermophysical properties of disordered binary alloys change as the alloy temperature is raised by heating.

Included in

Physics Commons

Share

COinS