Microscopic morphologies of monocomponent and binary solids growing from melts and vapors are simulated with a Monte Carlo model that accounts for bulk diffusion, interface attachment and detachment kinetics and surface diffusion. The interplay between transport and interface kinetics on morphologies and its effect on microstructures including lamellar spacing, tilted lamellar and oscillatory structures is investigated. An important aspect of these models is the length scale of the simulation.

A two dimensional equation has been solved which represents the heat transfer equation for the growth of single crystals system called Bridgman- Stockbarger method. Two variations were analyzed with and without an insulation between heater and cooler. System without an insulation shows stability problems because it's directly affected by the boundary between the cooler and heater region, in this case we obtained a discontinuity in this point. System with an insulation shows higher stability.

The atoms emitted from the single-crystal surface due to keV particle bombardment have been characterized using molecular dynamics to simulate the energy dissipation process. The model system in the calculation utilizes a Ag(100) microcrystallite of about 2000 atoms which is bombarded by as many as 1600 incident Ar atoms of 2-keV energy. The surface order is found to be preserved to a certain extent during the bombardment of static particle beams. A preferred direction of ejection is observed for the lower-kinetic-energy atoms (< I eV) sputtered from the second surface layer. Most of these atoms which take off late in the collision cascade originate from positions on the surface within three to four lattice spacings from the point of impact. The ejecting atoms are confined by the top layer atoms which are still located close to their lattice points when the collision cascade subsides. In addition, the angular pattern of ejection changes slightly as the emission energy of detection is varied. We found that the angular pattern of ejection did not vary with the emission energy in the low energy regime, in which the surface structure may be better determined.

Discrete computational models for the viscosities, sintering rates, and transport properties of sintering particle packings are presented. The packing is represented by a set of nodes (the particle centroids) connected by links (inter-particle contacts). The models for the mechanical behavior enforce equilibrium for each particle which leads to a set of simultaneous equations for the particle motion. Electrical or thermal transport through inter-particle contacts is modelled by imposing zero net flux at a node which also leads to a set of simultaneous equations for the value of potential at each particle center. The model is used to simulate the compaction of spheres to generate a threedimensional random packing. Statistical properties of the computed packing such as packing fraction, percolation threshold, and coordination number are compared with those of an experimental random packing. Results are also presented for the effective conductivity of mixtures of particles with very different conductivities.

By using isobaric-isothermal molecular dynamics and an n-body effective potential, we show that a dramatic elastic softening precedes the amorphization of NiZr2 occurring on chemically disordering the alloy. Above a critical level of chemical disorder, the shear elastic constant of the alloy increases suddenly up to a value comparable to the one of the amorphous solid obtained by quenching the liquid. This phenomenon occurs at the percolation threshold of the anelastically distorted regions surrounding the antisite defects. The percolation model describes satisfactorily experimental findings concerning the elastic softening and the loss of crystalline order in intermetallic compounds induced by irradiation or hydrogenation.

The method of direct configurational averaging (DCA) has been proposed to study the electronic structure of disordered alloys. Local density of states and band structure energies are obtained by averaging over a small number of configrations within a tight-binding Hamiltonian. Effective cluster interactions, the driving quantities for ordering in solids, are computed for various alloys using a tight-binding form of the linearized muffin-tin orbital method (TB-LMTO). The DCA calculations are used to determine various energetic and thermodynamic quantities for binary and ternary alloys.

In this paper we present results of a first-principles phase stability study of fcc-based Ti-Al alloys. In particular, the full-potential linear muffin tin orbital method has been used to determine heats of formation and other zero-temperature properties of 9 fcc ordered superstructures as well as fcc and hcp Ti, and fcc Al. From these results a set of effective cluster interactions are determined which are used in a cluster variation method calculation of the thermodynamic properties and the composition-temperature phase diagram of fcc-based alloys.

A statistical mechanical formulation is presented for the entropy of solution of simple molecules in water. The formulation is based on the Green-Wallace expansion for the entropy in terms of multiparticle correlation functions, which is derived here for rigid polyatomic fluids and for mixtures. With a factorization assumption for the solute-water correlation function we have been able to separate the translational and orientational contributions to the entropy of solution. This approach is applied to an infinitely dilute solution of methane in water. The required correlation functions are obtained by Monte Carlo simulation. The orientational contribution, which is due directly to the orientational asymmetry of water-water interactions, is found to be comparable to the translational contribution. We find that the large entropies and heat capacities of hydrophobic hydration can be accounted for by solute-water correlations alone and that large perturbations in water structure are not required to explain hydrophobic behavior.

A numerical model has been developed to predict the element distribution and some microstructural features in dendritically solidified binary and ternary alloys. Phase diagrams calculated on the basis of thermodynamic functions are taken into account. Thus, a minimum of assumptions is needed to describe the equilibrium at the solid/liquid interface, the partition coefficient of an alloying element is allowed to change as a function of temperature and concentration of other alloying elements.

The description of the equilibrium at the interface appears to be highly significant. Assuming a constant partition coefficient may be justified for some binary alloys, but the assumption has to be verified for each system. Small additions, like dopants or impurities, influence the position of equilibria in phase diagrams, thus requiring the treatment of basically binary alloys as multicomponent systems. In ternary or higher order alloys the assumption of constant partition coefficients leads to erroneous results. For microsegregation predictions in technical alloys realistic phase diagrams have to be taken into account.

We introduce a new semi-empirical method for calculating alloy properties. The method Is based on the concepts of equivalent crystal theory of defect formation energies in elemental solids. With this new method we predict heats of formation, lattice parameters, surface energies, segregation senergies and other properties of several binary alloys of fcc (Cu, NI, Ag, Au, Al, Fe, Pd and Pt) and bcc (Cr, Mo, Fe, V) elements. The method is characterized for its extreme computational simplicity and good agreement with experimental results. Several applications of the method are discussed.

A model of the energetics of bcc transition metals based on the low-order moments of the electronic density of states is presented. The new feature of the model is an additional energy term related to the fourth moment of the density of states. This term reflects the coarse shape of the density of states. The model is tested by the computation of point defect properties, phonon dispersions, structural energy differences and surface properties. The results are compared to experiment, ab-initio calculations and other model interatomic potentials. The results indicate that the inclusion of the fourth moment term in the energy does not significantly improve the description of properties of the bulk bcc metals. However, the fourth moment term substantially improves the description of large deviations from the bcc bulk such as surfaces and alternative crystal structures.

In the existing CVM (Cluster Variation Method) formulations, atoms are placed on lattice points. A modification is proposed in which an atom can be displaced from a lattice point. The displaced position is written by a vector r, which varies continuously. This model is treated in the CVM framework by regarding an atom at r as a species r. The probability of finding an atom displaced at r in dr is written as f(r)dr, and the corresponding pair probability is written as g(r1, r2)dr1dr2. We formulate using the pair approximation of the CVM in the present paper. The interatomic potential is assumed given, for example as the Lennard-Jones form. The entropy is written in terms of f(r) and g(r1, r2) using the CVM formula. The special feature of the present formulation which is different from the prevailing no-displacement cases of the CVM is that rotational symmetry of the lattice is to be satisfied by the f(r) and g(r1, r2) functions. After the general equations are written in the continuum vector form and in the integral equation formulation, an example of a single-component system is solved by changing integrals into summations over finite intervals. Further we construct simulations of displacement patters in such a way that the pattern satisfies the pair probability distribution which has been calculated as the output of the CVM analysis. The simulated pattern shows the wavy behavior of phonons. Future directions are discussed.

We discuss the validity of the usually used orbital radii capable to separate differently coordinated compounds in efficient structural maps. In order to account for the real dependence on the angular momentum, we redefine the atomic terms by including the d-states contribution, previously ignored. The choice of the right amount of the correction has been performed thanks to a series of structural maps obtained through a computer-aided code.

A method for calculating the thermodynamic properties of both perfect crystals and defects from a single zero-temperature energy minimization is described. The validity of the method is demonstrated by determining the free energy and the lattice parameter of a perfect Au crystal, as modelled by an embedded-atom method (EAM) potential. In addition, we determine the temperature dependence of the vacancy formation energy and the excess free energy of a (100) free surface.

In ordering alloy systems, the overall long-range order parameter does not distinguish the degree to which the component structures are delocalized for different configurations of equal randomness. We define a new parameter, which is sensitive to such differences, by considering the short-range order parameters as a set of statistical moments for geometric variation in the crystal. A computer model has been developed which uses this parameter to analyze finite alloy systems for configurational preferences by a Monte Carlo method. Initial studies have established the parametric range as a function of LRO and nonstoichiometry. Qualitative results correspond well to predictions for ideal systems.

This work presents a molecular dynamics simulation method designed to describe the processes of electron and lattice relaxation taking place in typical cascade volumes formed by high-energy implants. The simulation method is based on classical mechanics and includes the motions of electrons and nuclei. The results are in agreement with experiments.

Grain growth in polycrystals occurs by decreasing the total number of grains as a result of the disappearance of small ones. That is why the both the kinetic and topological details of shrinking of small grains are important.

In 2-D, “uniform boundary model” assumptions imply the von Neumann-Mullins law, and all grains having less than 6 neighbors tend to shrink. The mean topological class ef vanishing grains was found experimentally to be about 4.3. This result suggests that most vanishing grains are either 4- or 5-sided, not transforming to 3-sided ones.

Shrinking of 4- and 5-sided 2-D grains was studied experimentally on transparent, pure SCN, (succinonitrile) polycrystalline films and by direct computer simulation of grain boundary capillary motion together with triple junctions. It was found that the grains which are much smaller than their neighbors are topologically stable.