Descripción
|
|
---|---|
In this work, we formulate a theory to address simulations of slow time trans- port e ects in atomic systems. We rst develop this theoretical framework in the context of equilibrium of atomic ensembles, based on statistical me- chanics. We then adapt it to model ensembles away from equilibrium. The theory stands on Jaynes' maximum entropy principle, valid for the treatment of both, systems in equilibrium and away from equilibrium and on mean eld approximation theory. It is expressed in the entropy formulation as a varia- tional principle. We interpret atomistic equivalents of macroscopic variables such as the temperature and the molar fractions, wich are not required to be uniform, but can vary from particle to particle. We complement this theory with Monte Carlo summation rules for further approximation. In addition, we provide a framework for studying transport processes with the full set of equations driving the evolution of the system. We rst derive a dissipation inequality for the entropic production involving discrete ther- modynamic forces and uxes. This discrete dissipation inequality identi es the adequate structure for discrete kinetic potentials which couple the micro- scopic eld rates to the corresponding driving forces. Those kinetic potentials must nally be expressed as a phenomenological rule of the Onsanger Type. We present several validation cases, illustrating equilibrium properties and surface segregation of metallic alloys. We rst assess the ability of a simple mean eld model to reproduce thermodynamic equilibrium properties in systems with atomic resolution. Then, we evaluate the ability of the model to reproduce a long-term transport process in complex systems. | |
Internacional
|
Si |
ISBN
|
|
Tipo de Tesis
|
Doctoral |
Calificación
|
Aprobado |
Fecha
|