Download Free Monte Carlo Simulations Of Phase Transitions Book in PDF and EPUB Free Download. You can read online Monte Carlo Simulations Of Phase Transitions and write the review.

In condensed matter physics, researchers study the physical properties of condensed phases of matter, theoretically or experimentally. The fundamentally appealing topic in this research area is how to classify phases of matter and identify phase transitions between them.Different from traditional theoretical or experimental approaches, which relies on either complicated mathematical formulation or equally complex experimental equipment, Monte Carlo based stochastic methods, which are often treated as "computer experiments", introduce a relatively "cheap" but effective approach to study phases and phase transitions. In this dissertation, we employ the classical Monte Carlo simulation, which utilizes the Metropolis algorithm to evolve system configurations, and also the determinant quantum Monte Carlo simulation to study phases and phase transitions of model Hamiltonians, such as the Hubbard model, and the periodic Anderson model (PAM). In the 21st century, data driven machine learning techniques have proven to be an another research "engine" for detecting phases and phase transitions. In this dissertation, I explore potential usages of unsupervised machine learning techniques in phase transition. Specifically, I leverage the principal component analysis (PCA) to extract internal structures, which are fully reflected in leading principal components, of Monte Carlo generated configurations, and then quantify obtained principal components to distinguish phases and phase transitions. This technique is applied to study model Hamiltonians, such as the Ising model, the XY model, the Hubbard model and the PAM. The exact organization of this dissertation is as follows: In chapter 1, I first introduce basic concepts of phase transitions and related model Hamiltonians. In chapter 2, I talk about a variety of methodologies utilized. In chapter 3, I present studies of phase transitions in a spin-fermion model. In chapter 4, I explore phase diagrams of the PAM coupled with an additional layer of metal. In chapter 5 and 6, I discuss how to apply machine learning techniques, especially PCA, to distinguish phases and detect phase transitions in classical and quantum model Hamiltonians. In chapter 7, I summarize previous chapters and discuss potential future directions.
This book is based on research carried out by the author in close collabora tion with a number of colleagues. In particular, I wish to thank Per Bak, A. John Berlinsky, Hans C. Fogedby, Barry Frank, S. 1. Knak Jensen, David Mukamel, David Pink, and Martin Zuckermann for fruitful and extremely stimulating cooperation. It is a pleasure for me to note that active interaction with most of these colleagues is still continuing. The work has been performed at several different institutions, notably the Department of Chemistry, Aarhus University, Denmark, and the Depart ment of Physics, University of British Columb~a, Canada. I wish to thank the Department of Chemistry at Aarhus University for providing me with splen did research facilities over the years. From May 1980 to August 1981, I visited the Department of Physics at the University of British Columbia and I would like to express my sincere gratitude to members ofthe department for provi ding me with excellent working conditions. My special thanks are due to Professor Myer Bloom who introduced me to the field of phase transitions in biological membranes and in whose biomembrane group I found an extre mely stimulating scientific atmosphere happily married with a most agreeable social climate. During the last two years when a major part ofthis work was carried out, I was supported by AlS De Danske Spritfabrikker through their Jubilreumsle gat of 1981. Their support is gratefully acknowledged.
In this book, the thermodynamic observables of the classical one- and two-dimensional ferromagnetic and antiferromagnetic Ising models on a square lattice are simulated, especially at the phase transitions (if applicable) using the classical Monte Carlo algorithm of Metropolis. Finite size effects and the influence of an external magnetic field are described. The critical temperature of the 2d ferromagnetic Ising model is obtained using finite size scaling. Before presenting the Ising model, the basic concepts of statistical mechanics are recapped. Furthermore, the general principles of Monte Carlo methods are explained.
When learning very formal material one comes to a stage where one thinks one has understood the material. Confronted with a "realiife" problem, the passivity of this understanding sometimes becomes painfully elear. To be able to solve the problem, ideas, methods, etc. need to be ready at hand. They must be mastered (become active knowledge) in order to employ them successfully. Starting from this idea, the leitmotif, or aim, of this book has been to elose this gap as much as possible. How can this be done? The material presented here was born out of a series of lectures at the Summer School held at Figueira da Foz (Portugal) in 1987. The series of lectures was split into two concurrent parts. In one part the "formal material" was presented. Since the background of those attending varied widely, the presentation of the formal material was kept as pedagogic as possible. In the formal part the general ideas behind the Monte Carlo method were developed. The Monte Carlo method has now found widespread appli cation in many branches of science such as physics, chemistry, and biology. Because of this, the scope of the lectures had to be narrowed down. We could not give a complete account and restricted the treatment to the ap plication of the Monte Carlo method to the physics of phase transitions. Here particular emphasis is placed on finite-size effects.
"This thesis focuses on the effects of both correlated and non-correlated disorder on non-equilibrium phase transitions, specifically those found in the d-dimensional contact process. These effects are studied by means of extensive Monte-Carlo simulations. The scaling behavior of various parameters is evaluated for both cases, and the results are compared with theory. For the correlated disorder case, the stationary density in the vicinity of the transition is also examined, and found to be smeared. The behavior in both cases can be understood as the results of rare regions where the system is locally free of disorder. For point-like defects, i.e., uncorrelated disorder, the rare regions are of finite size and cannot undergo a true phase transition. Instead, they fluctuate slowly which gives rise to Griffiths effects. In contrast, if the rare regions are infinite in at least one dimension, a stronger effect occurs: each rare region can independently undergo the phase transition and develop a nonzero steady state density. This leads to a smearing of the global transition"--Abstract, leaf iv.
This status report features the most recent developments in the field, spanning a wide range of topical areas in the computer simulation of condensed matter/materials physics. Both established and new topics are included, ranging from the statistical mechanics of classical magnetic spin models to electronic structure calculations, quantum simulations, and simulations of soft condensed matter.
Monte Carlo simulations can be used to explore many unsolved systems in Statistical Mechanics. The behavior of two Monte Carlo Algorithms, the Metropolis Algorithm and Wolff Algorithm, was analyzed through simulations of the ferromagnetic Ising Model, and then these simulations were applied to studies of the antiferromagnetic Ising Model and Hard Squares Model. Further study would include looking at the Hard Hexagon system, the Hard-Constraint model of Hard Hexagons, and the Ising Model on a three-dimensional antiferromagnetic face-centered-cubic lattice. Study of the latter model might elucidate behavior around "super-degenerate points" that is believed to exist in this model. The studies done so far have encountered useful results. The Metropolis Algorithm was found to provide a good qualitative description of each system studied, although it often returned inaccurate individual measurements. The Wolff algorithm was found to yield excellent results with shorter run-times, but could not give good values for the order parameter at zero field. A Metropolis algorithm was used to explore the Hard Squares system, which is not analytically solvable and has only been studied through series expansions. The results that we obtained seem to duplicate the results obtained through series expansion. Further study along these lines promises to yield useful results.