Files in this item



application/pdf9010815.pdf (4MB)Restricted to U of Illinois
(no description provided)PDF


Title:Numerical studies of phase transitions and critical phenomena in fermionic and polymeric systems
Author(s):Cannon, Joel W.
Doctoral Committee Chair(s):Fradkin, Eduardo H.
Department / Program:Physics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Physics, Condensed Matter
Abstract:In this thesis we study three problems in phase transitions and critical phenomena.
We study the phase diagram of the Hubbard model using monte-carlo simulation, and a novel application of the Lanczos algorithm. We demonstrate the presence of a tri-critical point, and estimate the phase boundary to occur at roughly V = 2.92 $\pm$.01 at U = 5.5, and V = 1.65 $\pm$.05 at U = 3.0. We estimate the tri-critical point to occur above U = 3.0.
We study the finite-sized dependence of the energy of the ground and first excited states for various values of U to observe the crossover of the conformal anomaly, c, from free fermion to Heisenberg behavior. We have inferred c along this line using the finite-sized dependence of the ground state energy. We obtain the correct asymptotic values, but find a maximum in the value of c at approximately U = 2 in apparent contradiction to Zamolodchikov's "c theorem". We show that this behavior is explained by postulating that our trajectory slices the RG trajectory in such a way that we retrace the irrelevant operators back on their trajectories for a short distance, and the presence of these operators invalidates the equation$$E\sb0(N) = E\sb{0}(\infty) - {\pi cv\sb{f}\over 6}{1\over N\sp2}\eqno(0.1)$$
We calculate the probability distribution functions for linear, 3-, and 4-star polymers and find that the configurations become less prolate and more anisotropic as the number of rays increase. We find that the parameterization developed by Aronovitz and Nelson is a useful and consistent way of characterizing polymer shape.
I have for the first time used ground state eigenvectors, calculated with the Lanczos algorithm, to define probability distribution functions for the order parameter of a system.
Algorithmic improvements include: (i) development of an efficient data structure for performing monte-carlo simulations on discrete systems; (ii) A method to obtain the ground state eigenvector via the Lanczos method with a small increase in required memory; (iii) A data structure to allow efficient monte-carlo simulation of a lattice polymer; and (iv) An algorithm for simulation of cross-linked polymers. In addition, we obtain numerical solutions for the ground state of the finite size Bethe ansatz equations for the Hubbard model. (Abstract shortened with permission of author.)
Issue Date:1989
Rights Information:Copyright 1989 Cannon, Joel W.
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9010815
OCLC Identifier:(UMI)AAI9010815

This item appears in the following Collection(s)

Item Statistics