Quantum Chaos: Spectral Fluctuations and Overlap Distributions of the Three Level Lipkin-Meshkov-Glick Model

<p>We test the prediction that quantum systems with chaotic classical analogs have spectral fluctuations and overlap distributions equal to those of the Gaussian Orthogonal Ensemble (GOE). The subject of our study is the three level Lipkin-Meshkov-Glick model of nuclear physics. This model dif...

Full description

Bibliographic Details
Main Author: Meredith, Dawn Christine
Format: Others
Language:en
Published: 1987
Online Access:https://thesis.library.caltech.edu/11406/1/Meredith_DC_1987.pdf
Meredith, Dawn Christine (1987) Quantum Chaos: Spectral Fluctuations and Overlap Distributions of the Three Level Lipkin-Meshkov-Glick Model. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/7y5k-ex17. https://resolver.caltech.edu/CaltechTHESIS:02202019-120226199 <https://resolver.caltech.edu/CaltechTHESIS:02202019-120226199>
Description
Summary:<p>We test the prediction that quantum systems with chaotic classical analogs have spectral fluctuations and overlap distributions equal to those of the Gaussian Orthogonal Ensemble (GOE). The subject of our study is the three level Lipkin-Meshkov-Glick model of nuclear physics. This model differs from previously investigated systems because the quantum basis and classical phase space are compact, and the classical Hamiltonian has quartic momentum dependence. We investigate the dynamics of the classical analog to identify values of coupling strength and energy ranges for which the motion is chaotic, quasi-chaotic, and quasi-integrable. We then analyze the fluctuation properties of the eigenvalues for those same energy ranges and coupling strength, and we find that the chaotic eigenvalues are in good agreement with GOE fluctuations, while the quasi-integrable and quasichaotic levels fluctuations are closer to the Poisson fluctuations that are predicted for integrable systems. We also study the distribution of the overlap of a chaotic eigenvector with a basis vector, and find that in some cases it is a Gaussian random variable as predicted by GOE. This result, however, is not universal. </p>