Monte Carlo Simulation in Momentum Space

碩士 === 國立彰化師範大學 === 物理學系 === 101 === It is well known that the Monte Carlo simulation hardly converges near the phase transition due to the so-called critical slowing down. Cluster algorithm in which the whole cluster of spins are flipped instead of flipping one single spin was invented to get rid o...

Full description

Bibliographic Details
Main Author: 龔柏儒
Other Authors: 鄭智豪
Format: Others
Language:zh-TW
Published: 2013
Online Access:http://ndltd.ncl.edu.tw/handle/63438965640886253769
Description
Summary:碩士 === 國立彰化師範大學 === 物理學系 === 101 === It is well known that the Monte Carlo simulation hardly converges near the phase transition due to the so-called critical slowing down. Cluster algorithm in which the whole cluster of spins are flipped instead of flipping one single spin was invented to get rid of the critical slowing down near the critical point. In the thesis, we perform the Monte Carlo sampling in the momentum space. The sampling method is quite general for all lattice models of arbitrary dimensions. However, from our studies in depth, sampling in the momentum space could not get rid of the critical slowing down, even that more computational resource on fast fourier transform is required.