1: %%%%%%%%%% espcrc2.tex %%%%%%%%%%
2: %
3: % $Id: espcrc2.tex 1.2 2000/07/24 09:12:51 spepping Exp spepping $
4: %
5: \documentclass[fleqn,twoside]{article}
6: \usepackage{espcrc2}
7:
8: % change this to the following line for use with LaTeX2.09
9: % \documentstyle[twoside,fleqn,espcrc2]{article}
10:
11: % if you want to include PostScript figures
12: \usepackage{graphicx}
13: % if you have landscape tables
14: \usepackage[figuresright]{rotating}
15:
16: % put your own definitions here:
17: % \newcommand{\cZ}{\cal{Z}}
18: % \newtheorem{def}{Definition}[section]
19: % ...
20:
21: \def\ij{\langle ij\rangle}
22: \newcommand{\ttbs}{\char'134}
23: \newcommand{\AmS}{{\protect\the\textfont2
24: A\kern-.1667em\lower.5ex\hbox{M}\kern-.125emS}}
25:
26: % add words to TeX's hyphenation exception list
27: \hyphenation{author another created financial paper re-commend-ed Post-Script}
28:
29: % declarations for front matter
30: \title{Phase structure of $(2+1)d$ strongly coupled lattice gauge theories}
31:
32: \author{Costas G. Strouthos \address{Department of Physics, Box 90305,
33: Duke University, Durham, North Carolina 27708, USA.}
34: }
35:
36: \begin{document}
37:
38: \begin{abstract}
39: \noindent
40: We study the chiral phase transition in $(2+1)d$ strongly coupled $U(N)$ lattice
41: gauge theories with staggered fermions. We show with high precision simulations
42: performed directly in the chiral limit
43: that these models undergo a Berezinski-Kosterlitz-Thouless (BKT) transition.
44: We also show that this universality class is unaffected even in the large N limit.
45:
46: \vspace{1pc}
47: \end{abstract}
48:
49: \maketitle
50:
51: \section{Introduction}
52: The behavior of symmetries at finite temperature is one of the most outstanding problems
53: in field theory. In recent years we have witnessed revised interest in the chiral phase
54: transition in QCD. The problem of symmetry breaking and its restoration is
55: intrinsically non-perturbative and therefore most of our knowledge about the phenomenon
56: comes from lattice simulations.
57: However, computing quantities in lattice QCD with massless
58: quarks is a notoriously difficult problem, because most known algorithms break down in the
59: chiral limit. In addition, the simulations near $T_c$ must be performed on lattices with large spatial
60: sizes in order to control large finite size effects due to the diverging correlation
61: length.
62:
63: A useful simplification of QCD occurs in the strong coupling limit, which retains much of
64: the underlying physics of QCD except for large lattice artifacts.
65: In this limit chiral symmetry breaking and its restoration at finite temperature have
66: been studied using large $N$ and large $d$ expansions \cite{kawamoto}. However, since these
67: approaches are based on
68: mean field analysis they cannot help in determining the universality of phase transitions.
69:
70: Interestingly, lattice QCD
71: with one staggered fermion interacting with $U(N)$ gauge fields can be mapped into a monomer-dimer
72: system in the strong coupling limit \cite{rossi}. Here, we present numerical results for the finite temperature
73: critical behavior of the $(2+1)d$ model. Our data were generated with a recently developed very efficient
74: cluster algorithm \cite{adams},
75: which allows us to perform precision calculations in the chiral limit.
76: In agreement with expectations from universality and dimensional reduction,
77: we show convincingly \cite{shailesh} that the chiral phase trasition belongs to the
78: BKT universality class \cite{kosterlitz}.
79:
80: The partition function of the model we study here is given by
81: \begin{equation}
82: Z = \sum_{[n,b]}
83: \prod_{\ij} (z_{\ij})^{b_{\ij}}\frac{(N-b_{\ij})!}{b_{\ij}! N!}
84: \prod_i \frac{N!}{n_i!}m^{n_i},
85: \label{pf}
86: \end{equation}
87: and is discussed in detail in \cite{rossi,adams}.
88: Here $n_i=0,1,2,...,N$
89: refers to the number of monomers on the site $i$, $b_{\ij}=0,1,2,...,N$
90: represents the number of dimers on the bond $\ij$, $m$ is the monomer weight,
91: $z_{\ij}=\eta_{ij}^2$ are the dimer weights. Note that while spatial dimers
92: carry a weight $1$, temporal dimers carry a weight $T$. The sum is over
93: all monomer-dimer configurations $[n,b]$ which are constrained such that
94: the sum of the number of monomers at each site and the dimers that touch
95: the site is always $N$ (the number of colors).
96: In this work we choose $L_x=L_y=L$. One can study the thermodynamics of the model
97: by working on asymmetric lattices with
98: $L_t \ll L$ and allowing $T$ to vary continuously.
99:
100: \section{Results}
101:
102: \begin{table*}[]
103: \caption{Fits for $\chi_c$ and $\chi_w$ near $T_c$.}
104: \label{table:1}
105: %\newcommand{\m}{\hphantom{$-$}}
106: %\newcommand{\cc}[1]{\multicolumn{1}{c}{#1}}
107: \renewcommand{\tabcolsep}{1.5pc} % enlarge column spacing
108: \renewcommand{\arraystretch}{0.9} % enlarge line spacing
109: \begin{tabular}{|c|c|c|c|c|c|c|}
110: \hline
111: $T$ & $\eta$ & 2r & $\chi^2_1/$d.o.f & $1/(2c)$ & $\chi^2_2/$d.o.f \\
112: \hline
113: 1.00 & 0.222(5) & 0 & 0.2 & 0.2343(8) & 183.2 \\
114:
115: 1.02 & 0.235(5) & -0.3(2) & 1.5 & 0.2411(5) & 53.4 \\
116:
117: 1.04 & 0.251(5) & -0.07(2) & 0.4 & 0.2483(5) & 2.8 \\
118:
119: 1.06 & 0.249(5) & -0.03(2) & 0.5 & 0.2583(5) & 33.8 \\
120:
121: 1.10 & 0.388(5) & -0.12(2) & 3.6 & 0.2831(5) & 770.0 \\
122:
123: 1.14 & 0.569(6) & -1.24(3) & 480 & --- & --- \\
124: \hline
125: \end{tabular}
126: \end{table*}
127:
128: In this section we present the results for the $T \neq 0$
129: critical behavior of the model. The observables used in this work
130: are the chiral susceptibility $\chi_c$ and the winding number susceptibility
131: $\chi_w$. The latter is proportional to the helicity modulus and describes the response
132: of the system to a perturbation that distorts the direction of the spontaneous
133: magnetization. The winding number susceptibility has been used successfully in
134: to demonstrate BKT behavior in other models \cite{harada}.
135: We fixed $L_t=4$ and computed $\chi_c$ and $\chi_w$ as a function of $L$.
136: If $T_c$ is the critical temperature, then the BKT theory predicts
137: \begin{equation}
138: \chi_c \propto \left\{\begin{array}{lc}
139: L^{2-\eta(T)} & T<T_c \cr
140: L^{1.75}\ [\log(L)]^{0.125} & T=T_c \cr
141: \mbox{constant} & T>T_c \cr
142: \end{array}\right.
143: \end{equation}
144: and
145: \begin{equation}
146: \chi_W = \left\{\begin{array}{lc}
147: 1/[2\eta(T)] + \alpha_1L^{-\alpha_2} & T<T_c \cr
148: [2 + 1/\log(L/L_0)] & T=T_c \cr
149: \alpha_3 \exp{(-\alpha_4L)} & T>T_c \cr
150: \end{array}\right.
151: \end{equation}
152: in the large $L$ limit.
153: The critical exponent $\eta(T)$ is expected to change continuously
154: with T, but remains in the range $0\leq \eta(T) < 0.25$. In order to
155: confirm these predictions we computed $\chi_c$ and $\chi_w$ for lattices
156: ranging from $L=32$ to $L=750$ and for $N=1,...,32$.
157:
158: \begin{figure}[htb]
159: \centering
160: \includegraphics[scale=0.56]{fig1.eps}
161: \vspace{-1cm}
162: \caption{Plot of $\chi_c$ vs. $L$ for various $T$.}
163: \label{fig:all_Nf}
164: \end{figure}
165:
166: \begin{figure}[b!]
167: \centering
168: \includegraphics[scale=0.56]{fig2.eps}
169: \vspace{-1cm}
170: \caption{Plot of $\chi_w$ vs. $L$ for various $T$.}
171: \label{fig:all_Nf}
172: \end{figure}
173:
174: Let us first discuss our results for $N=1$.
175: In figures 1 and 2 we plot $\chi_c$ and $\chi_w$ as functions
176: of $L$ for different values of $T$. We find that $\chi_c$ fits well to the form
177: $bL^{2-\eta}(\log(L))^{-2r}$
178: when $T \leq 1.06$. We also find that the logarithmic term is unimportant for
179: $T \leq 1.0$, whereas at $T=1.04$ the value $2r=0.7(2)$ is close
180: to the BKT prediction which is $0.125$. The values of $\eta$,
181: $2r$ and the quality of the fits $\chi_1^2$/d.o.f are shown in table 1.
182: We also fit the data for $\chi_w$ to the form $(c+1/\log(L/L_0))$. The values of
183: $1/(2c)$ are also shown in the table.
184: Finally, using the fact that $\chi_w = (2+1/\log(L/L_0))$ is exactly valid at $T_c$
185: we fit the data for $\chi_w$ to this form for various values of $T$.
186: The values of $\chi_2^2$/d.o.f are shown in the last column of table 1.
187: Based on where the minimum in $\chi_2^2$/d.o.f occurs we estimate $T_c=1.040(5)$.
188: Our value of $\eta=0.250(5)$ at $T_c$ is in excellent agreement
189: with the BKT prediction.
190:
191: We also checked that $\chi_c$ and $\chi_w$ show similar evidence for a BKT
192: transition at larger values of $N$.
193: Using techniques similar to the ones we used for $N=1$ we
194: computed $T_c$ for various values of $N$. We find
195: that $T_c = 0.708(6) N + 1.40(4) - 1.07(4)/N$ fits our results very
196: well for all values of $N$ with a $\chi^2$/d.o.f of $1.1$. The dependence
197: of the coefficients of this polynomial on $L_t$ is still under investagation.
198: \begin{figure}[t!]
199: \centering
200: \includegraphics[scale=0.25]{figkt3.eps}
201: \vspace{-1cm}
202: \caption{Plot of $\eta$ as a function of $N$ for $T=1.0$.
203: The data fit very well to the form $\eta = 0.169(6)/N + 0.050(9)/N^2$ with
204: $\chi^2$/d.o.f$=1.2$.}
205: \label{fig:all_Nf}
206: \end{figure}
207:
208: \begin{figure}[htb]
209: \centering
210: \includegraphics[scale=0.56]{fig4.eps}
211: \vspace{-1cm}
212: \caption{Plot of $\eta$ as a function of $T/T_c$ for various values of $N$.}
213: \label{fig:all_Nf}
214: \end{figure}
215: With regards to the $N$ dependence of our results we find two interesting
216: obervations. Witten argued that when the symmetry is $U(1)$ the large $N$
217: analysis is still applicable since $\eta \sim 1/N$ at large $N$ \cite{witten}. Our results
218: for a fixed $T=1.0$, shown in figure 3, do agree with his conjecture.
219: Interestingly, as $N$ becomes large and $T/T_c$ is held fixed instead of $T$,
220: we find that $\eta \neq 0$ even in the large $N$ limit. Figure 4 shows that $\eta$
221: approaches an interesting function of $T/T_c$ as $N$ becomes large. Extending this
222: observation to QCD, we think that the t'Hooft limit (large $N$ with $g^2N$ held fixed)
223: may be quite similar \cite{thooft}.
224:
225: \section{Summary}
226: We presented high precision results from simulations of $(2+1)d$ $U(N)$ strongly coupled
227: lattice gauge theories at $T\neq 0$
228: and we showed convincing evidence that the models undergo a BKT phase transition.
229: In addition, we showed that this universality class is unaffected even in the large $N$ limit, implying
230: that the mean field analysis often used in this limit breaks down in the critical region.
231:
232:
233: \section*{Acknowledgements}
234: \noindent
235: This work was done in collaboration with Shailesh Chandrasekharan and it was supported
236: in part by the NSF grant DMR-0103003 and DOE grant DE-FG-96ER40945.
237:
238: \begin{thebibliography}{9}
239:
240: \bibitem{kawamoto}
241: N. Kawamoto and J. Smit, Nucl. Phys. {\bf B192} (1981) 100;
242: H. Klumberg-Stern, A. Morel and B. Petersson, Nucl. Phys. {\bf B215} (1983) 527;
243: P.H. Damgaard, N. Kawamoto and K. Shigemoto, Phys. Rev. Lett. {\bf 53} (1984) 527.
244:
245: \bibitem{rossi}
246: P. Rossi and U. Wolff, Nucl. Phys. {\bf B248} (1984) 105.
247:
248: \bibitem{adams}
249: D.H. Adams and S. Chandrasekharan, Nucl. Phys. {\bf B662} (2003) 220.
250:
251: \bibitem{shailesh}
252: S. Chandrasekharan and C.G. Strouthos, {\tt hep-lat/0306034}.
253:
254: \bibitem{kosterlitz}
255: V.L. Berezinski, Sov. Phys. JETP {\bf 34} (1971) 610; J.M. Kosterlitz and
256: D.J. Thouless, J. Phys. {\bf C6} (1973) 1181.
257:
258: \bibitem{harada}
259: K. Harada and N. Kawashima, J. Phys. Soc. Jpn. {\bf 67} (1998) 2768;
260: S. Chandrasekharan and J.C. Osborn, Phys. Rev. {\bf B66} (2002) 045113.
261:
262: \bibitem{witten}
263: E. Witten, Nucl. Phys. {\bf B145} (1978) 110.
264:
265: \bibitem{thooft}
266: G. t'Hooft, Nucl. Phys. {\bf B72} (1974) 461.
267: \end{thebibliography}
268:
269: \end{document}
270: