1: %%%\documentstyle[aps,prl,twocolumn,graphicx,floats]{revtex}

2: %\documentstyle[aps,draft,prl,graphicx,floats]{revtex}

3: \documentstyle[aps,prl,graphicx,floats]{revtex}

4: % \documentclass[eps,12pt]{article}

5: % \usepackage{graphicx}

6: %%%%%%%%%%%%%%%%%%%%%%%%%%%

7: % alves

8: %\setlength\textwidth{15cm}

9: %\setlength\textheight{22.5cm}

10: \voffset=1.8cm

11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%

12: 

13: \renewcommand{\thesection}{\arabic{section}}

14: \renewcommand{\thetable}{\arabic{table}}

15: \newcommand{\beq}{\begin{equation}}

16: \newcommand{\eeq}{\end{equation}}

17: 

18: 

19: \begin{document}

20: 

21: % SETS A DOUBLE SPACE

22: \baselineskip=18pt

23: 

24: %{\wideabs{

25: \begin{center}

26: {\Large\bf  Mixed initial conditions to estimate the dynamic critical

27: exponent in short-time Monte Carlo simulation}

28: 

29: \vskip 1.1cm

30: {\bf Roberto da Silva\footnote{E-mail: rsilva@dfm.ffclrp.usp.br},~~

31:      Nelson A. Alves\footnote{E-mail: alves@quark.ffclrp.usp.br}~and~

32:    J.R. Drugowich de Fel\'{\i}cio\footnote{E-mail: drugo@usp.br}}

33: \vskip 0.1cm

34: {\it Departamento de F\'{\i}sica e Matem\'atica, FFCLRP

35:      Universidade de S\~ao Paulo. Av. Bandeirantes 3900. \\

36:      CEP 014040-901 \, Ribeir\~ao Preto, SP, Brazil}

37: 

38: \vskip 0.3cm

39: \today

40: \vskip 0.4cm

41: \end{center}

42: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

43: \begin{abstract}

44:  We explore the initial conditions in short-time critical dynamics to propose

45: a new method to evaluate the dynamic exponent $z$. Estimates are obtained

46: with high precision for 2D Ising model and 2D Potts

47: model for three and four states by performing heat-bath Monte Carlo

48: simulations.

49: 

50: \vskip 0.1cm

51: {\it Keywords:} short-time dynamics, critical phenomena, dynamic

52: exponent, Ising model, Potts model, Monte Carlo simulations.

53: 

54: \end{abstract}

55: \vskip 0.1cm

56: {\it PACS-No.: 64.60.Fr, 64.60.Ht, 02.70.Lq, 75.10.Hk}

57: 

58: %\newpage

59: %\vskip 1cm

60: 

61: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

62: 

63: \nopagebreak

64: 

65: %\section{INTRODUCTION} 

66: %\vspace*{-0.5pt}

67: %\noindent

68: \indent

69: 

70: The description of static critical phenomena in terms of finite size scaling

71: (FSS) relations, developed by Fisher {\it et al.} \cite

72: {Fisher72,FisherLectures} has been extended by Halperin, Hohenberg, Ma and

73: Suzuki \cite{HH1,Suzuki} to include dynamical {\it properties} of the

74: system. Later, Janssen {\it et al.} \cite{Janssen1}, and independently Huse 

75: \cite{Huse} found evidence for an universal behavior far from equilibrium.

76: 

77: As discussed by Janssen {\it et al.} %%%   \cite{Janssen1}

78: one finds an universal behavior already in the early stages of the

79: relaxation process for systems prepared at an initial state characterized by

80: non-equilibrium values of the order parameter. As a consequence, they could

81: advance the existence of a new critical exponent $\theta $, independent of

82: the known set of static exponents and of the dynamic critical exponent $z$.

83: This new exponent characterizes the so called ``critical initial slip'', the

84: anomalous increasing of the magnetization when the system is quenched to the

85: critical temperature $T_{c}$.

86: 

87: That new universal stage has been exhaustively investigated to confirm

88: theoretical predictions and to enlarge our knowledge on phase transitions

89: and critical phenomena. In this sense, several models and algorithms \cite

90: {Universal97,Sch97,Zhang99} have been used, as toy models, in order to check

91: the ability of the new approach in obtaining dynamic and static critical

92: exponents. Results are in good agreement with pertinent results for static

93: exponents and seems to be confident even for the new critical exponent $%

94: \theta $. However, a reliable technique to obtain the dynamic exponent $z$

95: is lacking. A first proposal by Li {\it et al.} \cite{Li95} using a

96: time-dependent Binder's cumulant yields estimates with low precision when

97: compared with other techniques \cite{Grassberger,Nightingale}. An

98: alternative way which uses another kind of cumulant, proposed by Zheng \cite

99: {Review}, gives the right answer for the 2D Ising model but fails in

100: determining the value of $z$ for the 3-state Potts model \cite{Review} and

101: for the Ising model with multispin interactions \cite{Simoes}.

102: 

103: In this letter, we introduce and check a new technique to obtain the

104: exponent $z$, combining the behavior of the order parameter and its second

105: moment when the system is submitted to different initial conditions.

106: 

107: Before presenting our proposal, we shall review the main results in short

108: time dynamics. 

109: 

110: Although Halperin {\it et al.} \cite{HH1} have studied

111: systems with different dynamics, we consider only systems without

112: conservation laws, the so called Model A \cite{HH2} because our discussion

113: aims dynamics generated by heat bath dynamics. Therefore, we consider a

114: magnetic system prepared at high temperature ($T>>T_{c}$) with a small

115: nonzero magnetization $m_{0}$, (this can be achieved with a small external

116: magnetic field $h$) and quenched to the critical temperature $T_{c}$ without

117: any external magnetic field. If the system is allowed to relax towards

118: equilibrium with the dynamics of model A, the magnetization obeys the

119: following scaling relation (generalized to $k{\rm th}$ moment), 

120: \begin{equation}

121: M^{(k)}(t,\tau ,L,m_{0})=b^{-k\beta /\nu }M^{(k)}(b^{-z}t,b^{1/\nu }\tau

122: ,b^{-1}L,b^{x_{0}}m_{0})\,.  \label{magk}

123: \end{equation}

124: Here $b$ is an arbitrary spatial scaling factor, $t$ is the time evolution

125: and $\tau $ is the reduced temperature, $\tau =(T-T_{c})/T_{c}$. The

126: exponents $\beta $ and $\nu $ are the static critical exponents, while $z$

127: is the dynamic one. $M^{(k)}=\left\langle M^{k}\right\rangle $ are the $k%

128: {\rm th}$ moments of magnetization. This scaling relation depends on the

129: initial magnetization $m_{0}$ and gives origin to a new, independent

130: critical exponent $x_{0}$, the scaling dimension of the initial

131: magnetization, which is related to $\theta $. 

132: %% by $\theta = (x_0 -\beta/ \nu)/z$.

133: 

134: From Eq. (\ref{magk}) we can derive the power law increasing of the

135: magnetization, observed in the initial stage of the dynamic relaxation. For

136: this purpose, we consider large lattice sizes $L$ at $\tau =0$ with $%

137: b=t^{1/z}$. This leads to the scaling relation 

138: \begin{equation}

139: M(t,m_{0})=t^{-\beta /\nu z}M(1,t^{x_{0}/z}m_{0})  \label{mag0}

140: \end{equation}

141: for the first moment of the magnetization $M^{(1)}\equiv M$. By expanding

142: this equation for small $m_{0}$, we have the following power law, 

143: \begin{equation}

144: M(t)\sim m_{0}t^{\theta }\,,  \label{m0}

145: \end{equation}

146: where, as anticipated, we identify $\theta =(x_{0}-\beta /\nu )/z$. Here we

147: also have the condition that $t^{x_{0}/z}m_{0}$ is small, which sets a time

148: scale $t_{0}\sim m_{0}^{-z/x_{0}}$ \cite{Janssen1,Foundations} where that

149: phenomena can be observed.

150: 

151: On the other hand, it has been realized the existence of another important

152: dynamic process from an initial ordered state \cite

153: {Stauffer92,Stauffer93,Sch96}, which represents another fixed point in the

154: context of renormalization group approach. This leads to a different universal

155: behavior of the dynamic relaxation process also described by Eq.~(\ref{magk}%

156: ) with $m_{0}=1$. In particular, dealing with large enough lattice sizes at

157: the critical temperature ($\tau =0$), one obtains a power law decay of the

158: magnetization 

159: \begin{equation}

160: M(t)\sim t^{-\beta /\nu z}\,,  \label{m1}

161: \end{equation}

162: which follows from Eq.~(\ref{magk}) when we choose $b^{-z}t=1$ and shows the

163: average magnetization is not zero as would be expected from a disordered

164: initial state.

165: 

166: Equation (\ref{magk}) and their particular forms in (\ref{m0}) and (\ref{m1}%

167: ) can be used to determine relations involving static critical exponents and

168: the dynamic exponent $z$ \cite{Review,Foundations}.

169: 

170: The observables in short-time analysis are described by different scaling

171: relations according to the initial magnetization. In particular, the second

172: moment $M^{(2)}(t,L)$, 

173: \begin{equation}

174: M^{(2)}=\left\langle \left( \frac{1}{L^{d}}\sum\limits_{i=1}^{N}\sigma

175: _{i}\right) ^{2}\right\rangle =\frac{1}{L^{2d}}\left\langle

176: \sum\limits_{i=1}^{N}\sigma _{i}^{2}\right\rangle +\frac{1}{L^{2d}}%

177: \sum\limits_{i\neq j}^{N}\left\langle \sigma _{i}\sigma _{j}\right\rangle \,,

178: \label{second}

179: \end{equation}

180: behaves as $L^{-d}$ since in the short-time evolution with initial condition 

181: $m_{0}=0$ the spatial correlation length is very small compared with the

182: lattice size $L$. Thus we have \cite{Sch97,Review} 

183: \begin{equation}

184: M^{(2)}(t,L)=t^{-2\beta /\nu z}\,M^{(2)}(1,t^{-1/z}L)\sim t^{(d-2\beta /\nu

185: )/z}\,.  \label{increasemag2}

186: \end{equation}

187: Moreover, under this initial condition one can also define the

188: time-dependent Binder's cumulant at the critical temperature, 

189: \begin{equation}

190: U(t,L)=1-\frac{M^{(4)}(t,L)}{3(M^{(2)}(t,L))^{2}}\,,  \label{binder0}

191: \end{equation}

192: which leads at $T=T_{c}$ to the FSS relation 

193: \begin{equation}

194: U(t,L)=U(b^{-z}t,b^{-1}L)\,,  \label{r13}

195: \end{equation}

196: and the exponent $z$ can be independently evaluated through scaling

197: collapses for different lattice sizes \cite{Li95,Li96}.

198: 

199: In order to obtain more precise estimates for the dynamic exponent $z$,

200: another cumulant has been proposed \cite{Review}. It is given by 

201: \begin{equation}

202: U_{2}(t,L)=\frac{M^{(2)}(t,L)}{(M(t,L))^{2}}-1  \label{binder1}

203: \end{equation}

204: and should behave as 

205: \begin{equation}

206: U_{2}(t)\sim t^{d/z}  \label{binder11}

207: \end{equation}

208: when one starts from an ordered state. In this case, curves for all the

209: lattices lay on the same straight line without any re-scaling in time and

210: results in more precise estimates for $z$. However, application of this

211: procedure has not been successful in at least two well known models: the

212: two-dimensional $q=3$ Potts model \cite{Review} and the Ising model with

213: three spin interactions in just one direction \cite{Simoes}. The reason for

214: the above disagreement could be related to the value of the second

215: term of r.h.s in Eq.~(\ref{second}) when $m_{0}=1$.

216: 

217: A plausible way to circumvent this problem is to work with different initial

218: conditions. For this we decided to follow the evolution of the ratio $%

219: F_{2}=M^{(2)}/M^{2}$ but using different initial conditions to calculate

220: each one of the mean values. The reason is we know the behavior of the

221: second moment of the magnetization when samples are initially disordered $%

222: (m_{0}=0)$ and also the dependence on time of the magnetization of samples

223: initially ordered $(m_{0}=1)$. Under the above mentioned conditions the

224: ratio behaves as 

225: \begin{equation}

226: F_{2}(t,L)=\frac{\left. M^{(2)}(t,L)\right| _{m_{0}=0}}{\left.

227: (M(t,L))^{2}\right| _{m_{0}=1}}\sim \frac{t^{(d-2\beta /\nu )/z}}{t^{-2\beta

228: /\nu z}}=t^{d/z}\,,  \label{binder2}

229: \end{equation}

230: which has the same potential law that the cumulant mentioned before but

231: requires two independent simulations instead of one used for calculating $%

232: U_{2}$.

233: 

234: In short-time Monte Carlo (MC) simulations, the time scale $t$ is settled in

235: units of whole lattice updates, and does not depend on the initial

236: conditions. However,  the same dynamics (Metropolis, Glauber or heat-bath)

237: should be used in both simulations \cite{Universal97,Sch97}.

238: 

239: Now we present our estimates obtained for 2D Ising model, $q=3$ and $q=4$

240: Potts models.

241: 

242: We have performed independent heat-bath (HB) MC simulations for a large

243: lattice $(L=144)$ according to the required initial conditions and the

244: evolutions have been done until the maximum time $t=200$ MC sweeps with $%

245: N=10000$ samples. This gives averages for the magnetization and its second

246: moment. We have performed this kind of simulation 20 times to obtain our

247: final estimates for each moment in function of $t$. Since all runs are

248: independent due to the choice of random numbers, we have a total of 

249: 400 time series for Eq.~(\ref{binder2}).

250: 

251: Our results are shown on a log-log scale for the time interval [10, 100] in

252: Fig.1 and Fig.2, respectively for three and four-state Potts model (a

253: similar figure is obtained for 2D Ising model) and they appear to be clearly

254: consistent with a power law in that time interval. The error bars are also

255: presented in those figures but they are hardly seen in that scale.

256: 

257: Our estimates for $z$ comes from a least square fitting in the time interval 

258: $[t_{i},t_{f}]$. Due to our small statistical errors, we can make a

259: systematic study for the range in $t$ where we find acceptable

260: goodness-of-fit $Q$ \cite{Press}. As examples, we obtain for 2D Ising model, 

261: $z=2.1435(2)$ in the time interval $[10,200]$ with $Q=10^{-250}$, $%

262: z=2.1359(3)$ in $[50,200]$, with $Q=10^{-41}$, and the most acceptable $%

263: Q=0.99$ in $[30,90]$, which yields $z=2.1565(7)$. This value is presented in

264: Table 1, where we also include some estimates for comparison. Here we

265: observe that our result is in agreement with more recent estimates within two

266: standard-deviations.

267:  This indicates our statistical errors are presumably underestimated possibly

268: due to corrections to scaling. 

269: 

270: We complete the overview in Table 1 with data from \cite{Wang97}. Estimates

271: have been obtained from the long time behavior of the magnetization for the

272: square lattice (sq), $z=2.168(5)$, for the triangle (TP), $z=2.180(9)$, and

273: for the honeycomb (hc) lattice, $z=2.167(8)$, while from damage spreading in

274: short-time the authors quote $2.166(7)$, $2.164(7)$ and $2.170(10)$ for sq,

275: TP and hc lattices.

276: 

277: 

278: For $q=3$ Potts model, our study monitoring $Q$ gives $z=2.198(2)$ in the

279: interval [50, 90] with $Q=0.82$. This result agrees with $z=2.203(11)$ (Ref. 

280: \cite{Review}) obtained from the Binder cumulant in Eq.~(\ref{binder0}), and

281: with the estimate obtained in \cite{Sch97} with HB algorithm from the second

282: moment $M^{(2)}(t)$ in short-time analysis.

283: 

284: The value in \cite{Zhang99} refers to TP lattice, presenting further

285: numerical evidence (comparing \cite{Sch97} and \cite{Zhang99}) to the

286: dynamic universality. Reference \cite{Review} also presents the value $%

287: 2.14(3)$, obtained from Eq.~(\ref{binder1}), in clear disagreement with $%

288: 2.203(11)$ as commented in \cite{Review}. On the other hand, our estimate

289: from Eq.~(\ref{binder2}) gives a value in full agreement with the Binder

290: cumulant and collapse data analysis in short time.

291: 

292: The case $q=4$ has been less studied. Our analysis gives $z=2.290(3)$ in the

293: interval $[60, 90]$ with $Q=0.72$. Here, we stress the importance of

294: monitoring $Q$, since we may find values for $z$ as large as $z=2.3483(2)$

295: in [10, 200] but with unacceptable value, $Q=10^{-269}$ or $z=2.3532(3)(2)$

296: in [10, 100] with $Q= 10^{-202}$.

297: 

298: As a final comment, the suspected $z$ as been weakly \cite{Arcangelis} or

299: even independent on $q$ \cite{Bonfim} is not supported by the most recently

300: results presented in Table 1.

301: 

302: 

303: 

304: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

305: 

306: 

307: \newpage

308: \vspace{0.4cm} 

309: 

310: {\bf Acknowledgements} 

311: \vspace{0.4cm}

312: 

313: R. da Silva gratefully acknowledges support by FAPESP (Brazil), and 

314: N. Alves by CNPq (Brazil). Thanks are also due to DFMA for computer 

315: facilities at IFUSP.

316: 

317: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

318: 

319: 

320: \begin{references}

321: \bibitem{Fisher72}  M.E. Fisher and M.N. Barber, Phys. Rev. Lett. 28 (1972)

322: 1516; M.E. Fisher and A.N. Berker, Phys. Rev. B26 (1982) 2507; V. Privman

323: and M.E. Fisher, J. Stat. Phys. 33 (1983) 385; V. Privman and M.E. Fisher,

324: Phys. Rev. B30 (1984) 322.

325: 

326: \bibitem{FisherLectures}  M.N. Barber, in Phase Transitions and Critical

327: phenomena, v.8. Edited by C. Domb and J.L. Lebowitz, 1973 (Academic Press).

328: 

329: \bibitem{HH1}  B.I. Halperin, P.C. Hohenberg and S-K. Ma, Phys. Rev. B10

330: (1974) 139.

331: 

332: \bibitem{Suzuki}  M. Suzuki, Prog. Theor. Phys. 58 (1977) 1142.

333: 

334: \bibitem{Janssen1}  H.K. Janssen, B. Schaub and B. Schmittmann, Z. Phys. B

335: 73 (1989) 539.

336: 

337: \bibitem{Huse}  D. Huse, Phys. Rev. B40 (1989) 304.

338: 

339: \bibitem{Universal97}  U. Ritschel and P. Czerner, Phys. Rev. E 55 (1997)

340: 3958.

341: 

342: \bibitem{Sch97}  K. Okano, L. Sch\"{u}lke, K. Yamagishi and B. Zheng, Nucl.

343: Phys. B485 [FS] (1997) 727.

344: 

345: \bibitem{Zhang99}  J.-B. Zhang L. Wang, D.-W. Gu, H.-P. Ying and D.-R. Ji,

346: Phys. Lett. A 262 (1999) 226.

347: 

348: \bibitem{Li95}  Z.B. Li, L. Sch\"{u}lke and B. Zheng, Phys. Rev. Lett. 74

349: (1995) 3396.

350: 

351: \bibitem{Grassberger}  P. Grassberger, Physica A 214 (1995) 547.

352: 

353: \bibitem{Nightingale}  M.P. Nightingale, H.W.J. Bl\"{o}te, Phys. Rev. B62

354: (2000) 1089; Phys. Rev. Lett. 76 (1996) 4548.

355: 

356: \bibitem{Review}  B. Zheng, Int. J. Mod. Phys. B12 (1998) 1419, H.P. Ying,

357: Phys. A, 294 (2001) 111.

358: 

359: \bibitem{Simoes}  C.S. Sim\~{o}es and J.R. Drugowich de Fel\'{\i}cio, Mod.

360: Phys. Lett. B15 (2001) 487, L. Wang, J. B. Zhang, H. P. Ying and D. R. Ji,

361: Mod. Phys. Lett. B13 (1999) 1011. 

362: 

363: \bibitem{HH2}  P.C. Hohenberg and B.I. Halperin, Rev. Mod. Phys. 49 (1977)

364: 435.

365: 

366: \bibitem{Foundations}  K. Okano, L. Sch\"{u}lke and B. Zheng, Foundations of

367: Physics 27 (1997) 1739.

368: 

369: \bibitem{Li96}  Z. Li, L. Sch\"{u}lke, B. Zheng, Phys. Rev. E53 (1996) 2940.

370: 

371: \bibitem{Stauffer92}  D. Stauffer, Physica A 186 (1992) 197.

372: 

373: \bibitem{Stauffer93}  C. M\"{u}nkel, D.W. Heermann, J. Adler, M. Gofman and

374: D. Stauffer, Physica A 193 (1993) 540.

375: 

376: \bibitem{Sch96}  L. Sch\"{u}lke, B. Zheng, Phys. Lett. A 215 (1996) 2940.

377: 

378: \bibitem{Press}  W. Press {\it et al.}, {\it Numerical Recipes} (Cambridge

379: University Press, Londom, 1986).

380: 

381: \bibitem{Zheng98}  G.P. Zheng and J.X. Zhang, Phys. Rev. E 58 (1998) R1187.

382: 

383: \bibitem{Wang97}  F.-G. Wang and C.-K. Hu, Phys. Rev. E 56 (1997) 2310.

384: 

385: \bibitem{Wang95}  F. Wang, N. Hatano and M. Suzuki, J. Phys. A: Math. Gen.

386: 28 (1995) 4543.

387: 

388: \bibitem{Ito93}  N. Ito, Physica A 196 (1993) 591.

389: 

390: \bibitem{Jan92}  K. MacIsaac and N. Jan, J. Phys. A: Math. Gen. 25 (1992)

391: 2139.

392: 

393: \bibitem{Bonfim}  O.F. de Alcantara Bonfim, Europhys. Lett. 4 (1987) 373.

394: 

395: \bibitem{Arcangelis}  L. de Arcangelis and N. Jan, J. Phys. A: Math. Gen. 19

396: (1986) L1179.

397: \end{references}

398: 

399: 

400: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

401: \vspace{1.0cm}

402: 

403: {\huge Figure Captions:} \\

404: \begin{description}

405: \item[Figure 1.] Time evolution of $F_2(t)$ for 2D three-state Potts model.\\

406: \item[Figure 2.] Time evolution of $F_2(t)$ for 2D four-state Potts model. 

407: \end{description}

408: 

409: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

410: 

411: % TABLE 1

412: \begin{table}[ht]

413: \caption{\baselineskip=0.8cm Dynamic exponent $z$ for 2D Ising model, three

414: and four-state Potts model.}\renewcommand{\tablename}{Table}

415: \par

416: \begin{center}

417: \begin{tabular}{llll}

418: &  &  &  \\[-0.3cm] 

419: Reference (year) & Ising & $q=3$ & $q=4$ \\ 

420: &  &  &  \\[-0.35cm] \hline

421: &  &  &  \\[-0.3cm] 

422: This work & 2.1565(7) & 2.198(2) & 2.290(3) \\ 

423: \cite{Nightingale} ~(2000) & 2.1667(5) &  &  \\ 

424: ~~\cite{Zhang99}$^1$ (1999) & 2.153(2) & 2.191(6) &  \\ 

425: \cite{Review}$^2$ (1998) & 2.153(4) & 2.203(11) &  \\ 

426: \cite{Review}$^3$ (1998) & 2.16(2) & 2.14(3) &  \\ 

427: \cite{Zheng98} ~(1998) & 2.137(8) &  &  \\ 

428: \cite{Wang97} ~(1997) & 2.166(7) &  &  \\ 

429: ~~\cite{Sch97}$^4$ (1997) & 2.155(3) & 2.196(8) &  \\ 

430: \cite{Grassberger} ~(1995) & 2.172(6) &  &  \\ 

431: \cite{Wang95} ~(1995) & 2.16(4) &  &  \\ 

432: \cite{Ito93} ~(1993) & 2.165(10) &  &  \\ 

433: \cite{Jan92} ~(1992) & 2.16(2) &  &  \\ 

434: \cite{Bonfim} ~(1987) & 2.16(5) & 2.16(4) & 2.18(3) \\ 

435: \cite{Arcangelis} ~(1986) &  & 2.43(15) & 2.36(20)

436: \end{tabular}

437: \end{center}

438: \par

439: $^1$ 2D TP lattice. \newline

440: $^2$ from scaling collapse (Table 3). \newline

441: $^3$ applying Eq.(\ref{binder1}). \newline

442: $^4$ from HB algorithm, while $z=2.137(11) \,{\rm and}\, z=2.198(13)$,

443: respectively for Ising and $q=3$ model from Metropolis algorithm. 

444: %%\end{tabular}}

445: \end{table}

446: 

447: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

448: 

449: \newpage

450: \cleardoublepage

451: 

452: 

453: %FIGURE 1

454: \begin{figure}[b]

455: %\begin{figure}[!ht]

456: \begin{center}

457: \begin{minipage}[t]{0.95\textwidth}

458: \centering

459: %%\includegraphics[angle=-90,width=0.72\textwidth]{fig1.eps}

460: \includegraphics[width=0.72\textwidth]{fig1.eps}

461: \renewcommand{\figurename}{(Fig.1)}

462: \caption{Time evolution of $F_2(t)$ for 2D three-state Potts model.}

463: \label{Fig. 1}

464: \end{minipage}

465: \end{center}

466: \end{figure}

467: 

468: 

469: 

470: %\newpage

471: %\cleardoublepage

472: 

473: 

474: 

475: %FIGURE 2

476: \begin{figure}[b]

477: %\begin{figure}[!ht]

478: \begin{center}

479: \begin{minipage}[t]{0.95\textwidth}

480: \centering

481: %%\includegraphics[angle=-90,width=0.72\textwidth]{fig2.eps}

482: \includegraphics[width=0.72\textwidth]{fig2.eps}

483: \renewcommand{\figurename}{(Fig.2)}

484: \caption{Time evolution of $F_2(t)$ for 2D four-state Potts model.}

485: \label{Fig. 2}

486: \end{minipage}

487: \end{center}

488: \end{figure}

489: 

490: 

491: \end{document}

492: 

493: 

494: