summaryrefslogtreecommitdiff
path: root/aps_mm_2017.tex
blob: e85f277a2abe0e56554e7a86c2371a87433c07c4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
%
%  research_midsummer.tex - Research Presentation for the Topaz lab.
%
%  Created by Jaron Kent-Dobias on Tue Mar 20 20:57:40 PDT 2012.
%  Copyright (c) 2012 pants productions. All rights reserved.
%

\documentclass[fleqn]{beamer}

\usepackage[utf8]{inputenc}
\usepackage{amsmath,amssymb,latexsym,graphicx}
\usepackage{concmath}
%\usepackage{bera}
%\usepackage{merriweather}
\usepackage[T1]{fontenc}

\usecolortheme{beaver}
\usefonttheme{serif}
\setbeamertemplate{navigation symbols}{}

\title{Universal scaling and the essential singularity at the abrupt Ising transition}
\author{ Jaron~Kent-Dobias\inst{1} \and James~Sethna\inst{1}}
\institute{\inst{1}Cornell University}
\date{16 March 2016}

\begin{document}

\def\dd{\mathrm d}
\def\im{\mathop{\mathrm{Im}}}
\def\ei{\mathop{\mathrm{Ei}}}
\def\crit{\mathrm{crit}}

\begin{frame}
  \titlepage
\end{frame}

\begin{frame}
  \frametitle{Parametric Ising Models}
  \begin{columns}
    \begin{column}{0.4\textwidth}
      \includegraphics[width=\textwidth]{figs/fig7}

      \vspace{1em}

      \includegraphics[width=\textwidth,height=0.5\textwidth]{susceptibility.jpg}

      \tiny\texttt{a plot of susceptibility with precision parametric fit and it
        isn't
        very good at the abrupt transition}
    \end{column}
    \begin{column}{0.6\textwidth}
      Scaling forms of Ising variables that do well globally.

      \pause\vspace{1em}

      Incorporate the critical point in a natural way:
      \begin{itemize}
        \item singular scaling with the ``radial coordinate''
        \item analytic scaling with the ``angular coordinate''
      \end{itemize}

      \vspace{1em} \pause

      Typically do a very poor job near the abrupt transition at $H=0$.
    \end{column}
  \end{columns}
\end{frame}

\begin{frame}
  \frametitle{Renormalization and Universality}

  \begin{columns}
    \begin{column}{0.4\textwidth}
      \centering 
      \includegraphics[width=\textwidth]{figs/fig2}

      \tiny
      From \emph{Scaling and Renormalization in Statistical Physics} by John
      Cardy
    \end{column}
    \begin{column}{0.6\textwidth}
      Renormalization is an analytic scaling transformation that acts on
      system space.

      \vspace{1em}\pause

      Fixed points are scale invariant, corresponding to systems representing
      idealized phases or critical behavior.

      \vspace{1em}\pause

      Nonanalytic behavior---like power laws and logarithms---are preserved
      under {\sc rg} and shared by \emph{any} system that flows to the same
      point.

      \vspace{1em}\pause

      Not all nonanalytic behavior are power laws!
    \end{column}
  \end{columns}
\end{frame}

\begin{frame}
  \frametitle{The Metastable Ising Model}

  \begin{columns}
    \begin{column}{0.3\textwidth}
      \only<1-1>{\includegraphics[width=\textwidth]{figs/fig3}}
      \only<2->{\includegraphics[width=\textwidth]{figs/fig4}}

      \vspace{1em}

      \includegraphics[width=\textwidth]{figs/fig8}

    \end{column}
    \begin{column}{0.7\textwidth}
      Consider an Ising-class model brought into a metastable state.

      \vspace{1em} \pause\pause

      A domain of $N$ spins entering the stable phase causes a free energy
      change
      \[
        \Delta F=\Sigma N^\sigma-MHN
      \]

      \pause

      The metastable phase is stable to domains smaller than
      \[
        N_\crit=\bigg(\frac{MH}{\sigma\Sigma}\bigg)^{-1/(\sigma-1)}
      \]
      but those larger will grow to occupy the entire system.
    \end{column}
  \end{columns}
\end{frame}

\begin{frame}
  \frametitle{The Metastable Ising Model}

  The formation of a critical domain has energy cost
  \[
    \Delta F_\crit\sim MH\bigg(\frac{MH}{\Sigma}\bigg)^{-1/(1-\sigma)}
  \]

  \pause

  The decay rate of the metastable is proportional to the probability of
  forming a critical domain $e^{-\beta\Delta F_\crit}$.
  
  \pause \vspace{1em}
  
  Decay of the equilibrium state implies existence of an imaginary part in the
  free energy,
  \[
    \im F\sim e^{-\beta\Delta F_\crit}
  \]

\end{frame}
\begin{frame}
  \frametitle{The Metastable Ising Model}

  Near the Ising critical point, $\sigma=1-\frac1d$ and
  \begin{align*}
    M=t^\beta\mathcal M(h/{t^{\beta\delta}})
    &&
    \Sigma=t^\mu\mathcal S(h/{t^{\beta\delta}})
  \end{align*}
  with $\mathcal M(0)$ and $\mathcal S(0)$ nonzero and finite.

  \pause \vspace{1em}

  Therefore,
  \[
    \Delta F_\crit\sim\Sigma\bigg(\frac{MH}{\Sigma}\bigg)^{-(d-1)}
    =X^{-(d-1)}\mathcal F(X)
  \]
  for $X=h/t^{\beta\delta}$, and 
  \[
    \im F=t^{2-\alpha}\mathcal I(X)e^{-\beta/X^{(d-1)}}
  \]
\end{frame}

\begin{frame}
  \frametitle{The Essential Singularity}

  \begin{center}
    \includegraphics[width=.7\textwidth]{figs/fig1}
  \end{center}

      Imaginary free energy is nonanalytic at $H=0$.

      \pause\vspace{1em}

      This and its implications are therefore a universal feature of the Ising class.
\end{frame}

\begin{frame}
  \frametitle{The Essential Singularity}

  Analytic properties of the partition function imply that
  \[
    F(X)=\frac1\pi\int_{-\infty}^0\frac{\im F(X')}{X'-X}\;\dd X'
  \]

  \pause

  Only predictive for high moments of $F$, or
  \[
    f_n=\frac1\pi\int_{-\infty}^0\frac{\im F(X')}{X^{\prime n+1}}\;\dd X'
  \]
  for $F=\sum f_nX^n$.
\end{frame}

\begin{frame}
  \frametitle{The Essential Singularity}

  Results from field theory indicate that $\mathcal I(X)\propto X+\mathcal
  O(X^2)$ for $d=2$, so that
  \[
    \im F=t^{2-\alpha}\big(AX+\mathcal O(X^2)\big)e^{-\beta/X^{(d-1)}}
  \]

  \pause

  The resulting moments for $n>1$ are
  \[
    f_n=At^{2-\alpha}\frac{\Gamma(n-1)}{\pi(-B)^{n-1}}
  \]

  \pause

  Not a convergent series---the real part of $F$ for $H>0$ is also
  nonanalytic!
\end{frame}

\begin{frame}
  \frametitle{The Essential Singularity}

  In two dimensions, the Cauchy integral does not converge, normalize with
  $\lambda$,
  \[
    F(X\,|\,\lambda)=\frac1\pi\int_{-\infty}^0\frac{\im
    F(X')}{X'-X}\frac1{1+(\lambda X')^2}\;\dd X'
  \]

  \pause

  Exact result has form
  \[
    \begin{aligned}
      F(X\,|\,\lambda)&=\frac{A}\pi\frac1{1+(\lambda X)^2}\Big[
      Xe^{B/X}\ei(-B/X)\\
      &\qquad+\frac1\lambda\im(e^{-i\lambda B}(i+\lambda
      X)(\pi+i\ei(i\lambda B)))\Big]
    \end{aligned}
  \]

  \pause

  The Cauchy integral is only predictive for high moments.

\end{frame}
\begin{frame}
  \frametitle{The Essential Singularity}

  What about the susceptibility $\chi=\frac{\partial^2\!F}{\partial h^2}$?

  \pause \vspace{1em}

  Has a well-defined limit as $\lambda\to0$, simple functional form:
  \[
    \chi=t^{-\gamma}\mathcal X(h/t^{\beta\delta})
  \]
  where the scaling function is
  \[
    \mathcal X(X)=\frac A{\pi X^3}\big[(B-X)X+B^2e^{B/X}\ei(-B/X)\big]
  \]
  
  \centering
  \includegraphics[width=0.6\textwidth]{figs/fig9}
\end{frame}

\begin{frame}
  \frametitle{The Essential Singularity}

  Two parameter fit to simulations yields $A=-0.0939(8)$, $B=5.45(6)$, close
  agreement in limit of small $t$ and $H$!

  \vspace{1em}

  \only<1-1>{\includegraphics[width=\textwidth]{figs/fig6}}
  \only<2-2>{\includegraphics[width=\textwidth]{figs/fig5}}

  \vspace{1em}\pause
\end{frame}

\begin{frame}
  \frametitle{What's Next}

  We have an explicit form for a new component of the universal scaling forms
  near the Ising abrupt transition.

  \vspace{1em} \pause

  Hope to form a parametric scaling variables that include this, correct
  leading
  analytic corrections to scaling, and (maybe?) extend smoothly through the
  metastable region.

  \vspace{1em} \pause

  Remain on the lookout for other universal properties to incorporate.
\end{frame}

\begin{frame}
  \huge
  \centering
  {\sl Questions?}
\end{frame}

\end{document}