1: \documentclass[12pt]{article}
2: \input epsf
3: \usepackage{fullpage}
4: %\usepackage{psfig}
5: \usepackage{graphicx}
6: \newcommand{\be}{\begin{equation}}
7: \newcommand{\ee}{\end{equation}}
8: \newcommand{\bphi}{\mbox{\boldmath $\phi$}}
9: \newcommand{\pauli}{\mbox{\boldmath $\tau$}}
10: %%%%%%%If you do not have the msbm fonts, delete the following 10 lines
11: \font\mybb=msbm10 at 11pt
12: \font\mybbb=msbm10 at 17pt
13: \def\bb#1{\hbox{\mybb#1}}
14: \def\bbb#1{\hbox{\mybbb#1}}
15: \def\bZ {\bb{Z}}
16: \def\bR {\bb{R}}
17: \def\bE {\bb{E}}
18: \def\bT {\bb{T}}
19: \def\bM {\bb{M}}
20: \def\bC {\bb{C}}
21: \def\bA {\bb{A}}
22: \def\bP {\bb{P}}
23: \def\e {\epsilon}
24: \def\bbC {\bbb{C}}
25: \def\Z {\bb{Z}}
26: %%%%%%%%%%%%%%%%%
27: \renewcommand{\theequation}{\arabic{section}.\arabic{equation}}
28: \newcommand{\news}{\setcounter{equation}{0}}
29: %\newcommand{\newss}{\setcounter{equation}{0}}
30: \def\ben{\begin{equation}}
31: \def\een{\end{equation}}
32: \def\bea{\begin{eqnarray}}
33: \def\eea{\end{eqnarray}}
34: %\input amstex
35: %\input amssym.def
36: %\input amssym.tex
37: %\renewcommand{\baselinestretch}{1.67}
38: \begin{document}
39: \title{
40: %\begin{flushright}\ \vskip -2cm {\tiny{\em DRAFT}}\end{flushright}
41: \vskip 2cm Knots in the Skyrme-Faddeev model}
42: \author{Paul Sutcliffe\\[10pt]
43: {\em \normalsize Department of Mathematical Sciences,
44: Durham University, Durham DH1 3LE, U.K.}\\
45: {\normalsize Email: p.m.sutcliffe@durham.ac.uk}}
46: \date{May 2007}
47: \maketitle
48: \begin{abstract}
49: The Skyrme-Faddeev model is a modified sigma model in
50: three-dimensional space, which has string-like topological solitons
51: classified by the integer-valued Hopf charge. Numerical simulations
52: are performed to compute soliton solutions for Hopf charges up to
53: sixteen, with initial conditions provided by families of rational
54: maps from the three-sphere into the complex projective line. A large
55: number of new solutions are presented, including a variety of torus
56: knots for a range of Hopf charges. Often these knots are only local
57: energy minima, with the global minimum being a linked solution, but
58: for some values of the Hopf charge they are good candidates
59: for the global minimum energy solution. The computed energies are in
60: agreement with Ward's conjectured energy bound.
61: \end{abstract}
62:
63: \newpage
64: \section{Introduction}\news
65:
66: \ \quad Over thirty years ago Faddeev suggested \cite{Fa2} that in
67: three-dimensional space the $O(3)$ sigma model, modified by the
68: addition of a Skyrme term, should have interesting string-like
69: topological solitons stabilized by the integer-valued Hopf charge.
70: Ten years ago substantial interest was generated by the first
71: attempts at a numerical construction of such solitons \cite{FN, GH}
72: and the suggestion that minimal energy solitons might take the form
73: of knots \cite{FN}. This conjecture has been confirmed by numerical
74: results, which demonstrate that the minimal energy soliton with Hopf
75: charge seven is a trefoil knot \cite{BS5}. Substantial numerical
76: investigations by several authors \cite{FN,GH,BS5,HS,HS2,Wa5} has
77: produced a comprehensive analysis of solitons with Hopf charges from
78: one to seven, and it appears that the global energy minima have now
79: been identified for these charges, together with several other
80: stable soliton solutions that correspond to local energy minima.
81:
82: For Hopf charges five and six the minimal energy solitons form
83: links, but so far the charge seven solution remains the only known
84: knot, even including local energy minima. It is therefore currently
85: unknown whether this charge seven trefoil knot is the only knotted
86: solution or if there are many other knots, perhaps of various types,
87: which arise for higher Hopf charges. It is this question which is
88: addressed in the present paper. Solitons with Hopf charges up to
89: sixteen are constructed numerically and it is found that a variety
90: of torus knots exist at various Hopf charges. Often these knots are
91: only local energy minima, with the global minimum being a link, but
92: for some values of the Hopf charge it appears that a knot is a good
93: candidate for the global minimum energy solution.
94:
95: One of the difficulties in extending previous numerical studies to
96: higher Hopf charges is that, even at low charges, there are a number
97: of local energy minima with large capture basins. This makes it
98: difficult to fully explore the landscape of local energy minima and
99: hence determine the global minimum, with the severity of the
100: problem generally increasing as the charge increases. To overcome
101: this problem an analytic ansatz is employed which uses rational maps
102: from the three-sphere into the complex projective line. Although
103: this ansatz does not provide any exact solutions, it does allow the
104: construction of reasonable initial conditions for any torus knot for
105: a large range of charges, plus a wide selection of links and
106: unknots. It is therefore possible to start at a variety of locations
107: in field configuration space, in particular where one suspects that
108: a local energy minimum might be close by, and explore the energy
109: landscape around this point. Not only is this a fruitful approach
110: for finding new solutions, but it can also be used to provide strong
111: evidence that a solution of a particular type does not exist, by
112: starting with an initial condition approximating this configuration
113: and demonstrating that it changes dramatically under energy
114: relaxation.
115:
116: An analysis is made of the new knotted and linked solutions and some
117: understanding is obtained regarding the charges at which particular
118: knots are likely to exist.
119:
120: \section{The Skyrme-Faddeev model and low charge solitons}\news
121: \ \quad The Skyrme-Faddeev model involves a map $\bphi:
122: \bb{R}^3\mapsto S^2,$ which is realized as a real three-component
123: vector $\bphi=(\phi_1,\phi_2,\phi_3),$ of unit length,
124: $\bphi\cdot\bphi=1.$ As this paper is concerned only with static
125: solutions then the model can be defined by its energy \be
126: E=\frac{1}{32\pi^2\sqrt{2}}\int \partial_i\bphi\cdot\partial_i\bphi
127: +\frac{1}{2}(\partial_i\bphi\times\partial_j\bphi)
128: \cdot(\partial_i\bphi\times\partial_j\bphi)\ d^3x, \label{energy}
129: \ee where the normalization is chosen for later convenience. The
130: first term in the energy is that of the usual $O(3)$ sigma model and
131: the second is a Skyrme term, required to provide a balance under
132: scaling and hence allow solitons with a finite non-zero size.
133:
134: Finite energy boundary conditions require that the field tends to a
135: constant value at spatial infinity, which is chosen to be
136: $\bphi(\infty)=(0,0,1)={\bf e}_3.$ This boundary condition
137: compactifies space to $S^3,$ so that the field becomes a map $\bphi:
138: S^3\mapsto S^2.$ Such maps are classified by $\pi_3(S^2)=\bb{Z},$ so
139: there is an integer-valued topological charge $Q,$ the Hopf charge,
140: which gives the soliton number. Unlike most theories with
141: topological solitons, the topological charge is not a winding number
142: or degree of a mapping. Rather, it has a geometrical interpretation
143: as a linking number of field lines, as follows. Generically, the
144: preimage of a point on the target two-sphere is a closed loop (or a
145: collection of closed loops) in the three-sphere domain obtained from
146: the compactification of $\bb{R}^3.$ Two loops obtained as the
147: preimages of any two distinct points on the target two-sphere are
148: linked exactly $Q$ times, where $Q$ is the Hopf charge.
149:
150: The Hopf charge can be written as the integral over the three-sphere
151: domain of a charge density, but this density is non-local in the
152: field $\bphi,$ as the following construction demonstrates. Let
153: $\omega$ denote the area two-form on the target two-sphere and let
154: $F=\bphi^*\omega$ be its pull-back under $\bphi$ to the domain
155: three-sphere. The triviality of the second cohomology group of the
156: three-sphere implies that $F$ is an exact two-form, say $F=dA.$ The
157: Hopf charge is then given by integrating the Chern-Simons three-form
158: over the three-sphere as \be Q=\frac{1}{4\pi^2}\int_{S^3}F\wedge
159: A.\ee
160:
161: The energy bound \be E\ge c\, Q^{3/4},\quad \mbox{where} \quad
162: c=\left(\frac{3}{16}\right)^{3/8}\approx 0.534\ee has been proved
163: \cite{VK,KR}, and it is known that the fractional power is optimal
164: \cite{LY}, though it is expected that the above value of the
165: constant $c$ is not. Motivated by a study of the Skyrme-Faddeev
166: model on a three-sphere with a finite radius, together with an
167: analogy with the Skyrme model, Ward has conjectured \cite{Wa4} that
168: the above energy bound holds with the value $c=1$ (hence the choice
169: of normalization factor for the energy in (\ref{energy})), but this
170: has not been proven. As discussed shortly, the current energy values
171: known for low charge solitons are in good agreement with Ward's
172: conjectured energy bound, and later it will be demonstrated that the
173: energies for higher charge solitons are too.
174:
175: The simplest examples of Hopf solitons are axially symmetric and
176: their qualitative features may be described as follows. First,
177: consider the model (\ref{energy}) defined in two-dimensional space,
178: so that the boundary conditions now result in a compactification of
179: space from $\bb{R}^2$ to $S^2.$ The field is then a map
180: $\bphi:S^2\mapsto S^2$ and is classified by an integer $m\in\bb{Z}
181: =\pi_2(S^2),$ which is the usual winding number familiar from the
182: $O(3)$ sigma model itself. Strictly speaking, in order to have
183: two-dimensional solitons with a finite size in this model then a
184: potential term must also be included, upon which the two-dimensional
185: solitons are known as baby Skyrmions \cite{PSZ}, but this aspect is
186: not crucial for the discussion here. Returning to the
187: three-dimensional model, then a toroidal field configuration can be
188: formed by embedding the two-dimensional soliton in the normal slice
189: to a circle in space. The two-dimensional soliton has an internal
190: phase and this can be rotated through an angle $2\pi n,$ as it
191: travels around the circle once, where $n\in \bb{Z}$ counts the
192: number of twists. A field configuration of this type has Hopf charge
193: $Q=nm,$ and it will be denoted by ${\cal A}_{n,m},$ so the first
194: subscript labels the number of twists and the second is the winding
195: number of the two-dimensional soliton forming the loop.
196:
197: The minimal energy $Q=1$ soliton is of the type ${\cal A}_{1,1}$ and
198: has been studied numerically by a number of authors
199: \cite{GH,FN,BS5,HS,Wa5} using different algorithms.
200: Taken together, these results suggest that the charge one soliton has
201: an energy of around $E=1.21,$ and this is part of a
202: general pattern where the minimal energy soliton exceeds Ward's
203: conjectured bound by around $20\%.$ The numerical computations
204: reported in this paper produce a charge one energy of $E=1.204,$ and
205: hence appear to underestimate the energy by an amount of the order
206: of $1\%.$ For low charge solitons this accuracy could be improved by
207: increasing the resolution of the numerical grid, but it is
208: computationally too expensive to increase the resolution for all the
209: large number of simulations reported here for large charges.
210: Moreover, although the energy is a slight underestimate for a given
211: soliton solution, the relative energies between two given solutions
212: are much more accurate than this (changing grid resolutions and
213: sizes suggests an accuracy of around $0.1\%$) and it is more
214: important to preserve the relative energies than the absolute energy
215: of any single solution.
216:
217: The details of the numerical computations are similar to those of
218: Ref.\cite{BS5}, and full details of the related numerical code used
219: to study Skyrmions can be found in Ref.\cite{BS3}. Briefly, a finite
220: difference scheme is employed with a lattice spacing $\Delta x=0.1$
221: on a grid containing $(151)^3$ lattice points, with the field fixed
222: to the vacuum value $\bphi={\bf e}_3$ on the boundary of the grid;
223: this is preferable to other methods of dealing with the finite
224: volume simulation domain, such as the one used in Ref.\cite{BS5}
225: which led to a more substantial underestimate of energies, though
226: again relative energies had a high accuracy. The energy minimization
227: algorithm proceeds by evolving second order in time dynamics and
228: periodically removing kinetic energy from the system whenever the
229: potential energy of the system begins to increase. It is based on
230: the approach employed in Ref.\cite{BS5}, but is computationally more
231: efficient since the second order dynamics is derived from the
232: kinetic term of the sigma model only, rather than the full kinetic
233: term determined from the Lorentz invariant Lagrangian associated
234: with the energy (\ref{energy}). This subtle difference leads to a
235: more efficient algorithm since the leading term in the evolution
236: equations is now diagonal, so a costly matrix inversion can be
237: avoided.
238:
239: A sensible definition of the position of a soliton is to identify
240: where the field is as far as possible from the vacuum field. A Hopf
241: soliton is therefore string-like, since the soliton's position is
242: defined to be the closed loop (or collection of loops) corresponding
243: to the preimage of the point $\bphi=-{\bf e}_3=(0,0,-1),$ which is
244: antipodal to the vacuum value on the target two-sphere. To visualize
245: a field configuration this position curve is plotted, though for
246: clarity a tube around this position is displayed, given by an
247: isosurface of the form $\phi_3=-1+\nu,$ where generally the value
248: $\nu=0.2$ is chosen. If a pictorial representation of the Hopf
249: charge is also required then a second preimage curve can also be
250: plotted (again as a tube), and the linking number of these two
251: curves inspected. There is no natural choice for this second
252: preimage value (${\bf e}_3$ is not a very useful choice) and in this
253: paper when a second preimage curve is displayed it corresponds to
254: the point $\bphi=(\sqrt{2\mu-\mu^2},0,-1+\mu)$ where $\mu=0.1.$ In what
255: follows this curve will be referred to as the linking curve, and
256: since $\mu$ is relatively small then the linking curve remains
257: reasonably close to the position curve, making it fairly easy to
258: inspect the linking number.
259:
260: In Figure \ref{fig-low} the position (light tube) and linking (dark tube)
261: curves are displayed for the known lowest energy solitons with Hopf
262: charges from one to seven. Given the large number of numerical
263: simulations performed by different groups it seems reasonably
264: certain that these are the minimal energy solitons for these
265: charges. The energies of these solutions will be discussed later,
266: and as mentioned above they are generally around $20\%$ above Ward's
267: conjectured bound, but for now it is the structure of the solutions
268: that is of primary interest, and this is reviewed below.
269:
270: The solutions with charges one and two are both axially symmetric,
271: and are of the type ${\cal A}_{1,1}$ and ${\cal A}_{2,1}$
272: respectively \cite{FN,GH}, with one and two twists of the linking
273: curve around the position curve clearly visible. There is a solution
274: of the form ${\cal A}_{1,2}$ \cite{Wa4,Wa5,HS} but it has an energy
275: around $13\%$ above that of the minimal charge two soliton. This
276: solution has the interpretation of two $Q=1$ solitons stacked one
277: above the other, preserving the axial symmetry.
278:
279: The charge three soliton is basically of the type ${\cal
280: A}_{3,1}$ but the axial symmetry is broken as the position curve is
281: bent \cite{BS5}, so it will be denoted by $\widetilde {\cal
282: A}_{3,1}$ to emphasize that the structure is deformed. There is an
283: axial solution of the type ${\cal A}_{n,1}$ for any integer $n,$ but
284: for $n>2$ this solution is unstable to a coiling instability
285: \cite{BS5}; a well-known phenomenon in other settings such as
286: twisted elastic rods. The $Q=4$ soliton is of the type ${\cal
287: A}_{2,2},$ \cite{HS2,Wa4}, and may be thought of as two $Q=2$ solitons
288: stacked one above the other. For both $Q=4$ and $Q=5$ there are also
289: bent solutions \cite{BS5}, similar to the charge three soliton, and
290: therefore denoted $\widetilde {\cal A}_{4,1}$ and $\widetilde {\cal
291: A}_{5,1},$ but unlike the $Q=3$ case, both these solutions are only
292: local energy minima, with energies around
293: $2\%$ and $4\%$ above the global minimum, respectively.
294: \begin{figure}
295: \begin{center}
296: %\leavevmode
297: %\epsfxsize=16cm\epsffile{1to7.ps}
298: \includegraphics[width=16cm]{1to7.ps}
299: \caption{The position (light tube) and linking (dark tube)
300: curves for the known lowest energy solitons with Hopf
301: charges $1\le Q\le 7.$}
302: \label{fig-low}
303: \end{center}
304: \end{figure}
305: The minimal energy charge five soliton provides the first example of
306: a link, that is, a solution in which the position curve contains two
307: or more disconnected components. It consists of a charge one soliton
308: and a charge two soliton which are linked once \cite{HS2}, with the
309: result that each of these solitons gains an additional unit of
310: linking number to make a total of five. To denote a link of this
311: type the notation ${\cal L}_{1,2}^{1,1}$ will be used, where the
312: subscripts label the charges of the components of the links and the
313: superscript above each subscript counts the extra linking number of
314: that component, due to its linking with the others. Hence the total
315: charge is the sum of the subscripts plus superscripts. The charge
316: six soliton is similar to that of charge five, but now both
317: components of the link have charge two \cite{BS5}, so using the
318: above notation it is written as ${\cal L}_{2,2}^{1,1}.$
319:
320: Finally, the charge seven soliton is the first (and so far only)
321: example of a knot \cite{BS5}. The position curve is a trefoil knot,
322: which has a self-linking or crossing number of three (the crossing
323: number of a knot is not an invariant and its use in this paper
324: refers to the minimal crossing number over all presentations). An
325: examination of the linking curve in Figure \ref{fig-low} confirms
326: that it twists
327: around the position curve four times as the knot is traversed, so
328: the Hopf charge is indeed seven, being the sum of the crossing
329: number plus the number of twists. Any field configuration of a
330: trefoil knot will be denoted by ${\cal K}_{3,2},$ which refers to
331: the fact that the trefoil knot is also the $(3,2)$-torus knot (see
332: the following section for more information regarding torus knots).
333: Note that this notation does not display the Hopf charge of the
334: configuration, which is determined by the twist number as well as
335: the crossing number, but this should not cause any confusion in what
336: follows. Of course, for knots and links, as well as for unknot
337: configurations, the precise conformation of the position curve is
338: important in determining the energy, not just its topological type,
339: but its knotted or linked structure is an important feature in
340: classifying the solution. Thus, for example, the $Q=7$ trefoil knot
341: displayed in Figure \ref{fig-low} is not as symmetric as its typical knot
342: theory presentation, and it is significant that the energy is
343: lowered by breaking the possible cyclic $C_3$ symmetry.
344:
345: Having reviewed the known results for low charge solitons and
346: introduced the notation used to label their structural type, it is
347: time to turn to the main questions addressed in this paper. Given
348: the variety of solutions which appear, even at low charges, it is
349: difficult to make a confident prediction of the kind of behaviour
350: that might arise at higher charges. In particular, an interesting
351: open question is whether more knot solitons appear at higher charge,
352: and if so what types of knots arise and are they local or global
353: minima? It could have been the case that the charge seven trefoil
354: knot was the only knot soliton, though in this paper it will be
355: shown that this is far from true. Also, it would be useful to have
356: at least a qualitative understanding of the fact that $Q=7$ is the
357: lowest charge at which a knot first appears, and desirable to be
358: able to estimate the charges at which other knots might exist.
359:
360: In this section it has been mentioned briefly that at each charge
361: there are often local minima in addition to the global minimum, and
362: it is expected that the number of local minima generally increases
363: with the charge. As noted in the introduction, these local minima
364: can have large capture basins, making it difficult to fully explore
365: the space of solutions. For any relaxation computation this makes
366: the choice of initial conditions a crucial issue. In particular, it
367: is not obvious how to construct a variety of
368: reasonably low-energy initial field
369: configurations for all Hopf charges. One family of configurations
370: are the axial fields ${\cal A}_{n,1},$ which are all unstable for
371: $n>2.$ However, even for charges as low as four, initial conditions
372: which use small perturbations of these unstable solutions get
373: trapped in local minima. One successful approach for low charges
374: involves constructing links by hand \cite{HS,HS2}, using a numerical
375: cut-and-paste technique where various numerical field configurations
376: are sewn together in different parts of the simulation grid to
377: create a linked field with a given Hopf charge. Although this
378: approach was the first method to successfully yield the global
379: minima at charges four and five, it becomes more cumbersome for
380: larger charges and is not applicable for creating knot initial
381: conditions; it is also not very elegant mathematically, but this is
382: perhaps not a serious criticism. Another set of initial conditions
383: that have been used \cite{HS2} are based on the fields ${\cal
384: A}_{n,m},$ with $m>1,$ though these fields tend to have relatively
385: large energies and therefore require quite long simulation times
386: (except for the low charge examples of ${\cal A}_{1,2}$ and ${\cal
387: A}_{2,2}).$ In the following section an analytic ansatz is presented
388: that yields reasonable field configurations for a variety of charges
389: and describes all torus knots plus a wide range of links, in addition
390: to fields of the type ${\cal A}_{n,m}.$ The key ingredient of the ansatz
391: involves a rational map from the three-sphere into the complex
392: projective line. The ansatz allows the construction of initial
393: conditions corresponding to a variety of locations in field
394: configuration space; in particular, configurations can be created
395: which are reasonably close to suspected local energy minima. This
396: approach substantially reduces the computational effort required to
397: explore the energy landscape and makes it feasible to study knots
398: and links for quite large charges.
399:
400: \section{Rational maps, torus knots and links}\news
401: \ \quad Torus knots are classified by a pair $(a,b)$ of coprime
402: positive integers with $a>b.$ A knot is a torus knot if it lies on
403: the surface of a torus, in which case the integers $a$ and $b$ count
404: the number of times that the knot winds around the two cycles of the
405: torus. The simplest example is the trefoil knot, which is the
406: $(3,2)$-torus knot and has three crossings. In the standard knot
407: catalogue notation it is $3_1,$ where the number refers to the
408: crossing number of the knot and the subscript labels its position in
409: the knot catalogue. Other torus knots which will be of interest in
410: this paper include the $(5,2)$-torus knot, also known as Solomon's
411: seal knot, which has five crossings (knot $5_1$ in the catalogue),
412: and the $(4,3)$-torus knot which has crossing number eight (knot
413: $8_{19}$ in the catalogue). In general the $(a,b)$-torus knot has
414: crossing number $C=a(b-1).$
415:
416: Denote by ${\cal K}_{a,b}$ any field configuration in which the
417: position curve is an $(a,b)$-torus knot. Such a field configuration
418: will have Hopf charge $Q=C+T,$ where $C=a(b-1)$ is the knot crossing
419: number and the integer $T$ counts the number of times that the
420: linking curve twists around the position curve (only situations
421: where the orientation is such that $T$ is non-negative will be of
422: relevance here). The main aim of this section is to construct fields
423: of type ${\cal K}_{a,b}$ for a large range of charges $Q.$
424:
425: The first step is to recall the standard description \cite{BK} of a
426: torus knot as the intersection of a complex algebraic curve with the
427: three-sphere. Consider $(Z_1,Z_0)\in \bb{C}^2$ and the unit
428: three-sphere $S^3\subset \bb{C}^2$ given by $|Z_1|^2+|Z_0|^2=1.$
429: Then the intersection of this three-sphere with the complex
430: algebraic curve $Z_1^a+Z_0^b=0$ is indeed the $(a,b)$-torus knot.
431:
432: To use the above description to produce a field configuration, the
433: spatial coordinates $(x_1,x_2,x_3)\in \bb{R}^3$ are mapped to the
434: unit three-sphere via a degree one spherically equivariant map.
435: Explicitly, \be (Z_1,Z_0)=((x_1+ix_2)\frac{\sin f}{r},\cos
436: f+i\frac{\sin f}{r}x_3),\ee where $r^2=x_1^2+x_2^2+x_3^2$ and the
437: profile function $f(r)$ is a monotonically decreasing function of
438: the radius $r$, with boundary conditions $f(0)=\pi$ and
439: $f(\infty)=0.$ The precise form of this function will be specified
440: shortly.
441:
442: A Riemann sphere coordinate $W,$ is used on the target two-sphere of
443: the field $\bphi,$ determined via stereographic projection as \be
444: W=\frac{\phi_1+i\phi_2}{1+\phi_3}.\ee With these definitions, then
445: constructing a map $\bphi:\bb{R}^3\mapsto S^2$ is equivalent to
446: specifying $W(Z_1,Z_0),$ that is, $W:S^3\mapsto \bb{C}\bb{P}^1,$ a
447: map from the three-sphere to the complex projective line. This map
448: will be taken to be a rational map, which means that $W=p/q,$ where
449: both $p$ and $q$ are polynomials in the variables $Z_1$ and $Z_0.$
450:
451: For an $(a,b)$-torus knot the mapping is chosen to have the form \be
452: W=\frac{p}{q}=\frac{Z_1^\alpha Z_0^\beta}{Z_1^a+Z_0^b},
453: \label{ratmap}\ee where $\alpha$ is a positive integer and $\beta$
454: is a non-negative integer. A map of this form has the correct
455: boundary conditions at spatial infinity, since as
456: $r\rightarrow\infty$ then $(Z_1,Z_0)\rightarrow(0,1)$ and, because
457: $\alpha>0,$ this gives $W\rightarrow 0,$ which is the stereographic
458: projection of the point $\bphi=(0,0,1)={\bf e}_3.$ Furthermore, the
459: position curve is the preimage of the point $-{\bf e}_3,$ which
460: corresponds to $W=\infty,$ and hence is given by $q=0=Z_1^a+Z_0^b.$
461: As the coordinates $(Z_1,Z_0)$ automatically lie on the unit
462: three-sphere then the position curve is the $(a,b)$-torus knot, for
463: all values of $\alpha$ and $\beta.$
464:
465: The Hopf charge of the map determined by (\ref{ratmap}) is given by
466: the cross ratio \be Q=\alpha b+\beta a.\label{charge}\ee To see this
467: note that a rational map \be W:S^3\mapsto \bb{C}\bb{P}^1, \quad
468: \mbox{given\ by} \quad W(Z_1,Z_0)=\frac{p(Z_1,Z_0)}{q(Z_1,Z_0)},\ee
469: has a natural extension to a map \be (p,q):\bb{B}^4\subset
470: \bb{C}^2\mapsto \bb{C}^2, \quad \mbox{given\ by} \quad
471: (p(Z_1,Z_0),q(Z_1,Z_0)).\ee Here $\bb{B}^4$ denotes the 4-ball with
472: boundary $S^3,$ and is obtained by replacing the constraint
473: $|Z_1|^2+|Z_0|^2=1$ by the inequality $|Z_1|^2+|Z_0|^2\le 1.$ The
474: map $(p,q)$ is between manifolds of the same dimension and
475: has a degree, defined by counting preimages of a generic point of
476: the target space, weighted by the signs of the Jacobian. As all the
477: other mappings involved in the ansatz have degree one, and the
478: standard Hopf map $(p,q)\mapsto p/q$ is employed, then the Hopf
479: charge $Q$ of the combined mapping is equal to the degree of the
480: mapping $(p,q).$ Furthermore, as the mapping $(p,q)$ is holomorphic
481: then the sign of the Jacobian of all generic preimage points is
482: positive, so the degree is simply the number of preimages.
483:
484: For the mapping $(p,q)$ given by (\ref{ratmap}) take the generic
485: point in target space to be $(\epsilon,0),$ so the Hopf charge is
486: the number of solutions of the equation $(Z_1^\alpha
487: Z_0^\beta,Z_1^a+Z_0^b)=(\epsilon, 0).$ Writing $Z_0$ in terms of
488: $Z_1$ from the first component of this equation and substituting
489: into the second component yields the polynomial equation
490: $Z_1^{\alpha b+\beta a}=(-1)^\beta\epsilon^b,$ which clearly has
491: $Q=\alpha b+\beta a$ solutions, as stated.
492:
493: For any $(a,b)$-torus knot the above ansatz can be used to construct
494: a field configuration of the type ${\cal K}_{a,b}$ with Hopf charge
495: $Q,$ providing there are integers $(\alpha,\beta),$ with $\alpha$
496: positive and $\beta$ non-negative, for which $Q=\alpha b+\beta a.$
497: For example, in the case of the trefoil knot then $(a,b)=(3,2)$ hence any
498: $Q>3$ can be obtained. In fact a trefoil knot with $Q=3$ can also be
499: obtained by interchanging the roles of $Z_1$ and $Z_0,$ but a field
500: of this type is not relevant, as will be made clear later. Note that
501: for some values of $Q,$ and fixed $(a,b),$ there are multiple
502: solutions for $(\alpha,\beta).$ These multiple solutions produce
503: qualitatively similar fields, in that both describe the same
504: $(a,b)$-torus knot position curve, and the linking curve twists the
505: same number of times around the position curve; though the
506: distribution of twist may vary. In examples where such multiple
507: solutions were used as initial conditions, the energy relaxation
508: algorithm produced identical final results for all choices.
509:
510: Recall that a profile function $f(r)$ needs to be specified,
511: satisfying the boundary conditions $f(0)=\pi$ and $f(\infty)=0.$ In
512: fact, these are the correct boundary conditions in an infinite
513: domain but in the finite domain used for numerical simulations the
514: last boundary condition needs to be replaced by $f=0$ at the
515: boundary of the numerical grid. As the ansatz is only used to
516: provide initial conditions then most reasonable monotonic functions
517: will suffice. The simulations discussed in this paper were performed
518: on a cubic grid of side-length $2L,$ that is, $-L\le x_i\le L,$
519: where $2L=15.$ A simple linear profile function was used, given by
520: $f(r)=\pi(L-r)/L,$ for $r\le L$ and zero otherwise. For a given
521: rational map one could aim to minimize the energy of the ansatz over
522: all profile functions, but this has not been attempted for reasons
523: discussed shortly.
524:
525: As an illustration of the use of the ansatz, consider an
526: approximation to the minimal energy $Q=7$ trefoil knot. To construct
527: a field ${\cal K}_{3,2}$ with $Q=2\alpha+3\beta=7,$ requires
528: $(\alpha,\beta)=(2,1),$ resulting in the map
529: $W=Z_1^2Z_0/(Z_1^3+Z_0^2).$ The field generated from this rational
530: map is displayed in Figure \ref{fig-approx}a,
531: where both the position and linking
532: curves are shown. It is clear from this figure that the field has
533: similar qualitative features to the minimal energy soliton, in that
534: it is a trefoil knot with four twists. The main qualitative
535: difference is that the ansatz produces a more symmetric field, having a
536: cyclic $C_3$ symmetry. This symmetry is obvious from the form of the
537: rational map, as a rotation by $120^\circ$ around the $x_3$-axis
538: corresponds to the transformation $Z_1\mapsto e^{2\pi i/3}Z_1,$ upon
539: which the rational map changes by only a phase; which is simply an
540: action of the global $O(2)$ symmetry on target space.
541:
542: \begin{figure}
543: \begin{center}
544: %\leavevmode
545: %\epsfxsize=10cm\epsffile{approx76.ps}
546: \includegraphics[width=10cm]{approx76.ps}
547: \caption{The position (light tube) and linking (dark tube)
548: curves for initial conditions created using the rational map
549: ansatz: (a) $Q=7$ trefoil knot; (b) $Q=6$ link
550: ${\cal L}_{2,2}^{1,1}.$ }
551: \label{fig-approx}
552: \end{center}
553: \end{figure}
554:
555: Using the above field as the initial condition for an energy
556: relaxation simulation yields the minimal energy soliton very
557: quickly, and certainly requires far less computational resources
558: than computing the solution from a perturbed axial configuration.
559: Note that, in practice, the cyclic $C_3$ symmetry is slightly broken
560: by the cubic numerical grid, and in particular its boundary, so an
561: explicit symmetry breaking perturbation is not required. In examples
562: discussed later, where initial conditions have a cyclic $C_4$
563: symmetry, then the configuration is created with a slight
564: displacement from the centre of the grid to again slightly break any
565: exact symmetry.
566:
567: As mentioned above, the profile function could be optimized to find
568: the minimal energy configuration, for a given rational map, but this
569: has not been attempted. As with the above example, generally the
570: ansatz is more symmetric than the relaxed solution, so it is not
571: expected that a good estimate of the energy can be obtained in this
572: way. It might be possible to improve the approximation to a
573: quantitative level by minimizing over families of rational maps,
574: which include symmetry breaking terms, rather than using the
575: rational maps in this paper; chosen to be the simplest with the
576: correct qualitative features. In this case it would be worthwhile to
577: also minimize the profile function, but as the minimization problem
578: couples the rational map and the profile function together then this
579: is not a simple numerical task. In fact, this numerical problem
580: appears to have about the same level of difficulty as performing
581: energy relaxation via full field simulations. As the energy
582: relaxation code produces a solution very quickly and efficiently
583: from the simple ansatz, then there appears little need to try and
584: improve upon it. The profile function could be adjusted to a minimal
585: extent, for example, by including parameters to set the scale and
586: thickness of the knot, but again this has not been necessary.
587:
588: So far only knot initial conditions have been considered, but the
589: ansatz can also be used to construct a variety of unknots and links
590: by choosing suitable rational maps. For example, axial fields of the
591: form ${\cal A}_{n,m}$ are generated by a rational map
592: $W=Z_1^n/Z_0^m.$
593:
594: Links correspond to rational maps in which the curve determined by
595: the denominator is reducible. The torus knot maps with $(a,b)$
596: coprime degenerate to links if $(a,b)$ are not coprime. As an
597: example, links of the type ${\cal L}_{n,n}^{1,1}$ (which have Hopf
598: charge $Q=2n+2$) are derived from the rational map \be
599: W=\frac{Z_1^{n+1}}{Z_1^2-Z_0^2} =
600: \frac{Z_1^n}{2(Z_1-Z_0)}+\frac{Z_1^n}{2(Z_1+Z_0)},\label{links}\ee
601: where the partial fraction decomposition reveals the charges of the
602: constituent links by comparison with the axial maps. The $Q=6$
603: field constructed using the rational map $(\ref{links})$ with $n=2$
604: is displayed in Figure \ref{fig-approx}b.
605: As can be seen from this figure, the
606: field is qualitatively similar to the minimal energy $Q=6$ soliton,
607: which is of the type ${\cal L}_{2,2}^{1,1}.$ More examples of
608: rational maps associated to linked configurations will appear in the
609: following section.
610:
611:
612:
613:
614: \begin{table}
615: %\centering
616: \vskip -2cm
617: \begin{tabular}{|c|c|c|c|c|c|c|}
618: \hline
619: & \multicolumn{6}{|c|}{initial} \\
620: Q & \multicolumn{6}{|c|}{$\downarrow$} \\
621: & \multicolumn{6}{|c|}{final} \\
622: \hline
623: %
624: & ${\cal K}_{3,2}$& & & & &\\
625: 5 & $\downarrow$& & & & &\\
626: & ${\cal L}_{1,2}^{1,1}$& & & & &\\ \hline
627: %
628: & ${\cal K}_{3,2}$& & & & &\\
629: 6 & $\downarrow$& & & & &\\
630: & ${\cal L}_{2,2}^{1,1}$& & & & &\\ \hline
631: %
632: & ${\cal K}_{3,2}$&${\cal L}_{2,3}^{1,1}$ & & & &\\
633: 7 & $\downarrow$&$\downarrow$ & & & &\\
634: & ${\cal K}_{3,2}$&${\cal K}_{3,2}$ & & & &\\ \hline
635: %
636: &${\cal L}_{3,3}^{1,1}$ &
637: ${\cal L}_{2,2}^{2,2}$ &${\cal L}_{4,2}^{1,1}$ & ${\cal K}_{5,2}$ &
638: ${\cal K}_{3,2}$
639: & \\
640: 8 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$
641: &$\downarrow$ &\\
642: &${\cal L}_{3,3}^{1,1}$
643: &${\cal L}_{3,3}^{1,1}$ &${\cal L}_{3,3}^{1,1}$
644: &${\cal L}_{3,3}^{1,1}$ &${\cal K}_{3,2}$ & \\ \hline
645: %
646: & ${\cal L}_{1,1,1}^{2,2,2}$& ${\cal K}_{4,3}$ & ${\cal
647: K}_{3,2}$&${\cal K}_{5,2}$
648: &${\cal L}_{2,3}^{2,2}$ &${\cal L}_{3,4}^{1,1}$\\
649: 9 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$
650: &$\downarrow$
651: &$\downarrow$\\
652: &${\cal L}_{1,1,1}^{2,2,2}$ &${\cal L}_{1,1,1}^{2,2,2}$
653: & ${\cal K}_{3,2}$&${\cal K}_{3,2}$ &${\cal K}_{3,2}$&${\cal K}_{3,2}$
654: \\ \hline
655: %
656: & ${\cal L}_{1,1,2}^{2,2,2}$& ${\cal K}_{4,3}$ & ${\cal
657: L}_{3,3}^{2,2}$ &${\cal L}_{5,3}^{1,1}$
658: &${\cal K}_{5,2}$ &${\cal K}_{3,2}$\\
659: 10 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$
660: &$\downarrow$
661: &$\downarrow$\\
662: &${\cal L}_{1,1,2}^{2,2,2}$ &${\cal L}_{1,1,2}^{2,2,2}$
663: & ${\cal L}_{3,3}^{2,2}$& ${\cal L}_{3,3}^{2,2}$ & ${\cal L}_{3,3}^{2,2}$
664: &${\cal K}_{3,2}$
665: \\ \hline
666: %
667: & ${\cal K}_{4,3}$ & ${\cal K}_{5,2}$ & ${\cal K}_{7,2}$ &${\cal
668: K}_{3,2}$
669: & &\\
670: 11 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$ &
671: &\\
672: &${\cal L}_{1,2,2}^{2,2,2}$ & ${\cal K}_{5,2}$
673: & ${\cal L}_{3,4}^{2,2}$& ${\cal K}_{3,2}$& &
674: \\ \hline
675: %
676: %
677: & ${\cal K}_{5,3}$ & ${\cal K}_{3,2}$ & ${\cal K}_{5,2}$ &${\cal
678: K}_{7,2}$
679: & &\\
680: 12 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$ &
681: &\\
682: &${\cal L}_{2,2,2}^{2,2,2}$ & ${\cal K}_{4,3}$
683: & ${\cal K}_{5,2}$& ${\cal L}_{4,4}^{2,2}$& &
684: \\ \hline
685: %
686: & ${\cal K}_{3,2}$ & ${\cal K}_{5,3}$ & ${\cal K}_{5,2}$ &${\cal
687: K}_{7,2}$
688: & &\\
689: 13 & $\downarrow$&$\downarrow$ &$\downarrow$ &$\downarrow$ &
690: &\\
691: &${\cal K}_{4,3}$ & ${\cal X}_{13}$
692: & ${\cal K}_{5,2}$& ${\cal L}_{3,4}^{3,3}$& &
693: \\ \hline
694: %
695: & ${\cal K}_{4,3}$ & ${\cal K}_{5,3}$ &${\cal K}_{3,2}$ & & &\\
696: 14 & $\downarrow$&$\downarrow$ &$\downarrow$ & &
697: &\\
698: &${\cal K}_{4,3}$ & ${\cal K}_{5,3}$ & ${\cal K}_{5,2}$ & & &
699: \\ \hline
700: %
701: & ${\cal K}_{5,3}$ & ${\cal K}_{4,3}$ & ${\cal K}_{3,2}$ & & &\\
702: 15 & $\downarrow$&$\downarrow$ &$\downarrow$ & &
703: &\\
704: &${\cal X}_{15}$ & ${\cal L}_{1,1,1}^{4,4,4}$ & ${\cal K}_{5,3}$ & & &
705: \\ \hline
706: %
707: & ${\cal K}_{4,3}$ & ${\cal L}_{1,1,1,1}^{3,3,3,3}$ & ${\cal K}_{3,2}$ & & &\\
708: 16 & $\downarrow$&$\downarrow$ &$\downarrow$ & &
709: &\\
710: &${\cal X}_{16}$ & ${\cal X}_{16}$ &${\cal X}_{16}$ & & &
711: \\ \hline
712: \end{tabular}
713: %\begin{flushleft}\caption{Caption}\end{flushleft}
714: \vskip 1cm \ \hskip -0cm Table 1: Initial conditions and final
715: solutions.
716: \label{tab-relax}
717: \end{table}
718: \begin{table}
719: \vskip -26.7cm \ \hskip 10cm
720: %\centering
721: \begin{tabular}{|c|c|c|c|}
722: \hline $Q$ & type & $E$ & $E/Q^{3/4}$\\ \hline 1 & ${\cal
723: A}_{1,1}$ & 1.204 & 1.204\\ \hline 2 & ${\cal A}_{2,1}$ & 1.967 &
724: 1.170\\ \hline 3 & $\widetilde{\cal A}_{3,1}$ & 2.754 & 1.208\\
725: \hline 4 & ${\cal A}_{2,2}$ & 3.445 & 1.218 \\ \hline 5 & ${\cal
726: L}_{1,2}^{1,1}$ & 4.095 & 1.225 \\ \hline 6 & ${\cal
727: L}_{2,2}^{1,1}$ & 4.650 & 1.213 \\ \hline 7 & ${\cal K}_{3,2}$ &
728: 5.242 & 1.218 \\ \hline
729: 8 & ${\cal L}_{3,3}^{1,1}$ & 5.821 & 1.224 \\
730: 8 & ${\cal K}_{3,2}$ &5.824 & 1.224 \\ \hline
731: 9 & ${\cal L}_{1,1,1}^{2,2,2}$ & 6.360 & 1.224 \\
732: 9 & ${\cal K}_{3,2}$ & 6.385 & 1.229 \\ \hline
733: 10 & ${\cal L}_{1,1,2}^{2,2,2}$ & 6.905 & 1.228 \\
734: 10 & ${\cal L}_{3,3}^{2,2}$ & 6.923 & 1.231 \\
735: 10 & ${\cal K}_{3,2}$ & 6.973 & 1.240 \\ \hline
736: 11 & ${\cal L}_{1,2,2}^{2,2,2}$ & 7.391 & 1.224 \\
737: 11 & ${\cal K}_{5,2}$ & 7.502 & 1.242 \\
738: 11 & ${\cal L}_{3,4}^{2,2}$ & 7.533 & 1.247 \\
739: 11 & ${\cal K}_{3,2}$ & 7.614 & 1.261 \\ \hline
740: 12 & ${\cal L}_{2,2,2}^{2,2,2}$ &7.833 & 1.215 \\
741: 12 & ${\cal K}_{4,3}$ & 7.857 & 1.219 \\
742: 12 & ${\cal K}_{5,2}$ & 8.070 & 1.252 \\
743: 12 & ${\cal L}_{4,4}^{2,2}$ & 8.093 & 1.255 \\ \hline
744: 13 & ${\cal K}_{4,3}$ & 8.272 &1.208 \\
745: 13 & ${\cal X}_{13}$ & 8.462 &1.236 \\
746: 13 & ${\cal K}_{5,2}$ &8.574 & 1.252 \\
747: 13 & ${\cal L}_{3,4}^{3,3}$ & 8.633 & 1.261 \\ \hline
748: 14 & ${\cal K}_{4,3}$ & 8.761 & 1.210 \\
749: 14 & ${\cal K}_{5,3}$ & 8.807 & 1.217 \\
750: 14 & ${\cal K}_{5,2}$ & 9.124 & 1.261 \\ \hline
751: 15 & ${\cal X}_{15}$ & 9.290 & 1.219 \\
752: 15 & ${\cal L}_{1,1,1}^{4,4,4}$ & 9.404 & 1.234 \\
753: 15 & ${\cal K}_{5,3}$ & 9.408 & 1.234 \\ \hline 16 & ${\cal
754: X}_{16}$ & 9.769 & 1.221 \\ \hline
755: \end{tabular}
756: \vskip 1cm \hskip 9.5cm Table 2: Solution types and
757: energies.\label{tab-energies}
758: \end{table}
759:
760: \section{Higher charge solitons}\news
761: \ \quad In this section the results of a large number of energy
762: minimization simulations are presented. Hopf charges up to sixteen
763: are studied, with initial conditions consisting of a variety of
764: links and knots, created using the rational map ansatz discussed
765: above.
766:
767: Table 1 summarizes the simulations performed, by listing the type of
768: each initial condition together with the type of the resulting final
769: solution, obtained from the energy minimization algorithm. In Table
770: 2 the types and energies of the known lowest energy solutions, plus
771: some local energy minima, are presented for $1\le Q\le 16,$ ordered
772: by increasing energies. Also given in Table 2 are the energies
773: divided by Ward's conjectured bound, $E/Q^{3/4},$ from which it can
774: be seen that the minimal energy solution is consistently around
775: $20\%$ above the conjectured bound. The values of $E/Q^{3/4}$ are
776: plotted in Figure \ref{fig-energy} using a notation that
777: identifies the different
778: torus knots as follows. White circles denote configurations which do not
779: contain knots, that is, either unknots or links. The symbols
780: denoting knots are triangles for ${\cal K}_{3,2}$, diamonds for
781: ${\cal K}_{5,2},$ squares for ${\cal K}_{4,3}$ and stars for ${\cal
782: K}_{5,3}.$ Configurations denoted by black circles will be discussed shortly.
783: \begin{figure}
784: \begin{center}
785: \leavevmode
786: %\epsfxsize=12cm\epsffile{energy.ps}
787: \includegraphics[width=12cm]{energy.ps}
788: \caption{The ratio of the energy to the conjectured bound, that is,
789: $E/Q^{3/4},$ as a function of Hopf charge $Q$ for a variety of solutions:
790: unknots and links (white circles); ${\cal K}_{3,2}$ (triangles);
791: ${\cal K}_{5,2}$ (diamonds); ${\cal K}_{4,3}$ (squares);
792: ${\cal K}_{5,3}$ (stars); links which are not resolved (black circles).}
793: \label{fig-energy}
794: \end{center}
795: \end{figure}
796:
797:
798: The present study has produced 26 new solutions with $8\le Q\le 16.$
799: The position curves for each of these is displayed in
800: Figure \ref{fig-high}, but
801: for clarity the linking curves are not shown,
802: although they have been examined to confirm the correct linking
803: number identifications. Each plot is labeled by its charge and type,
804: with energies increasing first from left to right and then top to
805: bottom.
806:
807: \begin{figure}
808: \begin{center}
809: \leavevmode
810: %\epsfxsize=14cm\epsffile{8to16_low.ps}
811: \includegraphics[width=14cm]{8to16_low.ps}
812: \caption{The position curves for a variety of solutions with Hopf charge
813: $8\le Q\le 16.$ Each plot is labeled by its charge and type.}
814: \label{fig-high}
815: \end{center}
816: \end{figure}
817:
818: First of all, consider trefoil knots, that is, solutions which have
819: the form ${\cal K}_{3,2}.$ As mentioned earlier, the $Q=7$ minimal
820: energy soliton has this form and is obtained from the related
821: rational map initial condition. This is encoded in Table 1 as the
822: process ${\cal K}_{3,2}\rightarrow{\cal K}_{3,2}.$ The entry to the
823: right of this one in Table 1 reveals that the same trefoil knot
824: solution is also obtained from the linked initial condition ${\cal
825: L}_{2,3}^{1,1}.$
826:
827: There are no trefoil knots with $Q<7.$ Evidence supporting this is
828: presented in Table 1 for $Q=5$ and $Q=6,$ where it is seen that in
829: both cases initial conditions of the form ${\cal K}_{3,2}$ result in
830: the linked minimal energy solutions, which are ${\cal
831: L}_{1,2}^{1,1}$ and ${\cal L}_{2,2}^{1,1}$ respectively. A
832: reasonable interpretation of these results is that for $Q<7$ the
833: number of twists $T=Q-3$ is too low, given the preferred length of
834: the soliton in a trefoil knot arrangement. In other words, the twist
835: per unit length is too low to be an energetically efficient
836: distribution of the Hopf charge between crossing and twisting. A
837: more detailed discussion of this aspect will be given later, when
838: general torus knots will be considered. Note that, in particular,
839: there is no trefoil knot solution in which there is no twist, which
840: corresponds to $Q=3,$ and was the original suggestion for a knot
841: soliton \cite{FN}.
842:
843: Table 1 shows that there are also trefoil knot solutions for
844: $Q=8,9,10,11$ and in each case they are obtained from rational map
845: initial conditions of the type ${\cal K}_{3,2}.$ For $Q=9$ the
846: trefoil solution is also obtained from a variety of other initial
847: conditions, including links and the torus knot ${\cal K}_{5,2};$
848: this last process ${\cal K}_{5,2}\rightarrow{\cal K}_{3,2}$ provides
849: an example of knot transmutation, where Solomon's seal knot deforms
850: into a trefoil knot. In Figure \ref{fig-high} a comparison of the
851: plots $8{\cal K}_{3,2},9{\cal K}_{3,2},10{\cal K}_{3,2},11{\cal K}_{3,2}$,
852: emphasizes that the conformation of the knot is important, with each
853: of these trefoils having a very different structure to the others.
854: In particular, as the number of twists increases the knot
855: increasingly contorts and bends, in a manner similar to that seen in
856: the axial solutions ${\cal A}_{n,1},$ where for $n=3$ the response
857: to the twist is to lower the energy by bending to break the axial
858: symmetry. As mentioned earlier, for $n=4,5$ there are similar bent
859: solutions, which are local minima, where the deformation increases
860: with $n.$ Trefoil knots therefore appear to follow a similar
861: pattern, with a minimal number of twists, that is 4, required for
862: existence and increasingly deformed solutions existing with slightly
863: larger twists than the minimal value.
864:
865: The triangles in Figure \ref{fig-energy}, and the associated energy values in
866: Table 2, show that the ratio of the energy to the bound steadily increases for
867: the trefoil solutions as the charge increases. It is therefore not
868: surprising that for a large enough charge, which happens to be
869: $Q=12,$ a trefoil solution fails to exist. Table 1 confirms that a
870: $Q=12$ initial condition of the type ${\cal K}_{3,2}$ produces a
871: solution which is not a trefoil knot, and will be discussed later.
872: The triangles in Figure \ref{fig-energy} also confirm that only for the minimal
873: value $Q=7$ is the trefoil knot the global minimum energy solution,
874: and in all other cases a trefoil is only a local energy minimum.
875: However, for $Q=8$ the trefoil knot energy is extremely close to the
876: lowest energy found, which is a linked solution ${\cal
877: L}_{3,3}^{1,1}.$ In fact the energies differ by less than $0.1\%,$
878: which is probably smaller than the accuracy of the computations, so
879: in this case it is difficult to make a definitive statement
880: regarding which (if any) is the global minimum.
881:
882: In Ref.\cite{BS5} a linked solution of the type ${\cal
883: L}_{2,2}^{2,2}$ was reported, but this appears to be an artefact of
884: that study being at the limit of computational feasibility for
885: computing resources available at that time. A configuration of this
886: type does appear during a relaxation procedure, but after further
887: relaxation it changes its type. Moreover, an initial condition
888: of the type ${\cal L}_{2,2}^{2,2}$ can be constructed using the
889: rational map \be
890: W=\frac{Z_1^4}{Z_1^4-Z_0^2}=\frac{Z_1^2}{2(Z_1^2-Z_0)}
891: +\frac{Z_1^2}{2(Z_1^2+Z_0)},\label{map8}\ee and, as seen from Table
892: 1, the relaxation yields a solution of a different type, namely the
893: linked solution ${\cal L}_{3,3}^{1,1}.$ It therefore now seems
894: unlikely that a charge eight stable solution exists of the type
895: ${\cal L}_{2,2}^{2,2}.$ Note that the rational map $(\ref{map8})$
896: describes a link in which both components are linked {\em twice}
897: with each other, and this is because the irreducible factors of the
898: denominator are terms like $(Z_1^2-Z_0),$ rather than terms like
899: $(Z_1-Z_0),$ where the latter corresponds to only single links
900: between any two components. The partial fraction decomposition in
901: $(\ref{map8})$ reveals that each component of the link has charge
902: two, by once again reading off the power of the numerator.
903:
904: For a given fixed charge there is nothing to prevent multiple
905: trefoil knot solutions which differ in their conformations and
906: energies, but no evidence for this phenomenon has been found. For
907: example, Table 1 shows that the $Q=9$ trefoil knot is produced from
908: (at least) four different types of initial condition, but in each
909: case the resulting trefoil solution is the same one with an
910: identical conformation. This is the case for all the solutions
911: described in this paper, that is, when a particular solution is
912: obtained from various initial conditions it is always in the same
913: conformation, so its charge and type are sufficient to distinguish
914: it from any other solution.
915:
916: As the trefoil knot solutions for $Q=9,10,11$ are clearly only local
917: energy minima then it remains to describe the candidates found for
918: the global minima for these charges. In fact for all three of these
919: charges, and also for $Q=12,$ the lowest energy solutions found have
920: a similar form, which is a link with three components. The simplest
921: case, $Q=9,$ is displayed in Figure \ref{fig-high} as plot
922: $9{\cal L}_{1,1,1}^{2,2,2},$ and as this label suggests there are three
923: components to the link, each of which has charge one and links once
924: with each of the other two components. An initial condition
925: of the type ${\cal L}_{1,1,1}^{2,2,2}$ with $Q=9$
926: is obtained from the rational map
927: \be W=\frac{Z_1^3}{Z_1^3-Z_0^3},\label{map9}\ee and quickly relaxes
928: to the minimal energy solution with the same type. The initial
929: condition has a more planar arrangement of the three components than
930: the final solution, and the initial cyclic $C_3$ symmetry is also
931: broken, as can be seen from plot $9{\cal L}_{1,1,1}^{2,2,2},$
932: in Figure \ref{fig-high}. Note from Table 1
933: that this 3-component link is also obtained from a knotted initial
934: condition of the type ${\cal K}_{4,3},$ but that out of the six
935: $Q=9$ initial conditions used, four lead to the higher energy trefoil
936: knot solution. This is evidence that supports the fact that minimal
937: energy solutions may not be the easiest to find; hence the need to
938: employ the variety of starting configurations used in this study.
939:
940: The minimal energy solutions with $Q=10,11,12$ are similar to the
941: minimal $Q=9$ solution and correspond to increasing one, two and
942: finally all three, of the charge one components to charge two. In
943: other words they have the forms ${\cal L}_{1,1,2}^{2,2,2},{\cal
944: L}_{1,2,2}^{2,2,2}$ and ${\cal L}_{2,2,2}^{2,2,2}$ respectively.
945: These solutions are displayed in Figure \ref{fig-high}
946: and in each case there is
947: an associated rational map of the same type. Table 1 shows that they
948: can all be obtained from the relaxation of certain torus knot
949: initial conditions.
950:
951: So far the only knot solutions discussed have been trefoil knots. As
952: the torus knot with the lowest crossing number after the trefoil is
953: the $(5,2)$-torus knot, then this is the most likely candidate to
954: appear next as a soliton solution. Table 1 reveals that an initial
955: condition of the type ${\cal K}_{5,2}$ does not yield a solution of
956: this type for $Q=8,9,10,$ but it does for $Q=11.$ The $Q=11$
957: solution is displayed in Figure \ref{fig-high}
958: as plot $11{\cal K}_{5,2}.$ It is
959: only a local energy minimum, despite the fact that it is lower in
960: energy than the $Q=11$ trefoil knot. There are also solutions of the
961: type ${\cal K}_{5,2}$ for $Q=12,13,14$ and the pattern mirrors that
962: of the trefoil knots, in that the ratio of the energy to the
963: bound steadily increases with the charge; see the diamonds in
964: Figure \ref{fig-energy}.
965:
966: At $Q=12$ a solution of the type ${\cal K}_{4,3}$ appears (see
967: Figure \ref{fig-high}
968: plot $12{\cal K}_{4,3}$) and its energy is just above that
969: of the minimal energy ${\cal L}_{2,2,2}^{2,2,2}$ link, though
970: substantially lower than that of the ${\cal K}_{5,2}$ solution.
971: Solutions of the type ${\cal K}_{4,3}$ also exist for $Q=13$ and
972: $Q=14,$ and for these two charges they are the lowest energy
973: solutions found. The excess energy above the bound is also
974: reasonably low, so these two solutions are good candidates for the
975: global minima at $Q=13$ and $Q=14.$ The squares in
976: Figure \ref{fig-energy} denote
977: the ratio of the energy to the bound for ${\cal K}_{4,3}$ solutions, from
978: which it can be seen that (unlike the other torus knots)
979: the lowest charge at which this knot
980: appears is not the one that is closest to the conjectured bound.
981:
982: At $Q=13$ a solution exists (see Figure \ref{fig-high} plot $13{\cal X}_{13}$)
983: which is denoted by ${\cal X}_{13}$ because there is no unambiguous
984: interpretation as a particular link or knot. This is due to the fact
985: that a definition of a link or knot requires the position curve not
986: to self-intersect, but there is nothing to prevent this in the field
987: theory. For the ${\cal X}_{13}$ solution then either the position
988: curve self-intersects exactly, or there are parts of the curve
989: that are so close together that they can not be resolved with the
990: numerical accuracy currently employed. Some of the lower charge
991: solutions presented earlier may also appear to self-intersect, for
992: example the $Q=5$ solution displayed in Figure \ref{fig-low}, but in all these
993: lower charge solutions the apparent self-intersection can be
994: resolved by reducing the thickness of the tube plotted around the
995: position curve, together with a careful consideration of continuity.
996:
997: If there are no self-intersection points in the solution ${\cal
998: X}_{13}$ then there are two possibilities for the configuration
999: type, depending on how this self-intersection point is resolved. The
1000: first possibility is the link ${\cal L}_{1,2,2}^{2,3,3},$ and the
1001: second is a link with two components, where one of the components is
1002: a trefoil knot with five twists and the second component is approximately
1003: axial with a single twist. The latter resolution is essentially
1004: an intertwining of the charge eight trefoil knot solution with the charge
1005: one solution.
1006:
1007: The lowest energy solitons found for $Q=15$
1008: and $Q=16$ are denoted by ${\cal X}_{15}$ and ${\cal X}_{16}$
1009: respectively, and also have points which can not be distinguished from
1010: self-intersection points, as seen from plots $15{\cal X}_{15}$
1011: and $16{\cal X}_{16}$ in Figure \ref{fig-high}.
1012: In each case there is a
1013: resolution into a link with two components where one of the
1014: components is a trefoil knot. The conformation of both the knot
1015: and unknot components in these cases strongly suggests that this
1016: is the correct resolution, if indeed one is required.
1017: The $Q=16$ solution is the easiest to identify and consists of the
1018: charge eight trefoil knot intertwined with the minimal energy
1019: charge two solution, where the conformation of both components is very
1020: similar to that of the components in isolation.
1021: A field with the correct qualitative behaviour to describe this resolution
1022: is given by the rational map
1023: \be
1024: W=\frac{Z_1^5+Z_1^2Z_0^2+Z_1Z_0^3}{Z_0Z_1^3+Z_0^3}
1025: =\frac{Z_1Z_0^2}{Z_1^3+Z_0^2}+\frac{Z_1^2}{Z_0}.\label{map16}
1026: \ee
1027: Initial conditions generated using this rational map quickly
1028: leads to the solution ${\cal X}_{16}$ under energy relaxation.
1029:
1030: The solution ${\cal X}_{15}$ is similar to the solution ${\cal X}_{16},$
1031: except that the resolution involves the charge seven trefoil knot, rather
1032: than the charge eight trefoil knot.
1033: The energies of the solutions of type
1034: ${\cal X}_n$ are represented by black circles in
1035: Figure \ref{fig-energy}.
1036:
1037: Note
1038: that $Q=16$ is the lowest charge in which a link with four
1039: components each linking all the others might exist. However, as seen
1040: in Table 1, an initial condition of the type ${\cal
1041: L}_{1,1,1,1}^{3,3,3,3}$ relaxes to the ${\cal X}_{16}$ solution.
1042:
1043: The final torus knot solitons presented in this paper are of the
1044: type ${\cal K}_{5,3}$ and exist for $Q=14$ and $Q=15,$ but in both
1045: cases these solutions are only local energy minima. The energies of
1046: these solutions are represented by stars in Figure \ref{fig-energy}.
1047:
1048: In summary, it has been shown that there are a variety of torus knot
1049: solitons at various Hopf charges. For example, at $Q=14$ three
1050: different torus knot solutions have been obtained. Most of the knot solitons
1051: found are only local energy minima, but in some cases they are good
1052: candidates for the global minimum.
1053:
1054: \begin{figure}
1055: \begin{center}
1056: \leavevmode
1057: %\epsfxsize=12cm\epsffile{length.ps}
1058: \includegraphics[width=12cm]{length.ps}
1059: \caption{String length (circles) as a function of charge $Q$ for several
1060: solitons, together with the expected growth
1061: behaviour $\gamma Q^{3/4}$ (curve).}
1062: \label{fig-length}
1063: \end{center}
1064: \end{figure}
1065:
1066: For a string-like soliton it is expected that one of the
1067: contributions to the energy should have an interpretation in terms
1068: of the string length. As the soliton energy grows like $Q^{3/4}$
1069: then this suggests that the string length should have a similar growth.
1070: In Figure \ref{fig-length} the string length is plotted (circles)
1071: for several solitons, including examples of both knots and links.
1072: Also shown is a curve with the
1073: expected power growth, $\gamma Q^{3/4},$ where $\gamma$ has been
1074: computed by a least squares fit to be $\gamma=7.86.$
1075: This plot demonstrates a reasonable agreement with the expected
1076: $Q^{3/4}$ dependence of the string length. The greatest discrepancy
1077: from the expected string length occurs for the $Q=15$ link
1078: ${\cal L}_{1,1,1}^{4,4,4},$ which is much longer than expected.
1079: However, note that this solution has an abnormally large ratio of linking
1080: to twist number, with the contribution to the Hopf charge from linking being
1081: four times that from twisting. The solution with the next largest ratio of
1082: linking to twist is the $Q=9$ link ${\cal L}_{1,1,1}^{2,2,2},$ where linking
1083: contributes twice as much as twisting, and the length of this solution
1084: is also a bit above the expected value. It therefore appears that
1085: solutions which link much more than they twist require an increased string
1086: length to minimize energy.
1087:
1088: Recall from the earlier discussion of knot solitons,
1089: that there appears to be a critical twist,
1090: below which knot solitons of a particular type do not exist.
1091: A twist below the critical value seems to be an energetically
1092: inefficient distribution of the Hopf charge between twisting and crossing.
1093: For twists just above the critical value a knot solution continues
1094: to exist but as the twist increases further then again the
1095: distribution between twisting and crossing becomes inefficient, but this
1096: time due to too much twisting.
1097:
1098: A naive approximation is to assume that for all solutions there is a
1099: universal optimum twist per unit length. Taking this optimum value from
1100: that of the $Q=1$ soliton, and using the fact that the string length
1101: grows like $Q^{3/4},$ this suggests that an optimum value for the number
1102: of twists is approximately $T\approx Q^{3/4}.$ For a knot with crossing number
1103: $C$ this gives $C=Q-T\approx Q-Q^{3/4}.$ Using this assumption, a knot
1104: with crossing number $C$ is predicted to be energetically efficient
1105: at an integer charge near to the real value $Q_*,$ which is defined
1106: as the solution of the equation $Q_*-Q_*^{3/4}=C.$
1107: For example, for the trefoil knot
1108: $C=3$ and therefore $Q_*=7.56,$ which roughly agrees with the $Q=7$ and
1109: $Q=8$ trefoils being closest to the energy bound. For $C=5$ then $Q_*=11.07$
1110: and this explains why the $Q=11$ knot ${\cal K}_{5,2}$ is closest to
1111: the bound. For $C=8$ then $Q_*=16,$ so the naive approximation
1112: overestimates the most efficient value, which is seen to be $Q=13$
1113: for the ${\cal K}_{4,3}$ knot.
1114: There are clearly more subtle effects at work than the simple naive
1115: constant twist rate assumed in this calculation, but it does seem to
1116: produce numbers which are in the right ballpark, suggesting that it captures
1117: some qualitative aspects of knot energetics.
1118:
1119: \section{Conclusion}\news
1120: \ \quad The results presented in this paper reveal that there are many
1121: low-energy knot solitons of various types, for a range of Hopf charges,
1122: together with increasingly complicated linked solitons. The qualitative
1123: features of these solutions can be replicated using an ansatz involving
1124: rational maps from the three-sphere to the complex projective line, and
1125: futhermore this provides a good supply of initial conditions for
1126: numerical relaxation simulations. The complicated nature of the problem
1127: means that it is difficult to be certain that the global minimal energy
1128: solitons have been found, but certainly some good candidates have been
1129: presented, whose energies agree well with the expected values based on an
1130: earlier conjectured bound.
1131:
1132: The rational map ansatz allows the construction
1133: of any torus knot initial condition, but it is not clear how to extend
1134: this to non-torus knots. All the knot solutions found so far are torus
1135: knots, and it is unknown whether the lack of non-torus knot solitons
1136: is a result of not having suitable initial conditions, or whether there
1137: is some energetic reason to favour torus knots. For example, the figure eight
1138: knot has four crossings and hence a soliton with this form might be expected
1139: with a charge around $Q=9.$ As seen from Table 1, quite a few initial
1140: conditions have been employed for $Q=9$ and $Q=10,$ and no such solution
1141: has been found. It remains an open problem to understand the absence
1142: (or otherwise) of non-torus knots.
1143:
1144: Finally, given that there are many knot and link solitons then it would be
1145: useful if there was an approximate string model that could, even
1146: qualitatively, reproduce the field theory results. It has been shown that
1147: the string length has the expected behaviour with Hopf charge, so it is
1148: plausible that there may be an approximate description based on a
1149: string energy that includes contributions from properties of the string such
1150: as its length, twist and writhe, together with relevant interaction terms.
1151:
1152:
1153:
1154:
1155:
1156:
1157:
1158:
1159:
1160: \section*{Acknowledgements}
1161: Many thanks to Michael Farber for helpful discussions.
1162: This work was supported by the PPARC special programme
1163: grant ``Classical Lattice Field Theory''.
1164:
1165:
1166: \begin{thebibliography}{100}
1167:
1168: \bibitem{BS5} R.~A. Battye and P.~M. Sutcliffe,
1169: Knots as stable soliton solutions in a
1170: three-dimensional classical field theory,
1171: \textit{Phys. Rev. Lett.} \textbf{81}, 4798 (1998);
1172: Solitons, links and knots, \textit{Proc. R. Soc. Lond.} \textbf{A455},
1173: 4305 (1999).
1174:
1175: \bibitem{BS3} R.~A. Battye and P.~M. Sutcliffe,
1176: Skyrmions, fullerenes and rational maps,
1177: \textit{Rev. Math. Phys.} \textbf{14}, 29 (2002).
1178:
1179: \bibitem{BK} E. Brieskorn and H. Kn\"orrer,
1180: {\em Plane algebraic curves}, Birkh\"auser-Verlag (1986).
1181:
1182: \bibitem{Fa2} L.~D. Faddeev,
1183: Quantization of solitons,
1184: Princeton preprint IAS-75-QS70 (1975).
1185:
1186: \bibitem{FN} L. Faddeev and A.~J. Niemi,
1187: Stable knot-like structures in classical field theory,
1188: \textit{Nature} \textbf{387}, 58 (1997).
1189:
1190: \bibitem{GH} J. Gladikowski and M. Hellmund,
1191: Static solitons with nonzero Hopf number,
1192: \textit{Phys. Rev.} \textbf{D56}, 5194 (1997).
1193:
1194: \bibitem{HS} J. Hietarinta and P. Salo,
1195: Faddeev-Hopf knots: dynamics of linked un-knots,
1196: \textit{Phys. Lett.} \textbf{B451}, 60 (1999).
1197:
1198: \bibitem{HS2} J. Hietarinta and P. Salo,
1199: Ground state in the Faddeev-Skyrme model,
1200: \textit{Phys. Rev.} \textbf{D62}, 081701(R) (2000).
1201:
1202: \bibitem{KR} A. Kundu and Yu.~P. Rybakov,
1203: Closed-vortex-type solitons with Hopf index,
1204: \textit{J. Phys.} \textbf{A15}, 269 (1982).
1205:
1206: \bibitem{LY} F. Lin and Y. Yang,
1207: Existence of energy minimizers as stable knotted solitons in the
1208: Faddeev model,
1209: \textit{Commun. Math. Phys.} \textbf{249}, 273 (2004).
1210:
1211: \bibitem{PSZ}
1212: B.~M.~A.~G. Piette, B.~J. Schroers and W.~J. Zakrzewski,
1213: Multisolitons in a two-dimensional Skyrme model,
1214: \textit{Z. Phys.} \textbf{C65}, 165 (1995).
1215:
1216: \bibitem{VK}
1217: A.~F. Vakulenko and L.~V. Kapitanski,
1218: Stability of solitons in $S^2$ in the nonlinear $\sigma$-model,
1219: \textit{Dokl. Akad. Nauk USSR} \textbf{246}, 840 (1979).
1220:
1221: \bibitem{Wa4} R.~S. Ward,
1222: Hopf solitons on $S^3$ and $\bb{R}^3$,
1223: \textit{Nonlinearity} \textbf{12}, 241 (1999).
1224:
1225: \bibitem{Wa5} R.~S. Ward,
1226: The interaction of two Hopf solitons,
1227: \textit{Phys. Lett.} \textbf{B473}, 291 (2000).
1228:
1229:
1230:
1231:
1232: \end{thebibliography}
1233:
1234:
1235: \end{document}
1236: