Lissajous-toric knot

Written by

in

Lissajous-toric knot
Lissajous-toric knot with parameters 5, 6 and 22 in braid form (with z-axis in horizontal direction)

In knot theory, a Lissajous-toric knot is a knot defined by parametric equations of the form:

x ( t ) = ( 2 + sin q t ) cos N t , y ( t ) = ( 2 + sin q t ) sin N t , z ( t ) = cos p ( t + ϕ ) , {\displaystyle x(t)=(2+\sin qt)\cos Nt,\qquad y(t)=(2+\sin qt)\sin Nt,\qquad z(t)=\cos p(t+\phi ),} {\displaystyle x(t)=(2+\sin qt)\cos Nt,\qquad y(t)=(2+\sin qt)\sin Nt,\qquad z(t)=\cos p(t+\phi ),}

where N {\displaystyle N} {\displaystyle N}, p {\displaystyle p} {\displaystyle p}, and q {\displaystyle q} {\displaystyle q} are integers, the phase shift ϕ {\displaystyle \phi } {\displaystyle \phi } is a real number
and the parameter t {\displaystyle t} {\displaystyle t} varies between 0 and 2 π {\displaystyle 2\pi } {\displaystyle 2\pi }.[1]

For p = q {\displaystyle p=q} {\displaystyle p=q} the knot is a torus knot.

Braid and billiard knot definitions

Lissajous-toric knot
Lissajous-toric knot T(4,7,35) as a billiard knot, showing period 7

In braid form these knots can be defined in a square solid torus (i.e. the cube [ 1 , 1 ] 3 {\displaystyle [-1,1]^{3}} {\displaystyle [-1,1]^{3}} with identified top and bottom) as

x ( t ) = sin 2 π q t , y ( t ) = cos 2 π p ( t + ϕ ) , z ( t ) = 2 ( N t N t ) 1 , t [ 0 , 1 ] {\displaystyle x(t)=\sin 2\pi qt,\qquad y(t)=\cos 2\pi p(t+\phi ),\qquad z(t)=2(Nt-\lfloor Nt\rfloor )-1,\qquad t\in [0,1]} {\displaystyle x(t)=\sin 2\pi qt,\qquad y(t)=\cos 2\pi p(t+\phi ),\qquad z(t)=2(Nt-\lfloor Nt\rfloor )-1,\qquad t\in [0,1]}.

The projection of this Lissajous-toric knot onto the x-y-plane is a Lissajous curve.

Replacing the sine and cosine functions in the parametrization by a triangle wave transforms a Lissajous-toric
knot isotopically into a billiard curve inside the solid torus. Because of this property Lissajous-toric knots are also called billiard knots in a solid torus.[2]

Lissajous-toric knots were first studied as billiard knots and they share many properties with billiard knots in a cylinder.[3]
They also occur in the analysis of singularities of minimal surfaces with branch points[4] and in the study of
the Three-body problem.[5]

The knots in the subfamily with p = q l {\displaystyle p=q\cdot l} {\displaystyle p=q\cdot l}, with an integer l 1 {\displaystyle l\geq 1} {\displaystyle l\geq 1}, are known as ′Lemniscate knots′.[6] Lemniscate knots have period q {\displaystyle q} {\displaystyle q} and are fibred. The knot shown on the right is of this type (with l = 5 {\displaystyle l=5} {\displaystyle l=5}).

Properties

Lissajous-toric knot
Symmetries of the Lissajous-toric knot T(3,8,7): symmetric union (vertical axis), rotation into mirror image and palindromic property within Q (horizontal axis)

Lissajous-toric knots are denoted by K ( N , q , p , ϕ ) {\displaystyle K(N,q,p,\phi )} {\displaystyle K(N,q,p,\phi )}. To ensure that the knot is traversed only once in the parametrization
the conditions gcd ( N , q ) = gcd ( N , p ) = 1 {\displaystyle \gcd(N,q)=\gcd(N,p)=1} {\displaystyle \gcd(N,q)=\gcd(N,p)=1} are needed. In addition, singular values for the phase, leading to self-intersections, have to be excluded.

The isotopy class of Lissajous-toric knots surprisingly does not depend on the phase ϕ {\displaystyle \phi } {\displaystyle \phi } (up to mirroring).
If the distinction between a knot and its mirror image is not important, the notation K ( N , q , p ) {\displaystyle K(N,q,p)} {\displaystyle K(N,q,p)} can be used.

The properties of Lissajous-toric knots depend on whether p {\displaystyle p} {\displaystyle p} and q {\displaystyle q} {\displaystyle q} are coprime or d = gcd ( p , q ) > 1 {\displaystyle d=\gcd(p,q)>1} {\displaystyle d=\gcd(p,q)>1}. The main properties are:

  • Interchanging p {\displaystyle p} {\displaystyle p} and q {\displaystyle q} {\displaystyle q}:
K ( N , q , p ) = K ( N , p , q ) {\displaystyle K(N,q,p)=K(N,p,q)} {\displaystyle K(N,q,p)=K(N,p,q)} (up to mirroring).
  • Ribbon property:
If p {\displaystyle p} {\displaystyle p} and q {\displaystyle q} {\displaystyle q} are coprime, K ( N , q , p ) {\displaystyle K(N,q,p)} {\displaystyle K(N,q,p)} is a symmetric union and therefore a ribbon knot.
  • Periodicity:
If d = gcd ( p , q ) > 1 {\displaystyle d=\gcd(p,q)>1} {\displaystyle d=\gcd(p,q)>1}, the Lissajous-toric knot has period d {\displaystyle d} {\displaystyle d} and the factor knot is a ribbon knot.
  • Strongly positive amphicheirality:
If p {\displaystyle p} {\displaystyle p} and q {\displaystyle q} {\displaystyle q} have different parity, then K ( N , q , p ) {\displaystyle K(N,q,p)} {\displaystyle K(N,q,p)} is strongly positive amphicheiral.
  • Period 2:
If p {\displaystyle p} {\displaystyle p} and q {\displaystyle q} {\displaystyle q} are both odd, then K ( N , q , p ) {\displaystyle K(N,q,p)} {\displaystyle K(N,q,p)} has period 2 (for even N {\displaystyle N} {\displaystyle N}) or is freely 2-periodic (for odd N {\displaystyle N} {\displaystyle N}).

Example

The knot T(3,8,7), shown in the graphics, is a symmetric union and a ribbon knot (in fact, it is the composite knot 5 1 5 1 {\displaystyle 5_{1}\sharp -5_{1}} {\displaystyle 5_{1}\sharp -5_{1}}).
It is strongly positive amphicheiral: a rotation by π {\displaystyle \pi } {\displaystyle \pi } maps the knot to its mirror image, keeping its orientation.
An additional horizontal symmetry occurs as a combination of the vertical symmetry and the rotation (′double palindromicity′ in Kin/Nakamura/Ogawa).

′Classification′ of billiard rooms

In the following table a systematic overview of the possibilities to build billiard rooms from the interval and the circle (interval with identified boundaries) is given:

Billiard room Billiard knots
[ 1 , 1 ] 3 {\displaystyle [-1,1]^{3}} {\displaystyle [-1,1]^{3}} Lissajous knots
[ 1 , 1 ] 2 × S 1 {\displaystyle [-1,1]^{2}\times \mathbb {S} ^{1}} {\displaystyle [-1,1]^{2}\times \mathbb {S} ^{1}} Lissajous-toric knots
[ 1 , 1 ] × S 1 × S 1 {\displaystyle [-1,1]\times \mathbb {S} ^{1}\times \mathbb {S} ^{1}} {\displaystyle [-1,1]\times \mathbb {S} ^{1}\times \mathbb {S} ^{1}} Torus knots
S 1 × S 1 × S 1 {\displaystyle \mathbb {S} ^{1}\times \mathbb {S} ^{1}\times \mathbb {S} ^{1}} {\displaystyle \mathbb {S} ^{1}\times \mathbb {S} ^{1}\times \mathbb {S} ^{1}} (room not embeddable into R 3 {\displaystyle \mathbb {R} ^{3}} {\displaystyle \mathbb {R} ^{3}})

In the case of Lissajous knots reflections at the boundaries occur in all of the three cube’s dimensions.
In the second case reflections occur in two dimensions and we have a uniform movement in the third dimension.
The third case is nearly equal to the usual movement on a torus, with an additional triangle wave movement in the first dimension.

References

  1. See M. Soret and M. Ville: Lissajous-toric knots,
    J. Knot Theory Ramifications 29, 2050003 (2020).
  2. See C. Lamm:
    Deformation of cylinder knots, 4th chapter of Ph.D. thesis, ‘Zylinder-Knoten und symmetrische Vereinigungen‘, Bonner Mathematische Schriften 321 (1999), available since 2012 as arXiv:1210.6639.
  3. See C. Lamm and
    D. Obermeyer: Billiard knots in a cylinder, J. Knot Theory Ramifications 8, 353–-366 (1999).
  4. See Soret/Ville.
  5. See E. Kin, H. Nakamura and H. Ogawa: Lissajous 3-braids, J. Math. Soc. Japan 75, 195–228 (2023) (or arXiv:2008.00585v4).
  6. See B. Bode, M.R. Dennis,
    D. Foster and R.P. King: Knotted fields and explicit fibrations for lemniscate knots,
    Proc. Royal Soc. A (2017).



This article is adapted from “Lissajous-toric knot” on Wikipedia, written by its contributors, and used under CC BY-SA 4.0. Text on this page is available under the same licence. Source snapshot: Wikipedia via Kiwix, 2026-07-20.