Docsity
Docsity

Prepare for your exams
Prepare for your exams

Study with the several resources on Docsity


Earn points to download
Earn points to download

Earn points by helping other students or get them with a premium plan


Guidelines and tips
Guidelines and tips

Finite Rotation Groups in SO(3) and Their Orbits, Lecture notes of Mathematics

The properties of finite rotation groups in so(3), focusing on their actions on the set of poles and the number of orbits. Proofs of lemmas 3.5.1 and 3.5.2, which establish the relationship between the group order, the number of orbits, and the orders of the stabilizer groups. The document also presents corollary 3.5.3, which outlines the possible combinations of group orders and orbit sizes for finite subgroups of so(3).

What you will learn

  • How does the stabilizer group of a pole relate to the group action on the set of poles?
  • What are the possible combinations of group orders and orbit sizes for finite subgroups of SO(3)?

Typology: Lecture notes

2020/2021

Uploaded on 06/21/2021

rakshan
rakshan 🇺🇸

4.6

(18)

239 documents

1 / 4

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
3.5 Finite rotation groups
Let gSO(3) be a non-trivial rotation. Then gis the rotation around a
unique line `in R3. The two intersection points of `and the unit sphere
S={x2+y2+z2= 1} R3we refer to as poles.
Lemma 3.5.1. Let Gbe a non-trivial subgroup of SO(3), and let Xbe the
set of poles arising from gG, with g6= id. Then Gacts on the set X.
Proof. Let Pbe a pole, and gGa group element having Pas a pole. Let h
be an element of G. Then the action is natural, and we have that h.P =h(P)
is an element in R3, and in fact on the unit sphere. We need only to show
that h(P) is a pole, that is we need to see that h(P) is fixed by a non-trivial
element of G. Consider the element hgh1. Then hgh1(h(P)) = h(P), and
since g6= id we have that hgh1is a non-trivial element.
Proposition 3.5.2. Let Gbe a non-trivial, finite subgroup of SO(3). Let
mbe the number orbits |X/G|, where Xis the set of poles. Let ridenote
the order of the stabilizer group of an element Piin one orbit Bi(where
i= 1, . . . , m). Then we have the formula
22
|G|=
m
X
i=1
(1 1
ri
).
Moreover, the number of orbits m3is atmost three.
Proof. Let Pbe a pole. The stabilizer group of that point Pis all rotations
gGaround the line `defining P. This is a cyclic group of order r.
The opposite pole has the same stabilizator group. Moreover, each non-
trivial element gGis a rotation about one unique line `, and produce two
poles. By removing the identity element from that observation we obtain the
equality
X
PX
(rp1) = 2(|G| 1).
Let the orbits of Xbe B1, . . . , Bm. The cardinality of each orbit |Bi|=ni.
The order rof the stabilizator group Crof a point Pis independent of the
chosen point in orbit containing P. Therefore the above equality becomes
m
X
i=1
(ri1)ni= 2|G| 2.
We have from the Orbit Formula (2.1.18) that niri=|G|, for each i=
1, . . . , m. Using this when dividing the above formula with |G|establishes
the famous formula of the proposition.
56
pf3
pf4

Partial preview of the text

Download Finite Rotation Groups in SO(3) and Their Orbits and more Lecture notes Mathematics in PDF only on Docsity!

3.5 Finite rotation groups

Let g ∈ SO(3) be a non-trivial rotation. Then g is the rotation around a unique line in R^3. The two intersection points of and the unit sphere S = {x^2 + y^2 + z^2 = 1} ⊂ R^3 we refer to as poles.

Lemma 3.5.1. Let G be a non-trivial subgroup of SO(3), and let X be the set of poles arising from g ∈ G, with g 6 = id. Then G acts on the set X.

Proof. Let P be a pole, and g ∈ G a group element having P as a pole. Let h be an element of G. Then the action is natural, and we have that h.P = h(P ) is an element in R^3 , and in fact on the unit sphere. We need only to show that h(P ) is a pole, that is we need to see that h(P ) is fixed by a non-trivial element of G. Consider the element hgh−^1. Then hgh−^1 (h(P )) = h(P ), and since g 6 = id we have that hgh−^1 is a non-trivial element.

Proposition 3.5.2. Let G be a non-trivial, finite subgroup of SO(3). Let m be the number orbits |X/G|, where X is the set of poles. Let ri denote the order of the stabilizer group of an element Pi in one orbit Bi (where i = 1,... , m). Then we have the formula

|G|

∑^ m

i=

ri

Moreover, the number of orbits m ≤ 3 is atmost three.

Proof. Let P be a pole. The stabilizer group of that point P is all rotations g ∈ G around the line defining P. This is a cyclic group of order r. The opposite pole has the same stabilizator group. Moreover, each non- trivial element g ∈ G is a rotation about one unique line, and produce two poles. By removing the identity element from that observation we obtain the equality (^) ∑

P ∈X

(rp − 1) = 2(|G| − 1).

Let the orbits of X be B 1 ,... , Bm. The cardinality of each orbit |Bi| = ni. The order r of the stabilizator group Cr of a point P is independent of the chosen point in orbit containing P. Therefore the above equality becomes

∑^ m

i=

(ri − 1)ni = 2|G| − 2.

We have from the Orbit Formula (2.1.18) that niri = |G|, for each i = 1 ,... , m. Using this when dividing the above formula with |G| establishes the famous formula of the proposition.

The last stament follows as the stabilizer GP contains at least another element different from the identity, so the summand to the right is at least 1/2 whereas the sum to the left is strictly less than 2.

Corollary 3.5.3. Let r = (r 1 ,... , rm) be the order of the stabilizer groups where r 1 ≤ · · · ≤ rm, where m is the number of orbits X/G. Then the only possibilites that can occur are

  1. Two orbits, and r = (N, N ), where |G| = N ≥ 2.
  2. Three orbits, and r = (2, 2 , N ), where |G| = 2N ≥ 8.
  3. Three orbits, and r = (2, 3 , 3), where |G| = 12.
  4. Three orbits, and r = (2, 3 , 4), where |G| = 24.
  5. Three orbits, and r = (2, 3 , 5), where |G| = 60.

Proof. Let |G| = N. One orbit m = 1 is impossible since 2 − 2 /N ≥ 1 whereas 1 − 1 /r 1 < 1. With m = 2 the formula can be written as

2 /N = 1/r 1 + 1/r 2 ,

which is only true with r 1 = r 2 = N , proving the first case. With three orbits m = 3 we rewrite the formula as

2 /N = 1/r 1 + 1/r 2 + 1/r 3 − 1.

If r 1 ≥ 3 then the right hand side will not sum to a strictly positive integer as the left hand side is. Thus r 1 = 2. If also r 2 = 2 then r 3 = N/2, and since each ri ≥ 2 we have proven the second case. If r 1 = 2 then we note that r 2 ≥ 4 together with r 3 ≥ 4 is impossible since the right hand side sum needs to be strictly positive. Similarily, as 1/2 + 1/3 + 1/ 6 − 1 = 0, we are only left with three possibilities with

r = (2, 3 , 3), r = (2, 3 , 4) and r = (2, 3 , 5).

The group order N = |G| is in this three cases 12, 24 and 60, respectively.

Lemma 3.5.4. Let G be a finite subgroup of SO(3). Let Q be a pole, let H be the stabilizator group of Q. Let P be a pole defined by another line that defines Q. Then the orbit of P under the action of H consists of |H| poles.

Proof. By assumption the line defining P is different from the line′^ defining Q. Hence the intersection of the stabilizator groups GP ∩ H is the trivial group. The result then follows from the Orbit Formula 2.1.18.

In the case with (2, 3 , 5). The twelve points forming the orbit B 3 have all a regular pentagon of points in B 2 attached to it. The 20 points forming orbit B 2 have all regular triangles attached to it. The points in B 2 form the vertices of a dodecahedron, and the 12 points in B 3 form the vertices of an icosahedron. In these cases we get two platonic solids inscribed in the unit sphere. The symmetry group of one object equals the symmetry group of the other.

Corollary 3.5.6. The only regular polyhedra in three space are the Platonic solids; the tetraheder, the cube, the octaheder, the dodecaheder, and the icosa- heder.

Exercises. Finite rotation groups

  1. The twelve points (± 1 , ±α, 0), (0, ± 1 , ±α), (±α, 0 , ±1) fomr the ver- tices of a regular icosahedron if α is suitably chosen.

(a) Verify this and determine α. (b) Determine the matrix of rotation with angle 2π/5, about the line containing the point (1, α, 0).

  1. Prove Corollary 3.5.6 directly, with elementary methods.