28.11.2014 Views

Triangulated Surfaces and Higher-Dimensional Manifolds - TU Berlin

Triangulated Surfaces and Higher-Dimensional Manifolds - TU Berlin

Triangulated Surfaces and Higher-Dimensional Manifolds - TU Berlin

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Triangulated</strong> <strong>Surfaces</strong> <strong>and</strong> <strong>Higher</strong>-<strong>Dimensional</strong> <strong>Manifolds</strong><br />

Frank H. Lutz<br />

(<strong>TU</strong> <strong>Berlin</strong>)


----------------------------------------------<br />

Topology: orientable, genus g = 3<br />

Combinatorics:<br />

Geometry:<br />

vertex-minimal, n = 10 vertices,<br />

irreducible<br />

coordinate-minimal,<br />

general position


Combinatorics:<br />

triangulated<br />

2-manifold M :<br />

f-vector f = (n, f 1 , f 2 ).<br />

double counting 3f 2 = 2f 1<br />

Euler characteristic χ(M) = n − f 1 + f 2<br />

=⇒ f = (n, 3n − 3χ(M), 2n − 2χ(M))<br />

----------------------------------------------


----------------------------------------------<br />

Combinatorics:<br />

triangulated<br />

2-manifold M :<br />

f-vector f = (n, f 1 , f 2 ).<br />

double counting 3f 2 = 2f 1<br />

Euler characteristic χ(M) = n − f 1 + f 2<br />

=⇒ f = (n, 3n − 3χ(M), 2n − 2χ(M))<br />

number of edges: f 1 ≤ ( )<br />

n<br />

2<br />

i.e. n 2 − 7n + 6χ(M) ≥ 0


----------------------------------------------<br />

PROPOSITION HEAWOOD (1890)<br />

n ≥<br />

⌈<br />

1<br />

2 (7 + √ 49 − 24χ(M))⌉


----------------------------------------------<br />

PROPOSITION HEAWOOD (1890)<br />

n ≥<br />

⌈<br />

1<br />

2 (7 + √ 49 − 24χ(M))⌉<br />

THEOREM RINGEL (1955)/JUNGERMAN & RINGEL (1980)<br />

For every surface M, with the exception of<br />

M(2, +), M(2, −), <strong>and</strong> M(3, −),<br />

there is a triangulation with n vertices<br />

iff<br />

⌈<br />

n ≥ 1<br />

2 (7 + √ 49 − 24χ(M))⌉<br />

.<br />

(In the exceptional cases one vertex has to be added.)


----------------------------------------------<br />

Examples:<br />

χ(S 2 ) = 2: n ≥ 4<br />

χ(RP 2 ) = 1: n ≥ 6<br />

χ(T 2 ) = 0: n ≥ 7<br />

1<br />

1<br />

7 3 4 7<br />

3 4<br />

2<br />

1<br />

6<br />

1<br />

2<br />

4<br />

3<br />

5<br />

2 1<br />

3<br />

6<br />

2 5 2<br />

7 3<br />

4<br />

7


----------------------------------------------<br />

Enumeration/Realization:<br />

7 3 4 7<br />

[ 1, 2, 3 ], [ 1, 2, 4 ],<br />

[ 1, 3, 7 ], [ 1, 4, 5 ],<br />

[ 1, 5, 6 ], [ 1, 6, 7 ],<br />

[ 2, 3, 6 ], [ 2, 4, 7 ],<br />

[ 2, 5, 6 ], [ 2, 5, 7 ],<br />

[ 3, 4, 5 ], [ 3, 4, 6 ],<br />

[ 3, 5, 7 ], [ 4, 6, 7 ]<br />

1<br />

1<br />

6<br />

2 5 2<br />

7 3<br />

4<br />

7<br />

Möbius’ torus (1861)


----------------------------------------------<br />

Enumeration/Realization:<br />

7<br />

Coordinates:<br />

[ 1, 2, 3 ], [ 1, 2, 4 ],<br />

[ 1, 3, 7 ], [ 1, 4, 5 ],<br />

[ 1, 5, 6 ], [ 1, 6, 7 ],<br />

[ 2, 3, 6 ], [ 2, 4, 7 ],<br />

[ 2, 5, 6 ], [ 2, 5, 7 ],<br />

[ 3, 4, 5 ], [ 3, 4, 6 ],<br />

[ 3, 5, 7 ], [ 4, 6, 7 ]<br />

1: (3,-3,0)<br />

2: (-3,3,0)<br />

3: (-3,-3,1)<br />

4: (3,3,1)<br />

5: (-1,-2,3)<br />

6: (1,2,3)<br />

7: (0,0,15)<br />

5<br />

6<br />

1<br />

3<br />

4<br />

2<br />

Császár’s torus (1949)


----------------------------------------------<br />

Enumeration schemes:<br />

1. Generation from irreducible triangulations.<br />

2. Lexicographic enumeration.<br />

3. Strongly connected enumeration.


----------------------------------------------<br />

Enumeration schemes:<br />

1. Generation from irreducible triangulations.<br />

2. Lexicographic enumeration.<br />

3. Strongly connected enumeration.


[Royle, 2001]: n Types<br />

4 1<br />

5 1<br />

6 2<br />

7 5<br />

8 14<br />

9 50<br />

10 233<br />

11 1.249<br />

12 7.595<br />

13 49.566<br />

14 339.722<br />

15 2.406.841<br />

16 17.490.241<br />

17 129.664.753<br />

18 977.526.957<br />

19 7.475.907.149<br />

20 57.896.349.553<br />

21 453.382.272.049<br />

22 3.585.853.662.949<br />

23 28.615.703.421.545<br />

<strong>Triangulated</strong><br />

2-spheres<br />

with<br />

4 ≤ n ≤ 23<br />

vertices<br />

----------------------------------------------


----------------------------------------------<br />

edge contraction ↔ vertex split<br />

A triangulation is irreducible if there is no contractible edge.


----------------------------------------------<br />

THEOREM BARNETTE & EDELSON (1989)<br />

All 2-manifolds have finitely many<br />

irreducible triangulations.


----------------------------------------------<br />

THEOREM BARNETTE & EDELSON (1989)<br />

All 2-manifolds have finitely many<br />

irreducible triangulations.<br />

THEOREM NAKAMOTO & OTA (1995)<br />

Irreducible triangulations of a closed surface of genus g<br />

have at most O(g) vertices.


----------------------------------------------<br />

Numbers of irreducible triangulations:<br />

2-sphere: 1 [Steinitz, 1922]<br />

2-torus: 21 [Grünbaum, 1970, Lavrenchenko, 1984]<br />

M(2, +): 396.784 [Sulanke, 2005]<br />

M(3, +): ?<br />

RP 2 2 [Barnette, 1982]<br />

Klein bottle: 29 [Sulanke, 2004]<br />

M(3, −): 9.708 [Sulanke, 2005]<br />

M(4, −): 6.297.982 [Sulanke, 2005]<br />

M(5, −): ?


----------------------------------------------<br />

Enumeration schemes:<br />

1. Generation from irreducible triangulations.<br />

2. Lexicographic enumeration.<br />

3. Strongly connected enumeration.


----------------------------------------------<br />

Basic property:<br />

I :<br />

Every edge lies in exactly two triangles.


----------------------------------------------<br />

Ex.: n = 5 triangle-edge incidence matrix<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />


Ex.: n = 5 triangle-edge incidence matrix<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

2 2 2 0 2 2 2 2 2 2<br />

⎞<br />

⎟<br />

⎠<br />

----------------------------------------------


----------------------------------------------<br />

Ex.: n = 5<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />


Ex.: n = 5 backtracking!<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />

⎠<br />

1 1 0 0 1 0 0 0 0 0<br />

----------------------------------------------


Ex.: n = 5 backtracking!<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />

⎠<br />

2 1 1 0 1 1 0 0 0 0<br />

----------------------------------------------


Ex.: n = 5 backtracking!<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />

⎠<br />

3 1 1 1 1 1 1 0 0 0<br />

----------------------------------------------


Ex.: n = 5 backtracking!<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />

⎠<br />

2 2 2 0 1 1 0 1 0 0<br />

----------------------------------------------


Ex.: n = 5 backtracking!<br />

123<br />

124<br />

125<br />

134<br />

135<br />

145<br />

234<br />

235<br />

245<br />

345<br />

⎛<br />

⎜<br />

⎝<br />

12 13 14 15 23 24 25 34 35 45<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

1 1 1<br />

⎞<br />

⎟<br />

⎠<br />

lexicographic<br />

enumeration!<br />

2 2 2 0 1 1 0 1 0 0<br />

----------------------------------------------


Pseudomanifold with isolated singularities:<br />

----------------------------------------------


Pseudomanifold with isolated singularities:<br />

----------------------------------------------


----------------------------------------------<br />

Pseudomanifold with isolated singularities:<br />

II :<br />

The link of a vertex should be one circle.


Several components:<br />

----------------------------------------------


----------------------------------------------<br />

Several components:<br />

III :<br />

The surface should be connected.


----------------------------------------------<br />

<strong>Triangulated</strong> surfaces with up to 10 vertices [L. 2003]:<br />

n Surface Types n Surface Types n Surface Types<br />

4 S 2 1 8 S 2 14 10 S 2 233<br />

T 2 7 T 2 2109<br />

5 S 2 1 M(2, +) 865<br />

RP 2 16 M(3, +) 20<br />

6 S 2 2 K 2 6<br />

RP 2 1210<br />

RP 2 1 9 S 2 50 K 2 4462<br />

T 2 112 M(3, −) 11784<br />

7 S 2 5 M(4, −) 13657<br />

T 2 1 RP 2 134 M(5, −) 7050<br />

K 2 187 M(6, −) 1022<br />

RP 2 3 M(3, −) 133 M(7, −) 14<br />

M(4, −) 37<br />

M(5, −) 2


----------------------------------------------<br />

Isomorphism free lexicographic enumeration:<br />

<strong>Triangulated</strong> surfaces with 11 <strong>and</strong> 12 vertices<br />

[Sulanke & L., 2006].


Realization:<br />

THEOREM STEINITZ (1922)<br />

Every triangulated 2-sphere is realizable<br />

as the boundary of a simplicial 3-polytope.<br />

CONJEC<strong>TU</strong>RE DUKE/GRÜNBAUM (1970/1973)<br />

Every triangulated 2-torus is realizable in R 3 .<br />

THEOREM BOKOWSKI & GUEDES DE OLIVEIRA (2000)<br />

There is a non-realizable 12-vertex triangulation<br />

of the orientable surface of genus 6.<br />

----------------------------------------------


Realization:<br />

THEOREM STEINITZ (1922)<br />

Every triangulated 2-sphere is realizable<br />

as the boundary of a simplicial 3-polytope.<br />

CONJEC<strong>TU</strong>RE DUKE/GRÜNBAUM (1970/1973)<br />

Every triangulated 2-torus is realizable in R 3 .<br />

THEOREM SCHEWE (2006)<br />

For every surface of genus g ≥ 5<br />

there is a non-realizable triangulation.<br />

----------------------------------------------


----------------------------------------------<br />

Realization of vertex-minimal triangulations:<br />

g n min Types Realizable?<br />

0 4 1 yes<br />

1 7 1 yes [Császár, 1949]<br />

2 10 865 [L., 2003] yes [Bokowski & L., 2005]<br />

3 10 20 [L., 2003] yes [Hougardy, L., Zelke, 2006]<br />

4 11 821 [L., 2005] yes [Hougardy, L., Zelke, 2006]<br />

5 12 751.593 [Sulanke, 2005] yes/no [Hougardy, L., Zelke, 2006],<br />

[Schewe, 2006]<br />

6 12 59 [Bokowski, 1996] no [Bokowski & Guedes de Oliveira,<br />

2000], [Schewe, 2006]


Realization heuristics:<br />

----------------------------------------------


----------------------------------------------<br />

Realization heuristics:<br />

r<strong>and</strong>om realization <strong>and</strong> recycling of coordinates: 864 L. (2005)<br />

32768 3


----------------------------------------------<br />

Realization heuristics:<br />

r<strong>and</strong>om realization <strong>and</strong> recycling of coordinates: 864 L. (2005)<br />

32768 3<br />

geometric intuition: 1 Bokowski (2005)


small coordinates: Hougardy, L., Zelke (2005)<br />

----------------------------------------------


----------------------------------------------<br />

THEOREM HOUGARDY, L., ZELKE (2005)<br />

All 865 triangulations of the orientable surface<br />

of genus 2 with 10 vertices are realizable in the (4 ×4×4)-cube.


----------------------------------------------<br />

THEOREM HOUGARDY, L., ZELKE (2006)<br />

At least 17 of the 20 triangulations of the orientable surface<br />

of genus 3 with 10 vertices are realizable in the (5 ×5×5)-cube.


----------------------------------------------<br />

THEOREM HOUGARDY, L., ZELKE (2006)<br />

At least 17 of the 20 triangulations of the orientable surface<br />

of genus 3 with 10 vertices are realizable in the (5 ×5×5)-cube.<br />

THEOREM HOUGARDY, L., ZELKE (2005)<br />

All 20 triangulations of the orientable surface<br />

of genus 3 with 10 vertices are realizable.


----------------------------------------------<br />

intersection edge functional: Hougardy, L., Zelke (2005)<br />

E<br />

f<br />

a<br />

c<br />

u<br />

w<br />

b<br />

v<br />

d<br />

e


Example:<br />

----------------------------------------------


----------------------------------------------<br />

g = 4, n = 11:<br />

All 821 examples are realizable [HLZ, 2006].<br />

g = 5, n = 12:<br />

At least 15 of the 751.593 examples are realizable [HLZ, 2006]<br />

<strong>and</strong> at least 3 are not realizable [Schewe, 2006].


g = 4, n = 11:<br />

All 821 examples are realizable [HLZ, 2006].<br />

g = 5, n = 12:<br />

At least 15 of the 751.593 examples are realizable [HLZ, 2006]<br />

<strong>and</strong> at least 3 are not realizable [Schewe, 2006].<br />

CONJEC<strong>TU</strong>RE<br />

Every triangulated surface of genus 1 ≤ g ≤ 4<br />

is realizable in R 3 .<br />

----------------------------------------------


----------------------------------------------<br />

Enumeration schemes:<br />

1. Generation from irreducible triangulations.<br />

2. Lexicographic enumeration.<br />

3. Strongly connected enumeration.


----------------------------------------------<br />

Date: Thu, 3 Feb 2005 17:23:30 +0100<br />

From: John M Sullivan <br />

Subject: triangulations with low edge degree<br />

Dear Frank,<br />

...<br />

Consider triangulated 3-manifolds with every<br />

edge valence at most 5.<br />

Are there really (as I suspect) only a finite number<br />

of such triangulations, <strong>and</strong> are the manifolds then<br />

all spherical?<br />

?<br />

Do you perhaps already have such a list?<br />

If not, can we modify your programs to produce one?<br />

Best, John


----------------------------------------------<br />

Date: Thu, 3 Feb 2005 17:23:30 +0100<br />

From: John M Sullivan <br />

Subject: triangulations with low edge degree<br />

Dear Frank,<br />

...<br />

Consider triangulated 3-manifolds with every<br />

edge valence at most 5.<br />

Are there really (as I suspect) only a finite number<br />

of such triangulations, <strong>and</strong> are the manifolds then<br />

all spherical?<br />

Do you perhaps already have such a list?<br />

If not, can we modify your programs to produce one?<br />

?<br />

No :-(<br />

Best, John


----------------------------------------------<br />

Date: Thu, 3 Feb 2005 17:23:30 +0100<br />

From: John M Sullivan <br />

Subject: triangulations with low edge degree<br />

Dear Frank,<br />

...<br />

Consider triangulated 3-manifolds with every<br />

edge valence at most 5.<br />

Are there really (as I suspect) only a finite number<br />

of such triangulations, <strong>and</strong> are the manifolds then<br />

all spherical?<br />

Do you perhaps already have such a list?<br />

If not, can we modify your programs to produce one?<br />

Best, John<br />

?<br />

No :-(<br />

Yes :-)


Vertex-links: 2-spheres with face vector f = (n, f 1 , f 2 ).<br />

----------------------------------------------


----------------------------------------------<br />

Vertex-links: 2-spheres with face vector f = (n, f 1 , f 2 ).<br />

By Euler’s equation n − f 1 + f 2 = 2<br />

<strong>and</strong> double counting 2f 1 = 3f 2<br />

=⇒ f = (n, 3n − 6, 2n − 4)


----------------------------------------------<br />

Vertex-links: 2-spheres with face vector f = (n, f 1 , f 2 ).<br />

By Euler’s equation n − f 1 + f 2 = 2<br />

<strong>and</strong> double counting 2f 1 = 3f 2<br />

=⇒ f = (n, 3n − 6, 2n − 4)<br />

Vertex-degree ≤ 5: 2f 1 = ∑ deg ≤ 5n.<br />

Thus n ≤ 12


Enumeration (of vertex-links):<br />

----------------------------------------------


Enumeration (of vertex-links):<br />

----------------------------------------------


Enumeration (of vertex-links):<br />

----------------------------------------------


Enumeration (of vertex-links):<br />

----------------------------------------------


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .<br />

. . . . . .


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .<br />

. . . . . .


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .<br />

. . . . . .


----------------------------------------------<br />

Enumeration (of vertex-links):<br />

. . . . . .<br />

strongly connected!<br />

. . . . . .


Vertex-links:<br />

----------------------------------------------


----------------------------------------------<br />

Combinatorial types of 3-manifolds with edge valence at most 5:<br />

n Types n Types n Types n Types n Types<br />

5 1 19 230 33 115 47 15 62 2<br />

6 2 20 275 34 93 48 10 63 2<br />

7 5 21 240 35 76 49 15 64 3<br />

8 7 22 240 36 72 50 6 65 7<br />

9 16 23 214 37 55 51 9 69 1<br />

10 29 24 250 38 46 52 8 72 1<br />

11 37 25 220 39 37 53 11 73 1<br />

12 61 26 205 40 44 54 18 76 4<br />

13 77 27 191 41 54 55 2 80 2<br />

14 111 28 194 42 45 56 4 87 1<br />

15 134 29 174 43 33 57 3 98 1<br />

16 184 30 156 44 17 58 5 120 1<br />

17 202 31 115 45 23 60 3<br />

18 233 32 122 46 16 61 1


----------------------------------------------<br />

THEOREM L. & SULLIVAN (2005)<br />

There are exactly 4787 combinatorially distinct<br />

triangulated 3-manifolds with edge valence at most five:<br />

4761 × S 3 ,<br />

22 × RP 3 ,<br />

2 × L(3, 1),<br />

1 × L(4, 1),<br />

1 × S 3 /Q.<br />

THEOREM MATVEEV & SHEVCHISHIN (2005)<br />

Let M be a triangulated 3-manifold such that every<br />

edge has valence at most five.<br />

Then M is spherical <strong>and</strong> has at most 600 tetrahedra.


----------------------------------------------<br />

THEOREM L. & SULLIVAN (2006)<br />

There are exactly 41 distinct 3-dimensional combinatorial<br />

pseudomanifolds with edge valence at most five.<br />

They all are of type<br />

susp(RP 2 ).


----------------------------------------------<br />

THEOREM L. & SULLIVAN (2006)<br />

There are exactly 41 distinct 3-dimensional combinatorial<br />

pseudomanifolds with edge valence at most five.<br />

They all are of type<br />

susp(RP 2 ).<br />

THEOREM L. & SULLIVAN (2005)<br />

Let M be a triangulated d-manifold such that every<br />

codimension-two face has valence at most four.<br />

Then M is the join product of boundaries of simplices.


----------------------------------------------<br />

Recognition:<br />

Algorithms for the recognition of the 3-sphere:<br />

[Rubinstein, 1992], [Thompson, 1994], [S. Matveev, 1995], . . .


----------------------------------------------<br />

Recognition:<br />

Algorithms for the recognition of the 3-sphere:<br />

[Rubinstein, 1992], [Thompson, 1994], [S. Matveev, 1995], . . .<br />

. . . are exponential, difficult to implement.


----------------------------------------------<br />

Recognition:<br />

Algorithms for the recognition of the 3-sphere:<br />

[Rubinstein, 1992], [Thompson, 1994], [S. Matveev, 1995], . . .<br />

. . . are exponential, difficult to implement.<br />

Heuristics for the recognition of the 3-sphere:<br />

[Björner & L., 1997]<br />

. . . simulated annealing with bistellar flips.


----------------------------------------------<br />

Bistellar Flips/Pachner moves!<br />

2d:<br />

3d:


----------------------------------------------<br />

THEOREM CSORBA & L. (2005)<br />

Hom(C 6 , K 5 ) ∼ = (S 2 × S 2 ) #29 .<br />

f = (1920, 30780, 104520, 126000,50400)<br />

=⇒ f = (33, 379, 1786, 2300, 920)<br />

THEOREM CSORBA & L. (2005)<br />

Hom(C 5 , K 5 ) ∼ = V 4,2<br />

∼ = S 2 × S 3 .<br />

f = (1020, 25770, 143900, 307950,283200,94400)<br />

=⇒ f = (12, 66, 220, 390, 336, 112)


BUT. . .<br />

. . . there is a 3-sphere with no flips (16 vertices)<br />

[Dougherty, Faber, <strong>and</strong> Murphy, 2004]<br />

. . . there are non-shellable/non-constructible 3-spheres<br />

with n ≥ 13 vertices [L., 2004]<br />

. . . there are worse examples [King, 2004]<br />

. . . there is no algorithm to recognize 4-manifolds [Markov, 1958]<br />

. . . there is no algorithm to recognize d-spheres for d ≥ 5<br />

[Novikov, 1962]<br />

. . . there are non-PL spheres [Edwards, 1975], [Cannon, 1979]<br />

. . . there are non-PL d-spheres with d + 13 vertices for d ≥ 5<br />

[Björner & L., 2000]<br />

----------------------------------------------


----------------------------------------------<br />

THEOREM KING (2004)<br />

From any triangulation of S 3 , one can obtain<br />

an edge contractible triangulation of S 3<br />

by a sequence of at most 2 401f2 3<br />

successive expansions.


----------------------------------------------<br />

THEOREM KING (2004)<br />

From any triangulation of S 3 , one can obtain<br />

an edge contractible triangulation of S 3<br />

by a sequence of at most 2 401f2 3<br />

successive expansions.<br />

THEOREM LICKORISH (1991), ZIEGLER (1996), KING (2004)<br />

Every 3-manifold has infinitely many<br />

irreducible triangulations.


Why don’t we run into bad examples?<br />

----------------------------------------------


----------------------------------------------<br />

Why don’t we run into bad examples?<br />

Where do we get our examples from?


----------------------------------------------<br />

Why don’t we run into bad examples?<br />

Where do we get our examples from?<br />

1. Enumeration: bad examples need more space.<br />

2. Combinatorial constructions: mostly harmless.<br />

3. Topological constructions: we usually know the outcome in advance.


Construction of a non-realizable 3-ball with 16 vertices:<br />

----------------------------------------------


----------------------------------------------<br />

Construction of a non-realizable 3-ball with 16 vertices:<br />

15<br />

1<br />

3<br />

7<br />

12<br />

13<br />

10<br />

4 6<br />

5<br />

9<br />

8<br />

11<br />

14<br />

2<br />

non-realizable [non-constructible/non-shellable] 3-ball: (16, 75, 106, 46)<br />

[non-constructible/non-shellable 3-sphere: (17, 91, 148, 74)]


----------------------------------------------<br />

Construction of a non-realizable 3-ball with 12 vertices:<br />

1<br />

3<br />

10<br />

7<br />

4 6<br />

5<br />

8<br />

11<br />

12<br />

9<br />

2<br />

non-realizable [non-constructible/non-shellable] 3-ball: (12, 58, 84, 37)<br />

[non-constructible/non-shellable 3-sphere: (13, 69, 112, 56)]


----------------------------------------------<br />

COROLLARY L. (2003)<br />

There are non-realizable [non-constructible]<br />

d-balls with d + 9 vertices <strong>and</strong> 37 facets for d ≥ 3.


----------------------------------------------<br />

COROLLARY L. (2003)<br />

There are non-realizable [non-constructible]<br />

d-balls with d + 9 vertices <strong>and</strong> 37 facets for d ≥ 3.<br />

COROLLARY L. (2003)<br />

There are non-constructible/non-shellable<br />

d-spheres with d + 10 vertices for d ≥ 3.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!