ArticlePDF Available

A Unified Approach to Class I, II & III Geodesic Domes

Authors:

Abstract

This paper outlines a unified approach for generating all three classes of geodesic domes. The approach, which allows one method to generate configurations for all three classes, is presented in detail along with ten specific methods from which numerous actual configurations can be generated. The new approach is especially useful in generating the more difficult class III domes. Also a number of graphs are included to illustrate each of the given ten methods relationship of frequency to the number of different edges, and faces, which are important design criteria. Lastly, a number of older, independently developed methods, which are encompassed by the new approach, are cross-referenced.
Unified Approach to Class I, II & ru Geodesic Domes
Christopher J. Kitrick
Reprinted from
INTERNAIIONAL JOURNAL OF
SPACE, STRI.ICTTIRE,S
Volume 5 Nos. 3 &,4 1990
MUTTI-SCIENCE PTTBLISHING CO. TJTD.
107 HIGH STREET, BRENTWOOD, ESSEX CMI4 4RX LINITED KINGDOM
A Unified Approach to Glass l,ll &
lll Geodesic Domes
Christopher J. Kitrick
Wavefront Technologies, Inc., 530 E. Montecito St., Santa Barbara, California. 93103, USA
(Received 31st May 1989; revised version received l9th Rbruary 1990).
ABSTRACT: This paper outlines a unified approach for generating all three
classes of geodesic domes. The approach, which allows one method to generate
configurations for all three classes, is presented in detail along with ten specilic
methods from which numerous actual configurations can be generated. The new
approach is especially useful in generating the more difficult class III domes.
Also a number of graphs are included to illustrate each of the given ten methods'
relationship of frequency to the number of different edges, and faces, which are
important design criteria. Lasfly, d number of older, independently developed
methods, which are encompassed by the new approach, are cross-referenced.
1 . Geodesic Glassification
Current geodesic classification is the result of add-
ing two subscripts (D,c) to the Coxeter notation for
polyhedra {p,ql. Thus for the icosahedron the
notation reads {3,51u,r''',' . The values of b and c fatl
into three categories:
Class I b>0 c:0 (Fig. 1),
Class II b>0 c-b (Fig. 2),
Class III b>0 c)0 and c*b (Fig. 3).
Class III configurations are enantiomorphic;
exhibiting right and left handedness when the
values of b and c are reversed. In this paper all class
III (b,c) pairs will have b greater tharl c.
Most geodesic domes constnrcted are either
class I or II, due to the availability of methods
which generate tessellations with a linear increase
in complexity. Class I are the easiest to use
especially when near equatorial truncation is
desired. Class III have no easy tnrncation arcs,
making them require the most attention when a
tnrncation is desired.
Intemational Joumal of Space Structures Vol. 5 Nns'. 3&4 1990
Clas s I
Figure I
Class I I
Figure 2
Class I I I
Figure 3
lli
(b.c) - ({,0)
(b.c) - (2,21
tu.cl - (2.[ )
The number of vertices, edges, and faces for any
(b,c) pair is a function of r where:
t: b2 + bc + c2
2. The lcosahedron
The icosahedron is the typical base geometry used
for geodesic domes because its Schwarz triangle
(U6 of an icosa face) is the smallest single (enall-
tiomorphic) sphere section (L/120) that can be
repeated to cover the entire spherical surface (Fig.
4). The Schwarz triangle will herein be referred to
as the lcd (lowest common denominator) triangle.
Atl geodesic vertices on an icosahedron can be des-
cribed by their relationship to the lcd triangle. All
the concepts presented herein are also applicable
to the Schwarz triangles of both the regular
octahedron and tetrahedron.
Schw arz Triangle
l/6Icosa Face
Fig. 4.
The relationship of t to the number of vertices,
edges, and faces for the icosahedron is:
vertices
edges
faces
3. Methodology
+2,
The methodology presented in this paper is
designed to include all classes and frequencies (b,c
pairs). It is not intended to optimize any par-
ticular one.45
224
Unified Approach to Geodesic Domes
Grid subdivision of the lcd triangle
f = 12
Fig. 5.
Class I
Fig. 6.
Class I I
Fig. 7.
International Journal of Space Stntctures Vol. 5 Nos. 3&4 1990
: 10r
: 20t,
- 30r.
C.J. Kitick
The basic approach involves the modular sub-
division of the lcd triangle into a rectangular grid
(Fig. 5). The number of divisions along an edge of
the lcd triangle is denoted by frequency A.For
every (b,c) pair there is a corresponding frequency,
such that all vertices of the tessellation will falI on a
grid intersection. For simplicity we will consider
the planar lcd as it appears on the face of the
icosahedron as a (30", 60o, 90o) triangle where all
rectangular cells are the same size with propor-
tions of /3/3 to 1.
For Class I tessellations frequency equals b/2
and each triangle is two cells wide and three high
(Fig. 6). Class II tessellations have a frequency
equal to b and each triangle is one cell wide and
two high (Fig. 7).
Class III Ske\M Ratios
Fig. 8.
Class III tessellations fall at a skew angle to the
grid. The difficulty is in finding the frequency so
that all vertices fall on grid intersections. Fig. 8
illustrates the basic situation of class III. The
givens are b,c, r:l and the angle of I20'. The
following notation describes the procedure used to
determine the correct frequency (fl for any class III
(b,c) pair as well as the individual offsets (m, n, ffib,
rltg, ftb, nr) for a single triangle (Fig. 9).
International Journal of Space Stntctures Vol. 5 Nos. 3&4 1990
Skew Ratios for (3,5)2,1
.k Eb:4 \
Fig. 9.
ob=5
fla:3
d
cosB
cosC
sin^B
sinC
dx
dy
m'
n'
m
n
tt
a
mb
mc
nb
nc
f
J)
is:
Using the above terminology (m, n, ffib, tlts, N6, ns,
for classes I dnd II and a few examples of III
Table 1
Class (b,c) m n.,, m6
I (any)2 3 I
U (any)l 2 I
m (2,1)5 9 4
m (3,1) 7 t2 5
m (3,2) 8 15 7
m (4,1)3 5 2
m 9,2) 5 9 4
m (4,3) 1l 2r lo
mcn6ncl
230
lll
563
7e3
896
341
563
1l 12 9
b/2
b
7
l3
19
7
t4
37
(Fig.5)
(Fig.7)
":'""';""""
a-, I
."' ," \
att
.t' ,i ia
aaa
." ,' ta
.aI
.t' ,' ta
aat
,t-,t
aa
a,
."
aa
aa
aa
a
.a
.,
aa
.t
.,
aa
a
.t
.,
.,
,a
.a
.a
.,
.a
aa
.,
.a
tl
a
a
a
m.:5
= y'@, + c2/4 + bc + t/qcT,
= (b/Z + c)/d,
: (c/2 + b)/d,
= (\/3b)/(u),
= (t/tt)t(?d),
= cosc - cosB,
= tfi/ux,
= cosC(b-c)lax : Note: b)c,
= (sirrB + sinO(b-c)/dy : Note: b)c,
: m'/g[eatest common multtple (!fr', ft'),
= n'/gyeatnst common multiple (m', fi'),
= cosC/m,
= t/ltl F,
= cosB/p,
: cosC/p,
= sinB/d,
= sinC/d,
= (bm, + cm/Z = d/(Ztt).
225
Unified Approach to Geodesic Domes
Class I I I
(b,c) - (2,1)
All grid intersections are defined by a coor-
dinate pair denoted by (ij) where:
i+j<f
Each method described in this paper provides a
unique one to one mapping of (U) pairs to (x,y,z)
coordinates. Thus for a particular geodesic
tessellation (b,c) the icd triangle is divided to the
correct frequency (f) andthe appropriate (ii) coor-
dinates are located for all unique vertices. Next
using a specific method the (ii) coordinates are
mappecl to actual (x,y,z) from which edge and faces
can be calculated. For example {3,5}r,, (f:7) has
only two unique (ij) coordinates (7,0) and (2,3)
(Fig. 10). All other vertices can be determined from
simple transformations.
226
4. The Mappings
This section describes in detail the mapping
calculations necessary for mapping (ii) coor-
dinates into (x,y,z) coordinates.
4. 1 Nine Methods for a Right Spherical
Triangle
The spherical methods operate on the spherical
lcd triangle (Fig. 11). Notation used for all nine
spherical methods is:
a - opposite arc side,
b - adjacent arc side,
c - hypotenuse arc side,
A - angle opposite arc side a,
B - angle opposite arc side b,
International Journal of Space Structures Vol. 5 Nos'. 3&4 1990
C.J. Kirick
(ij) integer coordinate of a point on
the lcd grid,
"f - frequency of grid,
(x'yo),r,, angular equivalent of (U): sub-
script denotes side,
(x,y,z) 3d coordinates of (ij) and
(xo, /o)u*ir.
It must be noted that b andc here in the context of
these methods bears no relationship to the
{p,qI u,,.
Sphe rLcal lcd Triangle
dinatcs (x",! )rr.rzr and thcn convert to (x,y,z) (Fig.
t2).
4.1.1 Method ua
Side a is subdivided by J' into A,a arc segments. At
each La distance a perpendicular arc is projected
until it intersects side c. All grid intersections lie on
these projected arcs. Grid points with the same i
value are the same arc distance away from side a.
For any (ij) position the (x'.)r"),.* is (Fig. 13).
La - o/f,
xo - arctan (sin(iA a)tan B),
y" - jLa,
axis - y.
4.1.2 Method ab
Side a is subdivided by -f into La arc segments.
From side c arcs are dropped to be perpendicular
to side b such that the opposite sides of these
smallertriangles are multiples of La.All grid inter-
sections lie on the arcs perpendicular to side b.
cos(y")
cos ( y' )
A=
B=
a
b=
JO
60'
90'
20.9C5'
31.717"
37.377"
asin (sinAsinb/ sinB)
alanQ) /2
asin (sinb/sinB)
The naming convention for the methods is
derived from the following
Metho d aa(b)
The Iirst letter denotes which side of the lcd is
uniformly divided by .f.The second letter denotes
the first side used to project or drop perpendicular
arcs. The optional third letter denotes the last side
used to project or drop perpendicular arcs.
The convention for Figs. 13-21 is that solid lines
represent continuous arcs.
The procedure for finding (x,y,r) from (ij) is to
first determine the intermediate spherical coor-
International Journal of Space Structures Vol. 5 Nos'. 3&4 1990
X=
Y=
L
sln
sin
COS
xn
Y,
xn
srn
sin
COS
xo
Y"
xo
X=
Y=
Lcos ( x'
cos ( y'
211
aXlS : XaXlS : Y
Gricl ploints *'ith thc sarlrc./ r'alue rlru' thc' sanrc ilrc
distancc' a\\'a\ trom side b (nrultiplc's ot' Aa ). For
arU U.j) position the (x'r1''),,.r,, is (Fig. l4):
L,u - a/f.
x' - b arcsin (tan(Aa(f-i))ltanA),
,'' - j La.
uvis - x.
4.1.3 Method aab
Side a is subdivided bV "f into La segments. From
side a perpendicular arcs are projected at each L,a
L'niliecl Approuch to Geodesic Domes
iute'n'al. From where these arcs intersect side c
allother arc is dropped to be perpendicular to side
b. All grid intersections lie on these projected per-
pendiculr arcs. For any (ii) position the (x" ,yo),..r., is
(Fig. 15):
L,a - a/f,
y" : jLa,
xo - arctan (cosy" sin(iAa)
(90o -a*ih'a)),
axis : y.
tan B/sin
Fig. 13.
0,6
=f o,s
aal
+ 0,4
0,3
0,2
o
vo,l
0,5
AA
0,4
o,2
o,0 1,0 2,O 3,0 4,O 5,0 6,0
constant constant constant constant constant constant
Method ab
f :6
Fig. 14.
b
0,6
-+
CI
(irl I
LlI
CI
ol
:+
I
(!l
-l
u1 I
=l
ol
,:*
cl
.'J I
i, I
=l
a zl
-2f I
(=l
el
'J1 I
CI
9+
EI
'2 I
OI
:+
el
-l
dl
=l
ol
,+ o,o x-"
b
22ri Intentational Journal of Space Structures Vol. 5 l/os. 3&4 1990
0,6
t o,s
aa I
+04
C.J. Kiuic'k
4.1.4 Method ba
Side b is subdivided bV "f into Lb arc segments.
From side c arcs are dropped to be perpendicuar to
side a such that the adjacent sides of these smaller
triangles are multiples of Lb. All grid intersections
lie on these arcs perpendicular to side a. Grid
points with the same f value are the same arc dis-
tance away from side a (multiples of Ab). For any
(ij) position the (xo,yo),,,, is (Fig. 16):
International Journal of Space Structures Vol. 5 Nos. 3&4 1990
b
Fig. 15.
Method ba
f :6
3ab
5ab
constant constant constant constant constant constailt
Fig. 16.
b
Lb - bt,,
xo : i&b,
yo : a arcsin (tan (L,u(f' - .il /tanB)).
axis : y.
4.1.5 Method bb
Side b is subdivided byl'into Ab arc se-qlrlents. At
each Lb distance a perpendicular arc is projectccl
until it intersects side c. Nl grid intersections lic orr
lt9
['nili*l .1ppy1111c,lt ttt Geodtsit, Donres
Fig. I7.
Fig. 18.
these proJected arcs. Grict points with the same Jr
value are the same arc clistance away from side b.
For any (ij) position the (xo,yo)o,,:, is lnig . l7):
xo : i&h,
y" _ arctan (sin(7A a)tanA),
exis - x.
4.1 .6 Merhod bha
Side b is subdividecl by ,f into Lb arc segmenm.
From side b perpenclicular arcs are projected at
each Lb interval. From where these arcs intersect
230
b
side c another arc is dropped to be pe{pendicular to
side a. Allgrid intersections rie on these proJected
pe{pendicular arcs. For any (ii) posiiion the
(x" ,yo)*i, is (Fig. l g):
Lb : b/f,
xo : i&b,
yo : arctan (cosro sn(ia,b) tanA/sin(90 b +
jLb)),
axis : y.
4.1.7 Method ca
side c is subdivided by _f into Ac arc segments.
Intemational Journal of space smtctures vol. Sl/os. 3&4 lgg0
C.J. Kitick
0,0 l,o 3,O 5,0 6,0
2,O
d { { { ::d ::d :J
constant constant constant constant constant constdlt
b
0,6
4,O
x" l'0
-+
f,l
ul
tf, I
CI
OI
:+
EI
el
al
5l
:+
CI
(trl
ul
(rt I
CI
a 8V
u7f
EI
-, I
6l
CI
ol
:+
e.t
(t, I
el
(nl
c. l
ol
i+
6l
*)l
al
cl
ol
"+
Method
f :6 cb
0,5
Fig. 20.
AA
o,4
o,3
o,2
3,0
b
0,6
0,5
o.4
0,3
o,2
o
vo,l
Fig. 21.
J,0
b
International Journal of Space Structures Vol. 5 Nos. 3A4 D90 231
From side c perpendicular arcs are dropped at
each A,c interval to be pe{pendicular with side a.
Atl grid intersections lie on these arcs perpen-
dicular to side a. The arc distances from any adja-
cent i values remains the same at allT values. For
any (i j) position the vo, !o),.o:, is (Fig. 19):
L,c - c/t
xo - arcsin (sin(iAc) sin B),
y' - a arctan (tan(A c(f-j)) cosB),
axis - y.
4.1.8 Method cb
Side c is subdivided by -f into Lc arc segments.
From side c perpendicular arcs are dropped at
each Ac interval to be perpendicular with side 6.
All grid intersections lie on these arcs perpen-
dicular to side D. The arc distances from any adja-
centT values remains the same at all i values. For
any (ij) position the (x", !")o** is (Fig.20):
Lc - c/f,
y" - arcsin (sin(7Ac) siM),
xo - b arctan (tan(Ac(f-i)) cos A),
axis : x.
4.1.9. Merhod cab
Side c is subdivided bV f into Ac arc segments.
From side c two arcs are dropped at each Ac inter-
val to intersect both side o and side b perpen-
dicularly. All grid intersections lie on these arcs
perpdndicular to both side a and b. For any (ij)
position the (x", !")r.rr, is (Fig.2l):
Lc - c/t
y" - a arctan (tan(A c(f-j)) cosB),
x' - b arctan (tan(Ac(f-i)) cos A),
xo : arctan (tanx' cosy"),
axis : y.
4.2 Method Ten - Radial Projection
Ivlethod ten is included due to its simplicity and
previous description.6 It was also used to provide a
reference frame for the l/s ratio comparisons. The
lcd triangle is divided by f in its planar form and
each intersection (ij) is projected radially until it
232
(lnifieel Approach to Geodesic Dontes
reaches the sphere surface. Thke the (x'J'z') posi-
tion on the plane and divide each of its com-
ponents by the distance to the center of the sphere
to find the projected (x,y,z) coordinates on the
sphere surface (Fig. 22).
Method Radial
22.
5. Comparative Analysis
Once we have a number of tessellations methods,
how do we ciifferentiate between them? There are a
number of commonly used yardsticks to measure
the appropriateness of a particular method to a
specific design problem. One of the most common
measures is that of complexity, which manifests
itself as uniqueness or lack of. Each (b,c) pair
generates a tessellation that has a maximum
potential for complexity constrained by the sym-
metry of the base geometry. Typically we would
like to minimize that complexity. For any (b,c) pair
there is a maximum number of different vertices,
edges, and faces that cAn be generated. The
approximation for the rnaximum number of dif-
ferent faces ffA is il3; the maximum number of dif:-
ferent edges (eO is t/2.
Alrother criteria for measuring a tessellation is
by comparing the ratio of longest to shortest edges,
or l/s ratio. For a regular icosahedron the lower
limit on the l/s rutio is l.l7 56 or the ratio of the long
side of an isosceles 54" , 72o , 54" triangle to the
short side. Tessellations with high l/s rutios tend to
be visually disjoint along the edges of the lcd
triangle, due to the pairing of long and short edges
in a single triangle resulting in rather non-
Intemational Journal of Space Structures Vol. 5 Nos. 3&4 1990
C.J. Kitick
equilateral conditions.
The ten methods presented were used to find
appropriate three dimensional coordinates for all
necess ary vertices for tessellations from t: I
(b :1,c -0) through t:256 (b : l6,c-0) inclusive'
Graphed against t are the number of faces fffi,,
edges @A,and the l/s ratio. It should be noted that
the certain values ofr appear more than once in the
sequence: example t:49 from (b,c) pairs (5,3)
and (7,0).'
5. I Different Faces
Graphs 1-10, derived from the generated data in
Thble 4, show the number of different face s ffA for
all values ofr for all ten methods. The maximum/i/
appears nearly as a straight line on the graphs. For
metho ds aa and bb, classes I and ll.Jd is a linear
function of b (on the order of {t). Table 2 provides
the formulas for determining the value of Jd tor
classes I and II and these two linear methods.
Table 2 - Different Faces (fd)
Class II
| zh-l
I
0b
90
80
70
Method Class I
oa I : b-
(sb-4)lt :b)
bb $b-l)/r :b)
Note: use integer Portion
112 1 33 1 48 1 69 t 89 201 223 243
t
Graph 1.
Method ab
r 12 133 148 169 189 201 223 243
t
Graph 2.
50
fdoo
30
20
10
0
60
50
fd oo
30
20
t0
0
Method aa
International Journal of space stntctures vol. 5 Nos. 3&4 1990 233
Method aab
Unified Approach to Geodesic Domes
97 112 133
t
Graph 3.
Method ba
79 97 112 133
t
Graph 4.
Method bb
97 1 1 2 133 148 1 69 189 201 223 243
t
Graph 5.
International Journal of Space Stntctures Vol. 5 Nos. 3&4 1990
231
C.J. Kitick
t
Graph 6.
Method ca
Graph 7.
Method cb
97 112 133
t
Graph 8.
Method bba
International Journal of Space Structures Vol. 5 Nos. 3&4 1990 235
Method
Unified Approach to Geodesic Domes
79 97 trr, 133
GraPh 9.
Method radial
97 112 131
t
Graph 10.
Method aa
112 133 l4E I ti9 I ou 4v I lLo
t
Graph 11.
International Journal of space structures vol- 5 Nos' 3&4 1990
236
C.J. Kitrick
r00
120
112 133 148 169 r89 201 223 243
t
Table 3 - Different Edges (ed)
5.2 Different Edges
Graphs ll-20, derived from the generated data in
Thble 5, show the number of different edges (ed) for
all values ofr for all ten methods. The maximumed
appears nearly as a straight line on the graphs. For
methods aa and bb, classes I and ll, ed is a linear
function of b (on the order of 1/t). Table 3 provides
the formulas for determining the value of ed for
class I and II and these two linear methods.
5.3 l/s Ratios
Graphs 21-30, derived from the generated data in
Thble 6, show the l/svalues over all values of/. In all
cases the value is becoming asymptotic, reaching a
maximum value. The graphs also show that there
is a wide fluctuation of l/s as all three classes are
mapped for a single method. Method cb, class II,
t40
120
Graph 12.
Method
aa
bb
Class I
I : b- I
(7b-8)tt : I <b <8
(7b - tt)tl : D ) 8
b
Class II
2b
2b
100
Note: use integer portion
which has already been shown to exhibit
asymptotically the minimum possible l/s ratio for
all triangular tessellations on the spheres'8 (the low
spikes on graph 28), also shows very small change
across all three classes giving the best overall l/s
ratios. The disadvantage here is that lh rutios do
not guarantee any minimizing of other factors
making its practical use rather limrted.
112 1 33 I 48 1 69 1 89 201 223 243
t
Method ab
Method aab
International Joumal of Space Structures Vol. 5 Nos. 3&4 1990
Graph 13.
8,5,129
9,4.1 33
11,1.133
1 0,3, 1 39
12,C.'144
7,7 ,147
11 ,2,147
8,5,148
9,5,151
10,4,156
12.r,157
1 1,3,163
8,7,169
13,0,1 69
9,6,1 71
12.2,172
10,5,175
11,4,181
13.1,183
12.3,1 89
8.8, r 92
9,7,193
10,5,1.96
14.0. /36
1 3,2,1 99
1 1 ,5.201
12,1,208
14 .1 ,211
9,8,217
13,3,217
10,7 ,219
11,a,223
15,0,225
14,2,228
12,5.229
13,4,237
15,1,241
9.9.24 3
10,8,244
11 .7,247
14,3,247
12.5.252
16,C,256
Table 4 : Oiflerent Faces (ld)
1
1
2
3
3
2
5
5
7
7
6
3
r0
11
7
13
13
15
4
17
9
17
19
21
17
10
23
25
5
21
27
11
25
31
3t
31
aa
13
35
6
37
37
25
30
14
38
43
43
45
45
47
15
7
49
44
51
45
53
55
57
17
41
51
32
61
6l
43
8
65
57
18
67
67
43
71
lo
73
73
75
19
65
77
79
81
I
70
83
83
39
21
cab radial
11
238 Intemational Joumal of Space Structures Vol. 5l{os. 3&4 1990
b,c,t
1,0,1
1,1,3
2,0.4
2,1,7
3,0,9
2,2,12
3,1 ,13
4,0,16
3,2,19
4.1 ,21
5,0,25
3,3,27
4,2,28
5,1,31
6,0,36
1,3,37
5,2,39
6,1 ,43
4,4,48
5,3,49
7 ,A,49
6,2,52
7,1 ,57
5,4,61
6,3,63
8.0,64
7,2,67
8,'1,73
5,5,75
6.4,76
7.3,79
9.0,81
9,2,84
6,5.91
9.1,91
7,4,93
8,3,97
1 0,0,1 00
9,2.1 03
6,6.108
7,5.109
10.1 .1 1 1
8,4,112
9,3.1 17
1 1,0,121
10,2,124
7,6,127
8,5,1 29
9,4,133
11.1.133
10,3,139
12,0,144
7 ,7 ,147
11 ,2,1 47
9,6,148
9,5.1 5 1
1 0,4,1 55
12,1,157
1 1,3,163
8,7,1 69
1 3,0.1 69
9,6,171
12,2,172
10,5,175
11,4,181
13,1 ,1 83
1 2,3,1 89
8,8,192
9,7,193
10,5,195
14,0,196
13,2.199
11,5,201
12,4,208
14,1 ,211
9,8,217
13,3,217
10.7 .219
11,6,223
15,0,225
14,2,228
12,5.229
13,4,237
1 5.1 ,241
9,9,243
10 ,8,244
1't ,7,247
14,3,247
12,6,252
16,0.255
2
1
2
2
3
1
2
2
4
3
4
7
4
10
11
5
b
13
16
6
I
0
2
8
5
7
3
9
1
3
I
4
7
0
3
0
9
5
46
46
47
49
10
52
12
55
56
32
41
11
53
54
65
67
67
70
12
14
74
62
76
64
79
82
85
13
57
72
41
91
92
59
16
97
81
l4
100
101
57
06
08
09
10
't2
15
93
15
19
21
18
00
24
124
51
l6
UniJiecl Approach to Geodesic Domes
Table 5 : Oillerent Edges (ed)
aab bb ba bba ca cb radial
122
123
123
125
60
C.J. Kitick
97 112 133
t
Graph 14.
97 112 I 33
t
Graph 15.
1 13 28 48 63 79 97 112 133
t
Graph 16.
International Journal of space structures vol. 5 Nos. 3&4 1990
Method ba
Method bb
Method bba
239
I\{e thod cil
Liniliul .l\tprouch to Geoclesic Domes
97 112 13
t
Graph 17.
Method cb
97 112
Graph
Method
18.
cab
97 112 I 33 1 48 1 69 1 89 201 223 243
t
Graph 19.
Intentational Joumal of Space Structures Vol. 5 Nos. 3&4 1990
240
Method radial
C.J. Kitick
r40
120
100
80
60
ed
1.54
1.5
1.46
1.42
r.38
l/s1.34
1.3
1.26
1.22
1.18
1 .14
1.54
1.5
r.46
1.42
r.38
l/s
' 1.34
1.3
1.26
1.22
l.'18
1.14
t
Graph 20.
1 08 124 114
t
Graph 21.
52 7g 91 108 124 1 4/t
t
Graph 22.
Vol. 5 Nos. 3&4 1990
40
20
0148 169 189 201
r57 175 196 217 229 217
r75
r57 217
37
Method aa
Method ab
International Journal of space structures
196
241
Method aab
1.54
1.5
1.46
1.42
r.38
I / Sr.sr
1.3
1.26
1.?2
r.l8
l.t 4
1.54
1.5
1.46
1.42
r.38
I / Sr.gr
r.3
1.26
1.22
1.18
l.t 4
1.54
1.5
1.45
1.42
1.38
I / t,.rn
t.3
1.26
1.22
1.18
1 .14
1 08 121 1 44 157 175 1 96 217
t
Graph 23.
Method ba
Uni,fiecl Approach to Geodesic Donres
la4 157 175 196 217
Graph 24.
to8 121 144 157 175 196 217 229 217
t
Graph 25.
Intentational Jountal of space structures Vol.
Method bb
242 5 Nos. 3&4 1990
Method bba
C.J. Kitick
l/s
1.54
1.5
1.46
1.42
1.38
1.34
1.3
1.26
1.22
1 .18
t.l4
1.54
1.5
1.46
1.12
1.38
l/s
' 1.34
1.3
1.26
1.22
1.18
1 .14
108 124 144
t
Graph 26.
1 0E 124 144
t
Gruph 27.
157 .175 196 217 229 247
157 175 196 217 229 217
Method cb
1.54
1.5
1.46
1.12
1.38
l/s1.3,1
1.3
1.26
1.22
1.18
l.l4 t08 121 144 157
t
International Journal of Space Structures
Graph 28.
Vol. 5 Nos. 3&4 1990
Method ca
175 196 217 229 247
243
Method cab
Unified Approach to Geodesic Domes
r.5
1.46
1.12
1.38
I/sr.3a
r.3
1.26
1.22
1.r 8
1.14
r.54
r.5
1.t36
1.12
r.38
l/sr.34
t.3
r.26
1.22
6. Cross Reference
Over many years there have been numerous
individuals who have pursued the problem of
spherical subdivision. Much of the work has gone
unpublished and what has been published has
typically been rather isolated for long periods of
time. The nine spherical methods presented in this
paper are not a complete set of those available but
they do not include those most used in praciice
(Methods aa and bb). The set of nine are actually
the author's extension of the comm on aa and bb
methods when viewed in their spherical form and
in context of the underlying grid structure. The
following table is the author's attempt to cross-
reference previous published work in relation to
the ten presented.
24
91 r08 124 t{1 t57 175 196 217 229 247
t
Grupln29.
Method radial
I 08 121 I {'l 157 I 75 I 96 217 229 217
t
Graph 30.
aa
ab
aab
ba
bb
Table 7 - Cross Reference
Method Reference [Notation: Author (description) (classes)]
Scheele (Class II)
Clinton6 lMethod 6) (Class II)
:
Stuatlo (Class II)
Tarnail I (Class I)
Pavlovl'(o Method) (Class I&ID
Clinton6 6'tethod 6) (Class I)
bba
ca
cb Kitrick3 (Class II)
cab
radial Clintonl lMethod I&IV) (Class I&ID
Intemational Journal of Space Stntctures Vol. 5 Nos. 3&4 1990
C.J. Kitick
Table6:Usdata
b.c.t ail
1,0,1 1.00000
1,1,3 1.11359
2,0,4 1.13083
2,1 ,7 1 .17356
3,0.9 1 .20164
2,2,12 1.21851
3,1,13 1.21184
4,0,16 1.25956
3,2,19 1.25570
4.1 ,21 1.28220
5.0,25 1.30184
3,3,27 1.26792
4,2,28 1.29284
5,1 .31 1.3'14O4
6,0,38 1.32593
4,3,37 1.29368
5.2,39 1.31849
6.1 ,43 1.33630
4.4,49 1.29476
5,3,49 1.31876
7 ,O,49 1 .34629
6,2,52 1.33863
7,1,57 1.35408
5.4,61 1.31448
6,3,63 1.33726
8,0,64 1.38335
7,2,67 1.35507
8.1 ,73 1.36786
5,5,75 1.31147
6,4,76 1 .33339
7 ,9,79 1.35270
9,0,81 1.37433
8,2,84 1.36764
6.5,91 1.32747
9.1 .91 1 .37888
7,4,93 1.34820
9,3,97 1.36532
10,0,100 1.38436
9,2.103 1.37867
6,6,108 132293
7,5 ,1 09 1 .34279
10,1,111 1.38793
8,4,112 1.36069
9,3,117 1:37588
1 1 ,0,1 21 1 .39342
10,2,124 1 .38746
7,6,127 1.33632
8.5,1 29 1 .3551 I
9,4.133 1.37135
1 1 .1 .133 1.39561
10.3,139 1.38468
12,0,144 1.39963
7,? ,147 1.33105
11 ,2,147 1.39500
8,6,148 1.34921
9,5,151 1.36591
10.4.156 1.38034
12.1,157 1.40204
1 1 ,3,1 63 1.39237
8,7,169 1,34271
13,0.1 69 1 .40556
9,6.171 1 .35994
12,2,172 1 .40141
1 0,5,1 75 I .3751 0
1 1 ,4,181 1.38820
13,1 ,1 83 1.40762
12,3,'l 89 1 .39872
8,8.192 1.33727
9,7,193 1.35388
10,6,1 96 1 .36933
14,0,196 1.41118
13.2,199 1.40689
1 1,5,201 1.38320
12,4,208 1.39502
14,1,211 1 .41242
9,8,217 1.34755
13,3,217 1.40456
.10,7 ,219 1.36336
1 1 ,6,223 1.37760
15,0,225 1.41513
14,2,228 1.41175
12,5,229 1.39023
13,4,237 1.40091
15.1 .241 1 .41665
9,9,243 1.34214
10,8,244 1.35740
11 ,7,247 1.37179
14,3,247 1.40948
12,6,252 1.38485
16.0.256 1 .41905
ab aab
1.00000 1.00000
1.r1359 1.11359
1.13083 1.13083
1.18183 1.19856
1.19157 1.24946
1.16113 1.21851
1.21181 1.26452
1.22320 1.28427
1.19353 1.30365
1.23680 1.32964
1.24762 1.37c30
1.18923 1.277A7
1.21108 1.3206t
1 .24315 I .36684
1.24S5t1 1.38673
1.20977 1.34228
1.24383 1.38847
1.25917 1.40010
1.17202 1.31656
1.20654 1.38217
1.27726 1.43415
1.23899 1.39743
1.25913 1.12528
1.21802 1.37364
1.24339 1.40326
1.27429 1.43904
1.27157 1.43692
1.27672 1.44804
1.17818 1.34730
1.21106 1.38227
1.23958 1.41517
1.29328 1.46690
1 .26829 1 .44191
1.22025 1.38651
1.27877 1.45948
1.24941 1.42016
1.26717 1.44217
1.29105 1.46964
1.28830 1.46704
1.18737 1.36331
1 .21546 1.39640
1.29037 1.47419
1.24493 1.42596
1.26513 1.44955
1.30332 1.48793
1.28593 1.47007
1.22538 1.39977
1.247?5 1.42502
1.27029 1.45194
1.29137 1.48172
1 .28270 1.46827
1 .30162 1.48965
1.19412 1.37809
1.29926 1.48725
1.21951 1.40561
1.24396 1.43150
1.25717 1.45587
1.29958 1.4919E
1.28211 1.47309
1.22819 1.40607
1.31020 1.50255
1 .24901 1.43238
1.29748 1.48923
1.26S12 1.45274
1.28472 1.47439
1.30011 1.49729
1.29345 1.48668
1.19927 1.38685
1.22130 1.41251
1.24488 1.43679
1.30888 1.50371
1.30691 1.50165
1.26513 1.45703
1.28241 1,477'.13
1.30622 1.50483
1,22885 1.41332
1.29355 1.48999
1.24727 1.43458
1.26638 1.45679
1.31521 1.51331
1.305s3 1.50302
1.27973 1.47317
1.29506 1.49083
1 .30650 1.50877
1.20335 1.39529
1.22379 1.41775
1.24444 1.43906
1 .30126 1 .50025
1.26333 1.4601 1
1.31416 1.51414
bb ba bba
1.00000 1.00000 1.00000
1.11359 1.11359 1.11359
1.13083 1.13083 1.13083
1.14872 1.15205 1.17867
1.15561 1.15561 1.15581
1.24287 1.19636 1.21?87
1.207 12 1.23A77 1.25548
1.21330 1.16433 1.21330
1.2127 4 1.22739 1.28484
1.21eSo 1.25075 1.28257
1.200n 1.16837 1.21353
1.28284 1.30666 1.3734+
1.23411 1.26208 1.31974
1.25212 1.28739 1.36375
1.19183 1.19204 1.22989
1.25792 1.29925 1.37253
7.2*172 1.27431 1.33141
1.L*52 1.2es63 1.35218
1.30096 1.33039 1.,10695
1.25831 1.33205 .1.40630
1.242'.t9 1.21105 1.29437
1.27991 1.29669 1.97987
1.24699 1.31 108 1.36513
1.27920 1.32339 1.40508
1.27417 1.33333 1.40354
1.23332 1.22658 1.29400
1.27143 1.30729 1.37631
1.27664 1.31590 1.40210
1.31 132 1 .36706 1 .45028
1.2729A 1.34821 1.42944
1.29170 1.35468 1.44140
1.22682 1.23947 1.30031
1.26519 1.30s87 1.37509
1.29192 1.35545 1.44188
1.26661 1 .32541 1.39636
1.29384 1.35238 1.42968
1.28288 1.35216 1.42945
1.25867 1.25033 1.33946
1.28930 1.32079 1.41120
1.31793 1.37792 1.46475
1.28664 1.37837 1.46421
1.25826 1.32805 1.39292
1.29887 1.36102 1.45087
1.27809 1.36277 1.43609
1.24949 1.25950 1.33881
1.28035 1.323s2 1.40107
1.30050 1.36692 1.45705
1.30512 1.37345 1.45688
1.29105 1.36556 1.44846
1.28106 1.33425 1.42476
1.29817 1.36230 1.45356
1.24220 1.28760 1.33685
1.32251 1.39621 1.4861 1
1.27300 1.32816 1.40085
1.29603 1.38653 1.47529
1.30618 1.38831 1.48002
1.28907 1.36418 1.44505
1.27193 1.33624 1.41403
1.28981 1.36902 1.45091
1.30672 1.38485 1.477AO
1.26617 1.27457 1.36856
1.312s0 1.38588 1.47212
1.29141 1.33195 1.42509
1.296s6 1.38229 1.46973
1.30471 1.37198 1.46493
1.26430 1.34051 1.41214
1.28321 1 .36734 1.44365
1.32585 1.40240 1 .49410
1.30321 1.40260 1.49372
1.31093 1.39248 1.48592
1.25797 1.28069 1.38386
1.28335 1.33364 1.41713
1.29998 1.39170 1.17602
1.29674 1.37056 1.45555
1.28270't.34191 1.43426
1.31145 1.39188 1.48572
1.29929 1.37225 1.46587
1.31818 1.39593 1.48578
1.30219 1.39350 1.48309
1.25112 1.28e12 1.36227
1.27549 1.33706 1.41421
1.30973 1.38846 1.48288
1.29122 1.37376 1.45544
1.27483 1.34516 1.42718
1.32841 1.41335 1.50679
1.30849 1.40751 1.50021
1.31470 1.40798 1.SO2S2
1.29175 1.37108 1.45600
1.30726 1.39368 1.48136
1.27033 1 .29096 1.38580
cb cab radial
1.00000 1.00000 1.00000
1.11359 1.11359 1.11359
1.13083 1.13083 1.13083
1.15002 1.14972 1.22'.116
1.15663 1.15663 1.18300
1.15997 1,15997 1.26645
1.20104 1.21874 1.29539
1.17850 1.17850 1.28333
1.17827 1.21569 1.35273
1.20996 1.26795 1.31695
r.19603 1.22075 1.32022
1.1 6863 1 .25517 1 .37358
1.20125 1.?5732 1.38416
1.22220 1.31 034 1.36635
1.20582 1.24035 1.33255
1.18610 1.26659 1.41652
1 .21 101 1.27248 1.39182
1.22188 1 .3170S 1r3t1520
1.17166 1.27268 1.42782
1.20248 1.31079 1.43297
1.21 186 1 .26372 1.36421
1 .21636 1.30180 1.41838
1.22861 1.33952 1.39008
1.18633 1.28249 1.45316
1.21009 1.31773 1.43591
1.21587 1.27568 1.37828
1.22078 1.31252 1.42829
1.22628 1.34138 1.40906
1.17307 1.31451 1.46012
1.19832 1.31309 1.46309
1.21832 1.34390 1.45185
1 .21868 1 .29045 1 .3831 6
1.22326 1.32535 1.42ss1
1.18551 1 .31 144 1.47674
1.23105 1.35539 1.11704
1.20631 1.32102 1.46422
1.22123 1.34703 1.45743
1.22073 1.29847 1.39818
1.22665 1.33327 1.44127
1.17383 1.33165 1.48144
1.19626 1.33425 1.48334
1.22833 1.35539 1.41880
1.21239 1.33894 1.47464
1.22620 1.36345 1.45706
1.22228 1.30863 1.40548
't,22812 1.33974'.|.44570
1.18457 1.32482 1.49316
1.20326 1.33694 1.48371
1.21663 1.34708 1.47800
1.23212 1.36516 1.42863
1.22691 1 .36443 1.46468
1.22348 1.31437 1.40803
1.17429 1.35269 1.49653
1.22968 1.34568 1.44570
1.19364 1.33521 1.49784
1.21031 1.35709 1.49098
1.22036 1.35484 1.47696
1.22941 1.36440 1.43296
1.23037 1.37546 1.46709
1.18372 1.34121 1.50523
1.22443 1.32178 1.41674
1.20047 1.33989 1.49787
1.23062 1.34946 1.45217
1.21431 1.35846 1.49310
1.22333 1.36221 1.48212
1.23261 1.37173 1.43377
'a.23002 1.37532 1.46617
1.17459 1.36163 1.5A776
1.19202 1.34626 1.50872
1.20632 1.35425 1.50320
1.2252A 1.32609 1.42118
1.23134 1.35397 1.45460
1.21903 1.37497 1.49183
1.22523 1.36516 1.48337
't.23005 1.37063 1.43974
1.18299 1.34866 1.51447
1.23268 1.38320 1.47050
1.19816 1.34729 1.50859
1.21073 1.36433 1.50459
1.225W 1.33173 1.42273
1.23199 1.35651 1.45433
1.22122 1.37888 1.49546
1.22719 1.37286 1.48201
1.23283 1.37641'.t.44245
1.17480 1,37434 1.51644
1.19032 1.34785 1.51716
1.20425 1.36867 1.51254
1.23180 1.38247 1.47173
1 .21468 1.37646 1.50328
1.22636 1.33509 1.42841
1.00000
1.1 1359
1 .1 3083
1.1 493 1
1.15598
1.15997
1.2a6tt
1.16480
1,2A734
1.24t58
1.1 8731
1.25517
1.2450E
1.28344
1.21179
1,26211
1.25904
1.29402
1.28998
1.30137
1.231 18
1.28488
1.31 192
1.27882
1.30710
1.24683
1.29691
1.31788
1.30052
1.30745
1.32981
1.25970
1.30699
1.30204
1.32853
1 .31467
1.33335
1.27046
1.31 684
1.30712
1.32934
1.33230
1.32964
1.347'.17
1.27958
1.32106
1.30950
1.33087
1.33302
1.33932
1.34927
1.28740
1.32207
1.32908
1.33203
1.34894
1.34351
1.34190
1.35817
1.32208
1.29418
1.33612
1.33235
'!.34995
1 .35160
1.34685
1.35943
1.32577
1.34330
1.34850
1.3001 1
1.33743
1.36221
1.35276
1.34873
1.32625
1.36546
1.34344
1.35421
1.30535
.34043
.36275
.36003
.35239
.33462
.34475
.35868
.36623
.35965
.31000
International Journal of Space Structures Vol' 5 Nos' 3&4 1990 245
7. Conclusions
A practical approach has been presented for
understanding all three classes of geodesic domes
in one general framework. This framework gives
us the abiliry to extend class specific tessellation
methods to cover all classes. Also presented were
ten methods to generate tessellations, including
the ones most often used by designers. A small
comparative study was provided to illustrate the
effect of the different methods on the different
number of faces, edges and l/s ratios generated
for each.
References
l. COXETER, H.S.M., "Virus Macromolecules and
Geodesic Domes", A Spectntm of Mathematia, €d. J.C.
Butcher, Auckland University Press and Oxford
Universiry Press, 1972, PP. 98-107.
2. WENNINGE& M. J., Spheical Models, Cambridge
University Press, 1979, p. 98-100, 120'l2l-
3. KITRICK CJ., "Geodesic Domes",,Srructural Tbpologt,
University of Montreal, 11, 1985, pp. I 5-20.
Unt-liecl Approach to Geodesic Domes
TARNAI, T., "Geodesic Domes with Skew Networks",
Spheical Grid Structures, ed. T. Tarnai, Hungarian
Institute for Building Science. 1987. .
IVIAIQ{I, 8., "On Polyhedra with Approximately Equal
Edges", Spherical Grid Stntctures, ed. T. Thrnai,
Hungarian Institute for Building Science, 1987.
CLINTON, J.D., Advanced Stnrctural Geometry Studies,
Part I, Polyhedral Subdivision Concepts for Structural
Applications, NASA CR-1734, Sept. 1971.
GOLDBERG, M., '.[ Class of Muiti-Symmetric
Polyhedra", Tbhoku Mathematical Journal, 43, Tokyo,
Japan, 1937, pp. lM-108.
MAKAI, E., and TARNAI, T., "IJniformity of Networks
of Geodesic Domes", Spheical Gid Stntctures, ed. T.
Tarnai, 1987, Hungarian Instinrte for Building Science.
SCHEEL, H., "Geodesic Surface Division", The Cana'
dian Architect, 14, No. 5, 1969, pp. 61-66.
STUART, D.R., A Report on the Tiiacon Gridding Sys-
tem for Spherical Surface, Skybreak Carolina Corpora-
tion, 1952.
TA-RNAI, T., "spherical Grids of Tiiangular Nenvofk",
Acta Tbchnica Acad. Scl Hungar., 7 6, 197 4, pp- 307 -
336.
PAVLOV, G.N., "Compositional Form-shaping of
Crystal Domes and Shells", Spherical Gid Stntctures, ed.
T. Tarnai, 1987, Hungarian Institute for Building
Science.
4.
5.
6.
7.
8.
9.
10.
11.
12.
246 International Journal of Space Stntctures Vol. 5 Nos- 3&4 1990
... From another point of view, today, creating elegant surfaces by repeating a few different elements is of great importance in many novel architectural applications (Grünbaum & Shephard, 1987;Pottmann et al., 2015;Javan et al., 2021). Many studies have investigated the spherical subdivisions from the perspectives of the spherical polygons' shape, area, or edge length (Kitrick, 1990(Kitrick, , 2015Snyder, 1992;Teanby, 2006;Harrison, 2012;Massey, 2012;Popko, 2012;Lee & Mortari, 2017;Gáspár, 2021;Lee et al., 2022). Spherical polygons are created on the surface of a sphere from several great circle arcs. ...
... On the other hand, spherical subdivisions with identical edge lengths provide cost-effective construction through the mass production of repeating components (Kitrick, 1990(Kitrick, , 2015. ...
Article
Full-text available
Dividing a sphere uniformly into equal-area or equilateral spherical polygons is useful for a wide variety of practical applications. However, achieving such a uniform subdivision of a sphere is a challenging task. This study investigates two classes of sphere subdivisions through numerical approximation: (i) dividing a sphere into spherical polygons of equal area; and (ii) dividing a sphere into spherical polygons with a single length for all edges. A computational workflow is developed that proved to be efficient on the selected case studies. First, the subdivisions are obtained based on spheres initially composed of spherical quadrangles. New vertices are allowed to be created within the initial segments to generate subcomponents. This approach offers new opportunities to control the area and edge length of generated subdivided spherical polygons through the free movement of distributed points within the initial segments without restricting the boundary points. A series of examples are presented in this work to demonstrate that the proposed approach can effectively obtain a range of equal-area or equilateral spherical quadrilateral subdivisions. It is found that creating gaps between initial subdivided segments enables the generation of equilateral spherical quadrangles. Secondly, this study examines spherical pentagonal and Goldberg polyhedral subdivisions for equal area and/or equal edge length. In the spherical pentagonal subdivision, gaps on the sphere are not required to achieve equal edge length. Besides, there is much flexibility in obtaining either the equal area or equilateral geometry in the spherical Goldberg polyhedral subdivisions. Thirdly, this study has discovered two novel Goldberg spherical subdivisions that simultaneously exhibit equal area and equal edge length.
... Equal area representations are highly beneficial for geographical data collection and information monitor- ing, as the region of interest in a particular part may share the same proportion to all other parts of the globe [11,25,30]. Equal length representations offer design possibilities to realize cost-efficient constructions by assembling mass-produced repeated components [19,31,32]. However, realizing a complex form using a single length can be challenging. ...
Article
Full-text available
Dividing a 2-dimensional sphere uniformly into a large number of spherical polygons is a challenging mathematical problem, which has been studied across many disciplines due to its important practical applications. Most sphere subdivisions are achieved using spherical triangles, quadrangles, or a combination of hexagons and pentagons. However, spherical pentagons, which may create elegant configurations, remain under-explored. This study presents a new sphere subdivision method to generate a large number of spherical pentagons based on successively subdividing a module of an initial spherical dodecahedron. The new method can conveniently control the shapes of generated spherical pentagons through specified design parameters. Two optimization problems have been investigated: (I) dividing a sphere into spherical pentagons of equal area; (II) minimizing the number of different arc lengths used in the pentagonal subdivision. A variety of examples are presented to demonstrate the effectiveness of the new method. This study shows that treating the mathematical challenge of dividing a sphere uniformly into a large number of spherical polygons as an optimization problem can effectively obtain equal area or equal length sphere subdivisions. Furthermore, considering additional constraints on the optimization problem may achieve sphere subdivisions of specific characteristics.
... In order to limit the length of the frames in the geodesic network and control their excessive slenderness ratios as the diameter of the dome increases, the equilateral triangles are divided into a number of subdivisions. According to Kitrick (1990), a Class I subdivision was used in this work, with the dividing lines parallel to the edges of the primary frame network, as shown in Figure 2(b). A V2 subdivision was implemented, meaning that the edges of the original geodesic network were divided into two triangles, as shown in Figure 2(c). ...
Article
Full-text available
This paper studies the feasibility of using geodesic domes as expeditionary protective structures, since they are fast assembly lightweight structures and able to withstand very heavy loads due to their specific geometry. A 6 m standard ISO container is used as a benchmark structure for the study which is developed using the nonlinear explicit code LS-DYNA. The blast simulations were performed with the Load Blast Enhanced option, using the phenomenological Johnson-Cook constitutive model to reproduce the behavior of steel subjected to a large range of strain rates. The performance assessment of the protective structures under blast loading was based on the evaluation of the associated safety zones inside the protective shelter as well as on the characterization of the potential structural damage. With the additional insights provided by this paper it is possible to identify some of the competitive advantages shown by this type of structures which will hopefully foster new advances in the field.
... For hexagonal grids derived from Class I triangular tessellations, Kitrick [2], the (b,c) values require that b be divisible by 3 and c be equal to zero. There are three complimentary grids used: the base triangular, its dual triangular, and the resultant hexagonal. ...
Conference Paper
Full-text available
The paper presents an improved methodology for creating spherical configurations that are comprised entirely of plane hexagonal faces. Hexagonal based tessellations are of particular interest since they allow the construction of shells with considerably less edge joints, non-acute face angles, and smaller dihedral face angles. Plane hexagons eliminate material bending geometry that arise from un-optimized hexagonal configurations.
... Geodesic tessellations are typically derived from projective and spherical division methods that provide a wide range of component variability. The computational requirements are straightforward and can be applied to all three topological classes [1][2]. This paper introduces the concept of quasi-regular geodesic tessellations using chains of similar triangles or polygons. ...
Conference Paper
Full-text available
This paper introduces the concept of quasi-regular geodesic tessellations using chains of similar triangles or polygons. This method creates unique tessellations with a limited set of member components. The seed triangle or polygon when chained comprises a significant portion of the spherical surface. The interstitial areas between chains can be appropriately stitched to complete the tessellation tiling. The general approach can be used on regular polyhedron (multi-axis: icosahedron, octahedron, etc.) or single axis spherical configurations. This approach can generate a wide range of new tessellations that greatly expands the palette of possible spherical tiling. Chaining is the method of taking a seed triangle or polygon and extending it spherically by adding the same triangle or polygon such that faces are fully adjacent and maintain a fixed dihedral angle between all the faces of the chain. Given a sphere of radius r, the size of the seed determines how long the chain can be without overlapping itself. By adjusting the size of the seed, the chain can be made to meet symmetrical or overlap requirements. The chain can extend in multiple directions as long as overlap and planar conditions are avoided. When chaining is used on regular polyhedra the fixed spherical points of symmetry are maintained while the overall tessellation grid adjusts it orientation rotationally. This rotational effect results in tessellations that can be closed but are enantiomorphic (handed). On single axis spherical forms, tessellation grids based on chains can originate at the pole or be bound to the equator and are also enantiomorphic. The essential chaining paradigm can be extended to include the use of multiple seed polygons that are of different sizes and result in multiple dihedral angles. Multiple seed poylgons provide some additional flexibility to route the tiling of polygons across the spherical surface while restricting the number of different polygons. This paper provides both the approach and the basic computational methodology while illustrating some of the quasi-regular tessellations of both multi and single axis spherical forms as examples of what designs can be achieved. This quasi-regular geodesic patterning flexibility introduces a whole new aesthetic tessellation option to the architect, engineer, or designer to explore with color, lighting, and materials to enhance visual impact. Figure 1 illustrates three icosahedral based tessellations that utilize the chaining of pentagonal triangle groups with complete and incomplete interstitial grid stitching.
Conference Paper
Full-text available
The paper presents a trigonometric methodology for creating hexagonal spherical configurations that are comprised of a single edge length. Hexagonal based tessellations are of particular interest since they allow the construction of shells with considerably fewer edges that may be used for pneumatic nets or tensegrity based shells.
Article
Full-text available
The definition of the shape of the enclosures of large spaces by means of large span structures, where the absence of intermediate supports is a strong conditioning factor, is an interdisciplinary activity where geometry, biology, topology, architecture and engineering have complemented themselves. A relatively recent discipline, COMPUTATIONAL GEOMETRY, has permitted a new formulation of the geometric basis and the numerical procedures that permit to generate a spatial dome, regardless of its type. The geometric and topologic configuration of any Spatial Mesh or Structure (including typologies like Lattice, Geotangent or any other patented or published structural form) does not suppose anything else than the creation of a polyhedron that approximates the shape of the ideal surface. Thanks to the methods of Computational Geometry, we are able to demonstrate that this problem has a purely and exclusively two-dimensional nature and treatment.
Article
Full-text available
The double cubic lattice method (DCLM) is an accurate and rapid approach for computing numerically molecular surface areas (such as the solvent accessible or van der Waals surface) and the volume and compactness of molecular assemblies and for generating dot surfaces. The algorithm has no special memory requirements and can be easily implemented. The computation speed is extremely high, making interactive calculation of surfaces, volumes, and dot surfaces for systems of 1000 and more atoms possible on single-processor workstations. The algorithm can be easily parallelized. The DCLM is an algorithmic variant of the approach proposed by Shrake and Rupley (J. Mol. Biol., 79, 351–371, 1973). However, the application of two cubic lattices—one for grouping neighboring atomic centers and the other for grouping neighboring surface dots of an atom—results in a drastic reduction of central processing unit (CPU) time consumption by avoiding redundant distance checks. This is most noticeable for compact conformations. For instance, the calculation of the solvent accessible surface area of the crystal conformation of bovine pancreatic trypsin inhibitor (entry 4PTI of the Brookhaven Protein Data Bank, 362-point sphere for all 454 nonhydrogen atoms) takes less than 1 second (on a single R3000 processor of an SGI 4D/480, about 5 MFLOP). The DCLM does not depend on the spherical point distribution applied. The quality of unit sphere tesselations is discussed. We propose new ways of subdivision based on the icosahedron and dodecahedron, which achieve constantly low ratios of longest to shortest arcs over the whole frequency range. The DCLM is the method of choice, especially for large molecular complexes and high point densities. Its speed has been compared to the fastest techniques known to the authors, and it was found to be superior, especially when also taking into account the small memory requirement and the flexibility of the algorithm. The program text may be obtained on request. © 1995 by John Wiley & Sons, Inc.
Geodesic Domes with Skew Networks
  • T Tarnai
TARNAI, T., "Geodesic Domes with Skew Networks", Spheical Grid Structures, ed. T. Tarnai, Hungarian Institute for Building Science. 1987..
Advanced Stnrctural Geometry Studies, Part I, Polyhedral Subdivision Concepts for Structural Applications, NASA CR-1734
  • J D Clinton
CLINTON, J.D., Advanced Stnrctural Geometry Studies, Part I, Polyhedral Subdivision Concepts for Structural Applications, NASA CR-1734, Sept. 1971.
IJniformity of Networks of Geodesic Domes
  • E Makai
MAKAI, E., and TARNAI, T., "IJniformity of Networks of Geodesic Domes", Spheical Gid Stntctures, ed. T.
Geodesic Surface Division
  • H Scheel
SCHEEL, H., "Geodesic Surface Division", The Cana' dian Architect, 14, No. 5, 1969, pp. 61-66.
A Report on the Tiiacon Gridding System for Spherical Surface, Skybreak Carolina Corporation
  • D R Stuart
STUART, D.R., A Report on the Tiiacon Gridding System for Spherical Surface, Skybreak Carolina Corporation, 1952.
Compositional Form-shaping of
  • G N Pavlov
PAVLOV, G.N., "Compositional Form-shaping of