See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/280233816

Structures and Geometry in Design Processes

Article · January 2012

CITATIONS READS

0 1,722

1 author:

Cornelie Leopold
RPTU - Rheinland-Pfälzische Technische Universität Kaiserslautern Landau
123 PUBLICATIONS 295 CITATIONS

SEE PROFILE

All content following this page was uploaded by Cornelie Leopold on 05 March 2016.

The user has requested enhancement of the downloaded file.

Cornelie Leopold

Structures and Geometry in Design Processes

Structural thinking in architecture is based on structuring information with the help of
mathematical-geometric methods. Some of those methods are presented in this paper which
had been one of the background for the DAAD Summer School "Structure – sculpture" in
Buenos Aires, where students worked on the design task analyzing and redesigning the Ulm
Pavilion by Max Bill. An overview of several mathematical-geometric approaches will be
given accompanied by examples of teaching practice at Ulm School of Design in Germany,
an innovative international school experiment between 1953 and 1968, where those methods
had an important impact.

1. Mathematics as a Structural Science

Our approach in structural thinking is looking for the relations between the elements of a
system and tries to formulate rules of their combinations in the system. Therefore we have to
ask: What are our elements out of which system and how could we structure the given or
collected information? Gordon Best explained three stages of the design process:
unstructured information, structured information and design solution.

Figure 1: Three stages of the design process (1), see (2) p. 60

There had been interesting efforts to base design methods on comprehensible rational
methods. The development of mathematics as a structural science according the project
Bourbaki had been picked up in the 1960s to structure design processes systematically.

Particular Structures
mathematical special fields and models
Example
Euclidean Plane Geometry

Multiple Structures

Example Geometry

Mother Structures (Basic Stuctures)

order structures algebraic structures topological structures
Composition, Relations, Grid, ... Operation with characteristics connectedness, neighbourhood,
Group, field, vector space, ... ...

Figure 2: Mother, multiple and particular structures of mathematics according Bourbaki (3), for more
details see (4)

1
With Bourbaki's structural understanding of mathematics the general applicability of
mathematics especially in a methodological sense had been given. According Bourbaki,
mathematics refers to basic structures, so called the mother structures: structures of order,
algebraic structures and topological structures. Multiple structures and mixed structures then
create special structures, which represent the mathematical single disciplines. Mathematics
as a general structural science is founded on the notions set, relation and transformation. If
we define a relation in a set of elements we get an order structure. In geometry for example,
we are dealing with the elements points, lines and planes as mental facts. In the classical
Euclidean geometry these notions are combined with ideas of the space of our experience,
the real space. Later with David Hilbert it is enough to state three sets of things where
relations between them are defined (for example incidence, order, etc.).

The philosopher Max Bense, Professor at University of Stuttgart and at Ulm School of
design, used this understanding of mathematics to develop an exact aesthetics as a basis for
designing. Elisabeth Walther, also professor for philosophy at University of Stuttgart and
teacher at Ulm School of Design, described such aesthetics as fundamental for techniques
as well as architecture and art: "Aesthetics, as Bense brings it into play, is the principle of
order par excellence. Aesthetics is order, and order on the other hand is describable by
mathematics. Therefore aesthetics is important as structuring the world for techniques as
well as architecture, literature, etc., for all what will be created. When ever we take
something out of the chaos of existing und assemble it new, we need an aesthetic
foundation." (5)

We will have a look at different kind of mathematical structures usable for aesthetics in
design and architecture and start with the most general ones.

2. Graph Theory and Topology

A graph is an ordered pair G = (V, E) comprising a set V of vertices or nodes together with a
set E of edges or lines, which are 2-element subsets of V. There may be distinguished
various types of graphs, also called network types: tree-structures, centralized, decentralized
and distributed networks.

The wellknown problem of the Seven Bridges of Königsberg is a historically notable problem
in mathematics. Its negative solution by Leonhard Euler in 1735 laid the foundations of graph
theory and prefigured the idea of topology. Euler found that the multigraph on four vertices
and seven edges has no Eulerian cycle, which is a trail that starts and ends at the same
graph vertex. Euler showed without proof that a connected graph, that means there is a path
from any point to any other point in the graph, has a Eulerian cycle if and only if it has no
vertex of odd degree.

Figure 3: Bridges Problem of Königsberg and its graph (6)

Graph theory had been used in architecture for analysing floor plans by representing space
and room linkages with the help of graphs.

2
March and Steadman (7) showed in a fascinating example using fruitfully graphs for
architecture. The flour plans of three houses by Frank Lloyd Wright had been analysed and it
has been found out, that all three flour plans can be represented by the same graph and are
therefore topological equivalent. The abstraction to room linkages makes it possible to
develop a general applicable flour plan concept, which can be transformed in different forms.

Figure 4: Three house project by Frank Lloyd Wright a. Life, 1938, b. Ralph Jester House, 1938,
c. Vigo Sundt House, 1941 and graph of space and room linkages (7)

Such a way of graph theoretical analyse was taught at Ulm School of Design in the
foundation course in the subject "visual methodology" by Anthony Froshaug (8) in 1959. He
introduced in a systematic way the different possible grids as part of graph theory, not under
a metrical aspect, although they are developed out of the concept to fill the plane with one or
more kinds of polygons. In a graph theoretical aspect they are characterized by the edges
meeting in one vertex. He presented the regular and semiregular grids as the most important
ones.

Figure 5: Three regular and eight semiregular grids (8)

3
He explains the role of graphs for the visualization of circulations by the example of Casa
Curutchet in La Plata, Argentina, designed by Le Corbusier in 1954. After finishing the
building Le Corbusier had to remodel the house because the arrangement of the rooms was
inappropriate. Froshaug suggests writing down the possible ways of circulation in a matrix,
after numbering the rooms. Then it can be seen that the maximal row sum is five. Therefore,
a grid where at least five lines meet in each point is necessary to represent the circulations
by a graph. From all regular and semiregular grids shown in figure 5 the appropriate type
was selected and then used to draw the graph of the circulation ways in the house. Froshaug
states that the visualization of the room linkages by graphs during designing or to start with
the graph in the design process would be a good way to avoid the later necessary
remodelling.

Figure 6: Floor plans of Casa Curutchet, matrix and graph of circulations

4
The graph theory was the prefigured idea of topology, a discipline where only neighbourhood
and connectedness had been taken into account. Properties of objects that are preserved
under continuous deformations are studied in topology. Also topological thinking had been
applied in designing at Ulm School of Design. Tomás Maldonado, teacher and rector at Ulm
School of Design, explains that topology allows to look at structure in a completely other way.
He reflects in the essay "Wissenschaft und Gestaltung" 1964 (9) together with Gui Bonsiepe
the role of topology. The designer could be prepared by topology to approach a problem in
another way, that there are not only problems of dimensions, forms and positions, but also
those of order, continuity and neighbourhood. But they realised that the application of
topology in the practice of product design was not satisfying. The examples of non-orientable
surfaces remained exercises. Figure 7 shows examples of topological exercises in the
foundation course by Gui Bonsiepe. First the students had to create a table with typical
cutting forms in a plane figure. Then the students had to develop a non-orientable surface by
torsion like the example by student Marlies Matthies.

Figure 7: Topological exercise in the foundation course by Gui Bonsiepe, student Marlies Matthies,
Ulm School of Design 1966 (10)

Josef Albers referred already to topological thinking in the folding experiments in the foundation
course 1955 at Ulm School of Design (Figure 8).

Figure 8: Folding experiments by Josef Albers, Ulm School of Design 1955 (11)

3. Tessellations and Patterns

Another more classical geometric approach is based on metric Euclidean geometry, where
triangles, squares, in general polygons are defined. Classical methods to develop ornaments
for example in Islamic Art are subdividing a square as we can see for example in the
drawings of El-Said in Figure 9. The system is based on "root-2" proportion system.
Drawings from Hans G. Conrad, student at Ulm School of Design in 1954 show the use of
these methods with their transformations. This root-2 proportion system gives us the
background to understand the design of the Ulm Pavilion by Max Bill (Figure 10).

5
Figure 9: Classical development of ornaments based on root 2 system (12) and drawings by Hans
G. Conrad, Ulm School of Design 1954 (13)

Figure 10: Divisions of a square according root 2 system and top view of the Ulm Pavilion

The underlied tessellation is the regular tessellation by squares. Looking for all possible
regular and semiregular tessellations of the plane gives the geometric basis for designing.
Besides the regular or Platonian tessellations by hexagons, squares, and triangles we have
eight semiregular or Archimedian tessellations that are characterized by the order surround
each polygon vertex that all have the same order. The graphs used by Anthony Froshaug in
the foundation course at Ulm School of Design, shown in figure 5, may be now seen under
the view of added measurement in refer to tessellation theory. When we take the midpoint of
each polygon, replace the plane figure with a point and combine these points with lines we
get the dual tessellations. In this way we can develop dual tessellations out of the
tessellations. The number of lines in one point (vertex) produces the number of vertexes of
the resulted polygon. The consequence is that the dual tessellations of the semiregular
tessellations do not consist of regular polygons any more, for example Caro-tiling consists of
non-regular pentagons.

Figure 11: Regular and semiregular tessellations and their duals (14)

6
The students at Ulm of Design worked on tessellations and patterns and developed one
pattern in another, called net transformations or metamorphosis.

Figure 12: Net transformations by student Arno Caprez at Ulm of Design 1965/66, teacher William
S. Huff (17)

Tessellations and then more general all patterns can be characterized by symmetry
transformations like translation, rotation, reflection, and glide reflection. Then we get the
result of exactly 17 different plane symmetry groups, also called crystallographic groups.
Such a group fulfils the conditions of an algebraic group. The symmetry group of the pattern
is the set of all isometries (transformations translation, rotation, reflection, and glide
reflection) that map the pattern onto itself. The symmetry group of a periodic pattern
necessarily maps a lattice associated to the pattern onto itself. In refer to the symmetry
groups, the Ulm Pavilion can be related to symmetry group p4, where the translation cell is a
square, which we get by rotation of order 4 of the fundamental cell module. Figure 13 shows
additionally a wickerwork example according symmetry group p4.

Figure 13: Symmetry group p4 (15), two possible fundamental cell modules, top view of Ulm Pavilion
and wickerwork example (16)

The spatial structure of the pavilion can be also understood by rotation of order 4. The fundamental
spatial cell of the Ulm Pavilion consists of the floor part combined with the inclined roof part, which is
twisted by 90° in relation to the floor part.

Figure 14: Spatial element of Ulm Pavilion und result after rotation of order 4

7
4. From Plane Pattern to Spatial Configuration

Extrusion is the easiest way to come from a plane pattern to a spatial configuration. Figure
15 shows space units for residential buildings in the architectural design class at Ulm School
of Design 1961 by Bernd Meurer and Herbert Ohl.

Figure 15: Space units for residential buildings at Ulm School of Design by B. Meurer and H. Ohl (18)

Folding is an intuitive experimental way to come from a plane pattern to a spatial

configuration. Especially Josef Albers included folding exercises in his foundation course at
Ulm School of Design. We can explore folded structures by series of folding experiments with
a square: zig-zag folding, remaining in a plane, results out of parallel folding lines, whereas
reflecting the folding lines produces a bent object. By translating the first folding in one
square (diagonals) by half a unit we get a very flexible folded structure, which can produce
finally the magic ball. Tessellation origami can twist the figures, the structure remains plain,
but various layers are developed, over and under are defined and a p4 pattern like the floor
plan of the Ulm Pavilion is created.

Figure 16: Series of folding experiments

The artist Lygia Clark from Brazil developed her sculptures by folding in the sixties. She had
one of the first exhibitions in the Gallery at University of Stuttgart 1964, which had been
organized by Max Bense. She experimented with various spatial configurations out of the
same plane pattern.

8
Figure 17: "Bichos" pattern and sculpture by Lygia Clark and Max Bense opening the exhibition in
the Gallery at University of Stuttgart 1964

Another way to become spatial from plane figures is making solids out of polygons. There
are five possibilities to produce totally regular solids, the Platonic solids. Also in space we
can use the principle of duality to bring the Platonic solids in relations to each other pair wise.
According to tessellation of the plane we can then try to fill the space with solids. When we
try this with the platonic solids, we are only successful with the cube. When we take two
kinds of platonic solids, octahedron and tetrahedron, a space filling is possible. Anthony
Froshaug included in visual methodology also studies about spatial regular and semiregular
lattice structures. Besides the cubes we get according Froshaug the following regular
tessellations (Figure 18), where the same number of edges meet in each node.

Figure 18: Regular lattice structures by Anthony Froshaug, Ulm School of Design 1959

9
The mentioned tetrahedron-octahedron lattice belongs to the semiregular lattices (Figure 19).
Froshaug worked with these structures from the point of graph theory. The most important
characteristics are the edges in one point. This characteristic has to be used to develop
nodes for those lattices. An example of a node by teacher Walter Zeischegg at Ulm Scholl of
Design, 1962/63 and a lattice structure by a student in the course of Günter Schmitz,
1963/64 are shown in Figure 20.

Figure 19: Semiregular lattice structures by Anthony Froshaug, Ulm School of Design 1959

Figure 20: Node example by teacher Walter Zeischegg, Ulm 1962/63 and lattice structure by student
Jörg Stumpp, teacher Günter Schmitz, Ulm 1963/64

10
With graph theory we come again to the beginning of this paper. On the level of topology we
look only for neighbourliness. The classification of the different geometric approaches
according the transformations and their invariants with descending invariants points out the
respective mathematical system as illustrated in the table by March and Steadman.

Figure 21: Transformations and invariants by March and Steadman (7)

Max Bill took the lattices, especially the cube lattice for his Pavilion designs. Besides his
realized pavilions, he designed Swiss Pavilions for the international exhibition 1937 in Paris
and 1939 for the world exhibition in New York (Figure 22). The design for New York was a
sculptural steal lattice where boxes as space cells and circulation paths could be positioned
at some places.

11
Figure 22: Project by Max Bill of a Pavilion for New York, 1939 (20)

Eduardo Catalano (21) took another way to develop spatial structures out of plane patterns.
He took tessellations by quadrangles and assigned various heights to the nodes. By joining
the opposite sides of the quadrangle with lines, a warped surface, a hyperbolic paraboloid, is
created. Catalano developed in this way surfaces out of various patterns. He tried out
different hyperbolic paraboloids out of the same and different patterns. Figure 23 shows
examples based upon combinations of four and six structural units.

Figure 23: Structures based upon combinations of four or six structural units (22)

5. Conclusions

The examples showed different mathematical approaches using structures for design
developments. It gets clear that mathematical concepts as general and abstract structural
ideas give foundation for designing. Max Bense described the overall applicability of
mathematical structures in his book Aesthetica: "It is easy to see, that mathematical
structures are general and abstract, so that they are usable for the reproduction of physical
as well as aesthetic structures." (23)

The approaches using mathematics for designing led perhaps in Ulm sometimes to an
overestimation of mathematics and the projects became self-referential. But these had been
just new ideas at that time. Tomás Maldonado and Gui Bonsiepe had been aware of the
necessity of the interaction of various disciplines and that the creative process can not be
12
totally algorithmized. Maldonado described together with Gui Bonsiepe in the paper
"Wissenschaft und Gestaltung" 1964 the following mathematical disciplines as operable for
the product designer in his design practice:

1. Combinatorics (for modular construction systems and problems of measure coordination)

2. Group theory (as symmetry theory for construction of patterns and grids)
3. Theory of curves (for the mathematical treatment of transitions and transformations)
4. Polyhedral geometry (for the construction of regular, half regular and irregular solids)
5. Topology.

And they conclude: "The array of mentioned mathematical disciplines should not tempt to the
perhaps tendering conclusion, that creative thinking and acting can be totally algorithmized.
Because it is wrong, to reproduce the relationship of the designer to the problems to be
solved in a model of a simple determined system, …" (9).

The given overview wanted to show the use of mathematics as a structural science for
design processes on different levels and contexts. A concept of design mathematics is
presented and discussed at the same time. The references to Ulm School of Design illustrate
the innovative aspects of the way of teaching and learning in design disciplines, which can
be our initial reference point for actual mathematical methods in designing.

References

(1) Best, Gordon: Design Methods in Architecture. London 1969

(2) Ulmer Museum/HfG-Archiv (ed.): ulmer modelle - modelle nach ulm. Hochschule für
Gestaltung Ulm 1953-1968. Ulm, 2003.
(3) Nicolas Bourbaki: Éléments de mathématique. Paris 1939 ff.; L'Architecture des
Mathématiques. 1948.
(4) Leopold, Cornelie: Strukturelles Denken als Methode. In: Medina Warmburg, Joaquín;
Leopold, Cornelie (ed.): Strukturelle Architektur. Zur Aktualität eines Denkens zwischen
Technik und Ästhetik. Bielefeld 2012, pp.9-29.
(5) Interview with Elisabeth Walther. In: Barbara Büscher, Hans-Christian von Herrmann,
Christoph Hoffmann: Ästhetik als Programm. Max Bense / Daten und Streuungen. Berlin
2004, p.72, translated by Cornelie Leopold.
(6) http://mathworld.wolfram.com/KoenigsbergBridgeProblem.html (as of 07/2012)
(7) March, Lionel; Steadman, Philip: The Geometry of Environment. Cambridge (Mass.)
1974, p.27.
(8) Froshaug, Antony: Visuelle Methodik. ulm 4. http://ulmertexte.kisd.de (as of 07/2012)
(9) Maldonado, Tomás; Bonsiepe, Gui: Wissenschaft und Gestaltung. ulm 10/11, 1964.
http://ulmertexte.kisd.de/fileadmin/pdf/WissUndGest_CB_prt.pdf (as of 07/2012)
(10) Bonsiepe, Gui: Dreidimensionale formale Übungen. ulm 17/18. http://ulmertexte.kisd.de
(as of 07/2012)
(11) Krampen, Martin: observation and formulation. Foundation course with exercises, based
on a script by josef albers. DVD, Ostfildern-Ruit, 2009.
(12) El-Said, Issam; Parman, Ayse: Geometric Concept in Islamic Art. 1976
(13) Conrad, Hans G.: Drawings from the foundation course at HfG Ulm 1954. By René
Spitz: www.flickr.com/photos/renespitz/5691315102/in/set-72157628537116743/ (as of
07/2012)
(14) http://mathworld.wolfram.com/DualTessellation.html
(15) http://upload.wikimedia.org/wikipedia/commons/c/c0/Wallpaper_group_diagram_p4.svg
(as of 07/2012)
(16) http://en.wikipedia.org/wiki/File:Wallpaper_group-p4-4.jpg (as of 07/2012)

13
 (17) Lindinger, Herbert (ed.): Hochschule für Gestaltung Ulm. Berlin 1987, p.64.
(18) Ulmer Museum/HfG-Archiv (ed.): ulmer modelle - modelle nach ulm. Hochschule für
Gestaltung Ulm 1953-1968. Ulm, 2003. p.24f.
(19) Clark, Lygia: Bichos. http://sonhosealegrias.blogspot.de/2010/12/os-bichos-de-lygia-
clark-linha-para.html and http://www.bense-und-die-kuenste.de (as of 07/2012)
(20) Büttner, Carsten a.: Max Bill – Homo universalis. Heidelberg 2007. p.79.
http://www.gta-fh-heidelberg.de/inc/GTA-FHHD-Max%20Bill_Buettner_071030.pdf
(21) Catalano, Eduardo: La Constante. Diálogos sobre estructura y espacio en arquitectura.
Cambridge 1996.
(22) Catalano, Eduardo: Structures of Warped Surfaces: Combinations of Units of Hyperbolic
Paraboloids. 1960.
http://aqua-velvet.com/2011/08/structures-of-warped-surfaces-eduardo-f-catalano-1960/
(23) Bense, Max: Aesthetica. Baden-Baden 1965, 2nd edition 1982. p.173.

14

View publication stats