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

net/publication/256059313

Fractal Geometry and Architecture Design: Case Study Review

Article · January 2012

CITATIONS READS

18 1,412

3 authors, including:

Xiaoshu Lu Derek John Clements-Croome

Aalto University University of Reading
38 PUBLICATIONS   731 CITATIONS    290 PUBLICATIONS   3,801 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Whole Life performance of Engineering Assets View project

Flourish Model for designing health and wellbeing in buildings View project

All content following this page was uploaded by Derek John Clements-Croome on 09 September 2020.

The user has requested enhancement of the downloaded file.

 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012

Fractal Geometry and Architecture Design: Case

Study Review
Xiaoshu Lu1,2, Derek Clements-Croome3, Martti Viljanen1
1
Department of Civil and Structural Engineering, School of Engineering, Aalto
University, PO Box 12100, FIN-02150, Espoo, Finland
E-mail: [email protected]
2
Finnish Institute of Occupational Health, Finland
3
School of Construction Management and Engineering, Whiteknights,
University of Reading, PO Box 219, Reading RG6 6AW, UK

Abstract: The idea of buildings in harmony with nature can be traced back to ancient
times. The increasing concerns on sustainability oriented buildings have added new
challenges in building architectureral design and called for new design responses.
Sustainable design integrates and balances the human geometries and the natural ones.
As the language of nature, it is, therefore, natural to assume that fractal geometry could
play a role in developing new forms of aesthetics and sustainable architectureral design.
This paper gives a brief description of fractal geometry theory and presents its current
status and recent developments through illustrative review of some fractal case studies in
architecture design, which provides a bridge between fractal geometry and architecture
design.
Keywords: Fractal geometry, Architecture design, Sustainability.

1. Introduction
The idea of buildings in harmony with nature can be traced back to ancient
Egyptians, China, Greeks and Romans. At the beginning of 21st century, the
increasing concerns on sustainability oriented on buildings have added new
challenges in building architecture design and called for new design responses.
As the language of nature [1,2], it is, therefore, natural to assume that fractal
geometry could play a role in developing new forms of design of sustainable
architecture and buildings.

Fractals are self-similar sets whose patterns are composed of smaller-scales

copied of themselves, possessing self-similarity across scales. This means that
they repeat the patterns to an infinitely small scale. A pattern with a higher

_________________
Received: 29 June 2011 / Accepted 30 March 2012
© 2012 CMSIM ISSN 2241-0503
312 Lu et al.

fractal dimension is more complicated or irregular than the one with a lower
dimension, and fills more space. In many practical applications, temporal and
spatial analysis is needed to characterise and quantify the hidden order in
complex patterns, fractal geometry is an appropriate tool for investigating such
complexity over many scales for natural phenomena [2,3]. Order in irregular
pattern is important in aesthetics as it embraces the concept of dynamic force,
which shows a natural phenomenon rather than mechanical process. In
architecture design terms, it represents design principle. Therefore, fractal
geometry has played a significant role in architectureral design.

In spite of its growing applications, such works in literature are rather narrow,
i.e. they mainly focus on applications for fractal design patterns on aesthetic
considerations. Few works have related to a comprehensive and unified view of
fractal geometry in structural design, for example, as it is intended in this study.
We aim to fill in this gap by introducing fractals as new concepts and presenting
its current status and recent developments in architecture through an illustrative
review of some fractal case studies in design. The paper shows that
incorporating the fractal way of thinking into the architecture design provides a
language for an in-depth understanding of complex nature of architectural
design in general. This study distils the fundamental properties and the most
relevant characteristics of fractal geometry essential to architects and building
scientists, initiates a dialogue and builds bridges between scientists and
architects.

2. Basic Theory of Fractal Geometry

2.1. Basic Theory

The mathematical history of fractals began with mathematician Karl Weierstrass
in 1872 who introduced a Weierstrass function which is continuous everywhere
but differentiable nowhere [4]. In 1904 Helge von Koch refined the definition of
the Weierstrass function and gave a more geometric definition of a similar
function, which is now called the Koch snowflake [5], see Figure 1. In 1915,
Waclaw Sielpinski constructed self-similar patterns and the functions that
generate them. Georg Cantor also gave an example of a self-similar fractal [6].
In the late 19th and early 20th, fractals were put further by Henri Poincare, Felix
Klein, Pierre Fatou and Gaston Julia. In 1975, Mandelbrot brought these work
together and named it 'fractal'.
 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012 313

Fig. 1. Illustration of Koch Curve.

Fractals can be constructed through limits of iterative schemes involving

generators of iterative functions on metric spaces [2]. Iterated Function System
(IFS) is the most common, general and powerful mathematical tool that can be
used to generate fractals. Moreover, IFS provides a connection between fractals
and natural images [7,8]. It is also an important tool for investigating fractal
sets. In the following, an introduction to some basic geometry of fractal sets will
be approached from an IFS perspective. In a simple case, IFS acts on a segment
to generate contracted copies of the segment which can be arranged in a plane
based on certain rules. The iteration procedure must converge to get the fractal
set. Therefore, the iterated functions are limited to strict contractions with the
Banach fixed-point property.

Let (X, d) denotes a complete metric space and H(X) the compact subsets of X,
the Hausdorff distance is defined as

h(A,B) = max{d(A,B), d(B,A)} ∀ A, B∈ H(X) (1)

It is easy to prove that h is a metric on H(X). Moreover, it can be proved that

(H(X), h) is also a complete metric space [7] which is called the space of fractals
for X.

A contraction mapping, or contraction w: X α X has the property that there is

some nonnegative real number k∈[0,1), contraction factor k, such that

d(w(x), w(y)) ≤ k d(x,y) (2)

In non-technical terms, a contraction mapping brings every two points closer in

the metric space it maps. The Banach fixed point theorem guarantees the
existence and uniqueness of fixed points of contract maps on metric spaces: If
w: X α X is a contraction, then there exists one and only one x ∈X such that
w(x) = x.
314 Lu et al.

The Banach fixed point theorem has very important applications in many
branches of mathematics. Therefore, generalisation of the above theorem has
been extensively investigated, for example, in probabilistic metric spaces. The
theorem also provides a constructive method to find fixed-point.

An IFS [9] is a set of contraction mappings wi defined on (X, d) with contraction

factors ki for i = 1,2,…, N. We denote it as{X; wi, i = 1,2,…, N} and contraction
factor k = max{ki, i = 1, 2,…, N}. Hutchinson [9] proved an important theorem
on a set of contraction mappings in which IFS is based: Let {X; wi, i = 1,2,…,
N} be an IFS with contraction factor k. Then W: H ( X ) α H ( X ) defined as

N
W(B) = Υ w ( B)i ∀ B ∈ H (X ) (3)
i =1
is a contraction mapping on (H(X), h). From Banach's theorem, there exists a
unique set A ∈ H ( X ) , the attractor of IFS, such that

N
A = W(A) = Υ w ( A)
i (4)
i =1

It can be seen that A is self-similar since it is expressed as a union of

transformations (copies) of itself. The attractor A can be taken as a definition of
deterministic fractals.

2.2. Fractal dimensions

Mandelbrot [2] proposed a simple but radical way to qualify fractal geometry
through a fractal dimension based on a discussion of the length of the coast of
England. The dimension is a statistical quantity that gives an indication of how
completely a fractal appears to fill space, as one zooms down to finer scales.
This definition is a simplification of the Hausdorff dimension that Mandelbrot
used as a basis. We focus on this one and briefly mentions box-counting
dimension because of its wide practical applications. However, it should be
noted that there are many specific definitions of fractal dimensions, such as
Hausdorff dimension, Rényi dimensions, box-counting dimension and
correlation dimension, etc, and none of them should be treated as the universal
one.

For A ∈ H ( X ) , let n(A,ε), ε<0, denote the smallest number of closed balls of
radius ε needed to cover A. If
 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012 315

log n( A, ε ) (5)
D = lim
ε →0 1
log
ε

exists, then D is called the fractal dimension of A.

So n(A,ε) is proportional to ε-D as ε→ 0 or the exponent D is in n(A,ε) = ε-D

which is the power law relationship. A power law describes a dynamic
relationship between two objects which portrays a wide variety of natural and
man-made phenomena. A key feature of the power law is that the power law
relationship is independent of scales. A good example of intuition of fractal
dimension is a line with the length of ε, where ε is the measuring length.
Assume the line is divided in 3 equal parts and ε = 1 then the simplified
3
n(A,ε)= ε-D gives 3 = (1/3)-D with D = 1. Similarly, the Koch curve's fractal
log 4
dimension is D = = 1.26.
log 3

Practically, the fractal dimension can only be used in the case where
irregularities to be measured are in the continuous form. Natural objects offer a
lot of variation which may not be self-similar. The Box-counting dimension is
much more robust measure which is widely used even to measure images. To
calculate the box-counting dimension, we need to place the image on a grid. The
number of boxes, with size s1, that cover the image is counted (n1). Then the
number of a smaller grid of boxes, with size s2, is counted (n2). The fractal
dimension between two scales is then calculated by the relationship between the
difference of the number of boxed occupied and the difference of inverse grid
sizes [10]. In more chaotic and complex objects such as architecture and design,
more flexible and robust measures, such as range analysis, midpoint
displacement, etc, can be employed. For more detailed information, readers may
refer to Bovill's book [10].

2.3. Examples of IFS applications

Fractal geometry is at the conceptual core of understanding nature's complexity
and IFS provides an important concept for understanding the core design of the
natural objects as well as approximating the natural design. In this subsection
we outline the evolution of the idea of IFS with our calculation examples.

We know that the Banach's fixed-point theorem forms the basis of the IFS
applications. However, applying the theorem in practiced raises two central
questions. One is to find the attractor for a given IFS. The other is to find IFS
for a given attractor, an inverse problem of the first.
316 Lu et al.

For the first problem, the attractor can be obtained by successive

approximations from any starting point theoretically. From a computational
point of view, two techniques, deterministic and stochastic, can be applied. The
deterministic algorithm starts with an arbitrary initial set to reach the attractor.
The stochastic algorithm is often more complex but more efficient. A stochastic
algorithm attributes to the IFS system with a set of probabilities by assigning a
probability to each mapping, which is used to generate a random walk. If we
start with any point and apply transformations iteratively, chosen according the
probabilities attached, we will come arbitrarily close to the attractor. The
associated probabilities determine the density of spatially contracted copies of
the attractor. Therefore, the probabilities have no effect on the attractor but
influence significantly the rendering of its approximations.

The second problem, the inverse problem, can be solved by Barnsley's Collage
Theorem, a simple consequence of Banach’s fixed point theorem. Such
procedure was illustrated nicely through the 'Barnsley fern' in [9] and [11] using
four-transformation IFS with associated probabilities. Figure 2 shows our
calculation examples of fractals using four-transformation IFS with variations
and their associated probabilities produced by Matlab, where 20000 iterations
were set. These fractals actually have more than one attractor. In Figure 2, the
four-transformation matrices are

0 0 0 0.19 − 0.25 0  −0.15 0.28 0   0.76 0.04 0 

   
A = 0 0.18 0 B = 0.22 0.23 1.7 C= 0.24 0.25 0.44 D= −0.05 0.95 1.7
  (6)
0 0 1  0 0 1   0 0 1   0 0 1 

D has the probability 0.75 and others 0.083.

Fig. 2. Calculation examples of fractals using IFS with variations.

 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012 317

3. Applications of Fractal Geometry to Architecture Design

3.1. Applications of IFS

IFS provides wide range of architecture design applications in patterns and
structures. Very often, IFS codes are used to generate fractals. For example,
topology (layout) optimization has been proposed and is based on IFS
representations with various applications [12]. Chang [13] proposed a
hierarchical fixed point-searching algorithm for determining the original
coordinates of a 2-D fractal set directly from its IFS code. The IFS code cane be
modified according to the desired transformation. Figure 3 shows the Castle
with different reflection directions generated by the modified IFS codes.

Fig. 3. Castle example generated by the modified IFS codes [13].

3.2. Applications of fractal geometry

Fractal geometry has been applied in architecture design widely to investigate
fractal structures of cities [14] and successfully in building geometry [15,16]
and design patterns [10].

Early fractal building patterns can be traced to ancient Maya settlements. Brown
et al. analysed fractal structures of Maya settlements and found that fractals
exhibit both within communities and across regions in various ways: at the
intra-site, the regional levels and within archaeological sites. Moreover, spatial
organisation in geometric patterns and order are also fractals, which presents in
the size-frequency distribution, the rank-size relation among sites and the
geographical clustering of sites [17]

In Europe, fractals were found in the early 12th century buildings. The floor of
the cathedral of Anagni in Italy built in 1104 is adorned with dozens of mosaics
in a form of a Sierpinski gasket fractal (See Figure 4).
318 Lu et al.

Fig. 4. The floor of the cathedral of Anagni in Italy [18].

Fractals have been applied to many elevation structures to exclusively address

power and balance. Some very excellent examples of classical architecture can
be seen in many parts of the Europe, in the Middle East and Asia which have
effects of fractal elevations, for example, Reims’ cathedral and Saint Paul
church in France, Castel del Monte in Italy and many palaces in Venice (ca’
Foscari, Ca’ d’Oro, Duke Palace, Giustinian Palace) in Italy. Venice has been
one of the most talked about fractals [18] (see Figure 5). More vital evidence
shows that fractals exist in Gothic cathedrals in general. The pointed arch, an
impression of elevation, appear in entrance, at windows and the costal arch with
many scales and details [19]. Figure 6 displays the elevations of a five-floor
tenement building in the historical part of Barcelona which shows self-
organisation structure.

Fig. 5. Fractal building in Venice [18].

 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012 319

Fig. 6. A tenement house in the historical part of Barcelona, Spain: the

elevation's photograph from the 90-s (left part ); the geometric synthesis shows
the original architecture design (middle part) [20].

In the Middle East, fractal patterns have been adopted widely in designing
stucco, a typically Persian art form for the decoration of dome interiors. In
Figure 7, the pattern in the dome interior has four attractors surrounding the
main one at the center (Sarhangi).

Fig. 7. Stucco dome interior in a private house in Kashan [21].

320 Lu et al.

In Asia, architectures with fractal structures have also been found in Humayun’s
Mausoleum, Shiva Shrine in India and the Sacred Stupa Pha That Luang in
Laos. Fractals have been used to study Hindu temples. In China, some mosques
in the west were more likely to incorporate such domes which are fractals. One
important feature in Chinese architecture is its emphasis on symmetry which
conveys a sense of grandeur [22].

Besides geographical localities, in recent times, the concept of fractals has been
extended to many well known architectures including Frank Lloyd Wright’s
'Robie House', 'Fallingwater', 'Palmer house' and 'Marion County Civic', which
demonstrate that fractals have universal appeal and are visually satisfying
because they are able to provide a sense of scale at different levels. Wright is
probably the most representative of the organic architects. His designs grew out
of the environment with regards to purpose, material and construction [10].
Fractals have inspired many great modern designers such as Zaha Hadid, Daniel
Liebeskind, Frank Gehry and others with many notable fractal architectures
[20]. Indeed, according to Ibrahim et al, architects and designers started to adopt
fractals as a design form and tool in the 1980s [10]. Yessios et al. was among
the first utilising fractals and fractal geometry design in architecture [23]. They
developed a computer program to aid architecture using fractal generators. In
1990th, Durmisevic and Ciftcioglu applied fractal tree as an indicator of a road
infrastructure in the architecture design and urban planning [24].

Wen et al. established the fractal dimension relations matrix table analysis to
classify architecture design style patterns for the masterpieces of three modern
architecture masters: Frank Lioyd Wright, Le Corbusier and Mies van der Rohe
[25]. Figure 8 shows the results. It can be seen that the temporal trends of
individuals vary. The fractal dimensions of Frank Lioyd Wright are average
with low beginning in the early 1900th and end in the mid 1930th. The trend of
Le Corbusier goes downside with gentle slope from mid 1900th to mid 1950th.
For the period shown in the grape, the trend of Mies van der Rohe has the same
trend as that of Frank Lioyd Wright from the early 1900th to the mid 1930th. The
average trend of these three masters goes down in general starting from 1930th.
 Chaotic Modeling and Simulation (CMSIM) 2: 311-322, 2012 321

Fig. 8. Fractal dimensions for the masterpieces of three modern architects.

4. Conclusions
This paper has illustratively reviewed the fundamental concepts and properties
of fractal geometry theory essential to architecture design, as well as the current
state of its applications. Fractal geometry has important implications for
buildings. The representative review shows that architecture design is not made
to be isolated but to anticipate changes in the environment. Accumulation of
technological modernisations, destroying, adapting and many changes have
caused the design temporal and spatial diversity and complexity. More
specifically, sustainable development in a building can be looked upon as
adaptability and flexibility over time when it comes to responding to changing
environments. Chaos and many other nonlinear theories have explained that
extremely deterministic and linear processes are very fragile in maintaining
stability over a wide range of conditions, whereas chaotic and fractal systems
can function effectively over a wide range of different conditions, thereby
offering adaptability and flexibility. In this context, fractal geometry theory
offers an alternative for sustainable architectural design. This paper provides a
bridge between building engineering and architecture and fractal geometry
theory.

References
1. B. Mandelbrot. Fractals, Form, Chance and dimension, Freeman, San Francisco,
1977.
2. B.Mandelbrot. The Fractal Geometry of Nature. W.H. Freeman and Company, 1982.
3. P.A. Burrough. Fractal dimensions of landscapes and other environmental data. Nature
294: 240–242, 1981.
 322 Lu et al.

4. K. Weierstrass, On continuous functions of a real argument that do not have a well-

defined differential quotient, in G. Edgar, ed Classics on Fractals, Addison-Wesley,
Reading, Massachusetts, 3-9, 1993.
5. H. von Koch. On a continuous curve without tangents constructible from elementary
geometry, in G. Edgar, ed Classics on Fractals, Addison-Wesley, Reading,
Massachusetts, 25-45, 1993.
6. G. Cantor. On the Power of Perfect Sets of Points. in G. Edgar, ed Classics on
Fractals, Addison-Wesley, Reading, Massachusetts, 11-23, 1993.
7. M. Barnsley. Fractals Everywhere. Academic Press, 1988.
8. H.O. Peitgen.1988. Fantastic Deterministic Fractals, in: H.O. Peitgen, D. Saupe. The
Science of Fractal Images, 202, Springer, New York, 1988.
9. J.E. Hutchinson. Fractals and self similarity. Indiana Univ. Math. J. 30: 713–747,
1981.
10. C. Bovill. Fractal Geometry in Architecture and Design, Birkhauser, Boston, 1996.
11. M. Ebrahimi, E.R. Vrscay. Self-similarity in imaging, 20 years after "Fractals
Everywhere", http://ticsp.cs.tut.fi/images/3/3f/Cr1023-lausanne.pdf.
12. H.T. Chang. Arbitrary affine transformation and their composition effects for two-
dimensional fractal sets. Image and Vision Computing 22: 1117-1127, 2004.
13. H. Hamda, F. Jouve, E. Lutton, M. Schoenauer, M. Sebag, Compact unstructured
representations for evolutionary topological optimum design. Appl Intell 16: 139-
155, 2002.
14. M. Batty, P. Longley. Fractal Cities, London: Academic Press, 1994.
15. K. Trivedi. Hindu temples: models of a fractal universe. The Visual Computer 5: 243-
258, 1989.
16. N. Sala. The presence of the Self-similarity in Architecture: some examples.
Emergent Nature - Patterns, Growth and Scaling in the Sciences, World Scientific
273-282, 2002.
17. C.T. Brown, W.R.T. Witschey. The fractal geometry of ancient Maya settlement.
Journal of Archaeological Science 30: 1619-1632, 2003.
18. N. Sala. Fractal models in architecture: a case of study.
http://math.unipa.it/~grim/Jsalaworkshop.PDF.
19. W.E. Lorenz. Fractals and Fractal Architecture, Mater Thesis, Vienna, 2003.
20. P. Rubinowicz. Chaos and geometric order in architecture and design. Journal for
Geometry and Graphics 4: 197-207, 2000.
21. http://en.wikipedia.org/wiki/Islamic_architecture.
22. J.S. Cowen. Muslims in China: The mosque. Saudi Aramco World. Retrieved 2006-
04-08, 30-35, 1985.
23. C.I. Yessios. A fractal studio. ACADIA ’87 Workshop Proceedings, 1987.
24. S. Durmisevic, O. Ciftcioglu. Fractals in architectural design. Mathematics and
Design. Javier Barrallo the university of the Basque Country, 1988.
25. K-C. Wen, Y-N. Kao. An analytic study of architectural design style by fractal
dimension method, 22nd International Symposium on Automation and Robotics in
Construction ISARC 2005, Italy, 2005,
http://www.iaarc.org/publications/fulltext/isarc2005-63wen.pdf

View publication stats