Content uploaded by Leonardo Todisco
Author content
All content in this area was uploaded by Leonardo Todisco on Sep 02, 2016
Content may be subject to copyright.
Technical University of Madrid
School of Civil Engineering
Department of Continuum Mechanics and Theory of Structures
An integrated approach
to conceptual design of arch bridges
with curved deck
Author
Leonardo Todisco
Civil Engineer
Technical University of Madrid
Supervisor
Hugo Corres Peiretti
PhD MEng Prof. In Civil Engineering
Technical University of Madrid
E.T.S.I.C.C.P.
U.P.M.
-I-
ABSTRACT
An integrated approach to conceptual design of arch bridges
with curved deck
by
Leonardo Todisco
MS THESIS IN ENGINEERING OF STRUCTURES, FOUNDATIONS AND MATERIALS
Spatial arch bridges represent an innovative answer to demands on functionality, structural
optimization and aesthetics for curved decks, popular in urban contexts.
This thesis presents SOFIA (Shaping Optimal Form with an Interactive Approach), a
methodology for conceptual designing of antifunicular spatial arch bridges with curved deck in a
parametric, interactive and integrated environment.
The approach and its implementation are in-depth described and detailed examples of
parametric analyses are illustrated. The optimal deck-arch relative transversal position has been
investigated for obtaining the most cost-effective bridge.
Curved footbridges have become a more common engineering problem in the context of urban
developments when the client is looking for a strong aesthetics component: an appropriate
conceptual design allows to obtain an efficient and elegant structure.
Antifunicular, hanging models, form-finding, spatial arch bridges, structural design, geometry, Spatial
analysis, structural response, structural forms, curvature, curved deck, compression-only structures,
structural optimization.
Thesis supervisor: Hugo Corres Peiretti, Professor of Structural Concrete and Conceptual Design
at the School of Civil Engineering at the Technical University of Madrid
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-II-
RESUMEN
An integrated approach to conceptual design of arch bridges
with curved deck
de
Leonardo Todisco
TRABAJO FIN DE MÁSTER EN INGENIERÍA DE LAS ESTRUCTURAS, CIMENTACIONES Y
MATERIALES
El conceptos de funicularidad se puede extender a estructuras lineales espaciales como, por
ejemplo, los puentes arco con tablero curvo. Estas estructuras, especialmente pasarelas
peatonales, son consecuencia de la necesidad de encajar trazados exigentes y de dar respuesta a
nuevas demandas arquitectónicas. En las estructuras curvas el diseño conceptual juega un papel
absolutamente esencial. Siempre ha sido así, pero en el caso presente, cabe resaltar que una
errónea elección de la geometría conlleva una serie de problemas que se irán acumulando a lo
largo del proceso de proyecto, de la construcción y de la vida de la estructura.
En este trabajo se presenta SOFIA (Shaping Optimal Form with an Interactive Approach), una
herramienta capaz de, conocida la geometría del tablero, de buscar automáticamente la forma
del arco antifunicular correspondiente. El planteamiento seguido es conceptualmente el mismo
que el utilizado en la búsqueda de formas óptimas en estructuras en dos dimensiones: el arco
antifunicular es el que representa, para unas cargas dadas, el lugar geométrico de los puntos con
momento flector nulo.
La herramienta ha sido desarrollada en un entorno integrado, interactivo y paramético. Su
implementación está ilustrada y unos ejemplos de análisis paramétricos están desarrollados. La
posición transversal relativa entre tablero y arco ha sido investigada para obtener la configuración
del puente estructuralmente más eficiente.
Las pasarelas curvas se han convertido en un problema de ingeniería más común de lo habitual
en el contexto de los desarrollos urbanos cuando el cliente está buscando un fuerte componente
estético: un diseño conceptual adecuado permite obtener una estructura eficiente y elegante.
-III-
TABLE OF CONTENTS
ABSTRACT ...................................................................................................................................................................................................... I
RESUMEN ..................................................................................................................................................................................................... II
FIGURES INDEX............................................................................................................................................................................................ V
TABLES INDEX ........................................................................................................................................................................................... VI
NOTATION ................................................................................................................................................................................................ VII
1 INTRODUCTION AND MOTIVATION ............................................................................................................................ 9
1.1 Overview of the thesis .......................................................................................................................................................... 9
1.2 Objectives................................................................................................................................................................................. 9
1.3 Conceptual design: need for new design tools ............................................................................................................. 11
1.4 Organization and structure of the work .......................................................................................................................... 12
2 STATE OF THE ART: THE CONCEPT OF FUNICULARITY ........................................................................................ 14
2.1 Introduction ........................................................................................................................................................................... 14
2.2 Historical overview............................................................................................................................................................... 16
2.3 The funicular curve ............................................................................................................................................................... 20
2.4 The search of the funicular curve ..................................................................................................................................... 26
2.4.1 Method of moments ........................................................................................................................................................... 26
2.4.2 Graphic statics ....................................................................................................................................................................... 27
2.4.3 Physical models ..................................................................................................................................................................... 27
2.4.4 Numerical methods ............................................................................................................................................................. 29
2.4.5 Particle-spring systems ....................................................................................................................................................... 29
2.4.6 Final comments ..................................................................................................................................................................... 29
2.5 Historical and modern examples ..................................................................................................................................... 29
2.5.1 The Pantheon ......................................................................................................................................................................... 30
2.5.2 Salginatobel Bridge .............................................................................................................................................................. 32
2.5.3 The Tiemblo bridge .............................................................................................................................................................. 34
2.6 Final remarks .......................................................................................................................................................................... 35
3 ANTIFUNICULAR SPATIAL ARCH BRIDGES FORM FINDING USING AN INTEGRATED AND
INTERACTIVE APPROACH ...................................................................................................................................................................... 37
3.1 Introduction ........................................................................................................................................................................... 37
3.2 State-of-the-art of spatial arch bridges ........................................................................................................................... 39
3.3 Designing spatial arch bridges .......................................................................................................................................... 42
3.3.1 An interactive and integrated environmental workspace: Rhinoceros, Grasshopper, Karamba .................... 42
3.3.2 Form finding methodology for spatial antifunicular arches ...................................................................................... 43
4 PARAMETRIC ANALYSIS ON THE INFLUENCE OF DECK-ARCH RELATIVE TRANSVERSAL POSITION 48
4.1 Cases of study ....................................................................................................................................................................... 48
4.2 Discussion of results ........................................................................................................................................................... 60
4.3 Conclusions ............................................................................................................................................................................ 67
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-IV-
5 CONCLUSIONS AND FUTURE STUDIES ..................................................................................................................... 69
5.1 Summary of results ............................................................................................................................................................. 69
5.1.1 The funicular principle ....................................................................................................................................................... 69
5.1.2 Designing antifunicular spatial arch bridges.................................................................................................................. 70
5.2 Further studies ....................................................................................................................................................................... 71
5.2.1 Introduction ............................................................................................................................................................................ 71
5.2.2 Torsional behaviour of curved decks .............................................................................................................................. 72
5.2.3 Arrangement of layout of cables ...................................................................................................................................... 73
BIBLIOGRAPHY .......................................................................................................................................................................................... 75
PUBLICATIONS BY AUTHOR ................................................................................................................................................................ 80
-V-
FIGURES INDEX
Figure 1-1: Antifunicular spatial arch for C- and S-shaped deck ....................................................................................................... 11
Figure 2-1: Funicular and anti-funicular geometries for known distribution of loads ............................................................... 15
Figure 2-2: Inca suspended bridge. Taken from Squier (Squier 1877) ........................................................................................... 16
Figure 2-3:Examples of compression-only structures built B.C. .................................................................................................... 18
Figure 2-4: Hanging chain and correspondent inverted arch used by Poleni for the structural safety assessment of San
Peter dome in Rome. Taken from Poleni (Poleni 1982). .................................................................................................................. 20
Figure 2-5:. infinitesimal element of rope loaded with a set of vertical loads ............................................................................ 20
Figure 2-6: Horizontally distribution of load for a constant load per unit length for arches with different rises .............. 21
Figure 2-7: Maximum value of the horizontal projected load for different values of λ .......................................................... 22
Figure 2-8: Difference between catenary (dot line) and parabola (continue line) for different values of ...................... 23
Figure 2-9: Distribution of loads corresponding to different funicular geometries ................................................................. 24
Figure 2-10: Influence of the backfill disposition on the minimal thickness of a stable circular shell. Taken from Ramos
Casquero (Ramos Casquero 2011) ......................................................................................................................................................... 25
Figure 2-11: Funicular geometry for a general distribution of loads .............................................................................................. 26
Figure 2-12: Physical models used by Gaudí and Isler ....................................................................................................................... 28
Figure 2-13: Concrete density of Pantheon dome. ............................................................................................................................. 30
Figure 2-14:. Independent arches used for analysis ............................................................................................................................ 31
Figure 2-15: Line of thrust for different load cases ............................................................................................................................. 32
Figure 2-16: Salginatobel Bridge by Maillart ........................................................................................................................................ 33
Figure 2-17: Analysis of the bridge using graphic statics. Taken from Fivet (Fivet and Zastavni 2012). ............................... 34
Figure 2-18: Front view and cross section of the Tiemblo bridge. Courtesy of Fhecor Ingenieros Consultores ............. 34
Figure 2-19: Graphic statics applied for finding the thrust line of the Tiemblo bridge due to own weight. ..................... 35
Figure 2-20: Main Train Station in Berlin. Photo taken from SBP (SBP 2014). ........................................................................... 36
Figure 3-1: Examples of curved bridges ................................................................................................................................................ 38
Figure 3-2:sketch of the Golden Horn Bridge designed by Leonardo da Vinci in 1502 (Biblioteque Institute Paris ). .... 40
Figure 3-3:Ripshorst Footbridge by SBP (SBP 2014) .......................................................................................................................... 41
Figure 3-4:: spatial arch footbridge proposals .................................................................................................................................... 41
Figure 3-5:Description of the form-finding process SOFIA ............................................................................................................ 45
Figure 3-6: Resulting antifunicular arch for bridges with C- and S-shape curvature of the deck ......................................... 46
Figure 4-1: Cases of studies for the parametric analysis .................................................................................................................. 49
Figure 4-2: Antifunicular configuration for C-shape deck ............................................................................................................... 54
Figure 4-3: Antifunicular configuration for S-shape deck ................................................................................................................ 59
Figure 4-4: Axial forces along the arch ................................................................................................................................................. 61
Figure 4-5: Summation of axial force of cables for different values of y. .................................................................................... 62
Figure 4-6: Axial forces in the deck ....................................................................................................................................................... 63
Figure 4-7: Axial forces along the arch ................................................................................................................................................. 64
Figure 4-8: Summation of axial force of cables for different values of y. .................................................................................... 65
Figure 4-9: Axial forces in the deck ...................................................................................................................................................... 66
Figure 5-1: different arrangements of cables layout .......................................................................................................................... 74
-vii-
NOTATION
la span of the arch
ld span of the deck
fa rise of the arch
fd sag of the arch
N axial force
H horizontal component of axial force
q load along the curved geometry
p horizontally distribution of load
E elastic modulus
I area moment of inertia
s curvilinear coordinate of the arch
C torsion centre
G centre of gravity
Kt tangent stiffness
ΔF load step
Miy bending moment for the i-section around the y-local axis
Mix bending moment for the i-section around the x-local axis
ei value of the eccentricity for the i-section
e* maximum allowable value of the eccentricity
y deck-arch transversal distance
1. INTRODUCTION AND MOTIVATION
-9-
1 INTRODUCTION AND MOTIVATION
Architecture is too important to be left to architects.
Giancarlo de Carlo
1.1 Overview of the thesis
Spatial arch footbridges represent an innovative answer to the demands on functionality,
structural optimization and aesthetics for curved decks, popular in the urban context.
This work focuses on a sound conceptual design of arch footbridges with curved deck, using an
integrated and interactive approach developed in a parametric environment. A concatenation of
parametric design with structural analysis is proposed.
Curved structures are characterized for having the geometry as an essential factor involved in
their structural behaviour: a proper conceptual design of the structural shape is a key aspect for
obtaining a cost-effective bridge.
Like in a curved inverted chain, forces should flow in axial compression toward the supports
minimizing moments to obtaining structural efficiency: in this thesis a new integrate approach for
finding funicular geometries is presented in detail, some results are discussed and future works
are proposed.
Curvature can be symbol of grace: a designer who is able to understand the fundamental
relationships between shape and structural behaviour will take advantage of curvature in order to
design more efficient and sustainable structures.
1.2 Objectives
The aim of this work is to illustrate an integrated approach in order to perform form finding
analysis of arches characterized for supporting a curved deck.
It is not unusual that the footbridge deck has to adapt to a complex geometry in plan as well as in
elevation, due to regulations on disabled peoes, which often entails the use of long
ramps. Usually designers incorporate external ramps to decrease the slope as much as possible
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-10-
to fulfil the admissible slope limit. An alternative to that solution is to increase the length of the
deck, by making it horizontally curved.
Structurally, the curved deck could be supported by an arch, or can be cable-stayed or suspended
by a cable system. In this work only spatial arch bridges which support a curved deck using
secondary ties are considered. The proposed approach is valid for any geometry of the deck.
Differently to comprehensive previous works on the topic (Jorquera Lucerga 2007), a interactive
and integrated approach has been investigated in a parametric environment.
The conceptual design of an arch footbridge involves choosing an appropriate geometry among
many possibilities. A shape finding process, in which the geometry of the arch is generated, has
been developed in order to simulate numerically a chain which works as an compression-only
configuration in static equilibrium with the design loading.
Examples of resulting antifunicular arches for a C- and S-shaped decks are shown in Figure 1-1.
a) C-shaped deck
1. INTRODUCTION AND MOTIVATION
-11-
b) S-shaped deck
Figure 1-1: Antifunicular spatial arch for C- and S-shaped deck
The thesis includes an in-depth look at how funicularity of form and new computational tools can
unlock hidden architectural and aesthetic potential in order to explore new efficient geometries.
1.3 Conceptual design: need for new design tools
As it has been remarked in the previous section, in designing curved structures, i.e. shells, arches,
domes, vaults, the most important phase is the choice of the structural shape: conceptual design
is main issue for designers involved in spatial structures.
Conceptual design is the approach that creates an idea in order to find a solution to a new
proposal for a structure or solve a detail in a specific structure (Corres 2013). The aim of this
section is not to illustrate how conceptual design should be carried out because has it been
described by several authors during recent years (Schlaich 1996; Corres 2013; Muttoni 2011; Allen
and Zalewski 2009), and it is not the main topic of this thesis. The goal of this section is to point
out the existence and the need of new tools that can help designers to perform sound
conceptual designs of curved structures. A more general discussion on the topic is proposed in
Clune (Clune et al. 2012).
Structural engineers usually have strong engineering knowledge but, despite architects, little
creative capacity. Nevertheless, taking in mind 'classical proportions' and having previous design
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-12-
experience, new structural solutions can be found in order to obtain more structural efficiency
(Romo 2013).
In order to compare different solutions and to explore new ones, designers do not need
advanced structural analysis programs, but they need tools for helping to design and to image
new solutions having more freedom changing geometry, supports, cross-sections, etc.. in a three-
dimensional environment.
During the last few years new tools have been developed in order to correlate geometry and
structural behaviour in a dynamic and interactive way. An introduction to concatenating
parametric design with structural analysis was proposed by Georgiou (Georgiou 2011).
A combined use of these tools has been performed for finding the most efficient spatial arch
bridge, as shown in section 3.
There is a strong need of parametric tools that allow to better manage data approximations
typical in the first stage of a project and to perform quick structural evaluations obtaining the
most efficient structural solution to the problem. An integration of empirical, traditional and
innovative structural design techniques could be an advantage for designing cost-effective
structures in a truly rational way.
Experience, knowledge of structural engineering (spatial structures cannot exist without
mechanical basis (Sasaki 2005)), ambition and the use of new tools are without doubt the
elements valuables for increase creativity and innovation in structural design.
1.4 Organization and structure of the work
This dissertation presents a novel computational methodology to design antifunicular arch
bridges with a curved deck and its implementation with a parametric study. The thesis has been
divided in four main parts:
Part 1: State of the art.
In this section the concept of funicularity is presented not only for the design of new structures,
but for safety assessment of historic arches and vaults too. Literature on the topic has been in-
depth reviewed and methodologies, more or less contemporary, have been illustrated and
critically analyzed. Some case studies have been described and discussed.
Part 2: Designing antifunicular spatial arch bridges.
This part focuses on the methodology and its implementation. Traditional techniques used by
engineers of the 20th century have been extended and implemented in a parametric
environment. The used tool has been described in detail in order to evidence the advantages
related to that novel integrated approach.
1. INTRODUCTION AND MOTIVATION
-13-
Part 3: Examples of parametric analysis.
In this section some examples of parametric analysis using the new tool have been performed.
The influence of relative position between arch and deck on structural behaviour has been
investigated. The optimal position of a curved deck minimizing the axial forces of arch and cables
has been found.
Part 4: Conclusion and further studies.
In this section general conclusions have been remarked and further studies that are currently
under development are illustrated.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-14-
2 STATE OF THE ART: THE CONCEPT OF FUNICULARITY
The resistant virtues of the structure that we make depend on their form; it is through their form that
they are stable and not because of an awkward accumulation of materials. There is nothing more
noble and elegant from an intellectual viewpoint than this; resistance through form.
Eladio Dieste
2.1 Introduction
Curved structures, if they are the outcome of an appropriate design, have the capacity to carry
large loads and cover important areas saving costs and can come to represent examples of the
highest structural efficiency. In designing shells or arches, geometry plays a primary role, so
enough energy should be spent for a proper design of the structural shape.
Arches and cables seem to be structures apparently different, but they are very similar under the
structural behaviour point of view. As it is well known, the funicular is the geometry of a chain,
pure in tension and free of bending, loaded with a determined distribution.
The word funicular comes from the Latin word funiculus, diminutive of funis, meaning 'slender
rope' (Treccani 2014). The basic idea to design an arch is to use the principle of inversion,
adopting its dual form: the main difference is that the dual geometry is a compression-only
geometry.
In figure 2.1 different symmetric funicular and anti-funicular geometries are shown for some load
distributions.
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-15-
Figure 2-1: Funicular and anti-funicular geometries for known distribution of loads
A pedagogic way to find intuitively a compression-only geometry is to use a chain without
bending stiffness and to hang the chain with a known set of loads.
As previously shown in figure 2-1, to one distribution of loads corresponds an infinite series of
funicular and antifunicular curves. Changing the horizontal reaction, o similarly, the rise of the
arch, o its length it is possible to obtain the family of compression-only or tension-only
geometries corresponding to one set of loads.
The physical concept that arises from the principle of the funicularity is the lack of bending
moments: ropes, free bending, find automatically their unique equilibrium geometry.
The principle of funicularity is important for several reasons:
- to allow designing efficient structures in which the use of material is optimized. The choice of
the funicular is conditioned by functional, economic, technological, architectural aspects.
- to allow assessing the safety of an existing construction that work principally with compression
forces, in which funicular curves coincide with thrust lines. In this case a thrust line included in
the thickness of the structure (arch, columns, wall, foundation, etc..) has to be found (Heyman
1966).
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-16-
The funicularity principle and its implications should be indispensable part of the expertise of a
structural designer. These ideas are well known but an everyday more frequently use of model
computer analysis has contributing to eclipse this vital concept.
In the following section, the concept of funicularity is in-depth investigated and its historical
evolution is illustrated. Several tools, more or less recent, for finding that geometry are
presented. Finally some compression-only structures are described and critically discussed.
2.2 Historical overview
Curved structures (arches, shells, domes, etc..) are constructions in which, as previously stated,
the thrust line has to be included in the structural thickness in order for such structures to work
primarily in compression.
First funicular structures for crossing a gorges built over the centuries were tension-only
geometry because was easier to find long fiber ropes for crossing a river; furthermore there was
no need of geometrical-mechanical knowledge because, due to their no bending stiffness, ropes
adopt automatically their optimal and equilibrated configuration.
Figure 2-2 show an example of an Inca suspended footbridge, where technology was well suited
to the problem and their resources.
Figure 2-2: Inca suspended bridge. Taken from Squier (Squier 1877)
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-17-
Figure 2-3 shows some examples of ancient structures working only with compression paths of
forces. Compression-only three-dimensional structures were spread in Cyprus in 5500 B.C, in
1700 B.C. in Sardinia (Italy) and during the VII century B.C. in Lazio (Italy).
(a) Cyprus. Image taken form Fernandez Casado ndez Casado 2006).
(b) Sardinia. Image taken from Wikipedia
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-18-
(c) Etruscan civilization in Lazio (Italy). Image taken from Canino (Canino 2014)
Figure 2-3:Examples of compression-only structures built B.C.
The majority of curved structures build over two thousand years ago were corbelled structures,
where starting from the base, the stones that form the internal structure were arranged
projecting towards the inner part of the dome. On this first ring, other rings are added in the
subsequent projection, with smaller diameter, until reaching a minimum opening, which is closed
by a last stone. An extensive archive of corbelled domes is provided by Lobbecke (Lobbecke
2012). The corbelling construction technique needs a self supporting structure, so that the hoop
forces are mobilized in order to ensure equilibrium (Allen 1984). An interesting study on this kind
of structures has been developed by Como (Como 2007) for Mycenaean Tholoi and by Sanitate
(Sanitate et al. 2014) for trulli in Apulia (Italy).
Before the XV century the use of suitable only-compression or only-tension structure was based
on the empirical acceptation of its good structural behaviour without an explicit formulation of
its mechanical behaviour.
Starting from the XV century, the first documents on investigations about the structural
mechanics behind the behaviour of chains and arches started to arise: a theoretical
approximation began to combine the experimental evidence. The first attempts defining the
structural behaviour of curved structures were due to Leon Battista Alberti (1404-1472), Palladio
(1508-1580) and Leonardo da Vinci (1452-1519), who proposed the basis of the scientific analysis
attempting to develop a theory with the concept of static analysis. In the following are reported
some writing of Leonardo da Vinci (Da Vinci and Richter 1970). He realized the basis of the
structural mechanics.
"The arch is nothing else than a force originated by two weaknesses, for the arch in buildings is
composed of two segments of a circle, each of which being very weak in itself tends to fall; but as each
opposes this tendency in the other, the two weaknesses combine to form one strength."
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-19-
"As the arch is a composite force it remains in equilibrium because the thrust is equal from both sides;
and if one of the segments weighs more than the other the stability is lost, because the greater pressure
will outweigh the lesser."
"Next to giving the segments of the circle equal weight it is necessary to load them equally, or you will
fall into the same defect as before."
"An arch breaks at the part which lies below half way from the centre."
"If the excess of weight be placed in the middle of the arch at the point a, that weight tends to fall
towards b, and the arch breaks at 2/3 of its height at c e; and g e is as many times stronger than e a,
as m o goes into m n."
"The arch will likewise give way under a transversal thrust, for when the charge is not thrown directly
on the foot of the arch, the arch lasts but a short time."
The difference between funicular due to self weight (constant weight per unit length) and due to
a uniformly distribution horizontally load, respectively a catenary and a parabola, could be very
small: Galileo (1564-1642) got it wrong in his writing 'Dialogues Concerning Two New Sciences"
(Galilei 1638), where he stated that the funicular geometry formed by a chain was a parabola. The
difference between the two curves was demonstrated and published by Jungius (1587-1657) after
his death. The first scientific essay on the theory of arches was formed diffused by La Hire (1640-
1718) in "Traitè de Mecanique" in 1695.
The idea to use the inversion principle was published by Hooke (1635-1703) in his famous
anagram to find the ideal compression-only geometry for a rigid arch (Hooke 1676):
"Ut pendet continuum flexile sic stabit contiguum rigidum inversum"
("As hangs the flexible line, so but inverted will stand the rigid arch.")
After Hook, several scientific of the past studied the problem (Bernoulli, Leibniz, Huygens) and
Gregory published an important essay on the topic (Gregory 1697) where he described
mathematically the catenary geometry.
The idea of obtaining the geometry of the arch as the inversion of a hanged cable encouraged the
use of reduced models. It is interesting to evidence that the shape of an hanging chain is
independent of scale, as the statical equilibrium of compression structures (Adriaenssens et al.
2014).
The idea of using physical reduced models was adopted by Poleni (Poleni 1982) in his assessment
of the structural stability of San Peter dome in Rome as shown in figure 2-4.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-20-
Figure 2-4: Hanging chain and correspondent inverted arch used by Poleni for the structural safety assessment of San
Peter dome in Rome. Taken from Poleni (Poleni 1982).
2.3 The funicular curve
Funicular curves represent, as previously described, the shape of a hanging chain for a known set
of loads. These curves are compression- or tension-only. Infinite number of funiculars exist for
one distribution of loads. Each curve corresponds to one horizontal reaction, or one rise, or one
total length. An important property of that geometry is its independence from the typology of
supports. Supports become relevant when the funicular does not match the axis of the structure.
Figure 2-5 shows an infinitesimal element of rope loaded with a set of vertical loads: the
horizontal component (H) of the axial force (N) is constant along the curve because only vertical
loads are imposed.
q(x)
AB
x
y
q dx
a
b
dx
Na
H
Nb
H
Figure 2-5:. infinitesimal element of rope loaded with a set of vertical loads
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-21-
The differential equation of the funicular curve can be expressed as:
2
2()d y q x
dx H
(2.1)
Equation 2.1 is similar to the differential equation of the classical beam theory:
2
2()d y M x
dx EI
(2.2)
The previously analogy shows that the funicular shape has the same geometry of the bending law
for a beam with same span and loaded with the same loads (Timoshenko 1953).
An important aspect that should always taken into account is the real distribution of loads; the
difference between a constant weight per unit length and an uniformly distribution horizontally
load can be relevant in some cases. Figure 2-6 illustrates three arches, with the same thickness,
span and different inclination, on which there is a load q uniformly distributed along the axis. The
consequent horizontally distribution of load is far from a constant value, as higher is the rise of
the arch.
Figure 2-6: Horizontally distribution of load for a constant load per unit length for arches with different rises
Figure 2-7 illustrates a plot that allows to calculate the maximum variation of the horizontally
distribution of loads respect to a constant value obtained for a flat beam. A parabolic arch has
been used for plotting the graph and the value of the maximum load, in proximity of hanging
lines, has been plotted versus the parameter = l / f. For high values of , i.e. arch with small rise,
the maximum values of load coincides.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-22-
Figure 2-7: Maximum value of the horizontal projected load for different values of λ
Considering
2
1 ( )
dy
ds dx
(2.3)
and the equation of the arch as
2
4
( ) ( )
4
x
y x x
(2.4)
it is possible to evaluate that curve as
() ds
p x q dx
(2.5)
For low values of , as for example =2 (possible value in designing shells), the horizontally
distribution of loads can be 120% higher close to support respect to the centre of the span.
Parabola
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-23-
Parabola is the funicular curve corresponding to a horizontally constant distribution of load. If the
load is constant per unit length but the geometry has low inclination (high values of ), the
differences between parabola and real funicular can be neglected. Many times the parabola
equation has been used for designing suspended bridges: rigorously speaking parabola does not
represent the funicular curve for dead load and self weight for that kind of bridge but if loads due
to deck are really high compared to cables ones, the difference between the real funicular curve
and a parabola can be disregarded.
Catenary
The word catenary comes from the Latin word catenarus and it represents the funicular
geometry obtained with a load constant per unit length, as for example is the self weight if the
cross section is the same along the curve. Catenary represent the real geometry of an hanging
chain. The use of catenary for architectural purposes was diffuse in Middle East (Dome of the
Rock in Jerusalem). The catenary equation is well known:
//
cosh( ) ( )
2x h x h
xh
y h e e
h
(6)
Developing the equation using Taylor series, it can be observed that catenary and parabola share
the first three terms, with the differences only starting from the fourth. In figure 2-8 the
difference between catenary and parabola is shown for arches with different values of = span /
rise. For high values of , i.e. arch with low rise, the difference is minimal.
=10
=5
=3.3
=2.5
=2
=10
=5
=3.3
=2.5
=2
Figure 2-8: Difference between catenary (dot line) and parabola (continue line) for different values of
Circumference
Roman master builders designed curved structures, as arches and shells, starting from
geometries derived from the circumference a priori less efficient, but easier to built. There are
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-24-
symbolic reasons too: circumference represent the perfection, the unity, the harmony. The
distribution of loads for a circular arch in order to have a perfect compression-only behaviour
needs to be either constant in the radial direction or with a infinite value close to the supports as
is shown in figure 2-9.
qqq
parabola catenaria arco de circulo arco de circulo
q
a) parabola
b) catenary
qqq
parabola catenaria arco de circulo arco de circulo
q
c) arch of circumference
Figure 2-9: Distribution of loads corresponding to different funicular geometries
It seems that Roman master builders were aware that for these kind of geometries additional
loads had to be added close to the support. This necessity does not entail a strong obligation
because abutments and backfill were necessary to guarantee an horizontal slope of roads,
aqueducts and buildings: in this way, probably unconsciously, circumference was similar to
funicular of dead loads, absolute predominant in this kind of bridges.
Backfills have a key role: to increase loads, approach thrust line to barycenter axis and prestress
the structure in order to improve their structural behaviour in front of bending moments due to
live load (important for railway bridges, and starting from the XX century, for roadway bridges).
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-25-
The importance of filling has been studied by many authors. In Ramos Casquero (Ramos
Casquero and León González 2011) the different 'missions' of the fill are described, mainly to
guarantee the stability of the dome. As an example showing the importance of filling, the figure
2-10 (Ramos Casquero 2011) shows how an adequate disposition of fill allows that the minimal
thickness of a circular shell is stable for a fill high of 0.63 of the rise.
Figure 2-10: Influence of the backfill disposition on the minimal thickness of a stable circular shell. Taken from Ramos
Casquero (Ramos Casquero 2011)
The funicular curve for real structures
The funicular (or anti-funicular) curve is tension-only (or compression-only) if the distribution of
loads does not change, and if the geometry remains exactly the same, during the life of the
structure.
This condition is true only if the axial stiffness is infinite and there are not rheological effects. If
the real geometry is different from the funicular curve, bending moments arise.
In real structures the axial stiffness could have high values but never be infinite so if there are no
coactions (flexural stiffness) and loads do not change their position or magnitude, the deformed
curve will be similar to the starting one (there is no variation of curvature between sections) and
it is equally funicular (or anti-funicular), but with different rise (higher in funicular and lower for
anti-funicular).
General speaking, especially for concrete and masonry structures, flexural stiffness is not zero,
producing bending moments, in order to ensure the compatibility condition for fully restrained
structures, equal to the axial forces multiplied the distance between the barycentre axis and the
line of thrust. If the arch is fully restrained, the support rotation is not allowed, so compatibility
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-26-
moments arise close to supports. If the arch is simply supported, bending moments arise in order
to satisfy the condition of compatibility due to the fix length of the arch.
2.4 The search of the funicular curve
2.4.1 Method of moments
During last centuries different methodologies more or less efficient have allowed designers to
find the shape of the funicular for a known distribution of load.
Remembering the definition of funicular curve, bending moments in each point of the curve have
to be zero. If the rise is known, it is possible to find the points that generate the funicular
configuration. Bending moments, M(x) for a generic arch could be expressed as:
( ) ( ) ( )
o
M x M x Hy x
(2.7)
where Mo(x) is the moment produced by the same distribution of load for a simply supported
beam, y(x) is the vertical coordinate of points and H is the horizontal reaction. If the funicular
geometry is found, M(x) has to be equal to zero, so the vertical coordinate of point y(x) is:
()
() o
Mx
yx H
(2.8)
That result is of paramount importance because illustrates how the funicular configuration
corresponds perfectly to the moment diagram of a beam with the same span and distribution of
loads. In a more general case, if the distribution of load is variable, as shown in figure 2-11, the
vertical position of node y(x) corresponding to the funicular geometry is expressed by equation
2.9:
q(x)
M(x)
Figure 2-11: Funicular geometry for a general distribution of loads
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-27-
00
00
00
( ) ( )
1
( ) ( ( ) ) (1 ) ( ) ( )
( ) ( )
lx
lx
lx
q x xdx q x xdx
y x q x dx q x dx x
Hq x dx q x dx
(2.9)
An improvement of this methodology could be, for example, to take into account that the self
weight of the arch changes for different values of the rise. In that case, it should be necessary to
iterate in order to find the right funicular configuration.
2.4.2 Graphic statics
Varignon introduced the concept of funicular polygon and force polygon in his work "Nouvelle
Mécanique ou Statique" (Varignon 1725). Based on this principle, starting form the XIX century, a
new technique using the reciprocal relationship between funicular polygon and the force
polygon in order to find compression- and tensile-only geometries. Probably the most
comprehensive work on that topic is 'Graphical Calculus'(1890) by Cremona (1830-1903). Graphic
statics is an approach in which only equilibrium equation are taken into account and it is a very
powerful tool for designing new structures and to analyze the existing ones.
Important designers, i.e. Cullmann (1821-1881), Ritter (1847-1906), Maillart (1871-1940), Isler (1926-
2009) y Menn (1927), used that technique for designing elegant and efficient structures.
Graphic statics has been a little used during the last years because a variation of the geometry or
loads changes the whole model. However, during the last years new computational tools have
been developed in order to use the graphic statics in a very easy and intuitive way. Important
improvements have been reached thanks to works by Ochsendorf and Block, respectively at MIT
(Masonry Group 2009) and ETHZ (Block Research Group 2014).
2.4.3 Physical models
As it has been previously described, the idea to find the geometry of an arch using the inversion
approach of a hanging chain boosted the use of reduced physical models. Poleni used this
technique for a safety assessment of the Saint Peter's dome in Rome.
During the first decades of the XX century, some of the most important engineers and architects
faced the problem of finding a funicular geometry using physical models. The following figure 2-
12 show the models used by Gaudí for the Sagrada Familia in Barcelona, and by Isler for one of his
well known shells in Swizerland. Gaudí physical models were models in which a set of loads were
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-28-
suspended to a network of wires. Inverting and 'freezing' the system made it possible to obtain
compression-only structures (Popovic Larsen and Tyas 2003).
a) Physical models by Gaudí
b)Physical models by Isler
Figure 2-12: Physical models used by Gaudí and Isler
Main disadvantages of this approach are the impossibility to take into account variable thickness
and loads different from the vertical ones. Furthermore, it is not easy to obtain values of stress
and strain in different point of the hanging geometry and materials used for physical models are
different from real building materials. Today thanks to the incredible developing of 3D printer
and CNC machines, it is possible to promote a return to that technique.
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-29-
2.4.4 Numerical methods
Numerical methods have been developed in order to find the solution to a problem
characterized by a strong geometric non-linearity. They are numerical algorithms where an
iterative process stops when a static equilibrium is reached (Lewis 2003). An interesting overview
of these methodologies is illustrated in Veenendaal (Veenendaal and Block 2012). An in-depth
state-of-art review on these methods can be found in Ramm (Ramm 2004).
These methodologies could be divided in three different groups: methods using dynamic
relaxation, stiffness matrix and force density. An example of the utilization of computational
tools for finding the funicular geometry is included in the work of Galafel lez 2011)
where equilibrium equations and the theory of plasticity have been used for finding funicular
geometries. Differently to the previous described approaches, these tools cannot allow an easy
exploration of new geometries.
2.4.5 Particle-spring systems
Particle-spring systems are an interesting alternative for form finding tension-only geometries.
That approach was developed by Kilian (Kilian and Ochsendorf 2005). The use of spring system is
the basis of an algorithm able to find funicular geometries. Conceptually this methodology is
similar to the physical models with hanging elements. The stiffness of the springs corresponds to
the axial stiffness of the material used with reduced physical models. Changing the stiffness of
the spring it is possible to explore the infinite funicular configurations corresponding to one set
on loads. In the context of graphic statics that is equivalent to change the position of pole in
order to obtain different funicular polygons.
2.4.6 Final comments
Five different approaches for finding funicular geometries have been briefly described. All of
these methods allow to obtain different funicular configurations for one distribution of load.
Changing the horizontal reaction in the method of moments, moving the pole in graphic statics,
using materials with different properties in physical and numerical models, changing the constant
of elasticity in particle-spring systems, are all equivalent ways to explore different configuration
of compression or tension-only geometries.
2.5 Historical and modern examples
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-30-
Throughout history a large number of compression- and tension-only structures have been
designed, after a form finding process or simply using intuition or experience, in order to get an
efficient use the materials. In the following section few structures are described and an
implementation of some of the previously approaches is illustrated.
2.5.1 The Pantheon
The building devoted to all the gods, the Pantheon in Rome, represents the top of the
architectonic and structural revolution or Romans. It was rebuild, after a fire, between 118 AD and
128 AD, by Apollodorus of Damascus, and it has been, for different reasons, a revolutionary
construction. The most important are: the use of concrete, opus caementicium, the construction
of the 43-meters dome (record of span dome for several centuries), and the use of light
aggregates wisely distributed.
From a geometrical point of view, the Pantheon, has been built with a spherical geometry. That
geometry is far from being the funicular configuration due to self weight and it has been chosen
because it represents the heaven with evident reference to gods, to which the building is
devoted.
In order to improve the structural behaviour of the dome a series of strategies have been
adopted for changing the distribution of load: the variation of the dome thickness, the use of
concrete with different weight, the addition of external rings. As shown in figure 2-13, the density
of the concrete is between 1350 kN/m3 close to the crown dome and 1600 kN/m3 of bottom part
of the dome and walls. Romans had a large experience in the design of circular arches so they
were aware of the importance of adding more loads close to walls in order to 'verticalize' the
load.
Figure 2-13: Concrete density of Pantheon dome.
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-31-
The Roman Pantheon, as described by Robert (Robert and Hutchinson 1986), actually is cracked
along vertical meridians in the bottom part, creating independent meridian elements in that area;
actually its structural behaviour is a mix between independent arches in the bottom part and a
dome in the top part. In the top area the structural behaviour has to been considered in three
dimensions, otherwise it is not possible to have an oculus at the crown.
An approximated analysis has been performed, as if the structure is working as a series of
independent arches with a section corresponding meridian planes of 22,55° as shown in figure 2-
14.
Figure 2-14:. Independent arches used for analysis
In the following figure 2-15 three antifunicular geometries corresponding to different conditions
are illustrated. In the first case (1) a constant thickness has been considered and the different
density of concrete has not being taken into account. In the second case (2) the real thickness of
the dome has been considered, but ignoring the different density. In the third case (3) both
effects, related to different density and thickness, have been considered. Figure 2-15 shows the
three funicular geometries corresponding to the previous described cases. The eccentricity
between line of centroids and funicular is 1,07 m for the first case, 0,57 m for the second one and
0,54 m for the third one.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-32-
Blue line: Line of centroids
Red curve: case 1
Green curve: case 2
Violet curve: case 3
Figure 2-15: Line of thrust for different load cases
The previous comparative analysis about the distribution of loads on the structural behaviour
illustrates the great knowledge of Romans, who employed a new material in an efficient way with
an extraordinary intuition and talent for changing the thrust line in order to adapt it to the chosen
geometry.
Positioning loads at the top of elements addressed to receive horizontal thrust (walls, flying
buttress), with the aim of 'verticalize' the resulting axial force is an example of prestressing used
in roman and gothic constructions. The main difference, with the modern prestressed system, is
that in the latter case the system is auto-equilibrated, while the in first case the reaction are
increased.
2.5.2 Salginatobel Bridge
Arch bridges with hinges at the crown and at the springing lines have been developed firstly by
French and German engineers at the start of the second half of the XIX century in order to solve
the difficulties met designing and building fully restrained arch bridges.
In order to minimize bending in the arch due to dead load, the axis of the arch should be located
along the dead load pressure line (Menn 1990). If the bridge is hinged at the springing lines and at
the crown, the calculation of forces is easier and the funicular line is unique because three point
of its geometry are identified. Arch bridges, fully or partially restrained at the springing lines,
show different structural behaviour if the thrust line does not match with the axis of the arch.
Figure 2-16 illustrates a recent photo of the Salginatobel Bridge, a 132 meter span concrete bridge
hinged at the springing lines and at the crown. It was designed by Maillart [1871 1940] in 1929.
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-33-
Figure 2-16: Salginatobel Bridge by Maillart
The structural design of the bridge was faced by its author using graphic statics: a first geometry
of the arch was chosen using that approach and after several more in-depth analysis the
geometry did not change. A recent work has been developed by Fivet (Fivet and Zastavni 2012),
who, starting from original sketches of the designer, illustrated the importance of graphic statics
and its development with recent tools.
Figure 2-17a shows the distribution of permanent loads of the Salginatobel bridge, while figure 2-
17b illustrates the thrust line and its comparison with the axis of the bridge. There is a perfect
match close to springing lines, while when there is no perfect match, such as at the middle half of
the span, bending moments arise but the cross-section is bigger.
a) Distribution of permanent loads
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-34-
b) Axis line of the arch compared with thrust line.
Figure 2-17: Analysis of the bridge using graphic statics. Taken from Fivet (Fivet and Zastavni 2012).
The above briefly described bridge represents a fantastic example of beauty, efficiency and
economy. The previous analysis evidences that the funicular curve could not match perfectly the
axis of the arch: it is important to find a suitable solution according to theorems of plasticity.
2.5.3 The Tiemblo bridge
The Tiemblo bridge above the Burguillo river is a 165 meter concrete arch bridge supporting a
roadway. The total length is 268 meters. It has been designed by FHECOR Ingenieros
Consultores.
As it is shown in figure 2-18, the geometry of the arch evidences some singularities: variable cross
section with depth between 3,10 meters (l/53) at the springing lines and 1,75 meters (l/94) at the
crown, a constant width of 4.00 meters. The rise of the bridge is 22 meters and the ratio λ=l/f is
7.5.
Figure 2-18: Front view and cross section of the Tiemblo bridge. Courtesy of Fhecor Ingenieros Consultores
2. STATE OF THE ART: THE CONCEPT OF FUNICULARITY
-35-
Using graphic statics, as shown in figure 2-19, it is possible to find a line of thrusts obtained for
loads due to arch, girder and columns that approximately matches the axis of the bridge.
Figure 2-19: Graphic statics applied for finding the thrust line of the Tiemblo bridge due to own weight.
2.6 Final remarks
The definition of the ideal form for a given set of loads, known as the funicular or anti-funicular
curve, depending on the sign of the axial force to be mobilised, has been a very used principle
over last centuries in a more or less intuitive way.
In this section an historical modus operandi has been described: its utilization is revitalized thanks
to modern tools, that help designer to think about the structural behaviour, something that not
always happens due to the massive presence of advanced finite element software.
The Pantheon in Rome, as the roof of the Main Train Station in Berlin shown in figure 2-20,
designed by SBP (SBP 2014), represents the answer to the same problem: established a geometry
(a sphere and a three-centred arch, respectively, for cultural and functional reasons) it is possible
to change the distribution of the load in order to have the line of thrusts corresponding with the
line of centroids. The problem is the same, but it has been solved with different strategies. In the
first case adding external loads, changing the density of the material; in the second case adding
prestressed cables.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-36-
Figure 2-20: Main Train Station in Berlin. Photo taken from SBP (SBP 2014).
During last decades several interesting tools have been developed in order to perform form
finding analysis, saving computational cost, within friendly and innovative environments.
Analysis illustrated in this section have been performed using some of these modern tools,
Geogebra (GeoGebra 2013) and Grasshopper (McNeel 2014) that, working into an interactive
environment, allow the user to change parameters involved into the geometry and distribution
of loads, in order to evaluate different structural solutions.
Virtual tools, together with physical models, could have an enormous benefit to understand the
behaviour of different kind of structures. Physical models, totally disused during the last decades,
could be another time interesting tool thanks to the development of 3d printers and CNC
machines. The idea of exploring, designing, printing and testing a structure is a ambitious goal,
but probably achievable in few years.
The integration of conceptual design and CAD/CAM techniques can allow to build formwork in a
easy way and with a multitude of materials (wood, plastic, corrugated paper, that could be
recycled or not), allowing the development of novel construction systems that make these
fascinating structures more cost-effective.
Finally, the concept of funicularity could be extended easily to a three-dimensional problems: the
physical concept is exactly the same.
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-37-
3 ANTIFUNICULAR SPATIAL ARCH BRIDGES FORM FINDING USING AN
INTEGRATED AND INTERACTIVE APPROACH
El nacimiento de un conjunto estructural, resultado de un proceso creador, fusión de técnica con arte,
de ingenio con estudio, de imaginación con sensibilidad, escapa del puro dominio de la lógica para
entrar en las secretas fronteras de la inspiración. Antes y por encima de todo cálculo está la idea,
moldeadora del material en forma resistente, para cumplir su misión.
Eduardo Torroja
3.1 Introduction
Curved structures, if they are the outcome of an appropriate design, have the capacity to carry
large loads and cover important areas saving costs and can represent examples of the highest
structural efficiency. In designing shells or arches, geometry plays a primary role: enough energy
should be spent for a proper design of the structural shape. Acting loads and geometry are the
essential factors involved in the structural behaviour.
Pedestrian bridges are peculiar structures because they allow designer to create a contextualized,
unique, beautiful structure considering the human scale of the user, covering both functional and
aesthetical aspects (Stein 2010). It is not unusual that the footbridge deck has to adapt to a
gulations,
which entails the use of long ramps. Usually designers incorporate external ramps located to
decrease the slope as much as possible to fulfil the admissible slope limit. An alternative to that
solution is to increase the length of the deck, by making it curved in plant. Structurally the curved
deck could be supported by an arch, or can be cable-stayed or suspended by a cable system.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-38-
Curvature can be a symbol for grace and beauty: a designer who is able to take advantage of
curvature will design structures with efficiency of performance and elegance of form. Some
example are shown in figure 3-1.
a) Sassnitz Bridge, Germany, designed by SBP (SBP 2014).
b) Ponte del Mare, Italy, designed by Miranda. Taken from (Studio de Miranda Associati 2013)
Figure 3-1: Examples of curved bridges
The purpose of this section is to present a integrated and interactive approach to shape
antifunicular spatial arches that support a curved decks using secondary cables. Sarmiento-
Comesias (Sarmiento-Comesias et al. 2013) defines spatial arch bridges, as bridges in which
vertical deck loads centred on the deck induce internal forces not contained in the arch plane.
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-39-
To design efficient spatial arch bridges, the same concepts used for planar arch bridges,
described in the previous section, should be maintained: the most important is to choice the
geometry of the arch coinciding with one of the antifunicular curves.
As described above, the funicular curve is subjected to tensile loads only if the geometry and
loads do not change during the life of a structure; this situation is incompatible with real
structures showing the first limit of this design method: a geometry can be funicular only for a
unique distribution of loads. For each distribution of loads there are infinite funicular curves: the
variation of the horizontal reaction or of the rise or of the entire length allows to obtain infinite
configurations of compression-only geometries.
As already mentioned, this section illustrates an integrated approach to design antifunicular
spatial arch bridges developed in a parametric, three-dimensional and innovative environment.
In this work only spatial arches which support a curved deck using secondary ties will be
considered. In this section first the state-of-the-art of spatial arch bridges is presented; then, a 3-
dimensional design method and its implementation is described.
3.2 State-of-the-art of spatial arch bridges
The history of curved bridges is rather recent. Even more is so is the history of arch bridges with
curved decks. An in-depth overview on spatial arch bridges has been published by Sarmiento-
Comesias (Sarmiento-Comesias et al. 2013) covering over 80 spatial arch bridges designed and
built around the world.
The first idea of conceiving a spatial curved bridge came from the genius of Leonardo da Vinci
who in 1502 sent to Turkish Sultan Bayezid II his designing proposal to erect a bridge between
Galata and Istanbul with a span of 240 m, as shown in figure 3-2.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-40-
Figure 3-2:sketch of the Golden Horn Bridge designed by Leonardo da Vinci in 1502 (Biblioteque Institute Paris ).
The first constructed bridges of this nature were the Ziggenbach Bridge (1924), Landquart Bridge
(1930), Bohlbach Bridge (1932) and Schwandbach Bridge (1933), all designed by Maillart
(Laffranchi and Marti 1997).
Some of the best known contemporary designers of inclined arches are Santiago Calatrava, Jiri
Strasky, Wilkinson Eyre with Flint & Neil and Javier Manterola (Baus and Schlaich 2008). Several
curved pedestrian bridges have been constructed in UK before the end of the Millennium and
valid examples are the Millennium Bridges in York and in Gateshead (Davey and Forster 2007;
Curran 2003). Other examples in UK are the Merchants' bridge in Manchester by Whitbybird
and the Butterfly Bridge in Bedford by Wilkinson Eyre Architects and Jan Bobrowski & Partners.
Due to the strong need for urban bridges in Spain, several arch spatial bridges have been
designed by Calatrava (La Devesa footbridge in Ripoll, the Campo de Volatin bridge in Bilbao,
Port of Ondarroa bridge, Alameda Bridge in Valencia) and by Manterola, from Carlos Fernandez
Casado (the bridge over the Galindo River in Bilbao). Another example of spatial arch bridge is
the Footbridge Rari-Nantes in Padua, with span 75m, designed by Enzo Siviero. An outstanding
example of an existing funicular curved arch bridge with superior deck is the Ripshorst
Footbridge, shown in figure 3-3, designed by SBP (SBP 2014). As described in Schober (Schober
2003) its funicular geometry was found using a hanging model.
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-41-
Figure 3-3:Ripshorst Footbridge by SBP (SBP 2014)
The first in-depth theoretical studies on this topic have been developed by Jorquera (Jorquera
Lucerga and Manterola Armisen 2012; Jorquera Lucerga 2007; Jorquera Lucerga 2009), Lachauer
(Lachauer and Kotnik 2012) and by Sarmento-Comesias (Sarmiento-Comesias et al. 2013). Figure
3-4 shows two funicular spatial arch footbridges proposed, respectively, by Romo (Romo 2014)
for a competition in Salford and by SBP for a competition in Mettingen (Stein 2010).
a) Salford footbridge proposal by Romo. Taken from Romo (Romo 2014)
b) Mettingen footbridge proposal by SBP. Taken from Stain (Stein 2010)
Figure 3-4:: spatial arch footbridge proposals
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-42-
3.3 Designing spatial arch bridges
3.3.1 An interactive and integrated environmental workspace: Rhinoceros,
Grasshopper, Karamba
The procedure to design antifunicular spatial arch bridges has been developed thanks to a
combination of different software (Rhinoceros, Grasshopper and Karamba) in a parametric,
three-dimensional and innovative environment.
The development of generative design methods enables the designer to describe mathematically
complex form and explore new ones (Georgiou 2011; Georgiou et al. 2011).
Rhinoceros is a commercial NURBS-based 3-D modelling software; NURBS (non-uniform rational
B-splines) are mathematical representations that can accurately model any shape from a simple
2-D line, circle, arc, or box to the most complex 3-D free-form organic surface or solid. Because of
their flexibility and accuracy, NURBS models can be used in any process from illustration and
animation to manufacturing. NURBS geometry is an industry standard for designers who work in
3-D where forms are free and flowing (Robert McNeel & Associates 2013). During last decade
Rhinoceros has gained popularity in architectural design thanks to its plug-in Grasshopper
(Robert McNeel & Associates 2014).
Grasshopper provides for a new way for creating parametric geometries. It is a graphical
generative design editor that allows users, without knowledge of scripting, to generate
parametric complex geometries in a very intuitive way. This tool allows to create relations
connecting parameters and component with cables. Components are designed with input and
output, representing the input data and the output after performing script stored into the
component. The interface is enjoyable to use and encourages the development of new ideas. The
geometry generated using Grasshopper is transferred to Rhino in real time and this allows to
perform a fast visual check of the model. As briefly described in section 1, there is a strong need
for this kind of tools, devoted to design, to explore and not only to analyze structures.
Grasshopper, different to typical other software, is not a 'black box' because the user has total
control of what the software is doing because the user has programmed it.
Grasshopper allows to perform integrated finite element analysis (FEA) thanks to its plug-in
Karamba (Preisinger and Heimrath 2014). Karamba allows to interactively calculate the response
of three-dimensional structures under the action of external loads. Starting from generated
geometries in Grasshopper, Karamba is able to convert geometry into structural elements and to
assemble a model defining geometrical and mechanical properties of beams, loads and supports.
The FE Analysis performed by Karamba is linear elastic, but geometrical non-linear behaviour
could be taken into account using a "large deformation" component. This component performs
geometric non-linear analysis by an incremental approach: all external loads are applied in steps
starting from an initial geometry. Calculations and loads are not considered as a whole, but are
gradually increased until successive states of equilibrium are solved: after each step the model
geometry updates to the deformed state. Reducing the step size it is possible to have a better
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-43-
approximation of geometric non-linearity. Using the incremental method, inevitably the solution
will drift from the true equilibrium curve (Lewis 2003): considering this tool as a fast way to
produce shapes during conceptual design, these kind of approximations are acceptable because
they are conceptually correct.
The incremental approach makes use of the tangent stiffness Kt (instead of the stiffness matrix
used in static analysis) which relates small changes in load to small changes in displacement. At
each load step ΔF, Karamba calculates the incremented displacement Δu, which is:
t
F
uK
(3.1)
After each step the tangent stiffness matrix is taken as the one obtained from the previous step /
iteration. The procedure continues until all the load is applied. Convergence checks are made and
iterations are stopped as soon as the state of equilibrium is reached. The final solution is reached
when all (internal and external) forces reach equilibrium.
In real time software for structural applications, like Karamba + Grasshopper, user interaction,
visualization of the model and structural response occur simultaneously, without the need for
the user to explicitly run the analysis (Clune 2010). This kind of tools are valuable for exploring
new shapes and pushing a deeper understanding of structural behaviour. A geometric modeller
and an analysis software converge in a unique tool allowing the user to explore a wide range of
possible geometric configurations to design sustainable and efficient structures. The combined
use of these tools allows the user to explore new forms and shapes. Substantially there is a
revolution with respect to the classical procedure: create geometry - analyze - review results
because there is a real-time interaction between geometry and results.
3.3.2 Form finding methodology for spatial antifunicular arches
In this section, the methodology SOFIA (Shaping Optimal Forms with an Interactive Approach)
for the generation of antifunicular arch spatial bridges in a integrated approach is in-depth
described.
A well known method for finding the antifunicular shapes of an arch consists in loading a cable
(very low flexural stiffness compared with axial stiffness) with a known set of loads; the cables
have to be subdivided in elements connected with hinges, like a chain, in order to find a
compression-only geometry. This method is equal to the hanging models used by Gaudí and Isler
(Adriaenssens et al. 2014).
This kind of analysis is characterized by a strong geometric non-linearity. Starting from a generic
geometry, the tool, performing an incremental analysis, allows to obtain different funicular
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-44-
geometries depending on the applied load. Some tests have been performed for simple
configurations of loads. Starting from a simple circular arch with the following sets of loads:
projected distributed loads directed along the global z-axis, distributed loads directed along the
global z-axis, projected distributed loads directed along the local z-axis. Expected geometries,
respectively, a parabola, a catenary and a circular arch have been obtained.
The above illustrated methodology has been implemented for shaping three-dimensional
geometries with loads acting with any value and direction in order to find a funicular geometry
for a spatial arch bridge. Since the loads acting on the arch depend on shape of the deck, this
method is iterative. Usually after few iterations the solution is reached. The method is described
step-by-step in figure 3-5.
In the following paragraph the steps are described:
- the geometry of deck, the number and the position of hangers, and the initial geometry of arch
are necessary data for the first model of the bridge. The geometry of the deck is arbitrarily
curved. The starting geometry of the arch is arbitrarily chosen;
- loads are evaluated based on the dominant load case. All dead loads, related to arch and deck,
and one half of live loads are taken into account for designing the geometry of the arch. That
combination of loads is frequently used in order to decrease effects of live load during the life of
the bridge;
- forces acting on cables are evaluated taking into account an infinite stiffness of the hangars;
- the form finding analysis of the arch is performed taking into account the axial forces of the
hangars and the dead load of the arch. The final geometry has the total length of the initial one,
i.e. the an infinite axial stiffness has been considered. The analysis is stopped when a funicular
curve is reached, i.e. the maximum value of eccentricity ei is less than an arbitrary allowable value
of the eccentricity, e*. The eccentricity is calculated for each i section as:
22
yz
ii
ii
MM
eN
(3.2)
where Ni is the axial force in each section of the arch, Miy and Miz are the bending moments for
each section of the arch, respectively, around the y- and z-local axis. The methodology is
developed, not with real areas in order to avoid eccentricities due to the real deformability of the
structure; as already it has pointed out by Jorquera (Jorquera Lucerga and Manterola Armisen
2012) considering that there are no external torsional moments affecting the arch, the torsional
moments automatically vanish if the bending moments are set to zero. In other words: torsional
effects can be neglected;
- once the analysis has been completed, the bridge is regenerated with the deck and the updated
geometry of the arch. The funicular behaviour of the arch is checked. If the bending moments are
too big, the updated geometry of the arch is used for a new iteration. If the funicular
configuration has been reached the form finding analysis is completed.
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-45-
- once the final geometry of the arch is found, preliminary sizes of the members can be
determined using the computed forces; prestressing loads are imposed on cables in order to
obtain the small displacement of the deck determined from analysis with axially rigid members
adopting real properties of sections and materials.
Figure 3-5:Description of the form-finding process SOFIA
The resulting funicular arch is not only curved in elevation, but also double-curved in plan.
Examples of antifunicular geometries for bridges with 'C' and 'S' shape curvature of the deck are
shown in figure 3-6.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-46-
a) antifunicular arch for bridges with C-shaped deck
b) antifunicular arch for bridges with S-shaped deck
Figure 3-6: Resulting antifunicular arch for bridges with C- and S-shape curvature of the deck
Since hand calculations cannot be performed for these complex structures with sufficient
accuracy because curvature results in complication in computations, a validated finite element
3. ANTIFUNICULAR SPATIAL ARCH BRIDGES
-47-
software, Sofistik, has been used for checking the funicular geometries obtained with the new
design tool. There are no relevant differences between the results.
After the geometry has been created, since the funicular configuration of the arch has been
found for one set of loads, the designer should test results with all possible scenarios to consider
other variables such as real properties of materials, different distribution of loads, non-linear
behaviour, vibrations, construction stages and time-dependent effects.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-48-
4 PARAMETRIC ANALYSIS ON THE INFLUENCE OF DECK-ARCH
RELATIVE TRANSVERSAL POSITION
Più che parlare di rapporti tra Architetto, Ingegnere e Costruttore mi pare che si dovrebbe esaminare
come queste tre mentalità e competenze possano e debbano fondersi per raggiungere quella unità da
cui nasce la vera Architettura
Pier Luigi Nervi
4.1 Cases of study
As an example illustrating a case of SOFIA, parametric analyses have been performed for spatial
arch footbridges with C- and S-shapes of the deck. The starting geometries of the bridge before
form finding analysis are shown, in figure 4-1a and 4-1b, respectively, for deck with C and S
geometries.
a) C-shaped configuration
4. PARAMETRIC ANALYSIS
-49-
b) S-shaped configuration
Figure 4-1: Cases of studies for the parametric analysis
In its C-shaped configuration, the bridge has one span of 100m, while for the S-shaped
configuration the bridge has two spans, both of 50m. The rise (fa) for both bridges is 45m.
The geometry of the deck is usually fixed by functional constrains (Romo 2014) while the position
of the arch could be slight variable: the analysis has been carried on C- and S-shaped bridges
using as parameter the distance y between supports of deck and abutments of the arch. The
effect of the variation of y on deck, arch and cables has been evaluated.
Referring to figure 4-1, the main dimensions and parameters considered in the study are shown in
table 4-1.
Table 4-1: Main dimensions and parameters considered in the parametric analysis
[m]
C-Shape
S-Shape
da
100
100
fa
45
45
dd
100
50
fd
10
10
y
-10<y<3
-10<y<3
The bridge is only subjected to vertical loads, which are assumed to be uniformly distributed on
both deck and arch. The considered loads are 8.55 kN/m and 30 kN/m, respectively for arch and
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-50-
deck. Dead load of the arch corresponds to an arch with a steel hollow circular transversal section
with diameter of 2m and thickness of 17mm. The load of the deck corresponds to a steel deck
with a width of 5m, 3,50 kN/m2 of dead load and one half of live load, 2,50 kN/m2. The adopted
loads and geometries are similar to the real one of the Millennium Bridge in Gateshead (Davey
and Forster 2007).
The following figure 4-2 shows the starting geometry of the deck with C-shape and the
corresponding funicular geometry of the arch, obtained varying the value of y. The figures on the
left shows a perspective view while the images on the right show a lateral view.
y=-10
y=-9
4. PARAMETRIC ANALYSIS
-51-
y=-8
y=-7
y=-6
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-52-
y=-5
y=-4
y=-3
4. PARAMETRIC ANALYSIS
-53-
y=-2
y=-1
y=0
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-54-
y=+1
y=+2
y=+3
Figure 4-2: Antifunicular configuration for C-shape deck
4. PARAMETRIC ANALYSIS
-55-
Fgure 4-3 shows the starting geometry of the deck with S-shape and the corresponding funicular
geometry of the arch, obtained varying the value of y. The figures on the left shows a perspective
view while the images on the right show a lateral view.
y=-10
y=-9
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-56-
y=-8
y=-7
y=-6
4. PARAMETRIC ANALYSIS
-57-
y=-5
y=-4
y=-3
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-58-
y=-2
y=-1
y=0
4. PARAMETRIC ANALYSIS
-59-
y=+1
y=+2
y=+3
Figure 4-3: Antifunicular configuration for S-shape deck
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-60-
4.2 Discussion of results
The proposed form-finding method for shaping antifunicular spatial arch bridges, called SOFIA,
has been used to study the influence of the relative position between deck and arch on the
structural behaviour of the bridge. The aim of this analysis is to find the optimal position of the
arch's abutments with respect to the deck in order to minimize the cost of the bridge.
In figure 4-4 the axial forces along the arch have been plotted versus the curvilinear coordinate of
the arch for different values of y, respectively for C-and S-geometries of the decks.
a) C-shaped deck
4. PARAMETRIC ANALYSIS
-61-
b)S-shaped deck
Figure 4-4: Axial forces along the arch
The minimal axial force in the arch correspond to values of y between 7 and 5 for the C-shaped
deck. Figure 4-4b shows the asymmetrical distribution of the axial forces in the arch due to the
asymmetric position of the S-shaped deck. Only for the case y=0 the position of the deck is
symmetrical respect to the arch and therefore the distribution of the axial forces is symmetrical.
In figure 4-5 the summation of axial forces of all j cables has been plotted for different values of y,
respectively for C- and S-shaped deck.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-62-
a) C-shaped deck
b)S-shaped deck
Figure 4-5: Summation of axial force of cables for different values of y.
Figure 4-5 illustrates that the optimal position of the deck, corresponds to y=-6 and y=o for,
respectively, C- and S-shaped decks.
The values of axial forces in the C- and S-shaped deck are plotted versus the curvilinear
coordinate of the deck for different values of y in figure 4-6.
a) C-shaped deck
4. PARAMETRIC ANALYSIS
-63-
b)S-shaped deck
Figure 4-6: Axial forces in the deck
Figure 4-6a shows a symmetrical distribution of axial forces in the deck. The minimum absolute
values correspond to the situation in which y is included between 5 and 7. The difference
between the maximum and minimum axial force for each value of y, is almost constant for all
cases.
It is interesting to observe that the deck could be compressed or tensioned depending on the
direction of the cable respect to the curvature of the deck. If the cable direction is toward the
inner side of the curvature, the horizontal thrust compresses the deck, while for the opposite
case the horizontal thrust results in tension in the deck.
The previous comments of results illustrate that the optimal position of the arch's abutments to
design the most efficient spatial arch with C-shaped deck is around y=6. In this case the distance
between the centre of gravity of the deck and the vertical plane passing from the abutments of
the arch is minimum.
The same conclusion could be drawn for a spatial arch with S-geometry of the deck. The most
efficient solution coincides with y=0, the unique value of y for which the deck is symmetrical with
respect to the arch. Furthermore, the centre of gravity of the deck is contained in the vertical
plane passing through the abutments of the arch.
In order to validate the conclusions obtained for fd=10m, more parametric analysis have been
performed with fd=5m and fd=15m. Relevant plots only for fd=5m are shown in the following
figures.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-64-
Figures 4-7, 4-8 and 4-9 show, respectively, the distribution of axial forces in the arch, the
summation of axial force of cables and the distribution of axial forces in the deck for different
values of y in the case of fd=5m.
a)C-shaped deck
b)S-shaped deck
Figure 4-7: Axial forces along the arch
4. PARAMETRIC ANALYSIS
-65-
b)C-shape deck
b)S-shape deck
Figure 4-8: Summation of axial force of cables for different values of y.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-66-
b)C-shape deck
b)S-shape deck
Figure 4-9: Axial forces in the deck
The previous figures 4-7, 4-8 and 4-9 illustrate that for the C-shaped deck, the most cost-effective
bridge is obtained for y=3, while for the S-shaped deck, the best solution corresponds to y=0.
Both results point out that the deck optimal position minimizing the axial forces of arch and
4. PARAMETRIC ANALYSIS
-67-
cables corresponds to the situation for which the distance between deck's centre of gravity and
vertical plane passing from arch's abutments is minimum. The same conclusion has been
obtained with fd=15m.
4.3 Conclusions
In this section, SOFIA, a methodology to approach the conceptual design of spatial arch bridges
in order to obtain a funicular configuration of the main arch, has been presented. The main
parameters affecting the antifunicular shape of the arch are the geometry of the deck and the
arch-deck transversal relative position. Using SOFIA, parametric studies have been performed in
this section.
The presented parametric analyses allow designers to find the optimal position of the deck with
respect to the arch in order to minimize the cost of the bridge. Results show that the optimal
position of the deck minimizing the axial forces of arch, deck and cables corresponds to the
situation for which the distance between deck's centre of gravity and vertical plane passing from
arch's abutments is minimum.
The parametric analyses have been performed for different values of curvature for both, pointing
out the independence of conclusions from the curvature of the deck.
This design tool, in combination with profound knowledge, could be an efficient method for the
conceptual design of arch spatial bridges and a source of inspiration and creativity for the
designer.
5. CONCLUSIONS AND FUTURE STUDIES
-69-
5 CONCLUSIONS AND FUTURE STUDIES
The overall quality of many structures today leaves much to be desired. The rapid technological
progress does not reflect adequately in their variety, beauty and sensitivity. Too often structural
engineers neglect the creative conceptual design phase by repeating standard designs and not
sufficiently contributing with own ideas to the fruitful collaboration with architects. Engineers thus
often waste the chance to create building culture.
Jörg Schlaich
5.1 Summary of results
5.1.1 The funicular principle
The world is faced with challenges in all three dimensions of sustainable development: economic,
social and environmental. In cities of middle- and high-income countries, investment in
infrastructure, renewable energy, buildings, and improved electricity and water efficiencies is
important (DESA United Nations 2013). There is a great awareness on saving energy and
resources designing efficient structures.
Curved structures, if outcome of an appropriate design, have the capacity to carry elevate loads
and cover important areas saving costs and representing the highest examples of structural
efficiency: the prima donna among all structural typologies. Furthermore, curved structures
combine high structural performance and beauty, so are strongly used as architectural and civil
structures.
In designing shells or curved bridges, geometry plays a primary role, so a high attention should be
spent for a proper design of the structural shape. An inappropriate choice of the geometry
causes a series of problems that will be grow during the project, the construction and the life of
the structure.
The definition of the ideal shape for a given set of loads, known as the funicular or anti-funicular
curve, depending on the sign of the axial force to be mobilised, has been a very used principle
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-70-
over last centuries in a more or less intuitive way. That concept has been in-depth investigated
and illustrated.
Before the digital era, a lot of designers were interested in finding optimal geometries of curved
structures: using different form finding techniques, intuition and great knowledge of structural
behaviour they were able to construct great structures. Some examples of those constructions
have been shown in this work. An old modus operandi has been recovered, which actualization is
revitalized thanks to modern interactive tools, that invite users to think about structural
behaviour: something that does not always happen with more conventional tools, that use finite
element analysis.
During the last decades several interesting tools have been developed in order to perform shape
finding analysis, saving computational cost, within friendly and interactive environments. These
virtual tools together with physical models can have a positive effect on the degree of
understanding the structures behaviour. This way of investigating structures would also be a task
for education and research. It would be desirable that the Spanish Technical Universities, as well
as other Technical Schools in the world, such as MIT (Masonry Group 2009), Cambridge, USA,
ETH (Block Research Group 2014), Zurich and EPFL Lausanne (IBETON EPFL 2013), Switzerland,
strengthen teaching and management of these tools.
Engineers and architects design approaches should be carried out in parallel, in order to allow a
holistic solution which would be optimized in terms of both environmental and structural
performance (Shepherd and Richens 2012).
5.1.2 Designing antifunicular spatial arch bridges
In this work a new integrated and interactive approach for designing spatial antifunicular arch
bridges has been presented. Spatial arch footbridges represent an innovative answer to demands
on functionality, structural optimization and aesthetics for curved decks, quite popular in urban
contexts.
A new computational strategy, called SOFIA, has been developed in order to find an anti-
funicular arch configuration for any geometry of the deck: an extension of physical hanging
models in space. The resulting geometry has a double curved spatial configuration integrating
structural efficiency and an esthetical appeal.
The method is suited for the conceptual design phase of a project. After a funicular geometry has
been found for the predominant load case, more refined analysis must be performed in order to
check the structure for deflections, vibration, live loads, fatigue, time-dependent effects, etc...
Parametric analyses have been performed in order to evaluate the influence of deck-arch relative
transversal position on the structural behaviour of the arch. It has been show that the eccentricity
5. CONCLUSIONS AND FUTURE STUDIES
-71-
between the arch springings and the deck abutments plays a key role in finding the antifunicular
configuration.
The optimal position of the arch, in order to build a cost-effective bridge, has been found. Results
show that the best position of the deck minimizing the axial forces of arch and cables
corresponds to the situation for which the distance between deck's center of gravity and vertical
plane passing from arch's abutments is minimum.
The presented design tool, in combination with deep knowledge, could be an efficient method
for the conceptual design of antifunicular arch spatial bridges, and a source of inspiration and
creativity for the designer.
5.2 Further studies
5.2.1 Introduction
The developed interactive approach, SOFIA, has been used to study the influence of deck-arch
relative transversal position. It can be employed to analyze the effect on the arch due to the
variation of deck curvature. Superior deck antifunicular spatial arch bridges can be studied, with
an easy improvement of SOFIA.
In-plane and out-of-plane buckling, dynamic, temperature and fatigue effects have not been
taken into account in this study: they can be object of further researches.
In this work only uniformly distributed loads have been considered, however it can be interesting
to take into account point loads or distributed loads applied only in a limited part of the bridge.
Furthermore it can be interesting to study non-funicular spatial arch bridges, as for example
spatial arch bridges with imposed curvature (where the arch has the same curvature of the deck
(Sarmiento-Comesías et al. 2012)) or curved bridges characterized by a vertical arch.
In order to build a double curved arch the feasibility of manufacturing of its shape must be taken
into account. Fabrication aspects could be introduced as parameters in the form finding process.
A concrete spatial curved arch would require expensive formworks, complicated scaffolding and
manual casting; using steel it would be possible to use different kind of profiles that can be easier
changed in size and thickness. For these reasons concrete arches are rarely used in pedestrian
bridges (Keil 2013).
Nowadays there is an incredible development of CNC machines and 3D printers that use CAD-
CAE-CAM technology to create any geometry and shape; formworks could be directly printed
with any geometry, or real scale steel elements of the bridge could be produced. This is a very
promising field of research that merits strong consideration.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-72-
Two lines of research have been started to investigate the influence of deck cross-section, cables'
position and arrangement on the bridge behaviour.
5.2.2 Torsional behaviour of curved decks
The curved deck of an antifunicular spatial arch, as for classical cable-stayed bridges, can be
suspended on its outer or inner edge, or on both of its edges or on its central axis (Strasky 2005).
The cross section behaviour depends fundamentally on the supporting system, whether it is
placed inside or outside the curve, taking into account that a fluid route trajectory of pedestrians
and cyclist requires to have the supporting system placed on one side only (Keil 2013).
It is well-known that the statical behaviour of a curved decks is completely different compared to
straight ones. The transfer of loads cannot be separated clearly into longitudinal and transverse
actions because of three dimensional interactions: in curved bridges there is a reciprocal
relationship between torsion and bending. In the case that the radius of curvature is very large,
the two situations could become similar. In bridges with curved deck, torsional moments in
addition to longitudinal bending moments should be taken into account; for bridges with an
angle of curvature <30° coupling effect can be neglected (Kolbrunner and Basler 1969).
A clear understanding of the effects of torsion on structural members is essential to design
efficient curved structures. A structural element, in general, exhibits two ways to resist torsion.
The first, commonly called Saint-Venant torsion, generates a shear flow in the transversal cross
section, while the second, commonly named warping torsion, generates normal and shear
stresses.
In the majority of practical applications, only one structural mechanism is relevant, the other
could be neglected. In closed cross sections (solid or hollow) the warping mechanism could be
neglected, while for open sections that mechanism is the most relevant. It is important to point
out that the preponderance of one of the two structural mechanisms depends on cross section
properties, length of the structural element and support conditions.
There are two main strategies to design structures in which relevant torques are applied: the first
is the use of common cross sections adequately strengthened in order to resist torques, the
second consists in varying the geometry of the cross section (centre of torsion cannot match the
barycentre) and/or modifying the position of supports in order to generate torques for
decreasing torsional effects.
5. CONCLUSIONS AND FUTURE STUDIES
-73-
Transversal position of cables plays a key role in deck behaviour and cross section design. Varying
the cables transversal arrangement it is possible to change totally the structural behaviour of the
bridge. That topic is currently under development.
5.2.3 Arrangement of layout of cables
A research is currently ongoing to analyze analyzing the effects of hangers arrangement on the
antifunicular geometries including buckling effects and the effect of live loads.
The first investigated hanger configurations are shown in figure 5-1.
a) vertical hangers
b) radial hangers
c) hangers distributed according a growing linear law
d) hangers distributed according a decreasing linear law
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-74-
e) Nielson hangers arrangement
e) Nielson-network hangers arrangement
Figure 5-1: different arrangements of cables layout
PUBLICATIONS BY AUTHOR
-75-
BIBLIOGRAPHY
Adriaenssens, S., Block, P., Veenendaal, D., and Williams, C. (2014). Shell structures for architecture:
form finding and optimization. Routledge.
Allen, E., and Zalewski, W. (2009). Form and Forces Designing Efficient, Expressive Structures. John
Wiley & Sons.
Allen, E. (1984). Pietre di Puglia : dolmen, trulli e insediamenti rupestri. Adda, Bari.
Baus, U., and Schlaich, M. (2008). Footbridges construction, design, history. user, Basel;
Boston.
Biblioteque Institute Paris. "Paris Manuscript L, Folio 65v and 66r."
Block Research Group. (2014). "Web page of Block Research Group." http://block.arch.ethz.ch/
(05/15, 2014).
Canino. (2014). "Il portale culturale della Tuscia." www.canino.info (06/12, 2014).
Clune, R., Connor, J. J., Ochsendorf, J. A., and Kelliher, D. (2012). "An object-oriented architecture
for extensible structural design software." Computer and Structures, 100-101 1-17.
Clune, R. P. (2010). "Explorative structural design". PhD. Massachusetts Institute of Technology,
Boston.
Como, M. T. (2007). L'architettura delle "Tholoi" micenee: aspetti costruttivi e statici. Università degli
studi Suor Orsola Benincasa, Napoli.
Corres, P., Hugo, (2013). "Sound engineering through conceptual design according to the fib
Model Code 2010." Structural Concrete, 14(2), 89-98.
Curran, P. (2003). "Gateshead Millenium Bridge, UK." Structural Engineering International, 13(4),
214-216.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-76-
Da Vinci, L., and Richter, J. P. (1970). The notebooks of Leonardo da Vinci. Dover Publications, New
York.
P. Davey, and Forster, K. W. (2007). "Exploring boundaries the architecture of Wilkinson Eyre."
DESA United Nations. (2013). "World Economic and Social Survey 2013. Sustainable
Development Challenges." United Nations Publication .
Fernández Casado, C. (2006). La arquitectura del ingeniero. Colegio de Ingenieros de Caminos,
Canales y Puertos, Madrid.
Fivet, C., and Zastavni, D. (2012). "Robert Maillart's key methods from the Salginatobel Bridge
design process (1928)." Journal of the International Association for Shell and Spatial Structures,
53(171), 39-47.
Galafel González, J. (2011). "Estudio de elementos de fábrica por superficies antifuniculares".
E.T.S.I. Caminos, Canales y Puertos (UPM), Madrid.
Galilei, G. (1638). due nuoue scienze attenenti alla
mecanica i movimenti locali. Appresso gli Elsevirii, Leida.
GeoGebra. (2013). "GeoGebra Website." www.geogebra.org (06/07, 2013).
Georgiou, O. (2011). "Interactive Structural Analysis and Form Finding." IABSE IASS Conference .
Georgiou, O., Richens, P., and Shepherd, P. (2011). "Performance based interactive analysis."
Design Modelling Symposium Berlin .
Gregory, D. (1697). "Catenaria." Philosophical Transaction of the Royal Society, 19(231), 637-652.
Heyman, J. (1966). "The stone skeleton." Int.J.Solids Structures, 2(2), 249-279.
Hooke, R. (1676). A description of helioscopes and some other instruments. Printed by T.R. for John
Martyn, London.
IBETON EPFL. (2013). "i-structures. Intuitive understanding of structures." https://i-
structures.epfl.ch/ (02/07, 2013).
Jorquera Lucerga, J. J. (2009). "Structural behavior of spatial arch bridges." Evolutions and trends in
design, analysis and construction of shell and spatial structures. Proceeding of IASS Symposium.
Valencia. Editorial de la Univercnica de Valencia.
Jorquera Lucerga, J. J., and Manterola Armisen, J. (2012). "An iterative form-finding method for
antifunicular shapes in spatial arch bridges." Comput.Struct., 108-109 42; 42-60; 60.
PUBLICATIONS BY AUTHOR
-77-
Jorquera Lucerga, J. J. (2007). "Estudio del comportamiento resistente de los puentes de arco
espaciales". E.T.S.I. Caminos, Canales y Puertos (UPM).
Keil, A. (2013). Pedestrian bridges. - r Internationale
Architekturdnchen.
Kilian, A., and Ochsendorf, J. (2005). "Particle-Spring Systems for Structural Form Finding." Journal
of the International Association for Shell and Spatial Structures, (148), 77-84.
Kolbrunner, C. F., and Basler, K. (1969). Torsion in structures: an engineering approach. Springer
Verlag.
Lachauer, L., and Kotnik, T. (2012). "Curved bridge design" Computational Design Modelling,
Springer Berlin Heidelberg, 145-152.
Laffranchi, M., and Marti, P. (1997). "Robert Maillart's Curved Concrete Arch Bridges." Journal of
Structural Engineering, 123(10), 1280-1286.
Lewis, W. J. (2003). Tension structures: form and behaviour. Thomas Telford, London.
Lobbecke, R. (2012). Corbelled domes. ln.
Masonry Group. (2009). "Web page of Masonry Group at MIT." http://web.mit.edu/masonry/
(05/05, 2014).
R. McNeel. (2014). "Grasshopper generative modeling for Rhino. Computer software."
http://www.grasshopper3d.com/ (05/05, 2014).
Menn, C. (1990). Prestressed concrete bridges. user Verlag, Basel, Switzerland; Boston.
Muttoni, A. (2011). The art of structures ; introduction to the functioning of structures in architecture.
EPFL Press/Routledge, Abingdon, Oxford, UK; New York, NY.
Poleni, G. (1982). Memorie istoriche della gran cupola del tempio vaticano. INTEMAC, Madrid.
Popovic Larsen, O., and Tyas, A. (2003). Conceptual structural design: bridging the gap between
architects and engineers. Thomas Telford, London.
Preisinger, C., and Heimrath, M. (2014). "Karamba - A Toolkit for Parametric Structural Design."
Structural Engineering International Structural Engineering International, 24(2), 217-221.
Ramm, E. (2004). "Shape Finding of Concrete Shell Roofs." Journal of the International Association
for Shell and Spatial Structures, (144), 29-40.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-78-
Ramos Casquero, A. (2011). "Caracterización estructural de los rellenos en bóvedas históricas de
fábrica". Trabajo de investigación tutelado. E.T.S.I.C.C.P. de la Universidad Politécnica de Madrid.,
Madrid.
Ramos Casquero, A., and León González, J. (2011). "Importancia estructural de los rellenos en el
trasdós de bóvedas de construcciones históricas. ." Actas del V Congreso de la Asociación Científico
Técnica del Hormigón Estructural (ACHE). Barcelona, .
Robert McNeel & Associates. (2014). "Grasshopper generative modelling for Rhino. Computer
software. www.grasshopper3d.com." 0.9.0075.
Robert McNeel & Associates. (2013). "Rhinoceros 5 User's guide" .
Robert, M., and Hutchinson, P. (1986). "On the structure of the Roman Pantheon." The Art
Bulletin, 68(1), 24-34.
Romo, J. (2014). "Managing constrains in footbridges design: conceptual design and context."
Footbridge 2014. 5th International Conference Footbridges: Past, present & future.
Romo, J. (2013). "El desarrollo de la creatividad en la ingeniería estructural: herramientas para su
enseñanza." III Jornada Internacionales de la enseñanza de la ingeniería estructural de ACHE, 512-520.
Sanitate, G., Todisco, L., and Monti, G. (2014). "Effective assessment methodology for trulli in
Apulia, Italy." Proceedings of 9th International Masonry Conference, Guimares, .
Sarmiento-Comesias, M., Aparicio, A. C., and Ruiz-Teran, A. M. (2013). "State-of-the-art of spatial
arch bridges." Proceedings of the Institution of Civil Engineers: Bridge Engineering, 166(3), 163-176.
Sarmiento-Comesías, M., Ruiz-Teran, A., and Aparicio, A. (2012). "Structural Behavior of Inferior-
Deck Spatial Arch Bridges with Imposed Curvature." J.Bridge Eng., 17(4), 682-690.
Sasaki, M. (2005). Flux structure. Toto, Tokyo.
SBP. (2014). "sbp.de." www.sbp.de (06/03, 2014).
Schlaich, J. (1996). "On the conceptual design of structures: an introduction ." International
symposium on 'conceptual design of structures', Proceedings Volume 1. University of Stuttgart,
October 7-11, 1996, Stuttgart, Germany. Institut fur Konstruktion und Entwurf II, Universitat
Stuttgart, Stuttgart, Germany, 15-25.
Schober, H. (2003). "Steel Castings in Architecture and Engineering." NASCC proceedings.
Shepherd, P., and Richens, P. (2012). "The case for subdivision surfaces in building design." Journal
of the International Association for Shell and Spatial Structures, 53(4), 237-245.
PUBLICATIONS BY AUTHOR
-79-
Squier, E. G. (1877). Peru; incidents of travel and exploration in the land of the Incas, Harper &
Brothers, New York.
Stein, M. (2010). "Curved Pedestrian Bridge Straightforward Design." Structures Congress 2010,
American Society of Civil Engineers, 262-275.
Strasky, J. (2005). Stress ribbon and cable-supported pedestrian bridges. Thomas Telford, London.
Studio de Miranda Associati. (2013). "Studio de Miranda Associati Website."
http://www.demiranda.it/ (06/03, 2013).
Timoshenko, S. (1953). History of strength of materials : with a brief account of the history of theory
of elasticity and theory of structures. McGraw-Hill, New York.
Treccani. (2014). "Treccani. L'enciclopedia italiana." www.treccani.it (June, 02, 2014).
Varignon, P. (1725). canique ou statique 1. 1. Jombert, Paris.
Veenendaal, D., and Block, P. (2012). "An overview and comparison of structural form finding
methods for general networks." Int.J.Solids Structures, 49(26), 3741-3753.
An integrated approach to conceptual design of arch bridges with curved deck - Leonardo Todisco
-80-
PUBLICATIONS BY AUTHOR
- Todisco, L., (2011) "Test evidence for applying strut-and-tie models to deep beams and D-
regions of beams,"
October 2011.
- Sanitate, G., Todisco, L., and Monti, G. (2014). "Effective assessment methodology for trulli in
Apulia, Italy." Proceedings of 9th International Masonry Conference, Guimares.
- Reineck, K.-H.; Todisco, L.; (2014) -slender Reinforced
ural Journal. V. 111, No. 1-6, January-December 2014.
- Todisco, L.; Reineck, K.-H.; Bayrak, O.; -slender beams with
- Todisco, L.; Reineck, K.-H.; Bayrak, O.; Database of Shear Tests for Non-slender Reinforced
In preparation.
- la comprobación de
- . In
preparation.
- An integrated approach to conceptual design of arch bridges with curved deck
I.S.B.N.
E.T.S.I.C.C.P.
U.P.M.
