Computers and Structures 170 (2016) 1325

Contents lists available at ScienceDirect

Computers and Structures

journal homepage: www.elsevier.com/locate/compstruc

An optimization model for the design of network arch bridges

D. Bruno, P. Lonetti , A. Pascuzzo
Department of Civil Engineering, University of Calabria, Via P. Bucci, Cubo39-B, 87030 Rende, Cosenza, Italy

a r t i c l e

i n f o

Article history:
Received 21 August 2015
Accepted 28 March 2016
Available online 16 April 2016
Keywords:
Network arch bridges
Structural optimization
Finite element analysis
Design
Sizing optimization

a b s t r a c t
A new design methodology, which evaluates the optimum configuration of network arch bridge schemes
is proposed. A three-step optimization algorithm is implemented in a FE model, with the purpose to
evaluate the bridge optimum configuration, involving the lowest material quantity and the best strength
performance level in all structural members of the bridge. The stability and the efficiency of the formulation were verified with respect to several bridge configurations ranging from small to large spans.
Moreover, parametric results are presented to investigate the interaction between cable-system, girder
and arch, giving rise to specific analyses useful for design purposes.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Tied arch bridges can be considered as enhanced construction
schemes, which are able to provide aesthetic, structural and economic performances to overcome small, medium and large spans
[1]. Most of the existing bridge configurations consist of an arch
and a girder, whose internal transferring forces are guaranteed
by the cable system, typically formed by cable elements. The cable
arrangement plays a fundamental role in the structural behavior,
since it is able to strongly influence the internal stress distribution
as well as the deformability properties of the entire bridge structure [2]. Several hanger geometries, such as vertical, inclined
V-shaped or network, are frequently utilized for design purposes
in tied arch bridges. From the structural point of view, V-shaped
or network (inclined hangers with multiple intersections) arch
bridges are preferred to conventional vertical hanger bridge
typologies, since they are able to guarantee a high efficient
response, which minimizes bending effects in both arch and girder
[3]. However, the hanger arrangement, especially in network arch
bridges, can be considered as a complex structural system, whose
elements, i.e. the hangers, interact by means of tension only internal forces with girder and arch [4]. In particular, each element of
the cable-system is affected by geometrical nonlinearities arising
from cable sag effects, which strongly influence the actual stress
distribution in the bridge components. As a consequence, a fundamental task to be achieved is the evaluation of the initial configuration under dead loads in terms of internal stresses and strains of
Corresponding author.
E-mail addresses: d.bruno@unical.it
(D. Bruno),
(P. Lonetti), arturo.pascuzzo@unical.it (A. Pascuzzo).
http://dx.doi.org/10.1016/j.compstruc.2016.03.011
0045-7949/ 2016 Elsevier Ltd. All rights reserved.

paolo.lonetti@unical.it

bridge constituents, by means of a nonlinear field model, avoiding

unexpected and unrealistic stresses distribution in the cable due to
compressive forces. Moreover, it is required to identify the optimum design configuration consistently with a performance based
approach, in which a proper choice of hangers, arch and girder
dimensioning should be determined to take into account external
loads [5,6].
In the literature, most of the analyses are carried out for conventional vertical hanger bridge schemes, in which heuristic models and preliminary design rules can be adopted because of the
quite standard stress distribution in the bridge components. As a
matter of fact, under live loads, the design of the main constituents
is not complex at all, since the cable elements interact with the
arch and girder by means of uncoupled vertical forces. In this
framework, the evaluation of the initial configuration can be identified by using traditional zero displacement method, enforcing
the girder to remain straight under the application of the dead
loads [711]. The extension of the structural analysis in the framework of arch bridges with inclined hanger elements is not straightforward, especially in those cases in which a large number of
variables is involved in the analysis, i.e. in long span bridges. In this
framework, the determination of the initial cable configuration
under dead loads as well as the design of the bridge components
under the action of the external loads should be considered as an
important task to be achieved. A review of the literature dealing
with the analysis of network arch bridges denotes that although
such structures are receiving much attention in the last decades,
many points still remain to be addressed satisfactorily. Most of
the present studies on network arch bridges propose design
specifications and guidelines on bridge dimensioning by means

14

D. Bruno et al. / Computers and Structures 170 (2016) 1325

of preliminary design rules [3,4]. Notable parametric studies and

relevant guidelines useful for design purposes are proposed in
[2], in which investigations, in terms of arrangement of the hangers, arch configuration and geometrical characteristics of the
bridge constituents, are carried out. Moreover, fatigue behavior
of cable system elements in terms of hanger arrangements is investigated in [12], in which comparisons, in terms of hanger distribution based on radial, constant or constant change slope
configurations are proposed. However, most of the models available in the literature do not enter in detail in the calculation of
the bridge configuration under dead loads, in terms of both initial
cable force distribution and arch-tie geometric profile. The identification of such configuration is quite important, in relationship to
the nonlinear behavior of the cable-system elements, which could
be affected, under the external loads, by unexpected relaxation
effects of the hangers, producing overstressing in the adjoining
hanger elements and in both girder and arch [5,13]. Another
important issue is that preliminary design rules, available from
the literature, provide information on a reasonable dimensioning
of the structural components, but not on the best distribution of
material according to the Performance Based Approach (PBA) [14].
Currently, tied arch bridges are designed by using conventional
methodologies, which consist of heuristic procedures based on the
experience and expertise of the designer. Although parametric
studies are carried out on several classes of structures, the procedure to reach the best possible performance design solution still
remains extensive and quite difficult to be achieved. In the framework of tied arch bridges, to the best Authors knowledge only very
few models are concerned to investigate the optimum bridge configuration. Preliminary works are developed in [2,15], in which
parametric studies in terms of hanger arrangement with the purpose to minimize bending moments in both arch and girder are
presented. Moreover, a design methodology based on classical
structural optimization is proposed in [16], in which the optimum
configuration of vertical hanger bridge arrangement is discussed in
terms of hanger cable-force distribution and material quantity
involved in the bridge constituents. The optimum design configuration of network arch bridges is investigated in [17], in which
an optimization method is developed to minimize the cost of
superstructure (arch and hangers) in terms of geometric shape, rise
to span ratio, cross sections of arch and hangers. In this framework,
relevant results and guidelines for the design of network arch
bridges are proposed. Optimization methods are typically
employed to determine the bridge dimensioning, by minimizing
a convex scalar function, which combines variables concerning
the geometry of the structure and the internal stress/strain distribution. However, the use of pure optimization methods, especially
in the case of complex and large structures, such as the network
arch bridges, are affected by convergence problems in the solving
procedure, due to the large number of variables. Moreover, the
consistency of the solution is not guaranteed in the final optimum
configuration, since the solving procedure may lead to a local minimum or unpractical results from the engineering point of view
[10,14,18]. Alternatively, advanced formulations based on metaheuristic algorithms are frequently utilized in those cases in which,
multivariable or multicriteria affect the final optimum configuration. However, although the basic idea of heuristic methods is
conceptually simple, the application of a generic metaheuristic to
a generalized optimization problem requires a laborious implementation with specific guidelines, which introduce numerical
complexities in the formulation. Therefore, in order to avoid such
problems, in the present paper a design procedure based on an
iterative methodology developed in the framework of
performance-based optimization techniques is proposed. Despite
existing methods available from the literature, a simple and effective methodology, easy to be implemented in several FE software,

is proposed. The design variables are the post tensioning forces in

the hangers and the initial strains in the girder and arch, which
identify to the initial configuration under dead and permanent
loads. Moreover, under the external loads, bridge geometric characteristics, which involve the lowest possible material in the cable
system, girder, and arch and verify prescriptions arising from
external loads, are determined. In order to prove the effectiveness
of the proposed model, parametric studies in terms of cable system
configurations on several bridge schemes are proposed. The outline
of the paper is as follows. In Section 2, the formulation of the
design methodology, bridge modeling, together with the description of the iterative procedure is presented. In Section 3, numerical
details on the design method are reported, whereas in Section 4,
numerical comparisons and parametric results are presented.
2. Formulation of the procedure
2.1. Bridge modeling
The bridge typology, reported in Fig. 1a, refers to a generalized
tied arch scheme, in which arch and girder are connected between
themselves at the bridge extremities, whereas the girder is
assumed to be simply supported to the foundation system. Moreover, the hangers may be arranged in various configurations such
as vertical, V-shaped or network. Without loss of generality, the
arch and girder are assumed to be in steel, whereas hangers are
made of steel cable, requiring prestressing forces. The proposed
optimization method, is presented for a Network Arch Bridge
(NAB) scheme, which is, in comparison to the existing ones based
on vertical or V-shaped (Fig. 1b), the most complex configuration
to be analyzed. However, the theoretical formulation of the proposed model is quite general to be implemented also for conventional vertical or V-shaped cable arrangements.
The proposed strategy, despite existing optimization methods,
identifies the optimum configuration on the basis of a step-bystep procedure, in which the solution is enforced by using physically based expressions. The heuristic nature of the proposed
procedure does not ensure a priori that the predicted solution is
a global optima. However, the optimum solution is determined,
iteratively, by means of successive approximations of the final configuration by solving separate optimization problems implemented
in different substeps. The proposed design methodology is based
on a three-step analysis, described in the next three subparagraphs, in which the optimum solution is determined by using
an iterative procedure based on results obtained under Dead Loads
(DL) and Live Loads (LL) combinations. In the present approach the
design variables are the initial stresses in the hangers under DL, the
cross sections of the hangers, girder and arch, whereas the characteristics of the cable-system (angle, number of cables) are typically
assumed by the designer due to aesthetic requirements and thus
are not included as variables in the optimization procedure.
2.2. Analysis under the action of dead loads (STEP 1)
Under the action of DL, it is required to evaluate internal stresses and deformations, which enforce the prescribed design geometry, known in the literature as zero configuration. In this
framework, the unknown quantities are represented by the internal stresses of the cable-system elements, the initial position of
the arch and the girder, which are determined in such a way to
reproduce the design undeformed configuration. The hangers
should be designed in terms of post-tensioning forces to reproduce
the initial configuration under dead loads, i.e. zero configuration.
Moreover, the cross-sections are calculated to verify strength
prescriptions (maximum and fatigue stresses) under the action of

15

D. Bruno et al. / Computers and Structures 170 (2016) 1325

X2

h
X1

(a)

(b)
Fig. 1. Bridge configuration and representative geometric lengths (a) existing configurations of tied arch bridges (b).

the live loads. The use of an optimization algorithm to simultaneously design hangers under both dead and live loads may lead to
severe computational efforts with convergence problems in the
solving procedure. A design procedure based on different steps
seems to circumvent the numerical difficulties to achieve the final
solution. Moreover, the division in two substeps is quite consistent
to the current design approach in bridge engineering, in which at
first, the zero configuration and subsequently the behavior under
the action of live loads are evaluated. Finally, improvements on
the convergence behavior, obtained by using the iterative procedure, can be explained in relationship to the dominant truss behavior of the cable system, which partially reduces the coupling
behavior between each element of the cable system. It is worth
noting that the number of unknown quantities, represented by
post tensioning stresses in the cables, is larger than the number
of available constraint equations defined by enforcing zero vertical
displacements at the intersection points between cable and girder.
As a consequence, in order to determine the initial configuration, a
numerical procedure for solving an indeterminate system of equations is required. To this end, an optimization problem is implemented, in which the objective scalar valued function g is
defined as the norm of girder vertical displacements under the
action of dead and permanent loads and the control variables are
the post tensioning forces in the hangers SC . Moreover, constraint


equations are introduced for the final stresses of the hangers and
the deformations of the arch and girder. Therefore, the optimization problem can be expressed as follows:

8
T
>
min SC g kU G2 X 1 k1 ; with SC  fSL1 ; . . . ; SLNL ; SR1 ; . . . ; SRNR g
>
>


<
s:t: 0 6 SLi 6 SA ; 0 6 SRj 6 SA i 1 . . . NL ; j 1 . . . NR
>
>
>
: G G G
N E0 ; U 1  0; N A E0A ; hU 2A i 0

where k  kp is the Lp norm of the function  with p = 1, i.e.,

1=p
R LG
k  kp 0 j  jp dX 1 , hi is the notation utilized to indicate the
R LA
average value of the function , i.e. hi L1A 0 dS, with LG and

LA are the total length of the girder and the arch, X1 and S are the
spatial coordinates fixed on the girder and arch profiles, respec-

tively, U G2 X 1 is the function of girder vertical displacements,

SRi ; SLj and SRi ; SLj are final and initial stresses of the generic hanger
element of the cable system measured from the Right (R) or Left (L)
of the bridge extremities (Fig. 2), SA is the upper bound of the allowable stresses involved in the hangers and N R ; NL are the corresponding total number of hangers involved in the cable system
for the right (R) and left (L) orientations. Moreover, NG ; N A are


the constraint operators, which control the deformations in the arch

Fig. 2. Design variables involved in the optimization problem to determine the

initial configuration under DL.

and girder under DL. In particular, in order to avoid arch shortening

and tie lengthening, proper values of initial axial strains in the arch
EG0 and in the girder E0A are enforced to have zero horizontal
U G1 U G1 X 1 L and average vertical hU 2A i displacements, in the
girder and the arch, respectively. A synoptic representation of the
variables involved in Eq. (1) and the corresponding notation is
reported in Fig. 2.
2.3. Analysis under the action of live loads (STEP 2)
The design of structural components is developed on the basis
of common prescriptions arising from Ultimate, Serviceability
and Fatigue Limit State method, i.e. ULS, SLS, FLS, respectively.
The main aim of this step is to evaluate the optimum solution in
terms of the lowest possible material quantity involved in the
bridge components, taking into account both strength and
deformability prescriptions available from common codes on
bridge structures [1921]. In particular, the design methodology
should be able to verify, for each hanger, the following conditions
arising from the limit state design:

h
i
LR
max Sij

ULS

h
i
LR
6 SA ; max DSij

FLS

h
i
6 DSA ; max jU G2 j

SLS

6 dGA ;

where i = 1, . . . , NL or j = 1, . . . , NR are used to identify Left (L) or Right

(R) oriented hangers, respectively, SA and DSA are the maximum
(strength) or incremental (fatigue) allowable stress values for each
cable and dGA is the maximum displacement (vertical) of the girder.

16

D. Bruno et al. / Computers and Structures 170 (2016) 1325

The

evaluation

of

the

hanger

cross-section

vector

A fAR1 ; . . . ; ARNR ; AL1 ; . . . ; ALNL g, is developed, as described in Fig. 3,



by introducing two design factors for each element of the cable sysLR

LR

LR

tem, namely Uij and Xij . In particular, Uij modifies the cross
section of the generic hanger with the aim to verify recommendations on the admissible strength, namely Eq. (2).1 and Eq. (2).2. In
LR
ij

addition, the factors X

take into account deformability prescrip-

tions on the girder, i.e. Eq. (2).3, increasing at the intersection points
with the girder, the stiffness of those cables affected by displacements larger than the allowable ones. The evaluation of the vector
A is achieved by using a secant approach, in which the current value

one. Contrarily, in the case of a positive value, an increment of the

global bridge stiffness is required, which is performed by improving the stiffness of the cable elements. Such stiffness increment
will interact with the ones predicted in the arch and girder, as
LR

described subsequently. In particular, the definition of the Xij

for the generic element of the cable system is based on the following piecewise functions:

LR k
ij 

8
>
>
<

LR

LR k

LR k1

ALR ij Uij Xij  Aij

; i 1 . . . NL ; j 1 . . . NR

LR

The factors Uij are defined by means of a linear extrapolation

between the maximum applied stress values reached under LL and
the corresponding allowable strength. In order to verify prescriptions concerning both ULS and FLS, the worst design criterion is
considered in the definition of the generic performance factor, as
follows:

LR k
ij 

3
LR k
LR k
maxSLL ij maxDSLL ij
6
7
L
R
max 4 ULS
; FLS
5;i 1...N ;j 1...N
SA
D SA
4

where SLL and DSLL are the values of the stresses for the right (R) or
the left (L) oriented cables observed in the ULS or FLS combinations,
LR

respectively. The prediction of the factors Xij is strictly connected

to the following limit function, which indicates, at the generic
cable/girder intersection point, the ratio between the current value
of girder deflection and the allowable value:
k

k
g GLR X 1 ij

maxjU G2ij jLR

SLS

dGA

 1; i 1 . . . NL ; j 1 . . . NR

maxjU2ij j
>
>
: SLS G LR
G

dA

of the generic cross section is obtained as a function of the previous

estimate, at the k  1-th iteration, by means of the following multiplicative relationship:

if g GLR X 1 ij 6 0

LR

into account several cases, reported concisely in Table 1, depending

if ULS/FLS or SLS prescriptions are satisfied or not. However, the
fourth case excludes the over use of the material, since when both
strength and deformability requirements are verified a lower prediction of the cross-section is necessarily enforced depending from the
ratio between current stress and allowable quantity. Such condition
LR

is applied until the values of Uij differ from the unity. The design of
both girder and arch is considered by solving two uncoupled optimization problems, in which the lowest material quantity involved
in such bridge components is achieved. In particular, the objective
scalar valued functions, which are minimized during the solving
procedure, correspond to the required steel quantities involved in
the bridge components, i.e. Q G or Q A . Moreover, the control variables
are represented by the characteristic lengths, which describe the
cross section shape, i.e. t G1 ; t G2 ; . . . ; t GnG and t 1A ; t2A ; . . . ; tnAA . Finally,
constrain equations, concerning the design criteria adopted for the
arch or the girder, i.e. D A or DG complete the optimization problem,
which is described by the following expressions:

8
>
min Q GA kcGA AGA k1
>
>
< AGA
s:t: AG tG1 ; t G2 ; . . . ; t GnG > 0; A A t 1A ; t2A ; . . . ; t nAA > 0;
>
>
>
: A
Dj 6 aA ; DGj 6 aG ; j 1 . . . NULS

with
k

aG aG

k1

B
min @1;
0

LR

are necessary and thus the factors Xij are supposed to be equal to

LR

Therefore, during the iterations, Uij and Xij are evaluated taking

At a generic point of the girder, prescriptions on bridge

deformability are satisfied if the functions g GR or g GL are strictly negative; for these values, no increments of the cable system stiffness

if g GLR X 1 ij > 0

@1;

k
C
A; a A
maxjU G2 X 1 j
X 1 2G
1

d2A

A k1 B

1
dG2

maxjU 2A X 1 j

C
A

X 1 2A

where k  kp is the Lp norm of the function  with p = 1, cG;A is the

weight density of the girder or the arch AG and A A are the corre-

Fig. 3. Identification of the performance factors and the design criteria of arch and
girder.

sponding cross section areas, DjA and DGj represent the j-th values
of the design criterion involved in the bridge components and
NULS is the number of loading combinations defined by the design
recommendations. Moreover, aA and aG correspond to the limit
design thresholds, which modify the allowable maximum stress
levels in the bridge components on the basis of the requirement
achievements on bridge deformability. In particular, aA and aG are
equal to one, when bridge displacements are in agreement with
the serviceability displacement prescriptions; in such cases, the
design of the bridge components, i.e. arch and girder, are developed
essentially with respect to the maximum strength criterion.
Contrarily, when deformability prescriptions are not verified, an
increment of current stiffness is required and thus the design is performed consistently with the maximum displacement approach.
Such task is performed by reducing the maximum stress level,

17

D. Bruno et al. / Computers and Structures 170 (2016) 1325

Table 1
LR
LR
Synoptic representation of the procedure utilized to predict the factor Uij and Xij in the STEP 2.
Case

Verified

Predicted

Verified

Predicted

Predicted

Predicted

Verified

Predicted

(Strength) ULS/FLS

NO

ULR
ij > 1

NO

ULR
ij > 1

YES

ULR
ij = 1

YES

ULR
ij 6 1

(Deformability) SLS

NO

XLR
>1
ij

YES

XLR
=1
ij

NO

XLR
>1
ij

YES

XLR
=1
ij

reached in the bridge, below the value provided by the design stress
criterion. It is worth noting that Eqs. (6) and (8) enforce the design
conditions on strength or deformability on the basis of a secant
approximation, whose final optimum configuration is guaranteed
by the iterative nature of the proposed algorithm. Such assumption
appears to be quite reasonable in relationship to the global behavior
of the network arch bridges, which typically are considered as
prevalent truss structures with reduced bending moments [4] or
at least a continuous beam with multiple elastic supports with
relative stiffness proportional to the mechanical characteristics of
the hangers [10]. The connection between each hangers is dealt
with by means of a needle-eye device, which allows for hanger
passing through another. As a consequence the cable system is
based on a dominant truss behavior, in which a modification of
the geometry in the hangers, produce small perturbations in the
stress internal resultants of the adjoining ones.
2.4. Tolerance condition (STEP 3)
The tolerance conditions should be checked to verify if the current solution has reached the converged configuration. The present
stage should be considered after the evaluation of the initial configuration (STEP1) and the design of the bridge components under
the external loads (STEP 2). In particular, it is required to verify if
the design obtained at the current iteration (k), in terms of crosssections of the bridge components, does not differ from the one
obtained in the previous iteration step (k  1) by means of the
following tolerance condition:
9
8
>
>
>
>
>
>
2
3
>
>
>
LR k
LR k1
G k
G k1
A k
A k1 >
=
<
X
Aij  Aij
A


A


A

4
5
max
;
;
k1
k1
k1
LR
>
>
>
>
AG
A A
Aij
>
>
>
>
j 1;...;NR
>
>
;
:
;
L
i 1;...;N
6 toll:
9

If Eq. (9) is not satisfied, an update of the k-th current solution, in

terms of bridge geometry and initial stresses, should be carried out
and thus the algorithm continues to iterate. Contrarily, if tolerance
conditions are satisfied, the final solution is achieved. A synoptic
representation of the optimization procedure and the steps performed by the iterative procedure are reported in Fig. 4. Previous
procedure, presented for network systems, can be easily specialized
for other hanger arrangements, such as vertical or V-shaped arch
bridge schemes, because of the low complexities involved in these
cable systems. In particular, it is required to solve Eqs. (1) and (3) for
the evaluation of the initial configuration and cable dimensioning,
respectively, taking into account of the correct numbering of the
elements of the cable system. Moreover, governing equations concerning the optimization problems for the arch and girder, i.e. Eqs.
(7), should be solved separately, since they are uncoupled with the
analysis developed on the cable system element.
3. Numerical implementation of the model
The proposed formulation, presented in previous section from
theoretical point of view, is now described in its numerical imple-

mentation. In particular, the procedure consists of different steps,

which are executed iteratively:
(1) generation of the finite element formulation and use of
preliminary design rules;
(2) analysis under DL and identification of the initial configuration (STEP 1);
(3) calculation of the maximum stresses and displacements
under LL and prediction of the new geometry of the structural elements, i.e. arch, hangers and girder (STEP 2);
(4) check tolerance conditions and verify the consistency of the
design configuration (STEP 3).
The steps are reproduced by an external subroutine, which
interacts with Comsol Multiphysics and Matlab package [22,23].
In particular, the algorithm was implemented by means of proper
customized script files, which manage the parameters and the
results required by the iterative procedure. The proposed formulation is quite general to be implemented in several computational
frameworks, since it is based on data and results, which can be
easily extracted and managed from many standard commercial
FE software or by using conventional mathematical tools. In the
proposed analyses, the numerical formulation of the bridge is
based on a FE model, in which the arch and the girder are based
on Timoshenko beam formulation, whereas the hangers are discretized by using multiple truss elements in agreement with a
Multi Element Cable System formulation (MECS) [24]. The stiffness
reduction caused by sagging is accounted by using Green Lagrange
strain measure by expressing the global strains in tangential
derivatives and projecting the global strains on the cable edge.
Additional details on the approach here adopted to model nonlinear behavior of the cable elements can be found in [23]. The use of
1-D model, with respect to more enhanced analyses based on 2D or
3D FE formulations, appears to be more consistent with the purpose of the present study, mainly devoted to propose a simple consistent procedure from an engineering point of view. The initial
data, concerning the geometry and the mechanical properties of
the FE model are assumed by using preliminary design rules, which
consist of analytical expressions, obtained under very simplified
assumptions and structural schemes [15]. It is worth noting that
the use of such relationships is not mandatory, but it provides a
reasonable dimensioning for each structural element of the bridge,
reducing the number of iterations to evaluate the optimum solution. Once the FE model is generated, it is required to solve the
optimization procedure defined in Eq. (1), which corresponds to
the following nonlinear constrained optimization problem:

OBJ 1 ! Min GEG0 ; E0A ; SR ; SL ;



CE ! 0 6 SLi 6 SA ; 0 6 SRj 6 SA i 1 . . . NL ; j 1 . . . NR
EQ ! K U SLR SG0 S0A F DL ;
 

CE;

NA 0 ! N

CE;

NG 0 ! N

L
A R
E0A
0 Si ; Sj ; U

L
G R
EG0
0 Si ; Sj ; U


10

0;
0:

where G is the global objective function concerning the norm of vertical girder displacements as defined by Eq. (1), K is the stiffness


18

D. Bruno et al. / Computers and Structures 170 (2016) 1325

Fig. 4. Flow chart of design procedure.

matrix, U is the displacement vector, SG0 , S0A and SLR are the equiv

alent stress resultant vectors produced by the initial strain distribution in the girder, arch and cable system, respectively, and N0A ; NG0
are the constrain operator matrixes, which verify the initial bridge
configuration. Moreover, upper and lower bound constraints were
imposed on the design variables to ensure the accuracy of the explicit approximation. Since the proposed model is developed in the
framework of a NL formulation, the optimization problem requires
an iterative approach to determine the current solution. The governing equations concerning the STEP1 are solved in the framework
of gradient algorithms based on SNOPT method, in which the optimum solution is computed by the evaluation of the gradients of
both objective function and constraints by using numerical differentiation [23]. In particular, SNOPT uses a Sequential Quadratic Programming (SQP) algorithm, in which the objective function is
assumed to be a quadratic polynomial, whereas the constraints
are treated as linear. Finally, the optimum solution is derived iteratively by using the conjugate-gradient QP solver [25]. The analysis
under the action of LL was developed taking into account of the
stress and strain distributions arising from the DL configuration,
obtained by solving Eq. (10). However, in relationship to the

nonlinear behavior of structure, results under LL are obtained by

means of a restart or a continuation analysis, in which the initial
status of the structure coincides with the one obtained in the DL
configuration. The main aim of this step is to evaluate the worst
stress and displacement effects under LL, i.e. ULS, SLS, and to predict
the new values of hanger, girder and arch cross-sections. The
solution is performed for a proper number of loading conditions,
i.e. N LL , in which, for each bridge components, maximum stresses
and displacements are collected, by solving the following discrete
equations:

K U D U PLLi  P0 SR ; SL ; EG0 ; E0A i 1; . . . ; NLL

 

11

where K is the stiffness matrix of the discrete equations depending



on the current displacement vector U , P0 is the vector of nodal point



forces arising from the previous increment in element displacements and stresses from the dead load to the live load configurations, D U are the incremental displacement vector and PLLi is the


live load force vector of the current i-th loading combination. Starting from results obtained from Eq. (11), the design of arch, girder

D. Bruno et al. / Computers and Structures 170 (2016) 1325

and cable system elements is evaluated by using Eq. (3) or by

solving optimization problem described by Eq. (7). At this stage,
the analysis on the cable-system elements can be easily performed
LR

LR

in terms of optimization factors Uij and Xij , which are determined by means of Eqs. (4)(6) on the basis of maximum stress
and displacement values extracted by the LL combinations. Similarly, the design of arch and girder cross-sections is here performed
by solving the following optimization problems, which are executed
separately on the basis of the maximum stress resultants involving
the minimum safety factors in the arch or in the girder:
G

OBJ  1 ! MinAG GG A ;CE ! DG0 A 6 aG ;CVs ! A tG1 ;t G2 ;...;tGnG



OBJ  2 ! MinA A G A A A ;CE ! D0A A A 6 aA ;CVs ! A A t1A ;t 2A ;...;tnAA

12

where GG or G A are the global objective functions concerning the

volume involved in the Girder (G) or the Arch (A), DG0 or D0A corre

spond to the design criteria, here introduced as Constraint Equations (CE) for the ultimate limit state design of the girder (G) and
the arch (A) and the sets t G1 ; t G2 ; . . . ; t GnG or t 1A ; t 2A ; . . . ; t nAA are the
control variables of the optimization problem. Moreover, Eqs. (10)
and (12) refer to classical optimization problems expressed in terms
of minimum displacements or volume minimization involved in the
arch or in the girder. The optimum solution is determined
iteratively on the basis of previously converged values arising from
the k  1 iteration, which are considered as initial values in the next
substep, leading to relatively low computational efforts in the
solving procedure.
4. Results
4.1. Analysis for medium span tied-arch bridges
The consistency of the proposed model is investigated for an
arch bridge scheme with a medium span length. The main aim of
the present analysis is to verify, for a bridge scheme involving a
low number of variables, the convergence behavior of the iterative
procedure and the reliability of the optimum configuration. However, subsequently more complex cases concerning long span
bridges will be investigated. The bridge scheme, reported schematically in Fig. 5, presents a total length between vertical supports
and a width equal to 50 m and 10 m, respectively. Moreover, a
parabolic profile of the arch, with an aspect ratio H/L equal to
0.17 is considered (Fig.5). The arch and the girder are assumed to
be made of S420 steel material (fyk = 420 N/mm2), with a Rectangular Hollow Section (RHS). For the cable prestressing steel, an allowable stress corresponding to 45% of the ultimate tensile strength,
fpk = 1690 MPa, minimum fatigue strength equal to Dr = 200 MPa,
modulus of elasticity E = 200 GPa and specific weight c = 77 kN/m3
are considered [20]. The deck has concrete plate and transverse
beams, which are assumed to be simply supported by the longitudinal edge beams. The hangers are distributed by means of a radial
arrangement with a uniform distribution along the arch profile and
a constant radial angle / equal to 60 with a spacing step equal to
5 m. As a consequence, the total number of elements of the cable

19

system for each arch is 18, i.e. 9 for each orientation. The dead
loads concerning the secondary elements or Nonstructural Loads
(NSL), due to road pavement, concrete platforms and guardrail,
are equal to 80 kN/m and 50 kN/m, respectively. The live loads,
defined according to [21], present a transverse distribution based
on three lines of LL, whose equivalent load values, equal to
P
P
Q k 3i1 Q 1ki 1200 kN and qk 3i1 qki 42 kN=m, are applied
directly to longitudinal edge beams, considering several positions
to obtain the worst loading scenarios. The design criteria adopted
for the arch and girder are consistent with prescriptions concerning buckling strength analysis arising from the Method II of Annex
D by Eurocodes [19] and those recommended by parametric studies developed in [26]. Finally, only LL concerning traffic loads are
considered by using factored or unfactored loading combinations
equal to 1.35DL + 1.5NL + 1.5LL or to DL + NL + LL to analyze ULS
or SLS/FLS, respectively. However, the generalization of the proposed model for considering also the effects of seismic or wind
forces, can be developed introducing additional loading combinations. Moreover, at this stage, not much emphasis was considered
on rigorous recommendations arising from existing design criteria,
since the essential aim of the presented investigation is to verify
the efficiency of the optimization algorithm.
At first, convergence behavior is investigated in terms of representative variables of the cable systems, such as the cross section
area of the hangers and the stresses involved in the cables to predict the initial configuration. In particular, in Figs. 6 and 7, for each
iteration, the cross sections of the hangers for the Left (L) or Right
(R) orientations are reported as a function of the number of iterations (NIT) required to obtain the optimum solution. The analysis
denotes that, with respect to the initial values, the final solution
is strongly modified and presents a convergent evolution toward
the optimum configuration, with a relatively low number of iterations. Moreover, the distribution of the cross sections and the initial stresses under DL at the final design configuration are reported
in Fig. 8. The sets of the hangers with left or right orientations
denote their lowest or largest values in proximity of the arch
springing points, respectively, because of the rigidity of the connection between arch and tie. Such results are consistent with
existing studies available from the literature on network arch
bridge design, in which the inefficiency of the lateral cables in
the transferring forces between girder and arch was proved by
many investigations (see for instance [12]). The distribution of
the cross-section areas is not symmetric due to the presence of
the simply supported constraint at the right end of the girder.
The optimum solution is reached enforcing the lowest material
utilization in the design of the bridge constituents. Such task is
analyzed by the results, described in Fig. 9a and b, in which distributions of the optimization factors UR;L , related to design criterion
for representative elements of the cable system from the initial to
the final configurations, are reported. The results denote that all
elements of the cable system, in the worst loading scenario, reach
the maximum allowable strength value provided by the design
criterion, i.e. close to the unity. As a consequence, in such configuration, the steel quantity, involved in the cable system and
predicted by the optimization procedure, is the lowest possible,

Fig. 5. Synoptic representation and main data of the structural scheme.

20

D. Bruno et al. / Computers and Structures 170 (2016) 1325

0.325
0.300

0.275

ALSA/gL

0.250

0.075

0.050
(1),
(5),
(8),

0.025

(3)
(7)
(9)

0.000
0

10

12

14

NIT
Fig. 6. Convergence behavior of representative normalized Left (L) cross-section
areas as a function of the number of iterations (NIT).

Fig. 7. Convergence behavior of the normalized Right (R) cross-section areas as a

function of the number of iterations (NIT).

which ensures the best material utilization. The results denote that
the convergence behavior, in some hanger, appears to be irregular
toward the asymptotic value. Such phenomenon is produced by
the prevalence of the strength criterion utilized in the design of
the hangers, which can be related to the ULS (maximum stress)
or FLS (incremental stress) loading combinations. In order to verify
such occurrence, in Fig. 9a and b comparisons in terms of maximum values of the performance factors UR;L arising from both
design criteria, i.e. ULS or FLS, are reported. The results denote that
for the internal cables the worst criterion is the one concerning FLS
conditions, whereas, for the hangers located at bridge extremities,
ULS combinations produce the highest stress values. Moreover, the
evolution of the performance factors is not strictly convergent, but
it is affected by the prevalence of the worst design criterion, leading to discontinuities in the curves during the iteration steps.
The distribution of the cross-section characteristic lengths, the
steel quantities involved in the main constituents of the bridge
and the corresponding geometric properties are analyzed in Tables
2 and 3 and in Fig. 10. In particular, results reported in Table 2
show that the evolution of the cross-section lengths and the design
criteria presents an asymptotic convergent behavior toward the
optimum solution. Moreover, the values predicted by the design
criteria in the arch and girder reach the maximum allowable quantity, namely close to the unity, ensuring the lowest possible material quantity involved in such bridge components. From the results
reported in Fig. 10 or in Table 3, it transpires that the steel quantities, predicted in the arch and girder are quite comparable and are
much larger than the one involved in the cable system, i.e. within
the range between 12% and 15%. However, the importance of an
accurate design of the hangers should be considered not only for
structural reasons but also by the larger costs involved in such
elements than those required by girder and arch made of ordinary
steel, typically with ratios in the range between 3 and 7. Result
concerning geometric properties of the cross-sections denote that
dimensionless axial or bending stiffness ratios, i.e. IG/IA and AG/AA
respectively, differ within a range below than 15% (Fig. 10). Such
results are quite consistent with structural optimization achievements and current design approach developed in network arch
bridges, whose general aims are to minimize bending stresses
and girder/arch displacements, in such a way to reproduce a truss
structural scheme typically observed in the framework of network
systems. Moreover, since all the members under the worst design
scenario reach the lowest margin of safety, the optimum solution

Fig. 8. Distribution of the cross-section areas of the hangers and the stresses under DL as a function of the normalized girder position (X1/L).

D. Bruno et al. / Computers and Structures 170 (2016) 1325

21

Fig. 9. Distribution of the performance factor for the left (a) and right (b) hanger orientation as a function of the number of iterations (NIT) and the design criterion (FLS, ULS).
Table 2
Evolution cross section dimensions and the design criteria for the girder (DG) and arch
(DA) bridge components as a function of the number of iterations (NIT).
NIT

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

Girder

Arch

BG [m]

HG [m]

tG [m]

BA [m]

HA [m]

tA [m]

DG

DA

1.000
0.803
0.797
0.797
0.796
0.796
0.796
0.396
0.397
0.407
0.409
0.421
0.421
0.421
0.421

2.000
0.931
0.888
0.888
0.884
0.883
0.883
1.185
1.188
1.222
1.231
1.281
1.268
1.268
1.268

0.050
0.010
0.006
0.006
0.005
0.005
0.005
0.005
0.006
0.005
0.005
0.005
0.005
0.005
0.005

1.000
0.853
0.835
0.833
0.821
0.815
0.814
0.814
0.814
0.814
0.814
0.814
0.814
0.814
0.814

2.000
1.232
1.131
1.122
1.060
1.033
1.024
1.024
1.024
1.024
1.024
1.023
1.023
1.023
1.023

0.05
0.01657
0.00702
0.00512
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005

0.427
0.878
0.950
0.949
0.952
0.954
0.960
1.019
0.972
0.978
0.981
0.981
0.981
0.981
0.981

0.976
0.967
0.953
0.968
0.950
0.953
0.976
0.988
0.998
1.000
1.000
1.000
1.000
1.000
1.000
Fig. 10. Normalized material steel quantities involved in the Arch (A), Girder (G)
and Hangers (H), normalized ratios between cross-section areas (AG/AA) and inertial
moments (IG/IA) as a function of the number of iterations (NIT).

Table 3
Final values of the design variables related to the hangers.
Left

Right
2

Xi (m)

Si (MPa)

Ai (cm )

Xi (m)

Si (MPa)

Ai (cm2)

3.77
8.75
13.75
18.75
23.75
28.75
33.75
38.75
43.75

268
147
134
155
185
173
235
262
284

0.54
11.40
11.30
10.80
10.10
8.72
8.81
5.91
9.83

6.25
11.25
16.25
21.25
26.25
31.25
36.25
41.25
46.23

329
308
237
159
177
148
118
123
250

8.58
4.18
9.48
9.83
10.10
11.40
13.00
14.10
0.10

turns out to be a global minimum in the material quantity involved

in bridge constituents.
4.2. Analysis for long span bridge and comparisons with other bridge
configurations
In order to verify the reliability of the proposed methodology,
additional results are developed for a more complex bridge case,

involving a total of 180 m span, 25 m width and a rise to span ratio

equal to 1/6. In this framework, the analysis is extended also in
terms of cable system configurations taking into consideration
two different hanger arrangements, i.e. the network or vertical
(see Fig. 11). The main aim of the present results is to verify the
consistency of the optimization algorithm and to propose comparisons between different bridge typologies. The cable system of the
network arrangement is defined by a total number of 70 cables for
each arch and it is based on a constant radial angle / equal to 65
and a spacing step equal to 5 m. Moreover, in order develop a consistent comparison between the bridge configurations, the same
spacing step is also assumed for the bridge typology with vertical
hangers. The main constituents of the girder and arch are made
in steel with a RHS (Rectangular Hollow Section), whose geometric
values are considered by the optimization problem as design variables. Only the arch and girder cross sections enter in the design
procedure and not slab or transverse elements, which connect

22

D. Bruno et al. / Computers and Structures 170 (2016) 1325

the longitudinal beams. However, a generalization of the proposed

approach can be easily developed, also for girder typologies based
on reinforced concrete [15], just introducing the corresponding
strength criterion as required by the optimization algorithm
(see Eq. (12)). The dead loads are evaluated during the iteration
procedure because of the geometry changes involved by the
optimization algorithms, whereas the dead loads due to secondary
structural elements or nonstructural loads (road pavement, concrete platforms and guardrail) are equal to 170 kN/m and 50 kN/
m, respectively. Material properties concerning the main structural
elements, i.e. girder, arch and hangers, are assumed to be equal to
the previous investigated case concerning the bridge scheme
reported in Fig. 5. Similarly, according to the loading combination
defined in [21], live loads consist of two lines (NL) for each carriageway and are applied on the width of the bridge, with total loads
P L
P2NL
Q k 2N
i1 Q ki 2000 kN and qk
i1 qki 69 kN=m, to produce
the worst effects on the bridge components.
At first, the convergence behavior is analyzed in terms of
optimization factor evolution for the network system. In particular,
in Figs. 12 and 13, the relationships between right and left values
of UR;L as a function of the number of iterations (NIT) required by
the numerical procedure are presented. Similar results are proposed in Fig. 14, in which the distributions of the cross section area
in the hangers for a representative number of iterations are
reported. The results show that most of the hangers are designed
in such a way that, under the worst loading combination, reach
exactly the maximum strength, leading to the best performance
in the design and the lowest required material involved in the
cable system. Moreover, the cross-sections for the left and right
cable orientations denote approximately a symmetric distribution
along the girder projection length, which is basically altered by the
constrain conditions arising at the girder extremities. However, in
such regions, the algorithm predicts the lowest cross-section area,
which means that the efficiency of such hangers to the cable
system transferring forces is practically negligible. Similar observations can be drawn from the results proposed in Fig. 15, in terms of
stress distribution in the hangers under DL and LL configurations.
The results show that the prediction of the hanger cross-section
areas and initial stresses developed by the proposed model are able
to avoid unexpected relaxation effects and thus uncontrollable
concentrations of stress in the adjoining elements. Moreover, stress
levels reached in the hangers, under the worst loading condition,
present always a positive value, with margin of safety with respect
to unexpected compression status. Results concerning Vertical
Hanger Arch Bridge (VHAB) schemes are reported in Fig. 16 in
terms of cable system characteristics, i.e. cross section areas and

Fig. 12. Distribution of the performance factors for the hangers with left orientation
(UL) as a function of the number of iterations (NIT). Final distribution of the
performance factors (ULOPT) as a function of the dimensionless position of
hanger/girder intersection point on the girder (X1/L).

Fig. 13. Distribution of the performance factors for the hangers with right
orientation (UR) as a function of the number of iterations (NIT). Final distribution
of the performance factors (UROPT) as a function of the dimensionless position of
hanger/girder intersection point on the girder (X1/L).

Fig. 11. Synoptic representation of the structural scheme.

D. Bruno et al. / Computers and Structures 170 (2016) 1325

Fig. 14. Distribution of the cross-section areas of the hangers as a function of the
normalized girder axis (X1/L) and number of iterations.

23

Fig. 17. NAB: normalized ratios between cross-section areas (AG/AA) and inertial
moments (IG/IA), worst values of the design criteria in the arch (DA) and girder (DG)
as a function of the number of iterations (NIT).

Fig. 18. VHAB: normalized ratios between cross-section areas (AG/AA) and inertial
moments (IG/IA), worst values of the design criteria in the arch (DA) and girder (DG)
as a function of the number of iterations (NIT).
Fig. 15. Maximum/minimum stresses under ULS and DL conditions.

Fig. 16. VHAB: distribution of the cross-section areas of the hangers and the
stresses under DL as a function of the normalized girder axis (X1/L).

Fig. 19. Comparisons between NAB and VHAB: normalized material steel quantities
involved in the Arch (A), Girder (G) and Hangers (H) as a function of the number of
iterations (NIT).

24

D. Bruno et al. / Computers and Structures 170 (2016) 1325

Table 4
Final values of the design variables related to the hangers for the vertical and network arch bridge schemes.
VHAB

NAB
Left

Right

Xi (m)

Si (MPa)

Ai (cm2)

Xi (m)

Si (MPa)

Ai (cm2)

Xi (m)

Si (MPa)

Ai (cm2)

5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
155
160
165
170
175

372
405
389
382
383
386
388
387
384
378
371
369
374
382
389
389
368
394
367
389
390
387
387
387
388
389
389
388
387
384
382
382
388
400
367

1.66
20.40
21.20
19.60
17.60
16.30
15.50
14.90
14.20
13.30
14.30
14.80
14.10
13.50
15.70
13.80
7.15
15.40
6.85
13.70
15.50
14.70
13.50
12.90
13.10
13.60
14.20
14.80
15.60
16.30
17.90
19.10
21.10
20.60
1.68

3.75
8.75
13.75
18.75
23.75
28.75
33.75
38.75
43.75
48.75
53.75
58.75
63.75
68.75
73.75
78.75
83.75
88.75
93.75
98.75
103.75
108.75
113.75
118.75
123.75
128.75
133.75
138.75
143.75
148.75
153.75
158.75
163.75
168.75
173.75

303
377
394
393
387
382
374
367
361
355
353
353
345
337
336
333
329
326
322
321
317
313
311
309
313
317
321
329
330
335
323
303
268
213
127

0.01
0.47
1.24
5.57
8.73
8.90
7.50
5.63
9.96
14.30
9.88
8.59
8.35
7.66
12.70
11.70
10.00
9.80
11.90
11.40
10.10
10.40
13.40
13.10
12.70
11.10
13.00
13.20
14.10
13.70
14.20
14.40
17.00
17.20
7.26

6.25
11.25
16.25
21.25
26.25
31.25
36.25
41.25
46.25
51.25
56.25
61.25
66.25
71.25
76.25
81.25
86.25
91.25
96.25
101.25
106.25
111.25
116.25
121.25
126.25
131.25
136.25
141.25
146.25
151.25
156.25
161.25
166.25
171.25
176.25

133
212
243
249
248
250
244
246
241
244
251
258
267
275
284
288
290
294
298
301
304
305
313
320
321
326
333
339
345
354
357
359
355
335
274

9.77
14.70
14.10
12.50
11.70
12.20
12.60
12.10
11.30
10.20
11.30
12.40
12.20
10.10
9.06
10.20
10.80
9.34
7.95
11.20
12.50
6.84
7.24
7.66
8.93
8.32
6.60
5.21
5.14
4.25
4.10
1.84
0.70
0.06
0.01

Table 5
Final values of the design variables related to the arch and girder for the vertical and
network arch bridge schemes.
Scheme

VHAB
NAB

Girder

Arch

AG [m2]

IGz [m4]

IGy [m4]

AA [m2]

IAz [m4]

IAy [m4]

0.438
0.088

0.373
0.052

0.251
0.036

0.257
0.2866

0.0277
0.115

0.115
0.0225

stresses under DL, for different iteration steps. In particular, the

distribution of the hanger cross sections and stresses in the zero
configuration is quite smooth and regular, except for the hanger
ends, because of the girder/arch connections. Such distribution
points out how in VHAB, the cable system is mainly devoted to
transfer the internal forces between arch and girder by means of
uncoupled internal forces. As a matter of fact, in the case of VHAB,
the required steel quantity of the cable system is much lower than
one observed for NAB, since in the case of inclined hangers,
the cable system contributes notably to the global stiffness of the
bridge. Moreover, the shear forces are carried out also by the
cable-system and not as in the VHAB by the arch and girder only.
In order to verify such concept, results concerning girder and arch
dimensioning are presented in Figs. 17 and 18, in which the ratios
in terms of cross section characteristics and current values of the
design criteria for the arch and girder are reported. The analyses
denote that the relative stiffness ratios predicted for the VHAB
are much larger than those required for the NAB. This result is
quite consistent with current studies available from the literature

on arch bridge design, which point out enhanced stiffness properties due to the presence of inclined cables in NAB with respect to
conventional bridges based on vertical hanger arrangements [15].
Moreover, the proposed algorithm is able to evaluate the optimum
configuration, in which all the bridge components, i.e. arch, girder
and hangers, works under the worst scenarios to the design
strength. Finally, comparisons in terms of steel quantity for the
NAB and VHAB structural components are proposed in Fig. 19. In
particular, the current optimum solution predicted during the iterative procedure is described in terms of total material quantity
involved in girder, arch and hangers. From the results, it emerges
that, in NAB, a lower total steel quantity than the one involved in
VHAB is predicted, within a percentage error equal to 43.8%.
Although, in the VHAB, arch and cable system present values of
quantity of material lower than the ones predicted in the case of
NAB, i.e. almost 11% and 48% less, respectively, such reductions
of material volume are annihilated by a larger quantity involved
in the girder, which is the most prevalent in the total computation
of required volume material. Finally, the final value of design variables concerning the hangers, girder and arch are reported in
Tables 4 and 5.
5. Conclusions
An optimization model for network arch bridge schemes is proposed, in which the optimum solution is achieved by means of an
iterative methodology based on three-step algorithm. The proposed technique identifies the initial configuration under dead
loads in terms of post-tensioning forces in the hangers as well as

D. Bruno et al. / Computers and Structures 170 (2016) 1325

the initial deformations in arch and girder, reproducing the design

configuration under structural and nonstructural permanent loads.
Moreover, the evaluation of post tensioning forces and the optimum cross-sectional areas of the bridge components, i.e. arch
and girder, takes into account of maximum/minimum stress and
displacement effects produced by the external loads. In the present
optimization scheme, an elaborate and efficient iteration methodology is incorporated in a FE code to obtain an approximated solution based on minimum volume criteria. The proposed approach
enforces the minimization, which is not exactly imposed as in classical optimization procedure. Actually, the heuristic approach of
the procedure does not ensure, a priori, that the predicted solution
is a global optima; such configuration can be obtained by means of
complex optimization analyses involving all the variables, which
characterize the bridge design (geometry, initial stresses, loads,
design prescriptions, etc.). However, despite existing methodologies, based on pure optimization procedure, the proposed method
seems to be not affected by numerical convergence problems, since
the iterative procedure is able to reduce computational efforts, typically documented in standard optimization models. The proposed
method is able to design each element of the structure in such a
way that, under the worst loading combination, maximum allowable strength is achieved, leading to the lowest possible material
utilization. The optimum solution guarantees that, under the worst
loading design combinations, the design criterion for each bridge
components reaches the maximum allowable value, leading to
the lowest required material quantity involved in the bridge components. The iterative nature of the proposed methodology on
results arising from dead and live loads, controls the minimum
stress level in the hangers, avoiding as a result unexpected relaxation effects in the cable system elements and amplifications of
the stress resultants in the adjoining elements or in the arch and
girder. The proposed algorithm is based on a simple procedure,
which is based on data easily recoverable by using standard commercial FE software packages. Moreover, the proposed procedure is
able to solve min/max problem with implicit objective functionals,
reaching the global minimum in the feasible set, by means a
reduced number of iterations. The results show that the cables
which require the largest values of required steel quantity than
the rest of the elements are those located at the bridge extremities,
almost normal to the geometric axis of the arch; contrarily, the
cables approximately parallels denote inefficiency in the transferring forces between arch and girder and thus their contribution
is quite negligible. Comparisons in terms of bridge typology denote
that NAB, despite to VHAB, are able to reduce the required material
quantity involved in the bridge components, due to the presence
inclined cables, which improve the global bridge stiffness, leading
to a better distribution of the internal stress resultants in the
bridge components.

25

References
[1] Hu N, Dai GL, Yan B, Liu K. Recent development of design and construction of
medium and long span high-speed railway bridges in China. Eng Struct
2014;74:23341.
[2] Teich S. Development of general design principles for the hanger arrangements
of network arch bridges. Stahlbau 2011;80(2):10011.
[3] Tveit P. Considerations for design of network arches. J Struct Eng ASCE
1987;113(10):2189207.
[4] Tveit P. Optimal network arches for rail and road bridges. In: Proceedings of
the 4th international conference on new dimensions in bridges: flyovers,
overpasses & elevated structures; 2006. p. 2818.
[5] Martins AMB, Simes LMC, Negro JHJO. Optimization of cable forces on
concrete cable-stayed bridges including geometrical nonlinearities. Comput
Struct 2015;155:1827.
[6] Simes LMC, Negro JHJO. Optimization of cable-stayed bridges with boxgirder decks. Adv Eng Softw 2000;31(6):41723.
[7] Beyer WE. Preliminar analysis and hanger adjustment of tied arch bridges.
Doctoral dissertation, Montana State University; 1984.
[8] Qiao J, Wu H, Yang W. Optimal design of tied arch bridge. J Southwest Jiaotong
Univ 2015;50(2):300.
[9] Yongjun M, Haipeng B. Cable tension monitoring of suspender arch bridges
during cable tension adjustment stage basis on neural network algorithm. In:
International conference on remote sensing, environment and transportation
engineering (RSETE); 2011.
[10] Lonetti P, Pascuzzo A. Optimum design analysis of hybrid cable-stayed
suspension bridges. Adv Eng Softw 2014;73:5366.
[11] Chen DW, Au FTK, Tham LG, Lee PKK. Determination of initial cable forces in
prestressed concrete cable-stayed bridges for given design deck profiles using
the force equilibrium method. Comput Struct 2000;74(1):19.
[12] Pellegrino C, Cupani G, Modena C. The effect of fatigue on the arrangement of
hangers in tied arch bridges. Eng Struct 2010;32(4):11407.
[13] Lonetti P, Pascuzzo A. Vulnerability and failure analysis of hybrid cable-stayed
suspension bridges subjected to damage mechanisms. Eng Fail Anal
2014;45:47095.
[14] Lonetti P, Pascuzzo A. Design analysis of the optimum configuration of selfanchored cable-stayed suspension bridges. Struct Eng Mech 2014;51
(5):84766.
[15] Tveit P. Optimal network arches for road and rail bridges. In: Proceedings of
the 6th international conference on arch bridges, College of Civil Engineering,
Fuzhou University, Fuzhou China; 2011.
[16] Qiao J, Wu H, Yang W. Optimal design of concrete tied arch bridge. Xinan
Jiaotong Daxue Xuebao/J Southwest Jiaotong Univ 2015;50(2):3005.
[17] Islam N, Rana S, Ahsan R, Ghani SN. An optimized design of network arch
bridge using global optimization algorithm. Adv Struct Eng 2014;17
(2):197210.
[18] Christensen PW, Klarbring A. An introduction to structural optimization book
series solid mechanics and its applications, Springer series, vol. 153; 2008.
[19] BSI. Eurocode 3: design of steel structures. BS EN 1993-1-1:2005. London, UK;
2003.
[20] FIB. Bulletin no 30. Acceptance of stay cable systems using prestressing steels.
Lausanne, Switzerland; 2005.
[21] BSI. EN 19912. Eurocode 1 actions on structures. Part 2: Traffic loads on
bridges. European Committee for Standardization ed. London, UK; 2003.
[22] Maltab 6.1. The MathWorks Inc, Natick, MA; 2012.
[23] Comsol 4.4. Reference manual. COMSOL AB, Stockholm, Sweden; 2012.
[24] Greco F, Lonetti P, Pascuzzo A. Dynamic analysis of cable-stayed bridges
affected by accidental failure mechanisms under moving loads. Math Probl
Eng 2013:120 ID 302706.
[25] Gill PE, Murray W, Saunders MA. SNOPT: an SQP algorithm for large-scale
constrained optimization. Siam Rev 2005;47(1):99131.
[26] De Backer H, Outtier A, Van Bogaert P. Determining geometric out-of-plane
imperfections in steel tied-arch bridges using strain measurements. J Perform
Constr Facil 2014;28(3):54958.