Diplom Brunn Schanack

TECHNISCHE

UNIVERSITT HGSKOLEN
DRESDEN I AGDER

Berechnung einer zweigleisigen Eisenbahn-

Netzwerkbogenbrcke unter Einsatz des europischen
Normenkonzepts
Calculation of a double track railway network arch bridge

applying the European standards
Benjamin Brunn Frank Schanack
Technische Universitt Dresden Dresden University of Technology

Fakultt Bauingenieurwesen Department of Civil Engineering
Institut fr Tragwerke und Baustoffe Institute for Structures and Materials
Lehrstuhl fr Stahlbau Chair of Steel Structures
Betreuer Tutors
Prof. Dr.-Ing habil. W. Gra e
Doz. em. Dr.-Ing. P. Tveit
Dipl.-Ing. S. Teich
Grimstad, August 2003

The authors hereby declare that this diploma thesis is entirely the result of
their own work. Information derived from the published and unpublished
work of others has been acknowledged in the text and a list of references
given.
Hiermit erklren die Autoren, dass die vorliegende Diplomarbeit vollstndig

Resultat ihrer selbststndigen Arbeit ist. Informationen die aus der
verffentlichten und unverffentlichten Arbeit Dritter stammen sind im Text
gekennzeichnet und im Literaturverzeichnis nachgewiesen.
Benjamin Brunn Frank Schanack
Grimstad, 18 August 2003

Abstract
In this work the results of investigations into how to optimise a network arch for the special
demands of railway traffic are presented. Network arch bridges have inclined hangers that cross
each other at least twice. The beneficial structural behaviour leads to slender bridge members
mainly subjected to axial forces. Structural parts above the bridge deck are therefore more likely to
be tolerated. Furthermore, the high stiffness and therefore small deflections favour the application
of network arches for railway bridges. Stress ranges caused by the load character of railway
bridges require special considerations for the design. Adequate solutions are elaborated
considering as an example a double track railway bridge spanning 100 meters.
The arrangement of the hangers has considerable influence on the structural behaviour. It
decides on the forces and force variations within the network arch depending on many parameters,
as for example span, rise, number of hangers, loading or arch curvature. A new introduced type of
hanger arrangement is involved in an optimisation process with regard to the mentioned
parameters. This improved hanger arrangement provides a simple method of designing network
arches with small hanger forces and small bending moments in the chords.
The hanger connection details call for special attention to fatigue strains. The fatigue design
check is decisive for the hanger cross section. Various designs of hanger connection details for
circular hangers are tested by numeric analysis. A hanger connection detail is derived from the
results satisfying the special demands of slender arches as they are found in network arches.
The structural behaviour of a network arch favours a lower chord consisting of a concrete slab.
Its vertical deflections are limited to ensure passenger comfort and track stability. Alternatives with
and without transverse prestressing are compared considering the deformation behaviour and
economic differences. Several other construction details such as arch root point, bearings or
drainage are elaborated on and the solutions are presented. An erection method using a temporary
lower chord is assessed and described in detail.
The investigations confirm the suitability of network arches for railway bridges. Economic
advantages due to significant savings of steel compared to other arch bridges contribute to the
overall convincing performance.
Zusammenfassung
In dieser Arbeit werden die Ergebnisse von Untersuchungen zur Optimierung von
Netzwerkbogenbrcken entsprechend den speziellen Ansprchen aus Eisenbahnverkehr
vorgestellt. Netzwerkbogenbrcken sind Bogenbrcken mit geneigten Hngern die sich wenigstens
zweimal berkreuzen. Da ihr vorteilhaftes Tragverhalten schlanke Bauteile ermglicht, die
hauptschlich durch Normalkrfte beansprucht werden, sind tragende Bauwerkselemente ber der
Brckennutzflche eher tolerierbar. Die hohe Steifigkeit und demzufolge kleinen Verformungen von
Netzwerkbgen tragen zustzlich zu deren Anwendbarkeit als Eisenbahnbrcken bei. Der
Lastcharakter von Eisenbahnbrcken verursacht groe Spannungsschwingbreiten deren
besondere Bercksichtigung beim Entwurf gefordert ist. Am Beispiel einer zweigleisigen
Eisenbahnbrcke mit 100 Metern Spannweite werden passende Lsungen erarbeitet.
Die Anordnung der Hnger hat einen erheblichen Einfluss auf das Tragverhalten. In
Abhngigkeit vieler Parameter, wie zum Beispiel Spannweite, Bogenstich, Hngeranzahl, Lasten
oder Bogenkrmmung, entscheidet sie ber die Schnittkrfte und das Spannungsspiel innerhalb
des Netzwerkbogens. Eine neue Art der Hngeranordnung wird eingefhrt und unter
Bercksichtigung der genannten Parameter einem Optimierungsprozess unterzogen. Diese
verbesserte Hngeranordnung stellt eine einfache Methode fr den Entwurf von
Netzwerkbogenbrcken mit kleinen Hngerkrften und kleinen Biegemomenten in den Gurten zur
Verfgung.
Die Hngeranschlsse unterliegen in besonderem Mae Ermdungsbeanspruchungen, so
dass der Querschnitt der Hnger vom Ermdungsfestigkeitsnachweis seines Anschlusses
bestimmt wird. Mit Hilfe von FEM-Berechnungen werden unterschiedliche Hngeranschlsse
hinsichtlich ihrer Ermdungsfestigkeit untersucht. Aus den Ergebnissen wird eine verbesserte
Anschlussgeometrie abgeleitet, die den speziellen Anforderungen von schlanken Bgen, wie sie
bei Netzwerkbgen vorkommen, gerecht werden.
Das Tragverhalten von Netzwerkbogenbrcken begnstigt die Verwendung einer einfachen
Betonplatte als Untergurt, dessen Durchbiegung fr den Fahrgastkomfort und zur Sicherstellung
der Stabilitt der Gleise beschrnkt ist. Alternativen mit und ohne Quervorspannung werden
hinsichtlich des Verformungsverhaltens und wirtschaftlicher Unterschiede verglichen. Fr weitere
Konstruktionsdetails, wie Bogenfupunkt, Lagerarten oder Brckenentwsserung werden
Lsungsvorschlge vorgestellt. Als eine Mglichkeit der Bauausfhrung werden Montagezustnde
unter Verwendung eines temporren Untergurtes sorgfltig berechnet und detailliert beschrieben.
Die durchgefhrten Untersuchungen besttigen die Eignung von Netzwerkbgen fr den
Einsatz als Eisenbahnbrcken. Die Kosteneinsparungen gegenber anderen Bogenbrcken durch
das bedeutend geringere Stahlgewicht tragen zu dem berzeugenden Gesamteindruck bei.
Resumen
En este trabajo se presentan los resultados de las investigaciones realizadas sobre la

optimizacin de arcos network, con respecto a las exigencias especiales del trfico de
ferrocarriles. El puente en arco tipo network es un puente en arco, en el cual algunos tirantes se
interceptan al menos 2 veces entre s. El buen comportamiento estructural de este tipo de arco,
nos lleva a utilizar elementos esbeltos en los puentes, los cuales toman principalmente fuerzas
axiales, por esta razn los elementos estructurales sobre el tablero pueden ser tolerados de mejor
forma. Adems, la rigidez alta y por ende las deflexiones pequeas, favorecen la aplicacin de
arcos network para puentes de ferrocarril. El carcter de cargas de puentes de ferrocarril causa
altas variaciones de fuerzas interiores lo cual requiere consideraciones especiales para el diseo,
para esto se elaboran soluciones adecuadas utilizando como ejemplo un puente de ferrocarril con
dos vas y una luz de 100 metros.
La colocacin de los tirantes influye cuantiosamente en el comportamiento estructural. Las
fuerzas y la variacin de fuerzas en el arco network, dependen de muchos parmetros como por
ejemplo: la luz, la altura del arco, el nmero de tirantes, las cargas y la curvatura del arco. Un
nuevo tipo de colocacin de tirantes es introducido y incluido en un proceso de optimizacin con
respecto a dichos parmetros. Esta colocacin de tirantes mejorada pone a disposicin un mtodo
fcil de disear arcos network con fuerzas axiales pequeas en los tirantes y momentos flectores
pequeos en los cordones.
Los detalles de la conexin de los tirantes necesitan atencin especial referente a exigencias
de fatiga. Las pruebas del estado lmite de fatiga son decisivas para la seccin transversal de los
tirantes. Usando FEM-calculaciones, son estudiados distintos detalles de la conexin de tirantes
con seccin transversal circular, de los resultados es derivada una geometra del detalle, la cual
satisface las exigencias especiales de arcos esbeltos, como usualmente ocurren en los arcos tipo
network.
El comportamiento estructural de un arco network favorece a un tablero inferior construido en
hormign. Las deflexiones verticales de estos estn limitadas para asegurar el confort de los
pasajeros y la estabilidad de las vas. Alternativas con y sin cables postensados transversales son
comparadas con respecto al comportamiento de deformacin y las diferencias econmicas.
Soluciones para otros detalles del diseo como los extremos del arco, los portes o el desaguado,
son elaborados y presentados. Como un mtodo posible de la ereccin del puente los estados de
construccin utilizando un cordn inferior temporero son calculados y trazados detalladamente.
Las investigaciones comprueban la aptitud y el buen comportamiento de arcos tipo network
para puentes de ferrocarril, as como tambin las ventajas econmicas que se tienen por los
ahorros notables de acero en comparacin con otros puentes en arco, todo esto en conjunto
contribuye a la convincente impresin general.
Contents
Preface x
Acknowledgements xi
1 Task for the diploma thesis 1
2 Theses 3
3 Introduction 5
4 What is a network arch? 6
5 Bridge design 8
5.1 The arches 8
5.2 The hangers and hanger connections 8
5.3 The wind bracing 8
5.4 The bridge deck 9
5.4.1 Main design 9
5.4.2 Alternatives without transverse prestressing 10
5.4.3. Comparison 12
5.5 The end cross girder 14
5.6 Constructive design of the arch root point 14
5.6.1 Anchorage of the arch 14
5.6.2 Anchorage of the longitudinal tendons 15
5.6.3 Bearings 16
5.6.4 Transversal prestressing 16
5.6.5 Lower hanger connection 17
5.6.6 End cross girder and concrete shape 18
5.6.7 Alternative stilt bearing 18
5.7 The handrails 19
5.8 The drainage 20
5.9 Investigation of two different types of bearings 21
5.9.1 General 21
5.9.2 Pot bearings 21
5.9.3 Stilt bearing with compression support in the middle of the end cross girder 22
5.9.4 Conclusion 24
5.10 Summary of materials used 25
6 Optimisation of the hanger arrangement 30

6.1 Scope 30
6.2 What is an optimal hanger arrangement? 30
6.3 The parameters 31
6.4 Evaluation of the results 31
v
Diploma Thesis Brunn & Schanack Contents
6.5 Preliminary investigations 33

6.5.1 Variation of the lower hanger nodes by the node distances 33
6.5.2 Variation of the lower hanger nodes by the slope of the hangers 39
6.6 Advanced model to describe the hanger arrangement 46
6.6.1 The idea 46
6.6.2 Derivation of the proposed hanger arrangement 48
6.6.3 Variation of the lower hanger nodes using the advanced model 51
6.7 Investigation of other parameters 56
6.7.1 Influence of the number of hangers and the span 56
6.7.2 Influence of the rise of the arch 59
6.7.3 Influence of the ratio between live load and dead load 60
6.7.4 Influence of the ratio between the radii of ends and the middle of the arch 63
6.7.5 Further consideration regarding the hanger arrangement 64
6.8 Discussion of the results 65
6.9 Summary of Section 6 68
6.9.1 In words 68
6.9.2 In a scheme 69
6.9.3 Example of use 70
7 Investigation of hanger connection details regarding fatigue strains 74

7.1 General 74
7.2 Loading 75
7.2.1 Axial loads 75
7.2.2 Horizontal loads 75
7.2.2.1 Gust excitation 76
7.2.2.2 Vortex excitation 76
7.2.2.3 Aeroelastic instabilities and interference effects 76
7.2.3 Conclusion 77
7.3 Boundary conditions for hanger connection design 78
7.4 Fatigue caused by axial loads 79
7.4.1 Fatigue assessment based on nominal stress ranges 79
7.4.2 Fatigue assessment based on geometric stress ranges 82
7.5 Stresses caused by horizontal deflections 90
7.6 Evaluation 91
7.7 Connection to arch profile 96
7.8 Hanger connection along the tie 97
7.9 Notes on the FEM analysis 98
8 Erection of the bridge using a temporary lower chord 99

8.1 Why a temporary lower chord? 99
8.2 The design of the temporary lower chord and formwork 100
8.3 Special considerations concerning the end of the temporary lower chord 101
8.4 Sequence of erecting the bridge 101
9 Summary 104
10 Criticism and proposals for future investigations 105
List of figures 107
List of standards 113
vi
List of references 114
Annex A: Load assumptions A-1

A.1 Permanent actions including self-weight A-1
A.2 Traffic loads A-2
A.2.1 Actions due to railway operations A-2
A.2.1.1 Vertical loads A-2
A.2.1.2 Dynamic effects A-3
A.2.1.3 Horizontal forces A-3
A.2.2 Accidental actions A-4
A.2.3 Assessment of traffic loads on railway bridges A-5
A.2.4 Fatigue load models A-6
A.3 Wind forces A-7
A.3.1 Wind forces in y-direction (perpendicular to bridge centre line) A-7
A.3.2 Crosswind forces in z-direction A-8
A.3.3 Longitudinal wind forces in x-direction A-8
A.4 Actions due to temperature A-8
A.5 Combination of actions A-9
Annex B: Preliminary design B-1

B.1 Bridge deck B-1
B.1.1 Transverse direction B-1
B.1.2 Longitudinal direction B-12
B.2 Arch B-20
B.3 Hanger B-22
Annex C: FEM-calculations C-1

C.1 General C-1
C.2 Internal forces of the arch C-5
C.2.1 Ultimate limit state C-5
C.2.1.1 Collapse about the weak axis y-y C-5
C.2.1.2 Collapse about the strong axis z-z C-7
C.2.2 Fatigue stress spectra C-7
C.2.3 Serviceability limit state C-7
C.2.4 Composition of the design-relevant calculation results C-8
C.3 Hanger C-9
C.3.1 Axial forces C-9
C.3.2 Deflections perpendicular to the hanger axis C-10
C.3.2.1 Static deflection due to wind load C-10
C.3.2.2 Deflections due to dynamic wind excitation C-11
C.4 Internal forces in the wind bracing C-13
C.4.2 Fatigue C-15
C.5 Internal forces of the bridge deck C-17
C.5.3 Fatigue C-23
C.5.4 Alternatives without transverse prestressing C-24
C.6 Internal forces of the end cross girder C-25
vii
C.7 Actions on the bearings C-26
Annex D: Assessment of the bridge D-1

D.1 Materials and cross-sections D-1
D.2 Arch D-4
D.2.1 Ultimate limit state assessment D-4
D.2.1.1 Collapse about the weak axis y-y D-4
D.2.1.2 Collapse about the strong axis z-z D-6
D.2.2 Fatigue limit state assessment D-7
D.2.3 Serviceability limit state assessment D-9
D.2.3.1 Limitation of nominal stress for rare load combinations D-9
D.2.3.2 Limitation of nominal stress range for frequent load combinations D-10
D.2.3.3 Limits for clearance gauges D-10
D.3 Hanger and hanger connections D-11
D.3.1 General D-11
D.3.2 Ultimate and serviceability limit state assessment D-12
D.3.2.1 Hanger D-12
D.3.2.2 Hanger connections D-13
D.3.2.2.1 Connection type 1 D-13
D.3.2.2.6 Suggested hanger connection D-25
D.3.3 Fatigue assessment D-26
D.3.3.1 Fatigue assessment based on nominal stress ranges D-26
D.3.3.2 Fatigue assessment based on geometric stress ranges D-31
D.3.4 Patch test D-34
D.4 Wind bracing D-35
D.4.2 Fatigue limit state assessment D-41
D.4.3.1 Limitation of nominal stress for rare load combinations D-44
D.4.3.2 Limitation of nominal stress range for frequent load combinations D-44
D.5 The bridge deck D-45
D.5.1 Concrete cover D-45
D.5.2 Main design with transverse prestressing D-45
D.5.2.1 Ultimate limit state D-45
D.5.2.1.1 Bending and longitudinal force D-45
D.5.2.1.2 Shear D-49
D.5.2.1.3 Punching D-51
D.5.2.2 Serviceability limit state D-52
D.5.2.3 Fatigue D-55
D.5.2.4 The footpath D-58
viii
D.5.2.5 Summary of reinforcement D-59
D.5.3 Alternative design proposals without transverse prestressing D-60
D.5.3.1 Alternative design 1 D-60
D.5.3.1.1 Ultimate limit state D-60
D.5.3.1.2 Serviceability limit state D-64
D.5.3.1.3 Fatigue D-65
D.5.3.2 Alternative design 2 D-66
D.5.3.2.1 Ultimate limit state D-66
D.5.3.2.2 Serviceability limit state D-70
D.5.3.2.3 Fatigue D-71
D.5.3.3 Deflections D-72
D.5.3.4 Summary of reinforcement D-77
D.6 The end cross girder D-79
D.6.1.1 Bending and longitudinal force D-79
D.6.1.2 Shear D-81
D.6.1.3 Punching at bearings D-82
D.6.2.1 Limitation of stress D-83
D.6.2.2 Limit states of cracking D-83
D.7 The handrails D-85
D.8 The drainage D-86
D.9 Bearings D-87
D.9.1 Pot bearings D-87
D.9.2 Stilt bearing D-87
D.9.2.1 Vertical plate D-87
D.9.2.2 Middle compression support D-89
D.10 Deformations D-90
Annex E: Investigations on the temporary lower chord E-1

E.1 General E-1
E.2 Formwork sheet E-1
E.3 Timber formwork beams E-3
E.4 Transverse steel beams E-4
E.5 Wind bracing members E-6
E.6 Longitudinal steel beams E-8
E.6.1 Construction phase 1 Mounting the steel skeleton and displacement E-8
E.6.2 Construction phase 2 Preparations for casting the bridge deck E-8
E.6.3 Construction phase 3 Casting of the concrete edge beam E-9
E.6.4 Construction phase 4 Casting the main part of the bridge deck E-12
E.6.5 Assessment of the longitudinal steel beam E-12
Annex F: Data from optimisation process F-1

F.1 Comparison of influence lines F-1
F.2 Results from the variation of the lower hanger nodes by the node distances F-12
F.3 Results from the variation of the lower hanger nodes by the slope of the hangers F-25
F.4 Results from variation of the lower hanger nodes using the advanced model F-36
F.5 Results from the variation of the number of hangers and span F-38
ix
Preface
The present work was done by the authors at the end of their 5-year studies. It has been
submitted to the Faculty of Civil engineering of Dresden Technical University in order to obtain the
degree of a Diplom-Ingenieur.
The topic, Calculation of a double track railway network arch bridge applying the European
standards, deals with a very efficient structure. Its inventor and the most vigorous researcher in
this field is Dr.-Ing. Docent Emeritus PER TVEIT. The authors were fortunate to be offered the
opportunity to carry out their investigations and to write this thesis from May to August 2003 at the
office of this engineer. From the conversations with him and his answers to questions several
suggestions arose, which are separately marked.
Serving as a basis for the calculated network arch bridge was the one designed in the Diploma
thesis of UWE STEIMANN, also done with TVEITS collaboration, in Grimstad, Norway. It is mentioned
in the text when STEIMANNS ideas are adopted.
B. Brunn
F. Schanack
Grimstad, Norway
August, 2003
x
Acknowledgements
Not only while working on this Diploma thesis, but also during the years of studying there are
many persons and institutions to which the authors owe deep thanks. Without a surely incomplete
list, our gratefulness should be expressed hereby.
Special thanks go to PER TVEIT, for his permanent disposition to answer our questions, for his
constructive criticism of our results which encouraged us to further investigate and review, and last
but not least for his willingness to share his office with three students.
xi
Section 1
Task for the diploma thesis
1
Diploma Thesis Brunn & Schanack Section 1: Task for diploma thesis
2
Section 2
Theses
The following theses list relevant problems and results of the diploma thesis. Hopefully they willl
inspire scientific discussion.
1. Two intersecting hangers with adjacent nodes at the arch can be assumed to act as a pair at the
arch. A line can be drawn between the middle of their upper hanger nodes and their intersection
which shall be called direction of action.
2. Small hanger forces in one hanger pair are obtained if their direction of action is aligned to the
deflection of the arch at the centre point between the adjacent nodes.
3. Each hanger pair causes a resulting force at the arch. Small bending moments about the
horizontal axis in the arch are obtained if these resulting forces cause a line of thrust along the
centre line of the arch.
4. The following simplification to the first three points can be assumed:

1a. On an average the hanger forces are equal, i. e. their resulting force is collinear to
their direction of action.
2a. Small deflections in the centre range of a circular arch are radial for uniform
loading, i. e. for this load case the direction of action is to align to the radii of the
arch circle to obtain small hanger forces.
3a. The centre range of a clamped arch behaves like the centre range of a simply
supported arch, i. e. uniform forces acting radially on a circular arch will cause a
line of thrust along the centre line of the arch.
5. On the basis of these assumptions a hanger arrangement for circular arches can be derived from
1., 2. and 3. giving small bending in the arch and small hanger forces. It can be described as
follows:
All hangers in the centre range of the arch cross the arch with the same angle; the
upper hanger nodes are equidistant.
6. The clamping at the ends of the arch causes other conditions. In order to obtain small bending in
the arch and small hanger forces the hanger pairs must be oriented differently as along the radii
of the arch circle.
7. The variable in such a hanger arrangement is the cross angle between the hanger and the arch,
found to influence the forces and force variations in the structural members of the bridge. This
cross angle is to be optimised regarding desired attributes. A cross angle of 45 gives smallest
variation of bending moments in the arch.
8. The application of the derived hanger arrangement caused a significant decrease of bending in
the arch and hanger forces compared to hanger arrangements which used to be considered as
near optimal.
3
9. Hanger connections along the especially slender arches of network arch bridges are
considerably restricted in their dimensions. This implies that details which are appropriate for
arch bridges with vertical hangers might not be advantageous in respect of fatigue for network
arches.
10. All transitions in hanger connection details must be made continuous and smooth in order to
reduce stress concentrations.
11. For the fatigue assessment of details such as hanger connections it might be essential to
include finite element analysis in order to determine stress concentrations. Dangerous stress
peaks arise at geometrical discontinuities; their magnitude depends on the shape of the
member at that location and the nominal stress level. Either factor alone is not meaningful in
order to judge the fatigue performance.
12. There is a tendency that about 10 m wide non-prestressed bridge decks are more economical
than their counterparts with prestressing. Prestressed slabs stand out because of small
deformations and higher durability.
13. The adaptation of roller bearings by eliminating the parts of the roll that never have contact with
the top or bottom member allows large radii without having a giant roll. This bearing alternative
is suitable for narrow and straight bridges, where a large construction height of the bearing can
be tolerated.
14. The decisive loads for the longitudinal beams of the temporary lower chord are self-weight of
formwork, reinforcement, prestressing tendons and end cross girder (see Figure E.8).
15. The sequence of casting the bridge deck during erection of the bridge using a temporary lower
chord has to be determined for each project to avoid extensive hanger relaxation.
4
Section 3
Introduction
Bridges are considered to be masterstrokes of engineering. Light, wide spanning structures

especially attract attention. The worlds most slender and lightest arch bridge was erected over
Bolstadstraumen 60 km northeast of Bergen, Norway (TVEIT [45], page 7). Using the benefits of a
network arch, PER TVEIT designed this bridge in the early 1960s. Since then, surprisingly few
network arch bridges have been built.
The authors are convinced that this is not due to any weak points, but rather due to little
familiarity with and maybe doubt about the benefits of such a structure. In light of this, the authors
have written this diploma thesis in English so that many persons will have easy access to this
information. Furthermore, the application of the European standards (Eurocode) for assessment
provides a comparison basis within the European Union and may simplify relating the results to
other standards.
Working together, the authors acquired sufficient time for intensive research, besides the
assessment of the network arch bridge.
The fields of this research shall be presented briefly:
Force distribution in structural parts of a network arch bridge, especially in hangers and arches,
depends on the slope and arrangement of the inclined hangers. Since the structure is sensitive to
changes in the hanger arrangement, this topic calls for special attention. The authors searched for
predictions on best hanger arrangements by an optimisation process.
Railway bridges, like the one calculated in this work, are subject to fatigue strains. Hangers and
their connection details are especially at risk of failure. Known design solutions for tied arches are
not optimal for the more slender structural members of network arches. Investigations were carried
out to find how to best adapt the known connection details.
Erection is always an important topic when designing bridges. PER TVEIT suggested an erection
method using a temporary lower steel chord. Its applicability to the bridge, the object of this work,
was verified in detail.
As alternative designs a concrete tie without transverse prestressing and the possibility of a
stilt bearing (see Section 5.9) were investigated. Special attention was also given to the
constructive design of the arch root point.
Carrying out the investigations and discussing the results with PER TVEIT, the authors gained
knowledge about the structural behaviour of network arch bridges. Utilising this behaviour correctly
leads to very efficient structures. The authors would be delighted to see more network arch bridges
built using all the advantages they offer.
5
Section 4
What is a network arch?
The structure named network arch bridge was invented by PER TVEIT. He defines it as an arch
bridge in which some hangers intersect at least twice (TVEIT [46], page 3). A short description of
this type of bridge will be given. An extended manuscript containing 100 pages about network
arches can be found on the homepage of PER TVEIT [45]. On the front page, in Figure 5.29 and on
the cover page of the Annexes the network arch bridge calculated in this work is shown.
Characteristics
To achieve great efficiency with this type of structure the following characteristics should be
applied. The arch should be part of a circle as this makes fabrication easy and contributes to a
more constant axial force in the middle portion of the arch and even maximum bending moments
along the tie. The hangers should be spaced equidistantly along the arch and not merged in nodal
points. This decreases bending due to local curvature and gives more efficient support to the arch
in buckling. The lower chord is a concrete slab between small concrete edge beams. Longitudinal
prestressing of the edge beams takes the horizontal forces of the arch. Furthermore, the
prestressing increases durability of the concrete. For a width between the arches of more than
about 10 meters transverse prestressing is suggested as this gives a more slender tie. The number
of hangers is usually much higher than for tied arch bridges with vertical hangers. Their
arrangement is a central question, which we have attempted to answer in this work.
Advantages in the structural behaviour
Compared with tied arches with vertical hangers the network arch bridges feature the fact that
the chords are only subjected to very little bending. The bridge acts more like a simple beam and
shows therefore a high stiffness and small deflections. Figures 4.1 and 4.2 will help to explain the
reason.
Fig. 4.1. Tied arch with vertical hangers, structural behaviour with partial loading (TVEIT [39], page 16)
Fig. 4.2. Tied arch with one set of inclined hangers, structural behaviour with partial loading (TVEIT [39], page 16)
Partial loading of the span leads to a deflection of the upper and lower chord in the arch with
vertical hangers. This causes bending which is to be taken by big cross-sections of the arch and
6
Diploma Thesis Brunn & Schanack Section 4: What is a network arch
the tie. In network arches the inclined hangers restrict these deflections, and so bending only
occurs as a result of local loading and the arch and tie are mainly subjected to axial force. Figure
4.3 shows the comparison of the influence lines for bending moments in the chords between an
arch with vertical hangers and a network arch.
4
2 7m
0.0
2 I= Ha
m ng
91 er s
0. 0
30 m
A=
h A= 0. 0
A rc 01
2m2
Network arch proposal 1980. 2

Half lower chord A = 2.2 m I = 0.018 m 4
4
m Ha
22
0.0 n ge
2 I= rs
A
m =0
094
31.2 m
. .00
=0 64
hA to
A rc 0.0
12
m 2
200 m Half lower chord A = 0.24 m 2 I = 0.27 m4
Bridge at Straubing built 1977.
Bridge at
Influence lines for bending
1.04 m Network arch
Straubing 1.55 m
moments in the arches
0.57 m
0.89 m
1.3 m
Influence lines for bending

moments in the lower chords
Bridge at
Straubing Network arch
15.5 m
1.4 m
10.5 m
Fig. 4.3. Areas, stiffnesses and influence lines for the lower and upper chord of two
tied arches (TVEIT [45], page 14)
Due to the stiffness of the hanger web, the bridge deck spans between the planes of the
arches and does not have to take much longitudinal bending, therefore it can be slender. As a
result of the larger number of hangers their cross-section can be very small.
As a conclusion, the structural members of the network arch mainly take axial forces and the
compression member, the arch, is more supported in buckling. The cross-sections can be very
compact, which contributes to a more efficient use of material, to less steel weight and a better
design due to higher transparency of the structure, SEIDEL [34].
7
Section 5
The bridge design
5.1 The arches
Each arch consists of six segments, which are connected by butt-welding on the construction
site. The two lower segments are American Wide flange profiles W360x410x900, have a constant
curvature radius of 66.86 metres and a bow length of 17.44 metres. The four upper segments are
W360x410x634 and have a constant curvature radius of 83.58 metres. Their bow length is 18.06
metres. This gives the arch a rise of 17 metres. The smaller radius of the arch ends leads to a
smaller length of the wind portal frame, which decreases bending moments. The weak axis of the
profiles is horizontal. The distance between the arch planes is 10.15 metres. As material S 460 ML
is used. ARCELOR LONG COMMERCIAL S.A. [4] provides such profiles with a constant curvature.
5.2 The hangers and hanger connections
The bridge has 48 hangers per arch plane. 6 slip resistant high
7 strength bolts
The geometry of their arrangement follows the
1
d = 24mm
improvements found in Section 6, hence all
hangers cross the arch with an angle of 49. At
the ends of the spans a special arrangement
according to Section 6.7.5 was applied. For the
upper and lower hanger connection details the
1
solutions found to be best in Section 7 were 34

2.5 30
0
chosen (see Figure 5.1). Each set of hangers is
52
shifted half the diameter of the hangers out of 0 22
0
the arch plane. This allows them to cross without
deflections. The eccentricity causes torsional cross section 1-1
moments in the arch profiles, which are partially Fig. 5.1. Upper hanger connection detail
taken by the wind bracing. The direction of the
eccentricity changes from each hanger connection to the next, so the torsional moments counteract
each other. In the event of hanger relaxation several hangers in a line cause torsional moments in
the same direction, which will give decisive forces. In the bridge calculated in this work no relaxing
hangers occurred.
At their intersections the hangers are protected by a sheathing of slit open plastic tubes and
tied together with elastic rubber bands. This couples the deflections out of the arch plane and
therefore increases damping. All hangers consist of a smooth bar with a circular cross-section and
a diameter of 60 mm. As material S 460 ML is used.
5.3 The wind bracing
The truss of the wind bracing is shown in Figure 5.2. Instead of a bending resistant top cross
bar the wind portal frame is completed by a truss. Diagonal struts below this truss allow a shorter
length of the portal frame columns, because the truss can be drawn down alongside the clearance
gauge. For aesthetic reasons the rest of the wind bracing was chosen to be a K-truss. The distance
8
Diploma Thesis Brunn & Schanack Section 5: Bridge design
between the Ks along the arch is 6.95 metres. The diagonal VIII
members have a system length of 8.60 metres. For all

members the circular hollow section CHS 219.1x8 was
chosen. At the nodes II and VIII extra bending moments
occur because the connected members do not lie in one
plane, but on a cylindrical sphere. Therefore the straight
members at nodes II and VIII need to be CHS 219.1x10. The
II
joints within the wind bracing are welded joints between
hollow sections. The truss members are connected to the
arch profile by welded endplates bolted to the flanges on the
construction site. S 355 is used as material. CORUS TUBES
STRUCTURAL & CONVEYANCE BUSINESS [7] provides such
Fig. 5.2. System lines of the wind bracing,
profiles. half bridge
5.4 The bridge deck
5.4.1 Main design
The tie of the bridge consists of a concrete slab (C50/60) spanning 10.15 metres between the
hangers. It is prestressed in transverse direction to prevent cracking in serviceability limit state and
decrease vertical deflections. Additionally this increases durability and leads to a slender cross-
section. An alternative without transverse prestressing can be found in Section 5.4.2. The
prestressing in longitudinal direction mainly counteracts the horizontal thrust of the arches. It is
increased to compress the concrete for the same reasons as before.
1000
CHS 219.1x10
W 360x410x634
5075
2200
Tensioning jack CL
5725
5425
2x60
4900
Moveable supports 335 3900
Schematic illustration of post-tensioning

the transverse tendons
750
380
1100
0.0
700
1.5 %
150
270 1.5 %
540
430
458
374
608
Side view
Longitudinal tendons,
630 220 875 484 1516 2200
6 x DYWIDAG Type 6827
1095 4200
Transverse tendons, DYWIDAG
5925
threadbar 36D, s=270
Fig. 5.3. Cross section of the bridge at mid-span
9
The formwork of the bridge deck receives a camber, sized according to the vertical deflections
due to dead load. The additional reinforcement is shown in Figure 5.4. For transverse prestressing
370 DYWIDAG thread bars type 36D are placed every 27 cm along the tie. One side is dead
anchored and the other prestressed (see Figure 5.3). Afterwards the prestressed anchorage is
covered by a cap and protected by anti-corrosion agent. The tendons are put into place so that
every second one has the dead anchorage on the same side. The longitudinal tendons are six
DYWIDAG Type 6827 on each side of the bridge.
12, s = 20 cm Transverse tendons, DYWIDAG

10, s = 15 cm
12, s = 20 cm threadbar 36D, s=270
12, s = 15 cm
610
20, s = 9 cm
374
10, s = 15 cm
mid-span
10, s = 15 cm
430 at
10, s = 15 cm 10, s = 14 cm
Longitudinal tendons, 12, s = 15 cm 10, s = 15 cm
Fig. 5.4. Reinforcement of bridge deck with transverse prestressing
The main design has the disadvantage of requiring a large amount of compression
reinforcement. This is caused by the small depth of the slab and the prestressing force which
increases the height of the compressive zone in the concrete section. According to SCHNEIDER [26],
page 5.126, DAfStb-Heft 425 suggests that the compression zone depth to effective depth ratio
(x/d) shall not exceed the value 0.35 for concrete classes C 40/50 and higher. Otherwise
compression reinforcement is required. In the bridge calculated in this work, the x/d ratio of 0.601
at the decisive section exceeds the limiting value significantly (Annex D. Section 5.2.1.1) making
compression reinforcement of 35.6 cm2/m necessary (re-bars: 20, s = 9).
It might therefore be advisable to use a bridge deck of greater depth. Certainly, the increased
dead load increases the bending moment. But the higher effective depth and the increased lever
arm of the tendon counteracts the negative effect of the higher dead load. Therefore the required
additional depth will be moderate and the compression reinforcement can be made redundant.
Approximate calculations showed that a thickness of about 53 cm at the slabs mid-span would be
enough to eliminate compression reinforcement. Besides, a thicker tie improves the stability and
continuity of the track.
However, if desired, a thinner slab without compression reinforcement can still be achieved
with higher concrete strength. The concrete class used, C 50/60, is the highest class regulated in
EC 2.
5.4.2 Alternative design proposals (without transverse prestressing)
We wanted to investigate the alternative of a bridge deck without transverse prestressing. For
long spans in transverse direction prestressing is essential, especially for railway bridges which are
subjected to high loads. Shorter spans may be more economical without prestressing, but require a
larger amount of reinforcement.
The double track railway bridge calculated in this work has a transverse span of 10.15 m,
which lies in the range where both prestressed and non-prestressed solutions may be equally
efficient. For example, the railway bridge calculated in STEIMANN [37] uses transverse prestressing
at a span of 11.45 m. The distance between the arches of the bridge which is the object of attention
10
in TVEIT [42] measures 9.65 m and uses transverse prestressing as well. TEICH & WENDELIN [38]
calculated a network arch without transverse prestressing. Even though this bridge was designed
for road traffic and not for railway traffic, it still indicates the possibility and feasibility of a non-
prestressed solution.
Introduction of alternative design proposals
The required reinforcement and deflections are determined for two different variants:
For the first proposal a maximum reinforcing bar diameter of 25 mm with a minimum spacing of
9 cm was assumed, which is regarded as a lower boundary due to construction. Further assuming
a single layer of reinforcement leads to a structural depth at the mid-span of 610 mm, which is
equal to the depth of the edge beam. Compression reinforcement is not required.
The second proposal is based on a maximum ratio between compressive zone and effective
depth x/d of 0.35, which is the upper boundary for a member using C50/60 concrete without
compression reinforcement, SCHNEIDER [29], page 5.126. In this case, a structural depth of 470 mm
can be obtained. However, the reinforcement (25) has to be applied in two layers, since
otherwise the spacing would fall under the predetermined 9 cm. The spacing for the double layer
reinforcement measures 14 cm.
Alternative design proposals
1. Slab depth at mid-span: 610 mm (Figure 5.5)
2. Slab depth at mid-span: 470 mm (Figure 5.6)
12, s = 15 cm
12, s = 15 cm
12, s = 15 cm
610
14, s = 10 cm
mid-span
10, s = 15 cm 10, s = 15 cm 610 at
12, s = 10 cm
554
Longitudinal tendons, 25, s = 9 cm

Fig. 5.5. Cross section with reinforcement of alternative design 1
12, s = 15 cm
12, s = 15 cm
14, s = 10 cm
610
12, s = 15 cm
14, s = 10 cm
10, s = 15 cm
mid-span
10, s = 15 cm
470 at
14, s = 10 cm
414
Longitudinal tendons, 25, s = 14 cm in two layers

Fig. 5.6. Cross section with reinforcement of alternative design 2
11
Effects on the edge beams

The increased thickness of the bridge deck demands more longitudinal prestressing, because
the higher dead load increases the horizontal thrust in the arch and therefore the axial tensile force
in the tie. Furthermore the enlarged cross section area of the deck requires more prestressing to
limit concrete tensile stresses.
The number of additionally required tendons was approximated by using the calculations for
the preliminary design (Annex B, Figure B.9 and Figure B.12) with respectively increased cross
sections and dead loads. To achieve at least the same concrete compressive stress, one more
tendon per edge beam is necessary for both alternative designs. However, it might be possible to
use only 6 tendons per edge beam for alternative design 2 with an only slightly increased concrete
depth. Since this would have to be proven by a detailed assessment, 7 required tendons per edge
beam will be presumed.
The edge beam dimensions of both alternative designs are sufficiently large enough to
accommodate the additional tendon.
Effects on other structural elements

An increased depth of the bridge deck influences all other structural elements of the bridge.
Hangers, arch profiles, bearings etc. receive higher internal forces due to the increased dead load.
For example, the additional dead load of alternative design 1 leads to about 6% higher axial forces
in the hangers and the arch, compared to the main design. Alternative 2 causes about 1% higher
forces. Assuming that the clearance below the bridge is critical, the increased depth shifts the
clearance gauge above the deck upwards which increases the length of the portal frame. Thus,
transverse bending moments in the portal frame are also increased.
Deflections
The deflections of the alternative designs proved to be bigger than the deflections of the main
design with transverse prestressing. The maximum deflection of the main design occurs at one
edge beam, since only one track is loaded and the relative deflection of the slab is small. The
deflection requirement of the main design is satisfied. The non-prestressed solutions show higher
edge beam deflections (increased dead load), which however do not constitute locations of
maximum deflections. The locations of the total maximum deflection lie between the edge beams,
which is due to the large deflection of the deck relative to the edge beam.
Considering both alternative design proposals, design 1 gives a higher edge beam deflection,
but the relative deck deflection is smaller. The latter is the decisive factor and results therefore in a
smaller maximum deflection than for alternative design 2. Both designs satisfy the vertical
deflection requirement of the Eurocode (Annex D, Section D.10, in this work).
5.4.3 Comparison
The following comparison between the main bridge deck design with transverse prestressing
and the alternative design proposals without prestressing is based on approximate costs for
tendons, reinforcement and concrete including material and labour, taken from VERCH [50].
In Figure 5.7, the amount of structural elements and materials of the three different bridge deck
designs is listed together with respective costs. The main design with transverse prestressing
appears to be the most costly solution at a cost of 350,640 . Design proposal 1 shows hardly any
difference. The costs for the eliminated transverse prestressing are countervailed by the additional
reinforcement and concrete as well as the additional longitudinal tendons.
12
Design proposal 2 is the most economical solution at a cost of 339,837 . It shows that costs
due to the increased amount of reinforcement are less than costs saved by the reduced amount of
concrete. The longitudinal tendons contribute considerably to the total costs and savings will have
significant impact. As mentioned earlier, proposal 2 might work well with 6 tendons per edge beam,
which would have to be checked for in detail. A reduction of 1 tendon per edge beam will then lead
to a total sum of 323,271 , which is approximately 10% cheaper than the main design with
transverse prestressing.
Comparison of costs for different bridge deck designs

Weight/Volume Cost/Unit Costs
3 3
[t] [m ] [/t] or [/m ] []
Main design: Transverse prestressing
1)
Longitudinal prestressing 12 tendons 38,23 2600 99398
2)
Transverse prestressing 373 thread bars 32,45 2600 84370
Reinforcement 500 S 62,028 800 49622,4
Concrete C50/60 469 250 117250
Total: 350640
Proposal 1: h3) = 0.61 m, no transverse prestressing
1)
Reinforcement 500 S 95,527 800 76421,6
Concrete C50/60 640 250 160000
Total: 352408
Difference to main design: 1767
3)
Proposal 2: h = 0.47 m, no transverse prestressing
1)
Reinforcement 500 S 114,81 800 91851,2
Concrete C50/60 528 250 132000
Total: 339837
Difference to main design: -10803
Notes: additional reinforcement in end cross
1)
DYWIDAG Post-tensioning system, tendon type 6827 girders is neglected
2)
DYWIDAG Post-tensioning system, thread bars, type 36 D
3)
h: structural depth at mid-span
Fig. 5.7. Comparison of costs for different bridge deck designs
Immense compression reinforcement also contributes to the high cost of the main design.
Approximations showed that a concrete slab about 10 cm deeper would make the compression
reinforcement redundant. The application of only minimum reinforcement saves 27,647 , whereas
the additional concrete causes costs of about 20,000 . However, additional longitudinal tendons
would be necessary in order to countervail the increased dead load and cross section area. Thus,
the design with transverse prestressing seems to constitute the most costly solution, regardless of
whether compression reinforcement is used or not.
Design proposal 2 appears to be the most economical solution. However, the difference to the
prestressed solution is still slight and it is necessary to consider the disadvantages of a non-
prestressed bridge deck. The requirements for railway bridges regarding vibrations and deflections
are high and often decisive. It might be possible that a non-prestressed deck does not satisfy such
demands and therefore does not constitute a feasible solution. However, network arch bridges are
stiffer than for example arch bridges with vertical hangers; deflections are smaller.
13
Nonetheless, durability reasons might necessitate prestressing in order to reduce or even

avoid cracking. When deciding between prestressed and non-prestressed solution, it is therefore
necessary to consider all aspects. Economical considerations cannot be the only criterion.
5.5 The end cross girder
The 1.75 metre wide end cross girders are formed by increasing the tie by 35 cm at the bottom.
They transfer forces between the bearings, decrease the vertical deflection at the expansion joints
and complete the wind portal frame. More details about this part of the structure can be found in
Section 5.6.5.
5.6 Constructive design of the arch root point
The detail which deserves special attention is the point of the bridge which combines the
bearing, the root of the arch, the end cross girder and the anchorages of the transversal and
longitudinal tendons. In this section the design of this important detail will be explained, considering
as an example the bridge calculated in this work.
5.6.1 Anchorage of the arch
The profile of the arch reaches the end of the bridge at an angle of 39 between the horizontal
and the arch centreline. It transfers axial force, shear forces, and bending moments about 3 axes.
Axial force
A
The axial force is split into a vertical and a horizontal
force acting in the plane of the arch. The vertical force is
taken by the bearing with its centre underneath plate B and
F
the horizontal part is taken by the longitudinal prestressing
tendons. For this purpose the root of the arch profile is full B
penetration butt-welded to a vertical end plate B which is
D
supported by the horizontal plate C above the bearing and E
serves as an anchorage for the prestressed cables, as well.
Since railway bridges are subjected to large stress
variations, the flanges of the profile have to be broadened to
increase the length of the welds to plate B and lower shear
C
stresses. The assessment of this connection detail showed
that this is still insufficient to satisfy the fatigue check (Detail
category 56). Consequently the enlargement of the flanges
was extended, so that they are supported directly by C, as
Bl 505x260x10
well. The enlargement was realized by plates D, full
penetration butt-welded to the flanges of the profile, which
gives a reasonable detail category (80) for the fatigue check.
Additionally the perpendicular position of this weld prevents
too large shear stresses.
Shear forces 10
The occurring shear forces in the arch are small and

converted into horizontal and vertical forces acting in the
plane of the arch. They are borne like the axial force. Fig. 5.8. Steel construction at arch root point
Additionally a horizontal force exists, acting perpendicularly
to the plane of the arch which is taken by the concrete
14
pressure around the steel construction of the arch root and

then taken by the bearings. The transverse prestressing
takes the produced tensile stresses in the concrete and
prevents cracking.
Bending moments
To take all acting bending moments, couples of forces in
three axes are needed. In the transverse direction a large
part of the bending moment about the longitudinal axis of the
bridge is taken by the pair of pot bearings directly below the
arch. The rest is transferred through the end cross girder
and taken by the couple consisting of the bearings on each
side of the bridge. This bending moment will be transferred Fig. 5.8. Steel construction at arch root point
by concrete pressure around the steel construction. The bending moment about the transverse axis
of the bridge is taken by a clamping action of the plates D and C. It causes an angular rotation of
the bridge end and is borne by the couple consisting of the bearings at each end of the bridge.
The bending moment about the vertical axis has to be taken by the concrete pressure around
the steel construction, as well. It will be transferred by the shear-stiff bridge deck to the bearings.
Compressive stress in the concrete around the steel structure cannot transfer the forces alone. As
before, transverse and longitudinal prestressing form the counterpart, taking the tensile forces and
preventing cracks around the steel structure.
5.6.2 Anchorage of the longitudinal tendons
The longitudinal tendons are anchored to the end plate B. The limits for centre and edge
distances are given by the manufacturer DYWIDAG [10]. The minimum distances are larger at the
anchorages than in the bridge deck. Additionally the neutral axis of the tendons should meet the
neutral axis of the arch to avoid
A B
bending moments due to F
eccentricity. Therefore the E
prestressed strands have to be
diverted with regard to the
minimum radii. Their
compressive forces are C
D
transferred by B to the concrete
towards the middle of the span.
Parts of the forces are directly
taken by the arch flanges D
which cause bending moments
in plate B. Since there is always
concrete pressure behind B
when the tendons are
prestressed, the bending top view
moments do not demand
additional vertical stiffening side view
plates.
The primary tensile splitting
Fig. 5.9. Anchorage of the longitudinal prestressing strands
forces of the four outer tendons
15
are taken by helixes according to the permission certificate of the post-tensioned strands. The two
inner tendons distribute their primary tensile splitting forces to the plates D and C. To enclose them
completely the top plate F is needed. The web of A has to be cut out to let the inner tendons pass.
Since a significant part of the horizontal forces of the tendons are taken by the arch and the
rest is already distributed to a larger area by plate B, it might be possible to eliminate the helixes to
reduce the space required for the anchorages. This must be clarified with the manufacturer
DYWIDAG.
Secondary tensile splitting forces and the lateral forces caused by the diversion of the tendons
are taken by the transverse prestressing tendons.
5.6.3 Bearings
For this first draft two pot bearings were applied at each support of the bridge. An alternative is
described in Section 5.6.6.
The pot bearings require a steel plate C above them to receive the vertical load. To achieve
proper load distribution this plate must be levelled carefully. To avoid eccentricity of the vertical
load the centre of the pot bearings is placed underneath plate B. The horizontal forces do not
exceed the permitted limits set up by the manufacturer MAURER SHNE GmbH & Co. KG [20].
Therefore additional cams, tracks or bolts are not needed. The concrete behind plate B, towards
the bridge ends, is cast at a later point during construction, to leave space for the tensioning jacks
of the longitudinal tendons. Because of that, vertical stiffening plates E are needed to support plate
C in transferring the vertical loads to the bearings.
Fig 5.10. Pot bearings
5.6.4 Transverse prestressing
The transverse prestressing bars at the arch root point are arranged in the same manner as in
the rest of the bridge deck. With the increasing depth of the concrete deck towards the end cross
girder the tendons can be situated lower and the effective depth increases, as well. Therefore
fewer tendons are needed. The width of the concrete deck also increases, which requires longer
tendons. Where the tendons cross the arch root point holes have to be made in plates D to let them
pass (Fig. 5.8). The last transverse prestressed bar behind plate B is placed after prestressing the
longitudinal tendons. If the additional length of the bridge deck behind the bearings is reduced, it is
possible to eliminate this tranverse tendon.
16
The longer transverse bars cross the plane of

the arch where the concrete deck is supported by
the hangers. On the cantilever side the neutral
axis of the tendons lies above the neutral axis of
the concrete section so that the bending moment
caused by the prestressing counteracts the
bending moments caused by the loading.
540 540 540 540 540 540 540 541 670 410 539 626 555 825
150 150 150 150 150 150 150 150 150 150 150 150 150 150
390 390 390 390 390 390 390 391 389 390 389 476 420 390
145
Fig. 5.11. Position of the transverse tendons at the end of the bridge
5.6.5 Lower hanger connection
The detail of the lower hanger connection is

shaped as described in Section 7.8. Here, it will be
shown how this detail fits into the transverse
tendons. Where lower bars cross the hanger
connection detail they can be put through the
cut-out at the bottom of the hanger connection.
The upper tendon bars have to be moved slightly
sideways, which hardly influences the even
distribution of their compression force. There is no
influence on the design of the arch root
construction detail.
Fig. 5.12. Lower hanger connection with transverse

tendons
17
5.6.6 End cross girder and concrete shape
The task of the end cross girder is to form a stiff

beam between the two bearings at the end of the side view
bridge. It distributes eccentric vertical forces, bending

moments about the longitudinal axis of the bridge,
reduces deflections, and supports the edge of the
plate-like bridge deck.
The depth of the bridge deck was therefore front view
increased at the lower surface from 430 mm to
780 mm in the middle of the bridge over 1.742 m of the
bridge length. This additional space is also needed by
the anchorages of the longitudinal tendons. The
gradual diversion of the tendons requires a gradually
increased depth of the bridge deck at the lower
surface. For aesthetic and economic reasons this section
smooth transition is applied over the whole bridge

width.
On the upper surface more space is also
necessary to fit the anchorages of the longitudinal
tendons into the bridge deck. However, only the edge
beams were enlarged. This means there is an
inclination of 9% for the footpath, which is accepted
because it is non-public and if necessary the
inclination can be reduced.
isometric view
The width of the footpath should be 0.75 m. To
save self-weight the cantilevers have a width that
allows 0.75 m clearance between the hangers and the Fig. 5.13. Outer shape of the end of the bridge
handrail. To pass by the arches on the footpath, it was
necessary to enlarge its width too. Due to the overall enlargement of the bridge deck in this area,
the broadened footpath does not affect aesthetics.
At the inner corners of the arch profile there is a slight elevation of concrete, so that rain water
and dirt do not collect close to the steel.
The large vertical forces acting on plate C cause large punching shear forces in the ends of the
cross girder. The necessary reinforcement was calculated in Section D.6.2.3, but is not shown in
the figures.
5.6.7 Alternative stilt bearing
P. TVEIT proposes a stilt bearing as an alternative to the two pot bearings. It is described in
Section 5.9. The changes to the construction of the arch root point to match the new conditions are:
1. The vertical loads will only be transferred by the vertical plate B to the bearing, because
plate C cannot be used anymore for distribution of vertical forces. This raises again the problem
with fatigue checks in the vertical weld between plates D and B. To increase the cross section area
of the welds, plate B was slit and plates D receive full penetration butt welds on both sides.
Additionally plates D have claws which rest on plate B (see Figure 5.14).
18
2. Although plate C is not capable of transferring vertical forces, it is still necessary to ensure
that primary tensile splitting forces of the inner longitudinal tendons do not cause damage. But its
dimensions can be decreased.
3. Without the load distribution of plate C the area provided to resist punching shear is
decreased. Therefore more reinforcement is necessary. Furthermore, an interconnection between
the loaded bearing plates D and B and the concrete, for example with shear studs, has to be
ensured. Their assessment was omitted and therefore not shown in the figures.
4. The bearing takes place through the cylindrical bottom surface of plate B. In Section D.7.2
the necessary length was calculated. The thickness had to be adapted for the stilt bearing too. So
plate B is twice as thick as for the pot bearings.
5. Additional vertical stiffening plates E are not necessary for construction phases.
A
42 278 482
F 235 106
B
343
10
D
1071
1071
893
1311
C
40
240
240
1002
120
1560
Fig. 5.14. Adapted design of the steel structure at the arch root point
5.7 The handrails

CHS 60.3 x 3.2
To protect pedestrians on the non-public footpaths a 1.10 metre
high handrail is applied on both sides of the bridge. Figure 5.3 shows
4
the construction details. The posts have a distance of 3 metres. The
330
double hook concrete anchor provides 4 threads to connect the anchor

I120
plate with nuts. Structural steel S 275 is used.

1100
330
wire rope,
d = 10 mm
4
320
l=100
plate
130x200x20
Fig. 5.15. Handrail
19
5.8 The drainage
The drainage is an important point when designing a network arch bridge. Since the lower
chord is so slender despite the long spans, it is difficult to obtain the additional longitudinal incline
for hidden drainage pipes. Nevertheless, it is possible to solve this problem. It just has to be kept in
mind when designing the cross section of the tie. In the following some suggestions are given.
If the gradient of the track provides a

sufficient incline it is easiest to drain the
bridge deck directly within the ballast. Then
it might be a good idea to keep a clearance
1.5 % free at the lowest point so that water can
1.5 %
flow faster. This can be protected by a
perforated steel box, which is capable of
bearing the railway loads.
Without an existing incline the easiest

option is the application of spouts with inlets
at the lowest points of the bridge that
provide drainage by free falling water. The
1.5 %
authors decided on that option. One spout
1.5 %
with a diameter of 50 mm was arranged for
the 100 m2 bridge deck.
In some cases free falling water cannot

be tolerated, and the incline of the bridge
does not allow drainage of the whole length
directly. Then pipes can be arranged
underneath the cantilevers, which drain
1.5 % towards both ends of the bridge. This
1.5 %
requires the consideration of a spout
through the edge beam, already while
designing the cross-section of the tie. Such
pipes require a minimum incline due to the
danger of sedimentation. For an example of
the assessment of the longitudinal pipe see
Annex D, Section D.8.
If such an incline cannot be provided an
1.5 %
1.5 % alternative would be an open canal which
can be cleaned from time to time.
Fig. 5.16. Drainage options
Another alternative for increasing the incline is to apply a camber to the bridge deck in the
longitudinal direction. This would favour drainage to both sides of the bridge. Such a camber might
be employed anyway, because a horizontal lower surface seems, to the human eye, to sag.
20
5.9 Investigation of two different types of bearings
5.9.1 General
The bridge is supported at the four corners of the bridge deck. Due to creep and shrinkage of
the concrete tie and expansion and contraction caused by changes in temperature these supports
must allow a horizontal movement. The maximum longitudinal horizontal deflection is 102 mm
(Section C.7). Additionally the occurring horizontal forces have to be borne. Therefore moving and
fixed bearings were combined as shown in Figure 5.17.
x bridge deck - pot bearing
fixed
y-direction slide
y x-direction slide
free float
x bridge deck - stilt bearing
fixed
x-direction slide
y compression support
Fig. 5.17. Arrangement of bearings
The two different types of bearings investigated are 1. pot bearings and 2. stilt bearings.
The choice of pot bearings was influenced by STEIMANN [37], whose bridge corresponds
essentially with the bridge calculated in the present work. This bearing type represents a present-
day standard method. The second type follows the idea of PER TVEIT and was applied in the
network arches at Steinkjer and over Bolstadstraumen, both in Norway. In the following the two
different types are described and then a comparison is drawn.
5.9.2 Pot bearings
This type of bearing is designed to sustain large vertical

loads and still allow tilt/angular rotation. It consists of metallic
piston elements that contain an elastomeric disc. The free float
type bearings have a PTFE (poly-tetra-fluoro-ethylene) insert
sliding against a polished stainless steel plate, which provides
translational movement in any direction from the neutral
position. The addition of a guide steel bar to the polished Fig. 5.18. Pot bearing in situ
steel plate and a flute to the counterpart transforms this
bearing into the guide/slide type. Both guide and free float bearings provide for simultaneous
rotation and translational movement. Shear keys are provided to bottom members to achieve
retention within the abutment. The upper member will be welded to the horizontal plate of the arch
root.
For this work pot bearings from MAURER SHNE GMBH & CO. KG [20] were used. For fixed
bearings type TF-10, free floating bearings type TGa-10 and for guide/slide bearings type TGe-10
21
were used. The maximum diameter is 770 mm and for the sliding bearings the upper member has
a length of 1110 mm.
The large bending moments reaching the bridge supports from the arch should be borne
directly by the bearings and not distributed by the end cross girder. Therefore, a pair of pot
bearings was applied at each support with a centre distance of 870 mm. Furthermore the neutral
axis of the arch and the neutral axis of the bearings have a systematic offset of 62 mm. The
bending moment caused by this eccentricity counteracts the bending moment from the arch and
the end cross girder. This results in lower maximum vertical forces on the bearings.
The horizontal forces due to friction in
the bearings while temperature changes
should be decreased. The PTFE insert
has a coefficient of friction = 0.032. With
the vertical loads of approximately 10,000
kN on each bearing this gives a horizontal
force of 1280 kN for four bearings. This is
about 31 % of the maximum horizontal
force occurring due to the other different
actions on the bridge. The permitted
horizontal force on the pot bearings is
10 % of the maximum vertical force, which
means 4000 kN for four bearings.
Therefore, it is not necessary to apply Fig. 5.19. Pot bearings at the bridge in this work
additional cams, tracks or bolts bearing
the horizontal force.
5.9.3 Stilt bearing with compression support in the middle of the end cross
girder
This type is very similar to roller bearings, with the difference that the parts of the roll which
never have contact to the top or bottom member are eliminated. This allows for large radii of the
roller bearing without having a giant roll. In the fixed bearings the load is directly transferred to a
plate with a cylindrical upper surface which is attached to the abutment. This allows angular
rotations of the end of the bridge deck but constrains horizontal deflections. Horizontal deflections
transverse to the bridge cannot be provided by this type of bearing.
The stilt bearing used in
this work consists of the lower
face of plate A (as shown in
Figure 5.20), the stilt plate B
and the elements C. It is all
made of the same steel as
used for the arches and the
hangers, S 460 ML, but will be B D
examined ultrasonically for A
evidence of laminations. The C
stilt plate has cylindrical upper
and lower surfaces. The
height of this plate must be
twice the radius of these
surfaces to allow proper roll-
off properties. The required
Fig. 5.20. Isometric view of stilt bearing and compression support
22
bow length is the maximum horizontal deflection of the bridge deck plus the diameter of the contact
surfaces according to Hertz pressing plus the eccentricity due to angular rotation of the bridge deck
end, Annex D.9.2.1. With regard to the stress distribution below the contact surface an additional
safety length should be added. In this work a bow length of 103.3 mm was necessary.
The required radius and the width of the stilt plate can be calculated from the maximal actions
and the permitted Hertz pressing. In this work the width was taken as 1560 mm corresponding to
the pot bearings. The required radius is then 400 mm. The radius and the bow length lead to the
required thickness of the stilt plate. This was taken as 120 mm. Using Hertz pressing for the
assessment leads to non-conservative results as the stilt plate does not correspond to the full
cylinder presumed for the formulas according to Hertz, see Section 10.
With this geometry the occurring nominal stress in the stilt plate is far below the yield strength,
and a sufficient buckling resistance is provided. The distribution of the large vertical forces
concentrated in a plate to the concrete of the abutment demands a stiffened steel plate C with a
width of 510 mm.
As a locking feature the bearing receives cams.
The excellent behaviour of the stilt plate for longitudinal horizontal movements and rotations
faces the impossibility of rotations transverse to the bridge. To avoid strong eccentric strains, the
bending moments about the longitudinal axis of the bridge should be reduced. Therefore PER TVEIT
suggested a compression support D (Figure 5.20) in the middle of the end cross girder, which is
not connected to the bridge deck. For dead load only, the clearance between the bridge deck and
the compression support should be 3 mm. This means the compression support will only act when
live loads cause deflections larger than 3 mm and only take compression forces. The advantage is
that this kind of bearing can consist of a simple fixed bearing, because in the short time of contact
no horizontal deflections of the bridge will occur.
For D the elastomeric deformation bearing MAURER
Verformungslager Typ 1/2 was used, [19]. It consists of a
reinforced elastomeric cuboid with one steel anchorage plate at
the bottom, which is attached to the concrete of the abutment.
The bottom side of the bridge deck is strengthened by another
steel plate anchored with shear studs.
The maximum bending moments could be reduced from Fig. 5.21. Elastomeric reinforced
2345 kNm to 930 kNm before contact and 934 kNm with contact, deformation bearing
which is still considerable. As a further reduction the same eccentricity as for the pot bearings
between the neutral axis of the bearing and the neutral axis of the arch was used. The eccentricity
of the loads on the stilt plate was then sufficiently small.
Bearing D receives a maximum static vertical load of 1337 kN. Since, the bridge deck receives
considerable acceleration while a train is passing, the static part of the force will be increased by
impact. PER TVEIT suggested increasing the static part by 20 %.
Special attention is required for the design of the end cross girder. In this thesis it was only
designed for the pot bearings. Nevertheless problematic points will be mentioned. For dead load
the structural behaviour is much like that of a simply supported beam. On the other hand when the
compression support acts, the end cross girder is a continuous beam and bending moments above
the compression support change their direction. Transverse prestressing and reinforcement have
to be assessed for both cases. PER TVEIT suggested reducing this problem by the use of
compression bearings underneath each track.
The coefficient of friction for roller bearings made of structural steel is given as = 0.03,
PETERSEN [24], page 1156. Since this value is given for rolls with diameters of 100 to 200 mm, it is
believed that due to an increase of the lever arm with a diameter of 800 mm the coefficient of
23
friction can be reduced. It is assumed to be = 0.0075. With the same vertical loads as for the pot
bearings the maximum horizontal force due to friction in the bearings while temperature changes is
300 kN. That is 15 % of the rest of the horizontal forces.
5.9.4 Conclusion
Both types of bearings have advantages and disadvantages. For an evaluation the following
comparison will help.
Pot bearing Stilt bearing

Horizontal force due to friction: Horizontal force due to friction:
1280 kN 300 kN
Overall height: 175 mm Overall height: 1120 mm
Overall width: 1560 mm Overall width: 1560 mm
Overall length: 1110 mm Overall length: 510 mm
Free floating possible Only one direction sliding possible
Tilt about transverse axis / no
Free tilt/angular rotation rotation about longitudinal axis
allowed
low cost material / expensive
Expensive
manufacturing
Additional bearing in the middle of
the end cross girder needed
Danger of skew roll off
Assessment of end cross girder Assessment of end cross girder for
for one structural behaviour two structural behaviours
Usual static/dynamic internal Impact forces due to shock on
forces compression support
Easy replacement of stilt plate,
Difficulties for replacement
difficult replacement of members
No extra construction for bearing Extra construction for bearing of
of horizontal forces needed horizontal forces needed
The general performance of the pot bearing shows a clear transfer of forces, no restraining
loadings and easy handling. This bearing type is preferred by the authors.
The application of the stilt bearing is limited to narrow and straight bridges, where not disturbed
by a large construction height of the bearing. The advantage over the pot bearings might be less
expense if manufacturing costs are reduced by hiring qualified steel factories in developing
countries.
24
5.10 Summary of materials used
In figures 5.24 to 5.27 on

the following pages the weights
of the materials calculated in the
assessment can be found. The
alternatives stilt bearing and
bridge deck without transverse
prestressing are included there, 10
as well. The overall steel weight t/m
of the network arch calculated in 8
this work is 376 tons. The
temporary lower chord adds steel weight per track
6
and meter of span
another 60 tons, but is not
considered as steel weight of
4
the bridge. As a comparison to
other bridges already built,
2
Figure 5.22 shows several
railway bridges in a diagram
0
depending on their spans and 40 80 120 160 m 200
steel weight per track/span. It is main span
still obvious that a network arch Fig. 5.22. Steel weight of railway bridges with ballast
is an efficient and competitive
structure.
Steel Number Steel weight per

weight Span of Year of track per meter
Location Bridge type [t] [m] tracks construction span [t/m]
1 Tied arch 48 5
2 Truss 175 5.8
[17]
3 Tied arch 58 2.8

before 1975
4 Tied arch 120 7.5

5 Tied arch 43 3.7
Data fromH
6 Tied arch 75 3.2

Tied arch w ith
7 77 5.4
orthotropic steel deck
8 Tied arch w ith truss 186 9
9 Tied arch w ith truss 187 7
10 River Main at Steinheim Tied arch 1450 160 1 unknow n 7.9
River Elbe at Torgau, Tied arch w ith
11 440 80 1 1996 5.5
track Halle/S.-Guben orthotropic steel deck
165/36
12 Tw o-hinge arch bridge 5350 2 1908-10 7.27
River Rhein 8
Oder-Havel-canal at
13 Tied arch 300 65 1 1999 4.62
Zerpenschleuse
Spreebrcke Station Tw o-hinge truss
14 1600 55 6 1997-1998 4.84
Friedrichstrasse, Berlin bridge
Bridge over the 50.55/
15 Continuous beam 510 1 1984 4.06
Sachsendamm, Berlin 41/34
16 Bridge in this w ork Netw ork arch 376 100 2 none 1.88
Fig. 5.23. Bridges used for comparison
25
Steel weights
Weight Total Weight [t]
per unit number/length S460 St St
S355 S275 S 500
Arches ML 1080/1230 1500/1770
Low er segments W 360x410x900 900 kg/m -/69.76 m 62.78
Middle segments W 360x410x634 634 kg/m -/144.48 m 91.6
Hangers
Smooth bar d=60 mm A = 0.00283 m2 22.2 kg/m -/1650.24 m 36.63
Wind bracing
Bending members CHS 219x10 51.55 kg/m 28.908 m 1.49
Truss members CHS 219x8 41.67 kg/m 316.59 m 13.19
Arch root point
Enlargment flanges V = 0.09417 m3 739.23 kg 8/- 5.91
Vertical plate V = 0.08228 m3 645.9 kg 4/- 2.58
Horizontal plate V = 0.186 m3 1460.1 kg 4/- 5.84
Vertical stiffeners V = 0.0126 m3 98.75 kg 8/- 0.79
Cover plate V = 0.00133 m3 10.44 kg 4/- 0.042
Handrail
Top holm CHS 60.3x3.2 4.5 kg/m -/200 m 0.9
Posts I 120 5.57 kg/m -/83.33 m 0.46
Smooth bar d=10 mm A = 0.785 cm2 0.62 kg/m -/400 m 0.25
Anchor plate V = 0.00052 m3 4.09 kg 68/- 0.28
Transverse
prestress
DYWIDAG
Transverse tendons
threadbar 36D 8.27 kg/m -/3923.96 m 32.45
Anchor plates V = 0.00305 m3 23.94 kg 746/- 17.9
Longitudinal
prestress
DYWIDAG
Longitudinal tendons
Type 6827 31.86 kg/m -/1200 m 38.23
Reinforcement
Re-bars 10 0.617 kg/m -/21190.95 m 13.07
12 0.888 kg/m -/3466.67 m 3.08
20 2.47 kg/m -/13166.67 m 32.52
Stirrups 10 0.617 kg/m -/1333.33 m 0.82
sum of re-bars+stirrups 12 0.888 kg/m -/5000 m 4.44
increased by 15 % 20 2.47 kg/m -/952 m 2.35
Sum of each material

206.2 14.68 19.8 32.45 38.23 64.72
Sum prestress steel 70.68
Sum reinforcement 135.402
Sum structural steel 240.602
Sum of all persistent
steel 376.004
Temporary lower
chord
Longitudinal beams HEB 220 71.44 kg/m -/198 m 14.14
Transversal beams IPEa 550 91.85 kg/m -/407.1 m 37.39
Wind bracing L 120x10 18.21 kg/m -/399 m 7.27
Connection part HEB 200 61.31 kg/m -/1.2 m 0.074
Bracing at transition
point V = 0.0039 m3 30.62 kg/m 8/- 0.24
Sum of temporary
lower chord 59.11
Total sum 435.118
Fig. 5.24. Steel weight of the bridge, calculated in this work
Material for formwork

Weight Weight
Amount
per unit [t]
Plywood
Formwork sheets 2 2
t = 18 mm 12.6 kg/m 1357 m 17.1
Formwork timber Doka formwork
beams beam H 20 P 5.1 kg/m 4704 m 24
Concrete for bridge deck

3 3
C50/60 V = 251.9 m 2.5 t/m - 1259.4
Fig. 5.25. Formwork material needed for the erection
26
Steel weights of alternative "stilt bearing"

S355 S275 S 500
ML 1080/1230 1500/1770
As in Figure 5.24 but:
Arch root point
3
Enlargment flanges V = 0.11038 m 866.48 kg 8/- 6.93
3
Horizontal plate V = 0.026 m 204.1 kg 4/- 0.82
3
Vertical plate V = 0.173 m 1358.1 kg 4/- 5.43
Stilt bearing
Vertical bearing plate

V = 0.149 m3 1169.7 kg 2/- 2.34
Sum of each material

207.36 14.68 19.75 32.45 38.23 29.47
Sum reinforcement 100.15
Sum of all persistent
steel 341.94
Fig. 5.26. Changes in the steel weight due to the application of the stilt bearings
Ste e l w e ights of alte rnative w ithout trans ve rs e pre s tre s s ing, propos al 1; h = 0.61 m at m id-s pan
S355 S275 S 500
ML 1080/1230 1500/1770
Trans ve rs e
pre s tre s s ing
Transverse tendons -/0 m
A nchor plates 0/-
Longitudinal
pre s tre s s ing
DY WIDA G
Type 6827 31.86 kg/m -/1400 m 44.604
Re inforce m e nt
Re-bars 10 0.617 kg/m -/7433.33 m 4.59
12 0.888 kg/m -/13623.33 m 12.1
14 1.21 kg/m -/11850.0 m 14.34
25 3.85 kg/m -/11766.67 m 45.3
Stirrups 10 0.617 kg/m -/1333.33 m 0.82
sum o f re-bars+ stirrups 12 0.888 kg/m -/6666.67 m 5.92
increased by 15 %
20 2.47 kg/m -/952 m 2.35
Sum of e ach
m ate rial 207.4 14.68 1.89 0 44.604 98.23
Sum reinf orcement 142.837
Sum of all
pe rs is te nt s te e l 366.767
Fig. 5.27. Changes in the steel weight for alternative bridge deck, proposal 1
27
Ste e l w e ights of alte rnative w ithout trans ve rs e pre s tre s s ing, propos al 2; h = 0.47 m at m id-s pan
S355 S275 S 500
ML 1080/1230 1500/1770
Trans ve rs e
pre s tre s s ing
Transverse tendons -/0 m
A nchor plates 0/-
Longitudinal
pre s tre s s ing
DY WIDA G
Type 6827 31.86 kg/m -/1400 m 44.604
Re inforce m e nt
Re-bars 10 0.617 kg/m -/7433.33 m 4.59
12 0.888 kg/m -/1733.33 m 1.54
14 1.21 kg/m -/23740.0 m 28.73
25 3.85 kg/m -/15128.57 m 58.25
Stirrups 10 0.617 kg/m -/1333.33 m 0.82
sum o f re-bars+ stirrups 12 0.888 kg/m -/6666.67 m 5.92
increased by 15 %
20 2.47 kg/m -/952 m 2.35
Sum of e ach
m ate rial 207.4 14.68 1.89 0 44.604 117.5
Sum reinf orcement 162.134
Sum of all
pe rs is te nt s te e l 386.064
Fig. 5.28. Changes in the steel weight for alternative bridge deck, proposal 2
Fig. 5.29. Photo composition of the bridge calculated in this work
28
CHS 219.1x10
CHS 2
19.1x8
CHS 219.1x8
CHS
2 19.1
x 8
CHS 219.1x8
CH
S2
19.
1x8
W 360x420x634
CHS 219.1x8
CH
S2
19
.1x
8
CHS 219.1x8
CH
S 8
21 1x
9. 9.
1x 21
S
17000
17920
8
CH
CHS 219.1x10
C
H
S
21
9.
1x8
1
W 360x420x900
10
2200 2500
109
750
100 m
870
10150
12240
Fig. 5.30. Front view
Double track railway network arch bridge assessed for the

European standards.
Fig. 5.31. Side view

29
Section 6
Optimisation of the hanger arrangement
6.1 Scope
Section 6 constitutes one of the main fields of research in this work. It was to optimise the
number and arrangement of hangers regarding maximal hanger forces and stress ranges. In a
preliminary investigation two algebraic descriptions were introduced for hanger arrangements
similar to the ones considered as near optimal by former studies. Thereupon, it was possible to
vary the geometry within these descriptions and analyse the influence lines of the structure using a
3D-FEM-model by means of SOFiSTiK structural analysis software. The results of 850 different
hanger arrangements were compared searching for minimum internal forces.
Delving into this topic and studying theories of the optimisation of structures, the authors had
an idea of a new description of the hanger arrangement. This was also converted into a 3D-FEM-
model, and the numeric analysis of another 80 bridges showed considerably improved results.
With the new type of hanger arrangement, about 100 bridges more were calculated varying
other geometry parameters, such as span, arch rise, number of hangers and curvature of arch, to
develop further improvements.
The knowledge about the structural behaviour of network arches obtained from the
investigations is discussed and explanations are suggested. As a summary a preliminary scheme
is given which allows designing a network arch railway bridge according to the results found.
6.2 What is an optimal hanger arrangement?
According to GRAF/STRANSKY [13], page 1/1, an optimal structure is characterised by the

following attributes:
-safe/durable
-economic/inexpensive
-fast/easy to build
-functional, aesthetic, ecological
The complexity of these demands would cause extensive work to satisfy them, so the number
of attributes considered was reduced. Therefore, in the following the more appropriate word
improve is used instead of optimise. The adapted attributes are:
-Minimal maximum hanger forces
-Minimal maximum bending moments in the arch about the horizontal axis
-Minimal variation in hanger forces
-Minimal variation in bending moments in the arch about the horizontal axis
The bending moments in the arch about the horizontal axis correspond in a certain way to
those in the tie, so minimal bending moments in the arch will consequently give small bending
moments in the tie. Railway bridges, like the one calculated in this work, are subjected to fatigue
strains, so the force variation should always be considered. Fulfilling these demands saves
material for hangers and arches, which can mean a less expensive structure and easier erection
due to less weight. Furthermore it leads to more slender arches and hangers, which might be a
30
Diploma Thesis Brunn & Schanack Section 6: Optimisation of the hanger arrangement
criterion for aesthetics. So the limitation to the mentioned demands should still lead to important
improvements in the structure.
6.3 The parameters
The parameters considered in this work to influence the internal forces are:
1. Location of hanger nodes along the arch
2. Location of hanger nodes on the lower chord
3. Number of hangers and span of the bridge
4. Rise of the arch
5. Loading
6. Curvature of the arch
Due to the number of parameters and the complexity of their influences, it was decided to use
experimental improvement. As a further limitation of the extent the following simplifications were
assumed:
To 1. According to the proposals by TVEIT [45], page 26, the hangers were placed equidistantly
along the arch for all hanger arrangements, and only the hanger nodes at each arch end
were considered to be variable (see Section 6.7.5).
To 2. The location of the lower hanger nodes was obtained by algebraic/geometric descriptions
as explained in Sections 6.5.1, 6.5.2 and 6.6.3. Only the first few lower hanger nodes
were shifted manually (see Section 6.7.5).
To 3. For the calculations the number of hangers was chosen to be 44 and the span was
100 metres, as it was in STEIMANN [37]. Additional investigations were carried out varying
the span and the number of hangers (see Section 6.7.1).
To 4. A rise of the arch of 17 meters was assumed, which gives a rise/span-ratio of 0.17. In
Section 6.7.2 ratios between 0.14 and 0.18 were tested.
To 5. The bridges were loaded by prestressing and dead load. Furthermore, one load model
71 with a partial safety factor of 1.5 was applied on each track. For dead load a partial
safety factor of 1.35 was used. This causes loads that are close to the decisive design
forces for the ultimate limit state design checks.
The application of these partial safety factors gives too high stress ranges regarding
fatigue checks. Since it only influences the absolute value, the results still describe
correctly the tendency within the improvement process.
To 6. In TVEIT [43], page 2195, it is suggested that the arch is a part of a circle. This was
assumed for all investigations. The radius of curvature near the ends of the arch was
decreased in Section 6.7.4, because it gives benefits for the wind portal (TVEIT [48],
page 4).
6.4 Evaluation of the results
The calculations were performed with the help of a 3D-model in SOFiSTiK (see Annex C.1).
Resulting from the analysis of the hangers influence lines, the maximum and minimum hanger
forces were recorded by this software. So were the bending moments in the arch about the
horizontal axis. The following names are used in the following sections.
MaxN = maximum hanger force
the maximum axial hanger force occurring in the calculation of one bridge
31
AveN = average hanger forces

the arithmetic average of the maximum axial hanger forces reported for one bridge
N = maximum variation of hanger force
Variation means the difference between minimum and maximum axial hanger
forces of each hanger of one bridge. The maximum of these values is recorded.
AveN = average variation of hanger force
the arithmetic average of the axial hanger force variations
The same is valid for the bending moments in the arch about the horizontal axis along the whole
bow length:
MaxM, AveM, M, AveM
The average values were of interest because the arch root point forms a clamping; therefore,
hanger arrangements best for the middle of the span may cause unfavourable results for the ends.
The occurring maverick axial hanger forces deviated significantly from the rest. The maximum
bending moments in the arch are subject to deviations caused by these maverick hanger forces, so
the average bending moments will also be involved into the evaluation. It is known that the bending
moments are not uniformly distributed along the arch, especially not in arches with vertical
hangers. So the relevance of the average bending moments in respect of minimisation of the
maximal bending moments might be small. But it is believed that a hanger arrangement which
causes small average bending moments also leads to small maximum bending moments.
An improved hanger arrangement can be expected if all values are minimums. Since this did
not occur for one and the same hanger arrangement, a certain weighting of the results had to be
made. The maximum forces shall satisfy the ultimate limit state and the stress range the fatigue
limit state. The target would be to find a solution satisfying both limit states in a way that the
utilisation rate of all checks is at about the same level, ideally 100%.
Therefore, not only the internal forces, but also the design checks would be needed in the
improvement process providing indications about the weighting to be used. In turn, this would also
mean that all design-relevant load combinations have to be included. Their consideration would
extend the process drastically due to the very time-consuming data processing. Furthermore, the
evaluation of that data would go beyond the scope of this work. So, the following evaluations based
on intuition were used:
1. A search was conducted for the minimum of each data series (MaxN, AveN, N, etc.) of one
investigation, for example varying the span. Then each value of the data series was divided by
this minimum. That means the data series were scaled, so that the minimum of each equals 1.
2. Then the scaled data series were summed up.
SummaxN = MaxN + N
SumaveN = AveN + AveN
Summax = AveN + (0.5 AveM + MaxM)
Sumvariation = AveN + (0.5 AveM+M) Terms in brackets scaled another time
Sumall = Summax + Sumvariation
The hanger arrangement leading to the minimum value of Sumall is considered to be the best.
Regrettably, from the investigations carried out in Section 6.5.1 and 6.5.2 no values of the
bending moments in the arch were considered. It was not expected that considering only the axial
hanger forces would lead to results that are not desired.
In Annex F, diagrams with all data series can be found, so that one can search for an
improvement applying its own weightings.
32
6.5 Preliminary investigations
To obtain a first insight into the problems of the hanger arrangement, two algebraic
descriptions of the geometry were set up. They are suitable for obtaining hanger arrangements
similar to the ones considered as near optimal by former studies.
6.5.1 Variation of the lower hanger nodes by the node distances
The mathematical model

In TVEIT [45], page 8, and TEICH & WENDELIN [38], page 16, it is suggested to have equally
spaced hanger nodes along the middle of the tie. Towards the bridge ends the distances of the
nodal points of one set of hangers increase while the ones of the other decrease. So, for one set of
hangers the node distances are arranged as shown in Figure 6.1 along the tie. The other set of
hangers is the mirrored equivalent. That is why in the following only one set of hangers is
considered.
increasing hanger node distance along the lower chord

Fig. 6.1. One set of hangers
Starting from the right, the node distances increase with a certain increment. There is hardly a
change of the distances in the mid range, and further to the left end the distances increase again
with the same increment as on the right side.
The authors intended to calculate many different hanger arrangements which all follow this
scheme. For this purpose an algebraic description of the mentioned geometry was needed. The
same behaviour was then found in the curvature of an ellipse. The left part of Figure 6.2 shows the
necessary adaptations, so that one can find the node distances by the help of an ellipse.
ax ymm
ax ymm
is e
s
is e
s
of try
of try
a a
b
node distance along the lower chord
x xm x
xm
bridge span divided into n+1 equal parts
(n - number of hangers per hanger set) usable range of ellipse
Fig. 6.2. Node distance spaced by the help of an ellipse (schematic illustration)
33
Two variables make it possible to change the hanger arrangement. The first is the ratio
between semiminor axis b and semimajor axis a. The second is found, if not the whole usable bow
length of the ellipse is deployed (see right part of Figure 6.2). The ratio between the unused range
and the usable range will be the second variable.
b
ellip = ratio between semiminor and semimajor axis, ellip = [0..1]
a
x
range = utilisation of the ellipse, range = [0..1]
xm
Parameter ellip allows a variation between equidistant spacing (ellip = 0) and an extreme
increase from one distance to the next (ellip = 1). Parameter range can be used to exclude the
larger curvature changes at the beginning and the end of the usable range and obtain more even
node distances.
The analysis
As mentioned in Section 6.3 for this calculation the following fixed parameters were used.
Span of the bridge: s = 100m

Arch rise: f = 17m
Number of hangers per set: n = 22
The results from the calculation of 450 different arrangements are presented in figures 6.3 to 6.7. In
Annex F, Section F.2 large size illustrations of these diagrams are provided.
MaxN [kN] range

0
0.2
0.4
0.6
0.8
1
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
0.48
0.44
0.4
0.36
0.32
0.28
0.24
0.2
0.16
0.12
0.08
0.04
0
ellip
1174-1182 1182-1190 1190-1198 1198-1206 1206-1214
Fig. 6.3. Results for the maximum axial force of all hangers
AveN [kN] range

0
0.2
0.4
0.6
0.8
1
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
0.48
0.44
0.4
0.36
0.32
0.28
0.24
0.2
0.16
0.12
0.08
0.04
0
ellip
855-860 860-865 865-870 870-875 875-880 880-885 885-890
Fig. 6.4. Results for the average axial force of all hangers
34
N [kN] range
0
0.2
0.4
0.6
0.8
1
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
0.48
0.44
0.4
0.36
0.32
0.28
0.24
0.2
0.16
0.12
0.08
0.04
0
ellip
680-685 685-690 690-695 695-700 700-705 705-710
Fig. 6.5. Results for the maximum variation of axial forces of all hangers
AveN [kN] range

0
0.2
0.4
0.6
0.8
1
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
0.48
0.44
0.4
0.36
0.32
0.28
0.24
0.2
0.16
0.12
0.08
0.04
0
ellip
630-635 635-640 640-645 645-650 650-655 655-660
660-665 665-670 670-675 675-680 680-685 685-690
Fig. 6.6. Results for the average variation of axial forces of all hangers
Number of relaxed hangers range

0
0.2
0.4
ra n g e
0.6
0.8
1
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
0.48
0.44
0.4
0.36
0.32
0.28
0.24
0.2
0.16
0.12
0.08
0.04
ellip
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10
10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18-19 19-20
Fig. 6.7. Number of relaxed hangers
If we look at figures 6.3 to 6.7, the results appear to be very sensitive in respect of small
variations in the parameters ellip and range. Minimum values of all considered forces accumulate
along a curve, which ends in the vicinity of ellip = 0.2 and range = 0.2. Towards either side next to
this curve, values increase significantly, which represents the high sensitivity of the system. In
other words, the grid chosen for investigation is too coarse to make it possible to read sensible
results from the diagrams. However, the coarse diagram is sufficient enough to give reasonable
evidence about the existence of curves on which minima are found for each attribute of the
optimisation. They will be called curves of minima.
The curves of minima of each data series lie close together and the envelope will be called
valley of minima. Taking two different parameter combinations of ellip and range from the curve of
minima of maxN and plotting the respective hanger arrangement shows their similarity (figures 6.8,
35
6.9). These two hanger arrangements are virtually equal. Consequently, they result in virtually
equal values of maxN, which can be seen in the diagram again.
Fig. 6.8. Hanger arrangement with parameters ellip = 0.54 and range = 0.65, span: 100m, arch rise 17 m, number of
hangers per arch plane: 44
Fig. 6.9. Hanger arrangement with parameters ellip = 0.84 and range = 0.75, span: 100m, arch rise 17 m, number of
hangers per arch plane: 44
However, a slight variation still exists along the curves of minima. The results increase
towards the upper right corner in Figure 6.10. Except for that range, one and the same minimum
value can be assumed to lie everywhere on each respective curve.
Figure 6.12 N
or
af xN
1 ellip of
minim
m af
or
ma
0
e n i
rv mi
cu of
rve
less steep hangers cu
no relaxation of hangers a
inim
range
of m htly
y slig
valle es s
how s
lt
curv er resu
l along each curve high
results are virtually equa
steep hangers
frequent relaxation of hangers
1
Figure 6.9 Figure 6.8 Figure 6.11
Fig. 6.10. Minimum curves within the data range
The arrangement in Figure 6.11 is taken from the curve of minima of N. Again, as with maxN,
all parameter combinations lying on that curve result in virtually one and the same arrangement.
When looking at Figure 6.10, below the valley of minima, parameter combinations lead to
arrangements with fewer hanger crossings and steeper hangers. Further up in the diagram,
hangers become less steep, which leads to more hanger crossings. That means hanger
arrangements resulting from the curve of minima of maxN show steeper hangers than in the case
of N. Also, hanger relaxation appears not to happen above, but below the valley of minima,
which can also be found in TVEIT [45], page 27: Too steep hangers lead to too much relaxation of
hangers.
Figure 6.11 shows an example of an arrangement resulting in many relaxed hangers and
Figure 6.12 a hanger arrangement giving minimal stress ranges.
36
Fig. 6.11. Hanger arrangement with parameters ellip = 0.3 and range = 0.6, span: 100m, arch rise 17 m, number of hangers
per arch plane: 44
Fig. 6.12. Hanger arrangement with parameters ellip = 0.3 and range = 0.3, span: 100m, arch rise 17 m, number of hangers
per arch plane: 44
Since there seems to be a certain dependence between the parameters ellip and range and the
valley of minima exists, it is sensible to make further investigations along a fixed parameter ellip,
hence along a section through the valley of minima. This is necessary due to the coarse grid used
so far. Obviously the two curves of minima shown in Figure 6.10 represent improvements of the
respective goals, but for final conclusions refinement is needed. The fixed parameter for further
investigation will be ellip = 0.5.
Less steep hanger slopes Steeper hanger slopes

number of relaxed hangers
10
11
16
29
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
2
2
2
2
2
2
3
4
4
4
4
5
6
6
8
8
2100 4
1900 3.5
m axN
ave N
N
1700 ave N 3
Sum m axN
evaluation criterion
1500 2.5
N [kN]
1300 2
1100 1.5
900 1
700 0.5
500 0
0.3
0.32
0.34
0.36
0.38
0.4
0.42
0.44
0.46
0.48
0.5
0.52
0.54
0.56
0.58
0.6
0.62
0.64
0.66
0.68
0.7
0.72
0.74
range
Fig. 6.13. Results of a refined variation of parameter range, while ellip = 0.5
37
Figure 6.13 shows that the minima of maxN and N do not coincide, as already seen in Figure
6.10. Note the fact that hanger relaxation occurs very close to the minima of N. For maxN,
hanger relaxation even happens at the minimum. However, results with hanger relaxation have to
be looked at critically, as the computing of influence lines is carried out in linear fashion where
actual relaxed hangers take compression. This causes falsified internal forces for other structural
elements. Since influence on the tendency of the curve shape is slight and otherwise the majority
of the bridges calculated show none or only a few relaxed hangers, it is nevertheless assumed that
the results are comparable.
The global minimum of the maximum hanger forces found in Figure 6.13 is at 1162 kN,
whereas the global minimum of the maximum variation of the axial hanger force is found to be
685 kN.
SummaxN was calculated as described in Section 6.4 to make conclusions about an
improvement satisfying both goals, the minima of maxN and N. The minimum of the obtained
data series can be taken from Figure 6.13. Having chosen the refined investigation along
ellip = 0.5, the best hanger arrangement is found at range = 0.51.
The respective forces are as follows:
Maximum axial hanger force: maxN = 1189 kN
Maximum variation of axial hanger force: N = 685 kN
(at the end of Section 6.6.3 the best results of
all examined descriptions of hanger
arrangements are listed)
Remarkable is the fact that the hanger arrangement giving the minimum value for maxN also
gives the minimum of the SummaxN. This indicates that deviation from the best hanger
arrangement increases stress ranges more than maximal hanger forces.
The investigation shows that less steep hangers cause smaller stress ranges. Therefore less
steep hangers are preferable in respect of fatigue design checks. On the other hand, the hangers
must be steeper if smaller maximum hanger forces are desired.
In figures 6.8, 6.9, 6.11 and 6.12 the first and the last hangers do not fit in a regular pattern. It
needs to be mentioned that their geometry is not a mistake, but also the result of the mathematical
model providing node coordinates along the lower chord.
38
6.5.2. Variation of the lower hanger nodes by the slope of the hangers
This hanger arrangement is based on equidistant nodes along the arch from which the hangers
slope down with a certain inclination until they reach the tie. It was derived from the diagrams in
TVEIT [45], page 27, dealing with the connection between slope of hangers and relaxation of the
first hanger. Furthermore, in TVEIT [42], page 6, a network arch is presented which is characterised
by a constant angle change between two adjacent hangers.
The variables describing the arrangement are the start angle and the change of the inclination
from one hanger to the next.
start angle 1 2 1-2=
Fig. 6.14. Definition of start angle and angle change
The angle change, , can be described by any mathematical function. For simplicity the linear
function (x) = ax+b was chosen, where x is the number of the hanger; a and b are parameters
varying the hanger arrangement. This turned out to be sufficient and sufficiently extensive to vary
mainly the constant part of the angle change to achieve an improvement for this kind of hanger
arrangement description.
To gain an insight into the behaviour of forces with this hanger arrangement, 400 different
bridges were calculated. With this algebraic description it is possible to obtain extreme hanger
arrangements. Thus, only start angles and angle changes that give reasonable geometries were
investigated (start angles from 50 to 84 and constant angle change from -0.3 to 3.5).
Results of the analysis
The results are shown in the diagrams of figures 6.15, 6.17, 6.19, 6.21, 6.22 and 6.23. The
ordinate visualizes one value for each bridge.
39
maxN - 44 hangers, span 100 m, h = 17 m 1410-1420

1400-1410
1390-1400
1380-1390
1370-1380
1360-1370
1350-1360
1340-1350
1330-1340
1320-1330
1310-1320
1300-1310
1290-1300
1280-1290
84 1270-1280
1260-1270
81 1250-1260
78 1240-1250
1230-1240
74 1220-1230
70 1210-1220
1200-1210
68 1190-1200
1180-1190
66 1170-1180
64 start angle [] 1160-1170
N [kN] 1150-1160
62 1140-1150
60 1130-1140
-0.3
1120-1130
0.1
0.4
58 1110-1120
0.7
1100-1110
1.3
56
1.6
1090-1100
1.9
1080-1090
2.2
50
2.8
3.2 1070-1080
3.5
constant angle change [] 1060-1070
1050-1060
Fig. 6.15. Results for maxN
Figure 6.15 shows the maximal hanger

force occurring in each of the 400 bridges.
They have their minimum at a start angle of
about 68 and a constant angle change of Fig. 6.16. Hanger arrangement with a start angle of 68
0.6. The minimum value is 1055 kN. and constant angle change of 0.6
Assuming an equal cross section for every
hanger, this would be the force the hangers
are designed for.
40
N - 44 hangers, span 10 0m, h = 17 m 1010-1020

1000-1010
990-1000
980-990
970-980
84 960-970
81 950-960
940-950
78 930-940
920-930
74
910-920
70 900-910
890-900
68
880-890
66 start angle [] 870-880
860-870
64
850-860
62 840-850
830-840
60
820-830
58 810-820
800-810
56 790-800
N 50 780-790
-0.3
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
770-780
2.4
2.6
2.8
3
3.1
3.2
3.3
3.4
3.5
[kN]
760-770
constant angle change []
750-760
740-750
Fig. 6.17. Results for N
Figure 6.17 shows the maximal difference

between the maximal and the corresponding
minimal hanger force. Assuming equal
cross - sections for all hangers, this value Fig. 6.18. Hanger arrangement with a start angle of 70 and a
indicates the stress range for fatigue checks. constant angle change of 1.4
The minimum is 741 kN at a start angle of 70
and a constant angle change of 1.4.
In order to draw conclusions about an improvement, the claims of the maximal forces and the
maximal force variation will be combined. To achieve this, SummaxN was calculated according to
Section 6.4 and shown in Figure 6.19.
41
SummaxN - 44 hangers, span 100 m, h = 17 m
2.38-2.4
2.36-2.38
2.34-2.36
2.32-2.34
2.3-2.32
2.28-2.3
2.26-2.28
84 2.24-2.26
81 2.22-2.24
78 2.2-2.22
74
2.18-2.2
70
2.16-2.18
68
2.14-2.16
66
2.12-2.14
64 start angle []
2.1-2.12
evaluation 62
criterion 2.08-2.1
60
2.06-2.08
-0.3
-0.1
0
0.1
0.2
0.3
58
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
56
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
2.2
50
2.4
2.6
2.8
3
3.1
3.2
3.3
3.4
3.5
Fig. 6.19. Results for SummaxN
Here a new minimum is formed at a start

angle of 66 and a constant angle change of
0.9. The maximal hanger force is 1099.1 kN
and the maximal force difference is 765.1 kN. Fig. 6.20. Hanger arrangement with a start angle of 66 and
a constant angle change of 0.9
When we analyse the results from the calculation, it is seen that several hanger forces,
especially of the hangers at the ends of the bridge, seriously deviate from the rest due to a
disturbance range caused by the clamping of the arch to the tie. It is believed that these mavericks
can be adapted at a later point of the improvement process. So it is regarded as more meaningful
to look at the average values. Explanations for how to adapt the hanger arrangements to improve
the hanger forces near the ends of the arch are given in Section 6.7.5.
42
aveN - 44 hangers, span 100 m, h = 17 m
990-995
985-990
84 980-985
81 975-980
78 970-975
74 965-970
70 960-965
68 955-960
950-955
start angle [] 66
64 945-950
62 940-945
60
58
N [kN]
56
3.4
3.2
3
2.6
2.2
2
50
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-0.3
Fig. 6.21. Results for aveN
Figure 6.21 shows the average of all maximum hanger forces of each bridge. It is now
apparent that, on average, the minimum is not found at one concentrated point, but rather in an
expanded concave area with further decline at the end of the area examined.
aveN - 44 hangers, span 100 m, h = 17 m
708-710
706-708
704-706
84 702-704
81 700-702
78 698-700
74 696-698
70 694-696
68 692-694
690-692
66 start angle []
688-690
64
686-688
62
684-686
60 682-684
58 680-682
N [kN]
56
-0.3
0
0.2
0.4
0.6
0.8
1
1.2
50
1.4
1.6
1.8
2
2.2
2.6
3
3.2
3.4
Fig. 6.22. Results for aveN
43
By considering the average difference between the maximal and the corresponding minimal
hanger force we arrive at a similar expanded range of minimal values as in Figure 6.22. A slight
decline at the ends of the diagram can also be reported.
Again, to combine the demands of the design check for maximal stresses and the fatigue
check, SumaveN was calculated and shown in Figure 6.23.
SumaveN - 44 hangers, span 100m, h 17m
2.285-2.3
2.27-2.285
2.255-2.27
2.24-2.255
2.225-2.24
2.21-2.225
2.195-2.21
84
2.18-2.195
81
78 2.165-2.18
74 2.15-2.165
70 2.135-2.15
68
2.12-2.135
66
64 2.105-2.12
62 2.09-2.105
start angle [] evaluation
60 2.075-2.09
criterion
58 3.5 2.06-2.075
3.4
3.3
56
3.2
3.1
3
2.8
2.6
2.4
2.2
2.1
2.045-2.06
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
50
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
2.03-2.045
0.1
0
-0.1
-0.3

2.015-2.03
Fig. 6.23. Results for SumaveN
This diagram is found to be the most meaningful and shows that within the examined type of
hanger arrangements a wide range of geometries exists, that cause similar minimal forces and
force changes in the hangers. The minimum is found at a start angle of 50 and a constant angle
change of -0.3.
The results for this hanger arrangement are:
aveN: 937 kN
aveN: 682 kN
Optimal arrangement considering the hanger forces only
The decline at the ends of the concave minimum area in Figure 6.23 encouraged further
examination of hanger arrangements that are not similar to the common ones. It was figured out
that the best hanger arrangement considering minimal hanger forces and low change in hanger
forces due to live load is the one shown in Figure 6.24.
The results for this arrangement are: aveN: 577 kN
aveN: 220 kN
44
Fig. 6.24. Spoked wheel, span 100 m, rise of the arch 17 m, number of hangers: 44
The hangers are placed along the radii of the arch circle and spaced equidistantly along the
arch. By chance the authors became aware that PHILIPPE VAN BOGAERT, Ghent University, had
already built a tramway bridge and a high-speed railway bridge with this type of hanger
arrangement but with fewer hangers. Furthermore, in January 2002 the Spanish architect and
bridge designer SANTIAGO CALATRAVA was contracted by the City Council of Dallas, Texas for the
Trinity River Corridor Project. Among several bridges that are going to be built there are two arch
bridges planned with radial hangers [40]. There was another arch bridge built with radial hangers in
Hannover, Germany [35]. It spans 58 meters and uses 7 hangers per arch plane.
Regrettably, the structural behaviour of such an arch bridge is similar to that of arch bridges
with vertical hangers. These hangers lead to large bending moments about the horizontal axis in
the arch and the tie. Thus, the bending moments in the arch and the tie will also be involved as
attributes in the improvement process continued in Section 6.6.
45
6.6. Advanced model to describe the hanger arrangement
In this Section the hanger arrangement believed to be best is presented.
6.6.1 The idea

Bending in the arch about the horizontal axis is minimised if the line of thrust deviates very little
from the centreline of the arch TVEIT [43], page 2190. This can be achieved if uniformly distributed
loads act ideally in the radial direction on the arch, as it is known from the structural behaviour of
circular arches. The authors suggest imagining the resulting force of the hanger forces directed
along the connecting line of their intersections (Figure 6.25). This would be true if all hangers had
the same axial force. However, the axial forces are not the same, as they change due to moving
live loads. Nevertheless, as a good approximation the authors adhere to their suggestion.
Then it is self-evident that the resulting forces should be aimed radially towards the arch to
decrease its bending moments. This also leads to small hanger forces (see Section 6.6.2). One
example of the suggested hanger arrangement is shown in Figure 6.25.
'Resutling force'
Fig. 6.25. Schematic illustration of the suggested hanger arrangement
Many publications about optimisation of structures indicate that the claim is true. The authors would
like to give some examples.
The Australian mathematician JOHN HENRY MICHELL introduced a method for the optimisation of
structures. He proposed to align structural members with the principal stress trajectories. He
considered the resistance of the members to tension and compression as equal. This optimisation
is only applicable to one load case. Despite these restrictions, the Michell structures are still used
to evaluate optimisation software (SCHWARZ [31], page 2). Two of them are shown in Figure 6.26.
Similarities to the geometry proposed in Figure 6.25 cannot be denied.
Considered part
Fig. 6.26. Two Michell structures, found in BECKER [5], page 7
46
An optimal network of cables for bearing loads underneath a

parabolic arch is proposed in MCCULLOUGH [21]. It is shown in (c)
of Figure 6.27. It can be seen that the cables connected to the arch
are oriented vertically. Almost all intersections of the cables lie on
the centrelines of these vertical cables. For a parabolic arch the
thrust line is along the centreline of the arch if the loads are ideally
vertical, as achieved in (c) of Figure 6.27. If transferred to a circular
arch assuming that the line of thrust still coincides with the
centreline of the arch, a network would be obtained that is similar
to the one in Figure 6.25.
Fig. 6.27. An optimal load bearing

network of cables underneath a
parabolic arch, McCullough [21]
A network arch is considered to have a structural behaviour similar to a truss. In turn, optimised
trusses can be examined to gain knowledge about network arches. In the demonstrations of the
optimisation software FrameGA [28] several trusses with a circular upper chord can be found. Two
of them are shown in Figure 6.28
Fig. 6.28. Two optimised trusses with circular upper chord, found on the Internet [28]
For the optimisation a maximal cross-sectional area of the truss members was given. Both
trusses are optima for the loads and constraints indicated in the picture. It can be seen that the
diagonal truss members tend to connect the lower and upper chord with radial resulting forces.
The truss bridge in Figure 6.29 was found with

graphical methods (ZALEWSKI ET AL. [51]). The loads
B-G are equal. The axial force in the arch is constant
along the span. The tension forces of the interior
members are found to differ only slightly, as do the
forces within the lower chord. If three fixed member
sizes are standardized, one each for the top chord, the
interior members, and lower chord, all the material in
Fig. 6.29. Efficient truss, ZALEWSKI ET AL. [51]
the truss is working at or near capacity. Thus, this is a
very efficient form for a truss. Again, this confirms the suggestion to improve the network arch by
applying the loads to the arch as near as possible in a radial direction.
47
An example of an already built bridge where diagonal members are radial to the circular arch is
shown in Figure 6.30.
Fig. 6.30. Cascade bridge of the Erie Railway, USA around 1845, PETRASCHKA [25]
With all these examples of optimal or near optimal structures and the similarities with the
authors proposal, it should be worth applying the idea to the hanger arrangement of network
arches and examining the changes, even though this idea has never been mentioned before in the
sources about network arches known to the authors. In Section 6.6.2 the authors present their
ideas, why their proposal should lead to improved results.
6.6.2 Derivation of the proposed hanger arrangement
With the help of the following figures it will be pointed out why the introduced hanger
arrangement causes small bending moments in the arch and small hanger forces. For the
derivation symmetrical loading and a simply supported arch are assumed.
For circular arches a uniformly distributed load that is
directed along the radii of the arch circle causes no
bending moments in the arch.
Figure 6.31.
The hangers transfer forces at their nodal points to the

arch. Therefore, the uniformly distributed load is
substituted by single loads. They have to be equidistant
and their direction has to stay radial. With an infinite
number of single loads the line of thrust follows the centre
line of the arch.
Figure 6.32.
In arch bridges with the lane under the arch the loads
do not act on top of the arch but underneath. Vertical
loads can be transferred to the arch in a radial direction by
using radial hangers, Figure 6.33. The hanger forces are
all of the same size because of the constant curvature and
the equidistant hanger nodes. This is proven in
Figure 6.34.
Figure 6.33.
48
The same structure as in

I1 I2 For equidistant nodes along the
C2 arch and force V acting in the Figure 6.33 with fewer hangers is
C1
centre of the arch circle: S1=S2=S3 subjected to a vertical load in the centre
S2
S3 C2 of the arch circle. The force transfer
S1 S3
Force triangles: I2 between the upper hanger nodes is
C1-C2-V
C1-I1-S1
S2 V straight. The hanger forces are equal. It
C2-I2-S3 I1 is easy to imagine that this is true for a
I1-I2-S2 S1 C1
V
S1-S2-S3-V larger number of hangers as well.
Figure 6.34.
This hanger arrangement represents the same loading as the single loads obtained by
discretisation of the uniformly distributed load into the hanger nodes (Figure 6.32). Thus, with such
arrangement of hangers the bending moments of the arch are smallest. With an infinite number of
hangers no bending moments would occur at all.
Considering small deformations of such a structure it
can be seen that each point of the arch moves almost in a
radial direction (Figure 6.35). If the arch receives no
bending moments, other deflections are impossible.
Another example where deflections are only possible
in one direction is shown in Figure 6.36. The force
required to cause deflection is minimal if it acts
horizontally, hence along the direction of motion not
Figure 6.35. restricted.
It is obvious then, that smallest hanger forces are
achieved if the hangers are directed along the possible
F2
deflections of the arch circle. Otherwise their force would
F1 be increased due to the diversion between the centreline
of the hanger and the direction of deflection. Maybe a
F1<F2
Figure 6.36. slight outwards deviation of the hangers direction is
favourable.
The next step towards an arch bridge is shifting the
point of action of the vertical force upwards. The stiff
horizontal beam in Figure 6.37 distributes the single load
in Figure 6.35 evenly to all hangers. The structural
behaviour of the arch is not changed.
Figure 6.37.
As stated at the end of Section 6.5.2, the structural
behaviour of an arch bridge with radial hangers leads to
large bending moments in the chords with partial loading.
To achieve the benefits of a network arch, in the next step
the radial hangers are split into pairs, symmetrically to the
radii. This increases the hanger forces, but this increase
will be the smallest possible as long as their resulting
Figure 6.38.
forces stay along the radii.
In the following step the hanger crossings are placed
below the arch, since two hangers in one node at the arch
are not desired. Until now, bending moments in the arch
and hanger forces seem the smallest possible for inclined
hangers in a circular arch.
Figure 6.39.
49
Further towards a network arch bridge it is necessary

to introduce changes in the structural system. The
constraint conditions of the arch are changed due to a
clamping at the lower chord. With this change a radial
loading will cause bending moments in the ends of the
arch. However, for the main part of the span the structural
behaviour will almost be the same as before. A task for
future research would be to investigate the direction of
Figure 6.40. loading which causes a line of thrust still following the
centreline of the arch. Then the hangers have to be
aligned to this direction.
The final design of the network arch is obtained by
increasing the number of hangers following the introduced
geometry (Figure 6.41). Such hanger arrangement
transfers the vertical loads as far as possible in a radial
direction, which leads to small bending moments in the
arch. Furthermore, the hangers are almost aligned to the
Figure 6.41.
deflections of the arch. This results in small hanger forces.
On the basis of this reasoning it has been indicated that this hanger arrangement gives
smallest bending moments in the arch and small hanger forces for uniform and symmetrical
loading. The load characteristic of railway bridges leads to many different partial loadings. It is hard
to say which load case should decide the hanger arrangement.
The load direction at the ends of the arch causing a line of thrust along the arch has to be
calculated for every single bridge project again, but applying the radial direction is likely to always
give good results for the main part of the span.
The variable in this kind of hanger arrangement is the cross angle between the hangers and
the radii. If this angle is about 1 all hangers are radial to the arch circle; for increasing cross angles
the structure becomes a network arch. Thus, to obtain improved results for different load cases,
this cross angle can be used to vary the arrangement (see Section 6.6.3).
PER TVEIT suggested comparing the authors kind of hanger arrangement with one of the
network arches designed by him. It was decided on the kviksound network arch because many
influence lines have been calculated for it, JAY [18]. The authors applied a radial oriented hanger
web to the kviksound bridge; all other geometry, cross-sections and materials were retained
unchanged. On an average the values of the influence lines of the proposed hanger arrangement
were:
98 % (axial force in arch)
79 % (hanger force)
81 % (bending moments in the arch)
of the values of the kviksound network arch. Additionally internal forces and deflections were
compared for live load on the whole and on 54% of the span. Similar improvements were found
there as well.
The comparison can be found in Annex F, figures F.1 to F.10.
50
6.6.3 Variation of the lower hanger nodes using the advanced model
According to the idea mentioned in 6.6.1 another description for the geometry was used and
analysed including the bending moments in the arch about the horizontal axis as a further attribute.
As before equidistantly spaced upper hanger nodes were assumed. The hanger intersections lie on
radii of the arch circle. The only variable is the angle with which the hangers cross each other. Due
to the properties of the new hanger arrangement, corresponding hanger crossings lie on circles
concentric to the arch circle. All hanger crossings on one circle have the same cross angle.
For the investigations the variable parameter will be defined as the angle between the hanger
and the radius at the first hanger crossing below the arch (see Figure 6.42).
Fig. 6.42. The hangers cross symmetrically the radii with the same angle, the cross
angle defined as being the variable is marked grey
To visualize the connection between the cross angle and the hanger arrangement, some
examples are shown in Figure 6.43.
Fig. 6.43. Different hanger arrangement with cross angle of 0.835, 25 and 41
The introduced type of hanger arrangement gives reasonable results with cross angles
between 0 and 50 for the bridge calculated in this work (span 100 m, rise of arch 17 m, number of
hangers 44). 80 different bridges were calculated with angles varying from 0.835 to 50 and a
shorter graduation between 38 and 45, where the best result was believed to be found.
51
[kN] Number of hangers: 44, span 100m, rise of arch 17m
1850
1650
1450
1250
1050
850
maxN
650
aveN
maxN
450
aveN
250
0
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
cross angle []
Fig. 6.44. Results for the hanger forces
As it can be seen in Figure 6.44 the global minimum of the hanger forces occurs at a cross
angle of 0.8345. The appendant geometry complies with the spoked wheel shown in Figure 6.24,
which also gave best results for the type of arrangement of Section 6.5.2. But there is another local
minimum in the range from 30 to 45. To obtain the value of SumaveN in this range, the data
series for the average values are scaled with their minimums to 1 and then added with an
importance of 1:1, as explained in Section 6.4. The local minimum of SumaveN can be found at a
cross angle of 35.
The results are: aveN: 925 kN
aveN: 529 kN
To extend the criteria of the optimisation the bending moments in the arch will be examined now.
52
[kNm]
[kNm]
Number of hangers: 44, span 100m, rise of arch 17
5500.00 900.00
5000.00 800.00
4500.00 700.00
4000.00 600.00
m axM
3500.00 500.00
3000.00 400.00
m axM
2500.00 300.00
ave M
2000.00 ave M 200.00
1500.00 100.00
1000.00 0.00
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
cross angle []
Fig. 6.45. Results for bending moments in the arch about the horizontal axis
As can be seen in Figure 6.45 the bending moments in the arch for small cross angles, which
means hanger arrangements similar to vertical hangers are about 4 to 10 times larger than for
hanger arrangements in accordance with the theory of network arch bridges (see Section 4). The
stiffening effect of inclined crossing hangers deploys its impact at cross angles over approximately
25. Best results only considering bending moments are achieved with a cross angle of about 48:
maxM: 428 kNm
maxM: 280 kNm
In order to draw conclusions about an improvement satisfying both goals minimum hanger
forces and minimum bending moments, the values of Summax, Sumvariation and Sumall were
calculated out of the data series as described in Section 6.4. These three data series are shown in
Figure 6.46.
53
Number of hangers: 44, span 100m, rise of arch 17m

10
Sumvariation
Summax
9 Sumall
evaluation criterion, according to Section 6.4
2
0.83 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
cross angle []
Fig. 6.46. Sums of scaled values, result of improvement process
5.50
As can be seen in Figure 6.46 best 5.48
5.46
5.45
5.44
5.42
5.42
5.43
5.41
5.40
5.40
5.40
5.39
5.40
results considering small maximal forces 5.5

are found at a cross angle of 35.
5
In Figure 6.47 the results of the
evaluation criterion, see Section 6.4
refined investigation are shown. The

4.5 Sum variation
best results for minimal stress ranges
Sum m ax
are found at a cross angle of 41. The
4 Sum all
combination with equal weighting for
both leads to an improved hanger
3.5
arrangement at a cross angle of 41, as
well. 3
2.5
cros s angle []
2
38.25
38.75
39.25
39.75
40.25
40.75
41.25
41.75
42.25
42.75
43.25
43.75
44.25
44.75
45.25
Fig. 6.47. Results of the refined analysis
54
For the cross angle 41 the results are:

aveN: 986 kN
aveN: 531 kN
maxM: 489 kNm
maxM: 310 kNm
In Figure 6.48 the best results of the three different hanger arrangements are compared.
Hange r arrange m e nt from Se ction:
6.5.1 "e llips e "
6.5.2 "s lope " 6.6.3 "radial"
"s pok e d ove rall m inim al hange r
w he e l" be s t force
A verage maximum hanger f orce 885.3 kN 937.3 kN 577 kN 986.5 kN 896.5 kN
A verage variation of hanger f orce 633.05 kN 682.4 kN 220 kN 531.7 kN 576.8 kN
Maximal bending moment - - 4900 kNm 489.5 kNm 610 kNm
Max variation of bending moments - - 4850 kNm 310.8 kNm 450 kNm
Fig. 6.48. Comparison of the results
As can be seen in Figure 6.48, the biggest advantage of the new type of hanger arrangement
is found in a lower stress range. Due to the weighting and the inclusion of the bending moments as
criteria, the maximum hanger forces are higher for overall best than for ellipse and slope.
Therefore, the values for the hanger arrangement of the radial type with the minimal hanger
forces are also shown. It is obvious that the introduced hanger arrangement is a noteworthy
improvement.
Hanger relaxation
Relaxing hangers is an important topic for network arch bridges. For the kind of hanger
arrangement dealt with in Section 6.6, there are no relaxing hangers reported at cross angles
larger than 39. The spoked wheel does not lead to hanger relaxation either. Much hanger
relaxation occurs at cross angles from 1 to 25, which additionally enlarges the bending moments
in the arch due to the multiplication of the distance between points of supports and the changing of
the structural behaviour. The chords in the area of relaxed hangers are only connected to each
other by one set of hangers in tension. This part of a bridge functions similarly to a tied arch with
one set of hangers. The equilibrium of the bridge in this area is dependent on shear and bending in
the chords, and this can lead to large bending moments (TVEIT [43], page 2191).
It should be added here, that the relaxed hangers for cross angles from 35 to 39 occurred at
the ends of the span. In Section 6.7.5 it was found that this can be prevented by small changes in
their geometry. After accurately recalculating the bridge, the object of this work, there were no
relaxed hangers at all. The new type of hanger arrangement was also applied to the network arch
calculated in RCK [27], using a cross angle of 35. His calculations did not show relaxed hangers,
either.
Erection phases constitute an exception; see Section 8 and Annex E.
55
6.7 Investigation of other parameters
For all investigations in the following sections the type hanger arrangement found to be best is
used.
6.7.1 Influence of the number of hangers and the span
To examine the influence of the number of hangers and the possibilities to transfer the
cognitions found with the 100 m span bridge to bridges with different spans, another 80 bridges
were calculated varying the span and the number of hangers.
The spans 75 m, 100 m, 150 m and 200 m with the scaled arch rise of 12.75 m, 17 m, 22.5 m
and 34 m were chosen. This keeps the ratio of the span to the rise of the arch at 0.17. The number
of hangers was also scaled and varied around the achieved value. Additionally different cross
angles were applied to see if the results are near the optimum for each span. They were 38, 39,
40, 41 and 42.
Fig. 6.49. Bridges with spans of 75 m, 100 m, 150 m and 200 m
In reality it would not be so easy to increase the span just by scaling the geometry. Different
cross-sections of the structural members would have to be chosen. The problem is a changed
stiffness. The increased thickness of the tie hardly influences its stiffness. But the necessary
increase of the arch cross-section quickly leads to a multiplied moment of inertia. In later tests the
authors found that this significantly influences the choice of the optimal cross angle. Regrettably
this was not considered in the investigations. Indicating a tendency, the results of Section 6.7.1 are
still considered as meaningful.
56
Span = 75m 100m 150m 200m

number of hangers
30 32 34 36 40 42 46 48 58 62 66 70 76 82 88 94
7
2.8
Summax and Sumvariation, see Section 6.4

6.5 2.6
evaluation criterion for Sumall , see Section 6.4
2.4
6
2.2
2
5.5 Sum all
Sum m ax
1.8
Sum variation
5
1.6
evaluation criterion for

1.4
4.5
1.2
4 1
cross angle 38, 39, 40, 41, 42 for e ach numbe r of hange rs
Fig. 6.50. Diagram with scaled and added values for determination of improved results
For larger spans with scaled arch rise and scaled number of hangers, best results are achieved
with a similar cross angle as with a span of 100 meters. But a slight tendency can be seen towards
smaller angles, which means steeper hangers. This is not right for a smaller span. Obviously an
improved hanger arrangement concerning low hanger forces and bending moments in the arch is
achieved in this case with larger cross angles. That means less steep hangers.
It can also be seen that fewer hangers demand a larger cross angle to obtain improved values,
which means less steep hangers.
Figure 6.50 shows that an increased number of hangers reduces the forces in the hangers as
well as the bending moments in the arch. This behaviour does not seem to have a global extreme.
To conclude which number of hangers should be used, another data series was calculated using
the 100 m span bridge.
The results were again scaled and added as described in Section 6.4 to find best results out of
a combination of the hanger forces and bending moments in the arch.
57
span=100m, f=17m, cross angle 41

18
17.0
16 Sumvariation
Summax
13.5
14 Sumall
10.9
12 10.8
evaluation criterion
10 9.0
8.4
7.7
8 7.3
6.6
6.5
5.6
5.6
6 5.2
4.8 4.6
4.5
6.1
4.3 4.1
4.1
3.6 4.0
5.0
4
3.1
2.8
2.5
4.2
2.3
2.1
2.1
2.0
3.4
3.3
3.2
3.0
2
2.5
2.4
2.3
2.2
2.1
2.1
2.0
0
20 24 28 32 36 40 44 48 52 56 60 64 68 72
number of hangers
Fig. 6.51. Variation of the number of hangers, results scaled with their minimum to 1 and then added
As can be seen in Figure 6.51 the forces and bending moments become smaller with an
increasing number of hangers. But the decrease is not constant, and in this case it does not
significantly progress with more than 48 hangers. Maybe, considering aesthetic and economic
matters, more than 48 hangers should not be used in a bridge with a 100 meter span. 48 hangers
correspond to a distance of approximately 2.15 m between the hanger nodes along the arch and
an average support distance of about 2 meters for the lower chord.
58
6.7.2 Influence of the rise of the arch
This section examines the effect of different arch rises. The arch rise of already built arch
bridges lies mainly between 14% and 17% of the span, NAKAI [22], page 129. This is primarily due
to the appearance of the bridge, as rises above these values look rather ungraceful. The hanger
arrangement of Section 6.6 and the 100 m span bridge with 44 hangers were used for calculations.
The cross angle was 41. The results are presented in figures 6.52 and 6.53.
600
m axM
550
ave M
500 m axM
ave M
450
400
M [kNm]
350
300
250
200
150
100
14
14.2
14.4
14.6
14.8
15
15.2
15.4
15.6
15.8
16
16.2
16.4
16.6
16.8
17
17.2
17.4
17.6
17.8
18
arch rise f [%]
Fig. 6.52. Results for bending moments in the arch about the horizontal axis for different arch rises
1300
1200
1100
m axN
ave N
1000
m axN
N [kN]
ave N
900
800
700
600
500
14
14.2
14.4
14.6
14.8
15
15.2
15.4
15.6
15.8
16
16.2
16.4
16.6
16.8
17
17.2
17.4
17.6
17.8
18
arch rise f [%]

Fig. 6.53. Results for hanger forces for different arch rises
The larger the arch rise, the smaller the bending moments as well as the maximum axial
hanger force. Therefore, larger rises should always be considered and balanced against aesthetic
reasons.
59
6.7.3 Influence of the ratio between live load and dead load
Considering this parameter is necessary because the ratio between live loads and permanent
loads changes from bridge to bridge. Even within the calculation of one bridge the ratio varies due
to different load combination factors and partial safety factors for different design checks.
The self-weight of the arch and the hangers were not considered. Then the load used for the
other calculations in Section 6 was 223.1 kN/m for permanent loads. The single loads of load
model 71 are in conformity with 156.25 kN/m and the uniformly distributed part is 80 kN/m. For the
permanent loads the partial safety factor of 1.35 and for the live load of 1.5 were used. There was
one LM 71 applied on each of the two tracks. This leads to the ratios:
LL: 1.52156.25 kN/m = 468.8 kN/m and
DL: 1.35223.1 kN/m = 301.2 kN/m ratio single live load/dead load = 1.55
LL: 1.5280 kN/m = 240 kN/m and
DL: 1.35223.1 kN/m = 301.2 kN/m ratio uni. distr. live load/dead load = 0.8
Boundary conditions
For load model 71 the ratio between the single live load and the uniformly distributed load is
fixed at 1.95. So it is only necessary to vary one ratio. It was decided on the ratio between the
single live load and dead load and varied from 2 to 0.4. A bridge with a span of 100 m, arch rise of
17 m and number of hangers 44 was used for the calculations. The arrangement geometry found in
Section 6.6 was applied with a cross angle of 41.
Additionally loads common for road bridges were applied.
The results were scaled and added as described in 6.4 to find the best results out of a
combination of the hanger forces and the bending moments in the arch.
60
Span = 100 m, f = 17 m, number of hangers 44

num ber o f relaxed hangers
3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0 0 0 0 0 0
Sumall
4.12 2.12
Summax and Sumvariation , see Section 6.4

ratio live ratio live ratio live ratio live ratio live
load/dead load load/dead load load/dead load load/dead load load/dead load
2 1.6 1.2 0.8 0.4

4.1
2.1
4.08
2.08
Sum all
Sum m ax
4.06 Sum variation
2.06
4.04

2.04
4.02
4 2.02
39 40 41 42 43 39 40 41 42 43 39 40 41 42 43 39 40 41 42 43 39 40 41 42 43
cross angle []
Fig. 6.54. Influence of the ratio between live load and dead load on the hanger arrangement
As can be seen in the diagram the ratio of live load to dead load hardly influences the hanger
arrangement considering fatigue checks. A slight tendency to larger angles that means less steep
hangers for higher live loads can be seen.
In contrast, for higher live load the inclination of the hangers must be steeper considering low
maximum forces. It can also be seen that considering only this attribute would lead to significantly
different hanger arrangements for different ratios.
In the combination best results are to be found at about the same cross angle (41-42) for all
load ratios. Higher ratios need a slightly larger cross angle than lower ratios.
The enlargement of the number of relaxing hangers with likewise increasing load ratio shows
an increase of the compression forces in the hangers caused by the live load. The only two
relaxed hangers out of 44 are found at the end of the bridge which constitutes a disturbance range
due to the clamping of the arch. In examinations presented in Section 6.7.5 it was found how the
geometry of the first several hangers can be adapted to improve the results. After that no hanger
relaxations occurred, except in construction phases (see Section 8 and Annex E).
61
Special examination: Road bridge loads

To imitate the actions occurring on road bridges, two lanes were applied to the bridge.
According to ENV 1991-3:1995 Section 4.3.2 the first lane was loaded by four wheel loads of
150 kN each and a uniformly distributed load of 9 kN/m. The second lane was loaded with another
four wheel loads of 100 kN each and a uniformly distributed load of 2.5 kN/m. In addition the loads
from rails, sleepers and ballast were removed. Live loads were multiplied by a partial safety factor
of 1.5 and self weight by 1.35. The obtained results are presented, scaled and added as described
in Section 6.4.
span=100m, f=17m, number of hangers 44

numb er of relaxed hang ers
4 4 4 2 2 2 2 2 2 2 2 2 2 2
4.4
Sum all
2.45
4.3
Sum m ax
Summax and Sumvariation

Sum variation
4.2
2.35
Sumall
4.1
4 2.25
3.9

2.15
3.8
3.7
2.05
3.6
3.5 1.95
33 34 35 36 37 38 39 40 41 42 43 44 45 46
cross angle []
Fig. 6.55. Results for the calculation with loads similar to road bridge loads
For loads typical for road bridges, the optimal cross angle determined for railway bridges is not
appropriate. Obviously the ratio between live load and dead load is much smaller for road bridges
than for railway bridges. Therefore best results concerning small maximal forces are found at lower
cross angles, which means steeper hangers. However, the best result considering a small stress
range for a favourable fatigue check is found again at a cross angle of about 41 to 42.
The combination of both demands leads to a cross angle of 36 and thus, to steeper hangers.
The reduced self-weight and the necessary steeper hangers cause more relaxing hangers than for
railway bridges. But as before, they are only in the disturbance range of the arch clamping. In
Section 6.7.5 it is explained how these mavericks can be adapted for a special bridge project.
62
6.7.4 Influence of the ratio between the radii of ends and the middle of the
arch
The hanger arrangement from Section 6.6 with a cross angle of 41 was used for this
examination. The span was 100 meters and the number of hangers 44.
It is desired to examine the effect of a slightly smaller curvature at the ends of the arch, as this
reduces the length of the columns of the wind portal frame and the transverse bending moments in
it. Due to fabrication reasons, only circular curvatures are used, so that R will be the ratio
between the smaller radius R1 of the end section and the larger radius R2 of the mid section. The
transition point depends on the segment lengths in which the arch is delivered. A smaller curvature
radius is restricted to the first segment near the arch end. In this calculation the height of the
transition point measures 8.8m, which means a length of the first arch segment of about 17.5 m.
m a xN
1800 1800
aveN
m a xN
1600 a v e N 1600
m a xM
1400 aveM 1400
m a xM
a v e M
1200 1200
N [kN]
M [kNm]
1000 1000
800 800
600 600
400 400
200 200
0 0
1
0.96
0.92
0.88
0.84
0.8
0.76
0.72
0.68
0.64
0.6
0.56
0.52
R
Fig. 6.56. Results from the calculation
Figure 6.56 shows that reducing R to 0.76 hardly affects the hanger forces nor the bending
moments in the arch. The axial force in the arch near the bearings decreases due to the steeper
angle of the arch. A reduction of R beyond 0.76 increases the bending moments in the arch about
the horizontal axis. This is caused by the diversion of the axial force in the arch at the transition
point.
As a result it can be said that reducing the radius at the ends of the arch until a ratio of the radii
of about 0.8 is positive for the wind portal, but has no negative effects on hanger forces and
bending moments in the arch.
63
6.7.5 Further considerations regarding the hanger arrangement
Hanger node distances along the lower chord

The hanger node distances along the lower chord, which are used in this work, result from
mathematical or geometrical expressions. Therefore, some hanger nodes appear to be so close
together that they have virtually the same coordinates. They should be merged, due to the fact that
the size of the hanger connection detail requires minimum distances.
If a temporary lower chord is used for erection the distances between the hanger nodes and
the temporary transverse beams should not be too large, because this increases bending in the
temporary longitudinal beams. As far as possible the nodes should be shifted to gain benefits for
this point.
Due to the greater thickness of the end cross girder, the temporary longitudinal beam has a
kink at about 5 meters from the bearings. It would save cost if one hanger can be connected to this
kink to take the vertical forces caused by the diversion of the axial tension forces.
Disturbance area
The hanger arrangement suggested in this work leads to larger maximum forces in the hangers
nearest to the ends of the arch, compared to the rest. Naturally, the end of the arch constitutes a
disturbance range, which has to be treated separately. For each single project there are special
adaptations needed.
While calculating the bridge which is the object of this work and applying the improved hanger
arrangement to the bridge calculated in RCK [27], it was possible to improve the distribution of the
hanger forces in the disturbance range. The different methods are presented here.
The analysis of the deflection line in the bridge of this work showed a point of contraflexure and
a convex deflection directly above it at the bridge end (Figure 6.57). The hangers with upper nodes
in the convex area receive above average maximum forces. Therefore, hangers within and around
this zone should be spaced closer together, so that the higher forces to be transferred are
distributed into more hangers. That means shifting upper and lower hanger nodes towards the
convex deflection range (see Figures 6.63 and 6.65). As a result, the previous maverick hanger
forces are reduced to the average value.
line
ction
defle
on
c ti
e f le
xd
ve
c on
point of contraflexure
deflection line
Fig. 6.57. Locating the convex deflection zone by the deflection line
In RCKs bridge [27] the convex deflection zone was not visible due to the much stiffer arch
cross section. Improvements were gained by shifting the first few hangers slightly towards the
middle of the bridge. Still hanger forces in the disturbance range showed deviations. Then the
64
hangers with above average forces were made steeper and hangers with below average forces
vice versa. This gave almost satisfying results. The final elimination of maverick hanger forces was
achieved by changing the slope direction of the first hanger.
Network arch according to improved hanger arrangement:
span = 160 m, rise = 24 m,

number of hanger per arch plane = 60,
cross angle between radii and hangers
at first intersection below the arch = 35
Same network arch as above, but with

adapted arrangement at the ends,
(calculated in RCK [27])
CL
Fig. 6.58. Hanger arrangement used in RCK [27] for a network arch bridge with combined road and railway traffic and
four arch planes
As described, finding the arrangement of the first several hangers is an iterative process.
The necessity of this measurement leads to the conclusion that neglecting different node
distances along the arch may impede finding a satisfying hanger arrangement. This means for
future work, that geometrical or mathematical models have to be found considering the variability of
upper node locations, as well.
6.8 Discussion of results
On the structural behaviour

It is apparent that the ideal way of applying forces on a circular arch is if they act in a radial
direction. That is why in the spoked wheel in Figure 6.24 smallest hanger forces among all hanger
arrangements occur. The disadvantages of such a hanger arrangement are the large bending
moments in the arch and the lower chord. The authors assume that they are result of the low
stiffening effect of nearly vertical hangers and the small shear resistance of the area between the
upper and the lower chord (web). Consequently the sections of the arch and the lower chord have
to take additional bending moments and shear forces. An increase of the stiffness and
consequently a better use of the web is obtained by using the hangers to constrict deflections
between the arch and the bridge deck. This can be achieved by arranging pairs of hangers with
opposite slopes that form a triangle, which is known to be immobile (Figure 6.59 on next page).
65
g
nin
iffe
t st
efi cien n
to d ctio
ue efle
nd e dd
lect
io st rict
Def Con
Fig. 6.59. Stiffening effect of pairs of crossing hangers
For an improved network arch not only small bending moments are of interest, but inter alia
also small hanger forces. So, the idea was to combine the two effect principles. Thereto the forces
transferred by the hangers should act on the arch in a radial way. Keeping in mind the idea of
hangers acting as pairs the resulting force of each hanger pair should aim towards the centre point
of the arch circle. Assuming equal average hanger forces this is achieved best by crossing the
hangers on a radius, which intersects the arch in the middle between their upper nodes (see
Figures 6.42). Then, an unknown angle in the triangle formed by each hanger pair is best for
reducing the bending moments. With a certain angle all hangers will lie on the radii as in the
spoked wheel.
Out of these two principles, spoked wheel and triangular pair of hangers, there is a solution
found if the demands of both aspects are combined and weighted as desired from case to case.
The variable to determine is the cross angle of the hangers to the radius, defined in this work as
the cross angle between the radii and the hangers in the first hanger crossing below the arch.
On fatigue
In the type of hanger arrangement introduced all hanger crossings lie on the same radii. The
cross angle increases downwards. Independently of the cross angle at the first crossing below the
arch, the cross angles within the web vary with different hanger numbers and arch rises. It was
found that, only considering best results for fatigue, a range exists where the hangers cross the
radii with an angle of 45. Surprisingly this does not depend on the ratio between live load and
dead load.
This means that the cross angle at the first node below the arch has to be smaller for a larger
number of hangers, because with closer hangers the downward increase of the cross angle is
larger. This was proven in Section 6.7.1. The location of the 45-range is shown in Figure 6.60.
Fig. 6.60. Range where the hangers cross the radii at an angle of 45
It seems, that the range in Figure 6.60 divides the hanger web evenly in some way. So maybe
in order to obtain the smallest stress range, it would be best if all hangers cross each other
perpendicularly. Regrettably this is not possible if the resulting force should still aim towards the
centre of the arch circle.
66
It has to be mentioned that this range is the result of the combination of the demands of the
hanger force variation and the bending moment variation weighted 1:1. Only considering the
bending moments best results are obtained if all hangers cross the arch at an angle of about 45,
which means they distribute their force tangentially and radially in equal parts. This confirms the
theory of all hangers crossing each other at an angle of 90 to obtain best results only considering
fatigue checks.
On the number of hangers
Increasing the number of hangers reduces all internal forces. Since this reduction is not linearly
dependent on the number of hangers, it is not reasonable to increase the number of hangers
limitlessly. It has to be considered, that the saving of material due to the reduction of internal forces
front the higher costs due to more hangers, hanger connections and labour hours for erection. For
a span of 100 m it was found that the ratio between the number of hangers and the span
[in meters] should not exceed 0.48. For longer spans a lower ratio and for shorter spans a higher
one will give reasonable results.
On the span
For the bridges calculated in Section 6.6, different spans were applied with a scaled arch rise
and number of hangers. Therefore, the largest difference between them was a changed ratio
between the length of the bridge and the form of the live load. The load model 71 acts in longer
spans much more like a single load. The results say that the cross angle has to be smaller for
longer spans, which means steeper hangers for more concentrated loads. This conforms to
TVEIT [39], page 16.
On the rise of the arch
The variations of the arch rise showed the known results: With an enlarged arch rise the
distance between the upper and the lower chord increases, which lowers the axial force in the arch
and the hanger forces. Furthermore, the angle of the arch at the root point becomes steeper, which
results in a decreased axial force and clamping moment in the arch.
Since the distances between the upper hanger nodes and the curvature of the arch become
larger, local bending moments along the arch increase.
It shall be mentioned here that less steep hangers increase the axial force in the chords. This
can be explained by the bigger number of hanger cross sections in a vertical section through the
arch plane. The hangers contribute tension forces to the horizontal equilibrium in the vertical
section, which results in increased axial forces in the chords. Nevertheless, this enlargement is
small compared to the maximum axial arch force. With regard to the values of the internal forces,
the benefits of less steep hangers outweigh the disadvantages.
On the ratio of live load to dead load
For a smaller ratio of live load to dead load as it occurs with heavier bridge decks or road
instead of railway loads, the internal forces caused by the live load become proportionally smaller.
If the importance of small stress ranges and low maximal forces is kept up by 1:1, the best
combination results are shifted towards the best results for ultimate design checks, which means
low maximal internal forces. The hangers must be steeper to satisfy the demand for minimal forces
than for minimal variation of forces. That is why in this work it was found that the cross angle for a
smaller ratio of live load to dead load has to be smaller and vice versa.
67
6.9 Summary of Section 6
6.9.1 In words
1. The target of the improvement process was to find the best hanger arrangement in respect of
low maximum hanger forces and arch bending moments as well as low variations in both. Each
attribute was considered with the same weighting.
2. Considering only small hanger forces and small stress variations in the hangers, the spoked
wheel leads to best results.
3. The best hanger arrangements were achieved when hanger intersections lie on the radii of the
arch circle. The cross angle is variable and depends on the weighting of the optimisation
attributes and the characteristics of the loads and the geometry of the bridge.
4. A high arch rise is advantageous. It must be limited for aesthetical concerns.
5. A higher number of hangers leads to smaller internal forces. From a certain number of
hangers, the savings do not countervail the extra expenses.
6. For railway bridges, the load ratio hardly affects hanger slopes. Road bridges call for steeper
hangers.
7. Smaller curvatures of the arch near the ends are advantageous in respect of bending moments
in the wind portal. For a ratio of 0.8 between the arch radius at the ends and the middle best
results are expected.
8. The disturbance range near the ends of the arch leads to above average hanger forces. This
can be eliminated by shifting the upper and lower node of the first several hangers in an
iterative process.
9. Despite all statics, it is necessary to bear in mind boundary conditions, e.g. temporary lower
chord or hanger connection details, which demand special locations of the lower hanger nodes
especially at the ends of the span.
10. If small stress variations are more important than small hanger forces, hangers must be less
steep.
11. The smallest variation in arch bending moments has been found if all hangers cross the arch at
an angle of about 45. For the smallest variation in hanger forces, hangers have to be steeper.
For a combination of both, hangers cross a smaller circle than the arch circle at an angle of
45.
The scheme in Figure 6.61 has been composed to give assistance on how to find an
appropriate hanger arrangement according to the results obtained in Section 6. An example of use
is given in Section 6.9.3.
68
Preliminary suggestion for finding an appropriate hanger arrangement
for single and double track railway network arch bridges CL
n: number of hangers per arch plane

R2 / R1 = 0.8
R1
f
R2
Fig. 1. Geometry and definition of variables

100
n
90
1. Decide on the rise of the arch: f = 14..17% of span
80
2. Determine number of hangers n from Fig.2.
70 ge
3. Determine cross angle from Fig.3. n
ra
le
4. Determine dead load g in kN/m 60 ab
on
as
Number of railway tracks: t re
50
1.5 t 156.25 kN m
Calculate factor L: L= 40
1.35 g
34
75 100 125 150 s 175 [m] 200
For L < 0.8, subtract 1 from Fig. 2. Number of hangers
For L > 1.6, add 1 to
5. With R2 / R1 = 0.8 and equidistant upper hanger nodes acquire coordinates of
nodes by drawing:
Hanger intersections lie on radii going through CP (Fig.1).
: angle between radius and hanger at first intersections below the arch.
6. Run analysis of influence lines. Determine internal forces and deflection line
0.6 7. Adapt maverick hanger forces at the ends of the

n span. Examples given in Section 6.7.5.
s [m]
Check results.
0.55 8. Adapt lower hanger nodes to other boundary
conditions (e.g. temporary lower chord).
0.5
Span 75m
Note: This scheme is to be regarded as a summary of
the results so far. Especially for spans different to 100
0.45
meters, further investigation is needed.
100m
150m
0.4
200m
0.38
38 39 40 41 42 43 44 45 46
Cross angle []
Fig. 3. Cross angle

CP
BRUNN&SCHANACK
Fig 6.61. Section 6.9.2: Summary in a scheme

6.9 Summary of Section 6
6.9.1 In words
1. The target of the improvement process was to find the best hanger arrangement in respect of
low maximum hanger forces and arch bending moments as well as low variations in both. Each
attribute was considered with the same weighting.
2. Considering only small hanger forces and small stress variations in the hangers, the spoked
wheel leads to best results.
3. The best hanger arrangements were achieved when hanger intersections lie on the radii of the
arch circle. The cross angle is variable and depends on the weighting of the optimisation
attributes and the characteristics of the loads and the geometry of the bridge.
4. A high arch rise is advantageous. It must be limited for aesthetical concerns.
5. A higher number of hangers leads to smaller internal forces. From a certain number of
hangers, the savings do not countervail the extra expenses.
6. For railway bridges, the load ratio hardly affects hanger slopes. Road bridges call for steeper
hangers.
7. Smaller curvatures of the arch near the ends are advantageous in respect of bending moments
in the wind portal. For a ratio of 0.8 between the arch radius at the ends and the middle best
results are expected.
8. The disturbance range near the ends of the arch leads to above average hanger forces. This
can be eliminated by shifting the upper and lower node of the first several hangers in an
iterative process.
9. Despite all statics, it is necessary to bear in mind boundary conditions, e.g. temporary lower
chord or hanger connection details, which demand special locations of the lower hanger nodes
especially at the ends of the span.
10. If small stress variations are more important than small hanger forces, hangers must be less
steep.
11. The smallest variation in arch bending moments has been found if all hangers cross the arch at
an angle of about 45. For the smallest variation in hanger forces, hangers have to be steeper.
For a combination of both, hangers cross a smaller circle than the arch circle at an angle of
45.
The scheme in Figure 6.61 has been composed to give assistance on how to find an
appropriate hanger arrangement according to the results obtained in Section 6. An example of use
is given in Section 6.9.3.
68
6.9.3 Example of use
As an example, the scheme in Figure 6.61 will now be applied to the bridge which is the focus
of this work. It is a double track railway bridge with a span of 100 m. The result is the design used
for all sections apart from Section 6.
1. Decide on the rise of the arch:
Small internal forces were more important than better aesthetics, so the arch rise was chosen
to 0.17% of the span. f = 17 m
2. Number of hangers per arch:
From Figure 2 in the scheme (Figure 6.61). It was decided on the best number of hangers
n = 48
3. Cross angle:
From Figure 3 in the scheme (Figure 6.61) with h = 48/100[m] = 0.48 and s = 100 m
= 41
4. Ratio live load to dead load:
In a preliminary design a uniformly distributed load of 223.1 kN/m was determined as the self-
weight of the lower chord including rails, ballast and sleepers. The bridge is a double track
railway bridge. The factor to determine the load ratio is:
1.5 2 156.25kN / m
= 1.55 L=
1.35 223.1N / m
With factor L = 1.55, the cross angle does not have to be changed.
5. With the fixed ratio R2 and R1 of 0.8 and equidistant locations of the upper hanger nodes along
the arch, the coordinates of the lower hanger nodes were acquired by drawing (Figure 6.62).
21 22 23 24 25 26 27 28 29 30
17 18 19 20 31 32
15 16 33 34
35 36
13 14 37
12
11 38
10 39
9 40
8 41
7 42
6 43
5 44
4 45
3 46
2 47
1 48
Fig. 6.62. Hanger arrangement found by the geometric description, labelling of the hangers
6. This first approach was calculated with one load model 71 on each track and the partial safety
factors of G=1.35 and Q=1.5. The analysis of the influence lines gave maximal and minimal
hanger forces as well as bending moments in the arch as can be seen in Figures 6.63 and
6.64.
70
m ax N m in N N
157.9
151.2
141.4
128.2
153.0
143.8
176.3
1 0,0 0,0 0,0
171.1
171.1
160.8
479,5 182,3 297,2
122
2
117
82.9
.5
104.6
3 1277,4 821,6 455,8
.1
-62.9
13.1
94.
104.1
114
-8.0
4 893,5 406,2 487,3
109
-39.4
.4
.7
56
69.
5 1152,4 686,4 466,0
.7
1
66
89
-15.3
6 961,7 482,1 479,6
2.
.4
26
3
12
.3
7 1001,3 540,9 460,4
5.1
-53
-1 3
.2
8 878,5 466,4 412,1
7.1
-91
9 837,0 427,9 409,1

-29
.3
.3
848,4 451,7 396,7

-10
10
82
0.9
.4
11 830,1 415,6 414,5

12 892,2 453,6 438,6
-97.24
-134.8
-148.7
855,5 397,5 458,0
-102.1
13
-104.8
-123.2 -73.2
-126.1
-50.0
14 947,4 457,0 490,4
-89.2
-88.4
-78.9
-48.0
-86.0
-80.
15 915,4 378,6 536,8

-35.4
4
-157.17
16 962,9 445,6 517,3
13.
-10
17 950,1 348,3 601,8

-9.3
5.3
-24
-18
-11
18 936,3 396,2 540,1

-60.3
0.1
.8
-42.7
0. 4
9. 2
-93
19 947,7 327,4 620,3

-13
20
.2
37
.0
.6
-24
20 892,7 369,2 523,5

-5 6
1.
.2
7
.6
21 932,3 331,1 601,2

-97
22 896,3 367,2 529,1

.2
-11
45
23 924,6 343,5 581,1

4.1
.9
24 914,8 363,6 551,2

Fig. 6.63. Envelope of maximal and minimal bending moments in the arch Fig. 6.64. Hanger forces
7. Adapting maverick hanger forces

The maximal hanger forces of some hangers close to the end of the arch deviate from the rest.
Since the maximal hanger force in the mid section of the arch is about 965 kN, it is desired to
decrease the maverick hanger forces below this value. Thereto the deflection line of the arch
was consulted to find the disturbance range in the arch caused by the clamping (see Section
6.7.5). In this area the upper nodes of the first 7 hangers were shifted upwards along the arch,
so that they concentrate their impact in the disturbance range. After that the lower nodes of
these hangers were shifted horizontally for aesthetic reasons, as well. Figures 6.65 and 6.66
show the results.
71
162.0
151.8
155.1
136.7
149.6
144.2
165.6
m ax N m in N N
163.5
158.8
147.6
1 174,8 128,1 46,7
108
102
70.1
.4
2 912,4 420,7 491,7
85.8
.2
-54.2
17.0
68.
101.1
83.
3 666,8 390,1 276,7
-17.3
- 64
6
5
-51.
829,2 383,8 445,4
.6
4
96
54.
8
28 8
.3
2.
5
.9 5 779,9 396,3 383,6
23
-36.
-1651.4 .2
.9
- 5
.7
6 774,1 389,8 384,3

13
5
87
92.
.7
82 7 826,9 405,9 421,0

7
.8
8 790,1 423,3 366,8
9 813,1 406,9 406,2
23
10 857,8 458,0 399,8

4.
7
11 814,9 402,6 412,3

12 903,4 453,6 449,8
-81.7
-97.8
-65.3
-105.0
-78.0
13 835,3 380,8 454,5
-73.2
-79.8 -131.5
-56.0
-48.0
-59.0 14 943,6 446,5 497,1
-141.9
-33.
-123.2
-73.
15 903,5 370,5 533,0

-22.
-102
- 5 9 . -1
-120.9
7
16 949,4 432,8 516,6

-19
-93.1
0
-40
.4
-13
.3
.9
17 944,3 347,7 596,6

0.4
11.
14
-55 .2
.5
917,0 381,8 535,2

-58 3
18
-34
.5
-63 .2
7 .7
.
945,2 330,7 614,5

-947.2
19
.7
2
22
20 871,1 353,4 517,7

-10
.8
.0
21 931,9 336,4 595,5

22 871,9 346,9 525,0
23 924,2 349,1 575,1
80
.4
24 889,8 343,2 546,6
Fig. 6.65. Envelope of maximal and minimal bending moments in the arch, Fig. 6.66. Hanger forces
adapted hanger arrangement
Adjusting the hangers in the disturbance range is an empiric operation. Additionally, slight
changes have considerable influences on the internal forces and may influence hangers in the
mid section of the span. It is suggested to stopping the minimisation of the hanger forces when
the forces are lower than the maximum forces in the mid section.
8. Adapting lower hanger nodes to other boundary conditions
For construction it is best if lower hanger nodes lying close together are merged into one
hanger connection. Depending on the method of erection of the bridge there might be other
conditions where to locate the lower hanger nodes on the lower chord. In this particular case a
temporary lower chord will be used for erection. Therefore it is favourable not to have too large
unsupported distances at the lower chord. These demands were applied and the effects on the
internal forces shown in Figures 6.67 and 6.68
72
Section 7
Investigation of hanger connection details

regarding fatigue
7.1 General
Railway bridges are subjected to dynamic

loading, which makes the consideration of the CL
fatigue behaviour necessary. This is especially

important for hangers and hanger connections,
since they receive larger force variations than
other bridge members. Subjected also to
horizontal loading, hangers and their
connections are therefore significantly prone to
fatigue failure.
This section aims to investigate causes of
Fig. 7.1. Example of hanger connection
dynamic loading and their effects on network
arch bridges. A fatigue assessment is carried
out for various types of hanger connections,
resulting in a feasible design proposal, which is
then adopted for the network arch calculated in
this work. Two possibilities for the fatigue
assessment provided in the Eurocode were
applied:
1. Fatigue assessment based on nominal stress ranges (Section 7.4.1)
2. Fatigue assessment based on geometric stress ranges (Section 7.4.2)
The second method of assessment is necessary because the hanger connection details are
more complex than the test specimen with which the detail categories and fatigue strength curves,
such as in the Eurocode 3, were created. If the geometry and the loading differ significantly from
the listed detail categories, the nominal stress is not meaningful, and its application would lead to
wrong results. Therefore, local stress concentrations at geometric discontinuities were investigated.
This investigation is based on the specifics of network arch bridges, which makes the short
introduction of relevant characteristics necessary. In network arches hangers cross each other at
least twice. At their intersections, hangers are covered with plastic tubes and tied together by
means of elastic rubber bands. This couples the horizontal movement of the hangers and therefore
increases damping of the whole hanger web. Circular hanger sections are favourable due to their
slender appearance regardless of the view angle. Smooth circular steel hangers will be basis for
the design of the investigated hanger connection types.
74
Diploma Thesis - Brunn & Schanack Section 7: Fatigue investigation of hanger connections
7.2 Loading
7.2.1 Axial loads
Hangers are mainly subjected to axial loads, as they carry the bridge deck as well as the traffic
loads. The maximum hanger force in the ultimate limit states in this work was determined to be
Fmax = 1062 kN. For fatigue considerations, the maximum force variation for two load models 71
acting on the bridge was obtained with Fp.1+2 = 440 kN, whereas the respective number for one
load model was found to be Fp.1 = 295 kN. Effects due to the variation of axial loads are
investigated in Section 7.4.
The damage equivalence factor was calculated to (Annex D, Section D.3.3):
See also: ENV 1993-2: 1997 Section 9.5.3
1 = 0.6 EC Mix L = 100 m
2 = 1.04 Traffic per year: 30 106 t/track
3 = 1.0 Design life: 100 years
Fp.1 295kN
4 = 0.77 = = 0.67
Fp.1+ 2 440kN
= 0.6 1.04 1.0 0.77 = 0.48
7.2.2 Horizontal loads
In the recent past a considerable number of vertical hangers in arch bridges, for example in
Germany, had to be repaired because of serious fatigue damage. This damage was mainly caused
by rain-wind-induced vibrations, GNTHER ET AL. [15]. Thus, it is important to examine how the
hanger web in a network arch behaves under wind loading.
The gusset plate of the hanger connection detail is perpendicular to the plane of the arch. The
resistance against movement of the hanger within the arch plane is therefore small, and large
bending moments in the hanger connection do not occur. Furthermore, the hangers are tied
together at their intersections reducing the deflections within the arch plane. In a network arch
horizontal deflections of interest are therefore not the ones in the plane of the arch, but
perpendicularly to it.
Static wind loads certainly cause the hanger web to deflect. The amplitude depends on the
axial hanger force, since the virtual bending resistance of hangers increases with higher hanger
forces. The maximum static deflection caused by wind loading in ultimate limit state and the
absence of other live loads was determined to be 1.18 mm at a distance of 1 meter from the
hanger node (Annex C, Figure C.22).
Regarding fatigue, it is to be examined apart from static deflection, if the possibility of a
dynamic excitation of the hanger web exists. The analysis of the dynamic behaviour of rope-net-
structures has to be performed considering the whole net, because the deformed shape is a result
of a state of equilibrium of a rope-net, BRUGER [6]. Otherwise, considering each hanger alone, the
fastening of the hangers at their crossings would have to be ignored in order to be on the safe side
and the favourable influence on the dynamic behaviour of the hanger web would be lost.
For dynamic investigations the mode shapes are of interest. The first mode shapes are shown
in Annex C, Figure C.24. As mentioned above, oscillation amplitudes are larger for less loaded
hangers. It is likely that wind excited oscillations occur for long periods of time while no train is on
the bridge, so the mode shapes were calculated without additional live load.
75
The following paragraphs give an introduction to various types of wind excitations relevant for
slender structures, such as hangers of arch bridges. The authors will try to examine which types
might occur in network arches.
7.2.2.1 Gust excitation wind direction perpendicular to hanger web
This type of wind excitation is a separate excitation.

Gust excitation is caused by varying wind velocity
resulting from ellipsoid- globe- or boomerang-shaped
areas of various velocities in the wind stream, PEIL [23].
The first mode shape of the hanger web (Figure 7.2) leads
to the largest amplitudes and would therefore be decisive
for gust wind response.
DIN 4131 09.1991 - Antennentragwerke aus Stahl
suggests loading more than 10 m wide structures partly
with only 60% of the wind load. To obtain maximum
deflections, it is therefore necessary to apply full wind load Fig. 7.2. First mode shape of hanger web,
f = 0.97 Hz
to the centre range of the web and reduced wind load to
the outer areas.
Annex B of ENV 1991-2-4 provides the dynamic factor cd taking gusts into account as an
increase of the static wind pressure. Regrettably the given calculation methods are not appropriate
for the present structure. Besides, it is not stated if the forces calculated by these methods are
relevant for fatigue assessment. The Eurocode refers to specialist advice.
7.2.2.2 Vortex excitation wind along the hanger web
Wind flow around a slender structure leads to vortices inducing alternating lateral pressure.
This transverse excitation is a combination of separate excitation and self-excitation. Especially
slender structures with small stiffness and damping, such as cables, are prone to be affected.
This excitation can and probably will occur in the hanger webs of network arches, as well. But
each hanger will oscillate with a different frequency and amplitude, since they have different
lengths, slopes and axial forces. The advantage is that these hangers with different oscillations are
tied together. A dangerous excitation will only occur if the dynamic wind forces cause an oscillation
according to the mode shapes of the whole hanger web. Otherwise the hangers vibrations damp
each other. An excitation according to one of the mode shapes is considered as improbable.
In GNTHER [15], page 908, it is stated that vortex vibrations are not critical. That article refers
to single hangers, where this excitation is believed to happen more easily. Thus, it is not regarded
as critical in network arches.
7.2.2.3 Aeroelastic instabilities and interference effects
These types of effects are based on self-induced excitation of the structure and can be divided
into galloping and interference effects to crosswind oscillation (with one degree of freedom) and
flutter and divergence (with two degrees of freedom).
Galloping wind along the hanger web
Galloping occurs on flexible structures with non-circular shaped cross-sections. Cylindrical

hangers are therefore not affected, unless their cross section changes due to coating with ice.
76
This effect also arises when rain drops hit the hangers and generate a rivulet on the surface of
the hanger. The presence of flowing water on the hanger changes its cross section and
consequently the wind flow pattern. This induces a lift-force as a mechanism for vibration. The
additional danger of this type of excitation is that the rivulets change their position with the
oscillation, which leads to varying aerodynamic forces (VAN DE BURGH [49]). This phenomenon is
known as rain-wind induced vibrations.
In GNTHER [15], Figure 5, criteria are given whether single hangers are sensitive to rain-wind
induced oscillations or not. It is more probable for larger hanger diameters and smaller eigen-
frequencies. PER TVEIT confirms that for the whole hanger web it is improbable that rain-wind
effects or galloping cause fatigue relevant vibrations.
Interference effects due to crosswind oscillation
Interference effects occur at closely spaced slender structures and hangers are vulnerable to
such effects due to their in-line arrangement. ENV 1991-2-4: 1995, C.3.2 provides different
methods of assessment, depending on the distance between the structures, their diameters and if
they are coupled or not. The distances between hangers within a web vary constantly. It is doubtful
whether these calculation procedures provided by the Eurocode can be applied to network arches.
In case of doubts the European standard recommends specialist advice.
Flutter and divergence
Flutter and divergence are coupled bending-torsional vibrations (SEDLACEK [32]). Since
torsional vibrations cannot be induced in circular hangers by wind loading, these effects are not
applicable. Hangers with flat plate-like cross-sections are vulnerable to such dynamic effects. In
this work rectangular cross-sections for hangers are not used; thus, flutter and divergence cannot
happen.
7.2.3 Conclusion
The fastening of the hangers at their intersections, which couples their deflections
perpendicularly to the hanger web, is very advantageous in respect of resistance against dynamic
wind excitation. Looking at the mode shapes of the hanger network in Annex C, Figure C.24, we
can see that the largest oscillation amplitudes occur when the whole net moves into one direction.
The amplitudes of vibrations are much smaller if each hanger is excited separately (see Figure 7.3
for an example).
Excitation of the whole net is probable under gust
wind loading, but its occurrence with fatigue relevant
vibrations is very doubtful. Other dynamic wind effects
exciting a single hanger are not dangerous in terms of
fatigue, because a response according to one of the web
mode shapes is considered as very improbable.
For the determination of the aerodynamic behaviour
the Eurocode refers to specialist advice. A final th
Fig. 7.3. 40 mode shape of hanger web,
conclusion might only be found by wind-tunnel tests. f = 3.24 Hz
Referring to PER TVEIT, who confirms the authors assumption, hangers in network arches are
not subjected to wind excitation in a way that would be dangerous in terms of fatigue.
For the sake of interest and in case of a wrong assumption, stress distributions in hanger
connection details due to horizontal deflections are examined in Section 7.5.
77
7.3 Boundary conditions for hanger connection design
It is desirable to arrange each set of hangers in an individual plane. Provided that all hangers
are straight, the centre distance between both hanger sets is equal to the hanger diameter. To
accomplish this, the hanger connections need to be designed with an eccentricity of half the hanger
diameter to the arch profile. For network arch railway bridges with spans up to about 100 metres,
rolled profiles such as Universal Columns or American Wide Flange can be used for the arches.
Since the bridge calculated in this work is equipped with American Wide Flange profiles, only such
profiles are considered in the following investigation.
To connect hangers to the arch profile, intermediate members, such as gusset plates, are
required, which constitute the actual connection. Hence, it is necessary to connect the circular
hanger profile to a plate-shaped element. This will be treated separately from the connection of the
plate-shape element to the arch profile. The smooth circular hangers are suitable for butt joints to
the gusset plates that are equipped with a cut-out, in which the hanger is inserted and welded.
Slitted hangers, in which the gusset plate is inserted and welded by means of fillet welds, as it is
suggested in DS 804 [S.8] for rectangular hangers are not considered.
Five different types of connections were investigated, see Figure 7.4. From the results, another
hanger connection was derived (Figure 7.5).
Geometric boundaries for the hanger connection design taken from the bridge that is the object of
this work:
Arch profile: American Wide Flange W 360x410x634 and W 360x410x900. The
design of the hanger connections is based on W 360x410x634
representing the critical case with smaller dimensions.
Hanger: Smooth circular bar, diameter D = 60 mm
Material for arch profile, hanger and gusset plates: Steel grade S 460 ML
Type 1 Type 2 Type 3
Type 4 Type 5
Fig. 7.4. Investigated types of hanger connections Fig. 7.5. Suggested hanger connection
78
7.4 Fatigue caused by axial loads
7.4.1 Fatigue assessment based on nominal stress ranges
The dynamic loading of bridges is characterised by low loads and long lives, commonly
referred to as high-cycle fatigue with fatigue lives greater than 105 cycles. For high cycle fatigue,
the nominal stress range method using SN curves is best suited, ALMAR-NSS [3], pages 214-215.
These curves (also Whler curves) were found by experimental fatigue tests on specimens. The
fatigue strength in the form of constant stress ranges (S) is related to the number of cycles (N) until
rupture, SEDLACEK [33], page 4.
Eurocode 3 provides these SN curves in ENV 1993-1-1: 1992 Figures 9.6.1, 9.6.2 and 9.6.3,
which correspond to the detail categories listed in, for example, ENV 1993-1-1: 1992 Tables 9.8.
The reference stresses used in these curves are the nominal stresses. They are calculated with
respect to net cross sections excluding all stress concentration effects. The magnitudes of the
dynamic stresses are below the tensile elastic limit.
In the following, the hanger connections will be examined using the detail categories provided
in the Eurocode applying the nominal stress range method (ENV 1993-1-1: 1992 Section 9.5.2). It
should be mentioned that locations of interest at the hanger hanger gusset plate joints are named
(A, B ), whereas relevant locations at the hanger gusset plate arch profile joint are named
(I, II ).
Hanger connection type 1
This type consists of two hanger gusset plates, requiring only one
gusset plate at the arch profile and therefore providing a two-shear
connection. STEIMANN [37] found this type not to be appropriate due to
the low detail category 45 (ENV 1993-2: 1997 Table L.4/1) at the weld
connection between hanger and gusset plates (A). With a hanger force
variation of Fp.1+2 = 440 kN and a damage equivalence factor of B
= 0.48, the hanger diameter would have to measure at least 78 mm
to satisfy the fatigue check. Following the ultimate limit states, only
60 mm are required. Clearly, this particular point appears to be very
critical, but, however, it can be improved by introducing a sufficiently
large enough transition radius. A radius of 150 mm would make detail
category 90 applicable (ENV 1993-2: 1997 Table L.4/1), see
Figure 7.7.
A
Except for point (A), the clearance between the two gusset plates
is not accessible, so that at least one butt weld can only be executed
from one side. At point (B), this leads to detail category 36 (ENV 1993-
Fig. 7.6. Hanger connection
2: 1997 Table L.3/2). To be able to judge this detail category, the type 1
splitting of the axial force between transverse butt weld and
longitudinal butt welds needs to be determined. A splitting of the force
proportional to the weld lengths and a constant stress distribution were
assumed. With a total weld length of
2 10 mm (2 270 mm + 60 mm) = 12000 mm2 and a maximum
hanger force variation of Fp = 440 kN, the stress variation is
36.6 N/mm2, which is hardly greater than c = 36 N/mm2. However,
with = 0.48, the damage equivalent stress range becomes
E = 17.6 and the fatigue check is fulfilled.
79
This connection (Figure 7.7) improves the detail category at point

(A) from 45 to 90 by introducing a sufficiently large transition radius.
The radius here of 342.5 mm is even greater than the required 150 mm
in order to achieve a pleasing appearance and adopt the geometry of
hanger connection type 1.
As with type 1, the hanger end is bevelled to provide a flowing
transition to the gusset plate.
The fact that there are two gusset plates is still regarded as Fig. 7.7. Hanger connection
type 2
disadvantageous because at least one weld can only be executed from
one side. The application of a single gusset plate (Figure 7.8) improves
this weak spot, but to achieve a two shear connection, two gusset
plates are required at the arch profile. Here, the degradation of the
fatigue strength due to an execution of the welds from one side is
tolerable.
A further unfavourable point in connection with type 2 is the
transverse butt weld at the end of the hanger, which is not only due to
the relatively low detail category. The fabrication of this sharp 90
corner is rather inconvenient and prone to developing a crack skew
into the plate. If type 2 is used, a certain sequence should be followed,
in which the welds are executed. To keep residual stresses at a
relatively low level, it is necessary to start with the transverse butt weld
and proceed with the longitudinal welds. In any case, residual stresses
should be avoided, as fatigue strength is lower for members with high
residual stresses, because locations with residual stresses constitute
locations of stress peaks, PETERSEN [24], page 429.
In order to enhance this critical point, the hanger geometry
suggested by GNTHER ET AL. [15] was adopted (Figure 7.21). It uses a type 3
hole in the plate at the end of the bevelled hanger butt end. The hole
with a diameter equal to the dimension of the hanger is milled into the
gusset plate and the rest of the plate where the hanger will be inserted
can be cut out easily. Finally, the hanger is welded into the gusset
plate and the edges and transitions are ground. The hanger is also
equipped with a semi-circular cut-out. The diameter of this cut-out will
be smaller than the hanger diameter, so that the ends are not pointed.
The Eurocode does not provide a corresponding detail category with a
respective SN curve for that detail. Therefore, the fatigue assessment
based on nominal stress ranges fails here and calls for an assessment
based on geometric stress ranges (Section 7.5).
80
This type of connection is derived from the German code for

railway bridges DS 804 [S.8], which was used in STEIMANN [37]. The
first modification is the adjustment to a circular hanger cross section
instead of a rectangular one. The main part of the connection is
constituted by a widened cylindrical member, butt-welded onto the top
of the hanger end (C). Detail category 90 (ENV 1993-2: 1997 Table
A
L.3/1) is applicable.
The connection detail suggested in DS 804 [S.8] uses a slitted
rectangular widened hanger end in which the gusset plate is inserted
C
and connected by means of fillet welds. This gives detail category 80
(ENV 1993-2: 1997 Table L.5/3). Here, the second modification will be
a full penetration butt weld between the gusset plate, which comes in
two parts, and the widened hanger end. Although this leads only to
detail category 45 (ENV 1993-2: 1997 Table L.4/1), the case is not as
critical as with hanger connection 1, since the widened hanger end Fig. 7.9. Hanger connection
type 4
with its increased cross sectional area is accompanied with a lower
stress level.
The nature of connection type 4 does not imply the critical point (B)
of types 1 to 3.
This connection represents a further development of type 4 with

the application of a sufficiently large transition radius at (A), as
introduced with type 2. This makes the widened hanger end member
redundant and the gusset plates can be directly welded to the hanger.

type 5
Note
This fatigue assessment considers only the members constituting the actual connection. If the
gusset plate is welded to the arch, the case of a crack occurring in the arch is not covered here.
This check is carried out in Annex D, Section D.2.2.
81
7.4.2 Fatigue assessment based on geometric stress ranges
In this section geometric stress concentrations at geometric discontinuities are used for the
assessment. Such stress concentrations can be obtained with the help of stress concentration
factors (SCF) retrieved from diagrams for standard cases by carrying out finite element analysis or
using the photoelastic method (TIMOSHENKO & GOODIER [39], pages 138-145). The stress
concentration factor is defined as the ratio of the local stress at a relevant location to the nominal
stress in the corresponding net cross section. Following Eurocode 3, the fatigue assessment based
on geometric stress ranges is carried out in similar fashion to the assessment based on nominal
stress ranges. The nominal stress range is adapted to the damage equivalent stress range at
2 106 cycles, which is an approved concept in engineering praxis, HARTMANN ET AL. [16], page 7.
The main difference is using the geometric stress range instead of the nominal stress range.
Furthermore, the fatigue strength curves to be used shall be of detail categories 90, 71 or 36
(ENV 1993-1-1: 1992 Section 9.6.3).
It was decided to determine the stress distributions in the various hanger connection types with
the help of finite element analysis in NE/Nastran (Annex C, Section C.1). Local stress
concentration effects due to the weld geometry and discontinuities at the weld toe are to be
ignored, as they are accounted for in the SN curves (ENV 1993-1-1: 1992 Section 9.5.3 (1)).
Therefore, only the overall geometry of the hangers was modelled. The 3D modelling was carried
out in SolidWorks 3D CAD Software (Annex C, Section C.1).
The hanger was subjected to a static tensile force of 1000 kN, which is a representative value.
Constraints were applied to the contact surface around the slip-resistant high strength bolts in
connection types 1 to 3. Connection type 4 and 5 were constrained at the lateral front faces of the
gusset plates where it can be directly welded wither to the arch or to an end plate.
The analysis was based on linear elastic theory. The principal stresses were recorded and are
given in the form of stress distribution illustrations. Where relevant, maximum or/and minimum
principal stresses are considered.
82

B
A
506.8
472.2
437.6
402.9
368.3
333.6
299.0
264.4
229.7
195.1
160.5
125.8
91.21
56.57
21.94
-12.7
-47.33
2
Fig. 7.11. Hanger connection type 1: Maximum principal stress [N/mm ], deformation exaggerated
In this connection forces are almost equally distributed from the hanger into the gusset plate.
This can be seen on the virtually horizontal colour gradients above the point where the free edges
of the plate splay upwards. This is exactly the point with the highest stress peak of the whole
hanger connection, which results from the discontinuous transition (A). A further stress
concentration is found at the transverse joint between hanger butt end and the gusset plates (B).
Geometric stress concentration factors (SCF) for axial hanger force:
(A): 1.43 (The calculation of the stress concentration
(B): 1.39 factors can be found in Annex D.3.3.2)
With the help of the SCF the nominal stresses are scaled to obtain geometric stresses which are
used in the assessment.
Fatigue strength curve to be used:

Eurocode 3 provides detail categories 90 and 71 for locations where the local stress
concentration has negative effects on full penetration butt welds, and detail category 36 where the
local stress concentration has negative effects on partial penetration butt welds and fillet welds.
(A): This point marks the end of the full penetration butt weld. Therefore, the SN curve for detail
category 90 is taken for the assessment (ENV 1993-1-1: 1992 Section 9.6.3 (2)), assuming that
both the weld profile and permitted weld defects acceptance criteria are satisfied.
(B): Here a full penetration butt weld constitutes the transverse joint between hanger butt end and
gusset plate. SN curve for category 90 is taken.
The assessment shows that the stress range at (A) exceeds the fatigue strength at 119%,
whereas the check is fulfilled with 74% at (B). The summary of all results is found in Figure 7.26.
83
A
399.4
372.9
346.3
319.8
293.3
266.8
240.3
213.7
187.2
160.7
134.2
107.7
81.16
54.64
28.12
1.601
-24.92
2
This connection type constitutes a further development of type 1. On the one hand, the stress
peak at the formerly discontinuous transition between hanger and gusset plate is reduced by the
introduction of a larger transition radius (A). On the other hand, the distribution of forces from the
hanger into the plates is not as even as in type 1. The large transition radius is accompanied with
a reduction of material. Consequently, forces are essentially transferred into the plate further
upwards or further inside the plate. This can be seen on the curved colour gradients indicating
higher stresses in the hanger region. That means that a larger portion of the hanger force is
transferred through the butt end of the hanger, leading to higher stresses at that area (B),
compared with hanger connection type 1. Therefore, it is suggested to increase the insertion
length of the hanger into the plate in order to provide a longer longitudinal weld.
(A): 1.13
(B): 1.47
The same SN curve as in connection type 1 is applicable.
The fatigue checks are fulfilled, at (A) with 94%, as well as at (B) with 78%.
84
A
1217.0
1140.0
1063.0
986.3
909.4
832.5
755.5
678.6
601.7
524.8
447.9
370.9
294.0
217.1
140.2
63.25
-13.67
2
The edge of the circular cut-out at the butt end of the hanger marks the highest stress peak of
this connection (B). However, the fact that the transverse butt weld and therefore the clashing
between transverse and longitudinal butt weld is eliminated, is advantageous, as residual stresses
are reduced. Still, in terms of the detail categories given in Eurocode (ENV 1993-2: 1997 Table L),
a butt weld here would not be as harmful as with hanger connection 1 and 2, since only a single
gusset plate allows the execution of the weld from both sides. At point (A) the transition radius is
as large as in connection type 2.
(A): 1.16
(B): 3.89
Since both stress concentrations are located at full penetration butt welds, the SN curve for detail
category 90 is applicable.
The SCF at (A) hardly differs from the respective value in connection type 2 (SCF = 1.13). The
fatigue check is here also fulfilled with 96%.
The SCF of 3.89 at the edge of the hole (B) appears to be rather large compared to the
reference example hole in a plate, where the possible SCF lies between 2 and 3, GIRKMANN [12].
The reason for the difference is that the double symmetric plate-with-hole standard case cannot be
transferred to the hanger geometry, which is only single symmetric. The continuously fixed upper
end is faced by the concentrated force transmission by the hanger on the other side of the hole.
Moreover, the bevelled butt end of the hanger incorporates a discontinuity in cross section, which
also contributes to the higher stress concentration factor.
At (B) the circular cut-out represents a milled edge making detail category 125 applicable,
GNTHER ET AL. [15]. However, since the longitudinal full penetration butt weld ends at the stress
concentration location as well, it is sensible to reduce the detail category to 90 (ENV 1993-1-1:
1992 Section 9.6.3 (2)). The fatigue check is not fulfilled with 285%.
85
IIa
D1 A
D2
D2
C
656.5
610.6
564.7
518.8
472.9
427.1
381.2
335.3
289.4
243.6
197.7
151.8
105.9
60.04
14.16
-31.72
-77.6
1 D1
2
The overall body of hanger connection 4 is virtually free of stress

peaks. The hanger force is continuously transferred into the widened
end member and further into the gusset plates. The stress level is
generally low in these areas, compared with connection types 1 to 3.
A small stress concentration certainly occurs at the widening of the
hanger (C), since an increasing cross section is also a stress
concentrator, ALMAR-NSS, [3], page 216. The butt weld splice
where the widened end member is joined to the hanger should
therefore be located below point (C) with a sufficiently large distance
from that geometric discontinuity. Based on results from further
investigation (Section 7.5), a distance of three times the hanger Suggested location
diameter is suggested (Figure 7.15). of weld joint
3D
Stress is also concentrated at point (A), which corresponds to

connection type 1. However, in this case the nominal stress and
therefore the geometric stress are reduced because of the widened D
member and are not critical.
The highest stress concentrations occur at (D). At these points,
Fig. 7.15. Location of weld joint
radii of 30 mm (D1) and 40 mm (D2) were introduced so that the for connection type 4
lower edges of the gusset plate contact under an angle of 90 to the
constraint. That allows continuous welding around the plate. The
diversion of forces at the radii causes the stress concentration. At
(D1) stresses appear to be higher, corresponding to an overall higher
stress level on this side of the hanger. This is because the gusset
plate on that side provides less cross sectional area.
86
Further stress concentrations are found at points (IIa) and (IIb), representing the corners of the
continuous constraint, where the gusset plate is welded to the arch profile or an end plate. The
absolute value of the minimum principal stress at point (IIb) appeared to be greater than the one of
the maximum. The minimum principal stresses are illustrated in Figure 7.16.
IIa
IIb
192.4
143.8
95.1
46.44
-2.214
-50.87
-99.52
-148.2
-196.8
-245.5
-294.1
-342.8
-391.5
-440.1
-488.8
-537.4
-586.1
2
Fig. 7.16 Hanger connection type 4: Minimum principal stress [N/mm ], deformation exaggerated
The calculation of the SCF calls for the maximum absolute value of principal stresses. Thus, the
minimum principal stress is critical for point (IIb).
(C): 1.14
(A): 1.81
(D1): 2.90
(D2): 3.26
(IIa): 4.85
(IIb): 2.96
Even though the absolute value of the stress concentration at (D1) is higher than at (D2), the
calculation of the SCF shows inverse results. It can be seen here, that the pure value of a SCF
does not give sufficient information about the actual stress. Moreover, the size of the investigated
member has great influence on the nominal stress, which is the second factor besides the SCF.
Taking two specimens, equal in geometry, but different in size and subjected to one and the same
load, the stress concentration factor is the same for both specimens, although, the actual value of
the local stress is higher at the smaller specimen (See Section 7.6).
(C): If this point is the location of the weld, detail category 90 is to be used (full penetration butt
weld). But if the splice is as recommended chosen to be sufficiently distant from the widening,
point (C) could be categorised in detail category 160 (ENV 1993-2: 1997 Table L.1/1). However, a
reduction of the fatigue strength is necessary since (C) constitutes a geometric discontinuity.
(A): This point marks the end of the full penetration butt weld. Therefore, detail category 90 is taken
for the assessment (ENV 1993-1-1: 1992 Section 9.6.3 (2)), assuming that both the weld profile
and permitted weld defects acceptance criteria are satisfied.
(D): The stress concentration does not affect any weld. According to Eurocode 3, the fatigue
assessment procedure using geometric stress ranges only specifies SN curves for welded
87
connections (ENV 1993-1-1: 1992 Section 9.6.3). In this case, the stress concentration occurs on a
gas cut edge and detail category 125 is taken (ENV 1993-2: 1997 Table L.1/1). A similar
assumption is made in GNTHER ET AL. [15].
(IIa): The ends of the constraints represent locations near fillet welds and are therefore to be
classified with detail category 36 (ENV 1993-1-1: 1992 Section 9.6.3 (2)). This low detail category
and the high stress peak will not fulfil the fatigue criteria. These points are certainly dangerous for
the weld joints around the plate at that location.
The utilisation of the fatigue strength at the various points is as follows:
(C): 95%
(A): 67%
(D1): 111%
(D2): 68 %
(IIa): 385%
(IIb): 343 %
It was found that subjected to purely axial load, the whole connection experiences an
inclination due to the eccentricity. The hanger is tilted and induces a bending moment around the
strong axis of the hanger connection. Stresses due to the bending moment are contained in the
stress distribution illustrations and therefore considered in the assessment.
IIb
D2
D1
IIa
585.5
544.8
506.0
467.2
428.4
389.7
350.9
312.1
273.3
234.6
195.8
157.0
118.2
79.47
40.7
1.926
-36.85
2
88
For point (A) the beneficial introduction of a transition radius has greater effect on the SCF
than the disadvantageous elimination of the widened hanger end member of connection type 4.
The SCF improved from 1.81 (type 4) to 1.27. However, the absolute stress ranges show opposite
results: The geometric stress range for (A) in type 4 was calculated to 125.09 N/mm2, whereas
type 5 gives a value of 197.48 N/mm2. This contradictory effect is discussed in Section 7.6.
In contrast to connection type 4, the lower edge of the gusset plate shows a larger transition
radius and connects to the constraint surface with an angle smaller than 90 (D). The stress
concentration around the circular shaped edge could therefore be decreased.
A further difference is the greater length of the constraint, which reduces the stress
concentration directly at the edges of the constraint (IIa) and (IIb). See also Figure 7.18 for
minimum principal stresses.
IIb
181.6
150.6
119.6
88.57
57.55
26.54
-4.48
-35.5
-66.51
-97.53
-128.5
-159.6
-190.6
-221.6
-252.6
-283.6
-314.6
2
Fig. 7.18. Hanger connection type 5: Minimum principal stress [N/mm ], deformation exaggerated
The stress concentration at (IIa) is greater than at (IIb). Since for both points the nominal
stress for the fatigue check is the same, only the maximum is of interest.
(A): 1.27
(D1): 2.10
(D2): 2.36
(II): 3.53
(A), (D1), (D2) and (II): See connection type 4.
(A): 105 %
(D1): 56 %
(D2): 43 %
(II): 343 %
89
7.5 Stresses caused by horizontal deflections
According to the assumption made in Section 7.2.2, horizontal deflections of hangers occurring
in network arch bridges due to wind load are not expected to cause fatigue damage in the hanger
connections. Nevertheless, forces caused by railway traffic may cause fatigue critical vibrations in a
horizontal direction. Therefore, it is desirable to investigate their impact. The structural analysis of
the 3D bridge model showed a maximum horizontal deflection perpendicularly to the arch plane of
1.18 mm, one meter below the transition between hanger and gusset plate. This value is based on
dead loads and lateral wind loads. This combination leads to maximum horizontal deflections;
higher hanger forces due to additional live loads cause smaller deflections.
To obtain information about the stress behaviour in the various connection types, the 3D
models were subjected to the predetermined 1.18 mm deflection causing bending about their
strong axis.
20.38 17.37 18.12
19.08 16.28 16.97
17.77 15.19 15.83
16.47 14.09 14.68
15.16 13.0 13.54
13.86 11.9 12.39
12.55 A 10.81 A 11.25

A 11.25 9.72 10.1
9.94 9.62 8.96
8.64 7.53 7.81
7.33 6.14 6.66
6.03 5.34 5.52
4.72 4.25 4.37
3.42 3.15 3.23
2.11 2.06 2.08
0.8 0.97 0.94
-0.5 -0.13 -0.21
type 1 type 2 type 3
2
Fig. 7.19. Maximum principal stress due to horizontal deflection [N/mm ], connection types 1, 2 and 3
An analysis of the stress distribution of connection types 1 to 3 in Figure 7.16 shows highest
stress peaks at the previously introduced point (A). This point already marks one of the most
critical stress concentrations due to axial loading (Section 7.4). Again, connection type 1 appears
to be most vulnerable with a stress peak of 20.38 N/mm2. The introduction of a transition radius
(type 2) improves the weak spot, giving 17.37 N/mm2. Connection type 3 shows a slightly worse
result than type 2 because of the different number of gusset plates.
90
22.87 15.85
21.4 14.86
19.93 13.86
18.45 12.07
16.98 11.87
15.51 10.88
14.04 9.88
A
12.56 8.89
11.09 7.89
9.62 6.90
C
8.15 5.90
6.67 4.91
5.2 3.91
3.73 2.92
2.26 1.92
0.78 0.93
-0.69 -0.07
type 4 type 5
2
Fig. 7.20. Maximum principal stress due to horizontal deflection [N/mm ], connection types 4 and 5
The stress peak magnitudes of connection types 4 and 5 differ significantly, which is however
caused by the different geometry. Remarkable is the fact that the highest stress concentrations
occur at different locations, namely (C) in type 4 and (A) in type 5. Evidently, the lower end of the
widened end part constitutes a critical area and demands, as already found in Section 7.4, that the
weld joint is located further below. It was found that a distance of about three times the hanger
diameter is sufficient, where the stress concentration is decayed to an acceptable value
(Figure 7.15).
The stress concentration at (A) in connection type 5 marks the lowest value of all types
examined.
7.6 Evaluation
Having carried out the nominal stress range as well as the geometric stress range method
using finite element analysis, the following results can be used to decide on the best hanger
connection among the various types tested. Since hanger connection types 1 to 3 are of similar
geometry, they will be treated separately from types 4 and 5.
Connection types 1 to 3
Advantages and disadvantages can be found in each of the types 1 to 3. The first advantage
to mention is the transition radius between hanger and hanger gusset plate. The positive effect
can also be found in ENV 1993-2: 1997 Table L.4/2, detail (2), where fatigue strength increases
with larger transition radii. Secondly, the application of only one gusset plate is beneficial, because
the welds can be executed from both sides over the whole length.
In contrast, it was found that the circular cut-out at the hanger butt end (type 3) is not suitable
for a hanger connection with such small dimensions as used in this work. GNTHER ET AL. [15]
found the geometry shown in Figure 7.21 advantageous for bridges subjected to rain-wind induced
vibrations. In that article the hanger connection was assessed for an axial force of 1311 kN.
91
302.2
126
283.2
264.2
245.2
225.2
207.1
188.1
169.1
874
m
1m 150.1
52
=1
1260
R 131.1
437
112.1
93.04
74.03
276 t = 25
386
55.01
185
35.99
16.97
1017
-2.043
2
Fig. 7.21. Hanger connection suggested in GNTHER ET AL. [15], maximum principal stress [N/mm ], deformation
exaggerated
In this connection detail, the maximum principal stresses due to an axial hanger force of
1000 kN show smaller overall stresses, as well as a smaller stress concentration around the hole
compared to connection type 3. The reason for the difference is a greater cross sectional area,
hence smaller nominal stresses. The connection in Figure 7.21 shows a ratio between the hole
diameter and the gross cross sectional area 126 / (2276 + 126) / = 0.19, whereas in connection
type 3 this ratio is calculated to 60 / 220 = 0.27. This indicates an even smaller stress concentration
factor for type 3, but the higher nominal stress has more impact and makes the connection
impractical.
This proves that the SCF alone does not give enough information. The simple example of a
hole in an infinitely long plate will be taken to explain the behaviour. It can be found in
GIRKMANN [12], page 144.
The diameter of the hole shall be d and the width of the infinitely long plate w. Two
extreme cases exist for the hole diameter. For d w (case 1), the stress concentration
factor is determined to SCF = 2, whereas for d 0 (case 2), the factor is SCF = 3.
Assuming the same uniformly distributed load p for both cases, the actual stress at the
hole in case 1 will be infinitely high, due to the infinitely small net sectional area. Hence,
in spite of the smaller SCF the actual stress is higher in case 1 than in case 2.
p w d p
Fig. 7.22. Infinitely long plate with hole, subjected to uniformly distributed load
It is found that the application of a hole is only a feasible solution for bridges where the arch
profile allows hanger connections with large dimensions. The wide acceptance and application of
the feature hole shows its importance and relevance. Examples can be found in Germany:
- Elbe bridge near Dmitz (semi-circular cut-out)
- Teltow-channel bridge (kidney-shaped cut-out)
- Oder-Havel-channel bridge (semi-circular cut-out)
92
The circular cut-out was originally adopted for connection type 3 because
the sharp 90 edges as well as the transverse butt weld of the hanger end to
the gusset plate constitute weak points. Firstly, the edges mark possible
starting points for notches and cracks. Secondly, the existence of both
transverse and longitudinal welds is a reason for additional residual stresses.
To improve the first, it is suggested to equip the corners with a small r
transition radius, which requires the equivalent at the hanger end. The round
corner can then more easily be ground to a smooth surface (Figure 7.23). To
reduce residual stresses, it is recommended to follow a certain sequence for
the welding process. The transverse weld at the hanger end should be
executed first, followed by the longitudinal welds. It is supposed that this
method of fabrication contributes to the feasibility of this connection detail.
Fig. 7.23. Smooth round
corner in cut-out of
Connection types 4 and 5 gusset plate
Very high stress concentrations were found at points (IIa) and (IIb) in connection types 4 and 5
(Figure 7.16), and the fatigue strength using the SN curve for category 36 (fillet welds) is far too
much exceeded. Using the nominal stress method, the appropriate detail category appears to be
category 80 (ENV 1993-2: 1997 Table L.5/3), with which the fatigue check is fulfilled with 65% (IIb).
It is therefore uncertain how to judge the FEM result.
The widened hanger end member proved to be advantageous for the assessment of location
(A) and can generally be suggested if the connection is welded to the arch, as in STEIMANN [37]. It
is recommended to locate the weld joint between hanger and widened hanger end at a distance of
three times the diameter below the transition. In that way it does not lie in the stress concentrated
range.
Suggested hanger connection

The similar overall geometry of connection types 1 to 3 recommends combining the
advantageous features and abolishing the weaknesses. Figures 7.24 and 7.25 show the
suggested connection detail.
450.0 17.01
417.8 15.94
385.7 14.87
13.80
353.5
12.73
321.4
11.66
B 289.3
10.59
257.1 9.52
225.0 8.45
192.8 7.38
6.31
160.7 Left: Subjected to axial load of
1000 kN 5.24
128.6
4.17
A Right: Subjected to horizontal
96.42 deflection of 1.18 mm 3.10
64.28 2.04
32.14 0.97
-0.11
0.0
2
Fig. 7.24. Suggested hanger connection: Maximum principal stress [N/mm ], deformation exaggerated
93
The connection detail consists of a single gusset plate, in which the hanger is inserted and
joined by means of full penetration butt welds executed from both sides. The weld should be
ground flush to the surface to provide smooth transitions between the adjacent members. The
above suggested round corner in the gusset plate cut-out will be applied as well.
The beneficial transition radius between hanger and plate is also adopted and leads to an
acceptable low stress concentration.
The suggestion of a greater insertion length (Section 7.4.2, hanger connection type 2) is
adopted resulting in a smaller stress peak at (B) compared to connection type 2.
Geometric stress concentration factors SCF due to axial load:
(A): 1.14
(B): 1.34

(A): 94 %
(B): 72 %
cross section 1-1
fillet joint,
1
a = 7 mm
W 360x410x634
6 slip resistant high

1
strength bolts
d = 24mm
r=
220
34
Single bevel 2.5
mm 75 70 75
both sides
40 70 70 40
154.9 70 94.9
60
t = 10 mm
60
65
227
185
65
300
t = 20
60
119.9 70 69.8 60.1

520
319.8
293
220
80 60 80
Fig. 7.25. Dimensions of suggested hanger connection
94
Summary
The obtained data on stress distributions Com par is on of r e s ults

(utilis ation of fatigue s tr e ngth in %)
in the connection details confirm certain
method of assessment
stress concentrations at points considered connection
as critical by detail categories in the type location nominal stress geometric stress
Eurocode. These are points (A), (B) and (C), 1 A 166 119
since they constitute geometrical B 41 74
discontinuities. 2 A 83 94
B 41 78
In contrast, the stress concentration at 3 A 83 96
(D) had not been expected. It is B - 285
questionable if the stress peak at (D) is 4 C 93 95
relevant for fatigue or not, since no weld A 74 67
joint is directly affected. If the plate edge D1 - 111
contains notches, a fatigue failure could be D2 - 68
IIa - 385
expected. The material should be machined
IIb 65 343
after gas-cutting to reduce the danger of
5 A 83 105
fatigue failure. D1 - 56
In Figure 7.26 results from both D2 - 43
II - 343
assessment methods are compared. It can
s ugge s te d A 83 94
be seen that the assessment based on
type B 41 72
geometric stress ranges is beneficial for
location (A) in connection type 1. Using the Fig. 7.26. Comparison between results from nominal stress and
geometric stress ranges
most appropriate detail category from
Eurocode 3 is evidently too far on the conservative side. The actual result obtained from the FEM
analysis shows a significant improvement from 166% to 119% utilisation of fatigue strength. If
connection type 1 is designed with greater dimensions than in this work, it may certainly result in a
feasible solution satisfying the fatigue assessment.
Comparing the results for (A) in types 2 and 3 leads to different conclusions. The assessment
based on nominal stress ranges underestimates the stress peaks obtained from the FEM analysis.
This is even more the case for location (B) in types 1 and 2. One explanation is that the detail
categories used assume slightly different conditions and do not entirely correspond to the actual
situations. For example, the chosen detail category 36 (ENV 1993-2 Table L.3/2) used for location
(B) in types 1 and 2 assumes a butt splice of two plates with free edges, whereas the hanger is
welded to the gusset plate on three faces. Moreover, category 36 presumes equal and constant
plate thickness for both plates. The bevelled hanger end incorporates a slope and therefore a kink
in the surface, which is a geometric discontinuity that increases the stress concentration.
In addition, the low category 36 results from the fact that a butt weld is carried out from one
side. Using the SN curve of detail category 90 for the assessment based on geometric stress
ranges does not distinguish between butt welds carried out from one side and welds carried out
from two sides.
A further reason might be simplifications made for stress distributions. The stress distribution
along the transverse and longitudinal butt welds in type 1 and 2 was taken as constant, which is an
approximation and implies deviations from the actual distribution.
It was not possible to assess the stress concentration at the hole (B) in connection type 3 with
the help of nominal stress ranges. Eurocode 3 does not contain any detail category which
corresponds to that particular point. Even adopting the standard example hole in a plate
underestimates the actual conditions of the specific hanger connection detail, and the FEM
analysis appeared to be essential for judging this connection.
95
7.7 Connection to arch profile (based on nominal stress ranges)
The connection between the circular hanger profile and the gusset plate(s) is clearly the most
critical detail of the whole connection. However, it is still important to examine the fastening of the
gusset plate to the arch profile.
It is advantageous to have bolted connections which can easily be fastened on site. According
to Eurocode 3 for steel bridges (ENV 1993-2: 1997 Section 6.4.3), either fitted bolts or preloaded
bolts shall be used in shear connections. Connections with fitted bolts demand higher
accurateness for assembling than preloaded bolts in holes with standard clearance. The
Eurocode 3 further recommends designing preloaded bolts to be non-slip at the ultimate limit
states.
The geometry of the suggested hanger connections (Section 7.6) allows a bolted connection
to a gusset plate which is welded between the flanges of the H-profile (Figure 7.27). This leads to
detail category 112 (ENV 1993-5: 1997 Table L.1/2) for a double sided connection with preloaded
high strength bolts (I).
The shop-welded joint between gusset

plate and arch leads to two detail
fillet weld
a = 7 mm
categories. One predicts a crack in the
fillet weld (ENV 1993-2: 1997 Table
III L.5/3, detail category 80), the other
I
II predicts a crack in the arch profile (ENV
1993-2: 1997 Table L.4.2, detail
category 80). If the arch segments are
6 slip resistant
high strength joined by means of full penetration butt
bolts, d = 24 mm
welds giving detail category 90 (ENV
1993-2: 1997 Table L.3/1), the fillet weld
at (II) marks the most critical point for the
fatigue assessment. In cases where this
would lead to the application of a larger
Fig.7.27. Welded connection to arch profile profile, an alternative joint is proposed
where the gusset plate is connected to
the arch with bolts (Figure 7.28).
II
I

strength bolts 8 slip resistant
d=24mm high strength bolts
d = 22 mm
Fig.7.28. Bolted connection to arch profile
74
96
The one sided connection with preloaded high strength bolts (II) gives detail category 90 (ENV
1993-2: 1997 Table L.1/2), which is better than the welded connection. The two U-shaped plates
need to fit exactly between the flanges to avoid restraining forces. Therefore, it is suggested to
prefabricate this bolt connection, whereas the connection at (I) can be fastened on site.
Types 4 and 5 are not suitable for bolt connections such as in Figure 7.28. The continuous
hanger does not allow the application of U-shaped plates between the flanges. Moreover, the two
gusset plates on either side of the hanger do not provide enough space to accommodate bolts in a
connection with L-profiles. Here PER TVEIT suggests bolted end plate connections which would still
have to be examined (Figure 7.29).

strength bolts
d = 22 mm
fillet joint
Fig.7.29. Bolted end-plate connection for connection types 4, 5
Instead of end-plate connections on either side which would have to be fitted exactly between
the flanges on site, only one end-plate connects to the flange, whereas the other is bolted to the
web.
The bolts in the flange end-plate connection are subjected to shear, similar to the previously
discussed connections. The web end-plate connection is subjected to tension as well as shear,
which is due to the inclination of the hanger. Therefore, preloaded bolts are essential here. Fitted
bolts are not an option.
Due to the eccentricity of the hanger force, a torsional moment arises. This is the case for all
connection types and the torsional moment can undoubtedly be taken by the arch profile.
However, the geometry and the locations of fastenings of this hanger connection make it difficult to
assess for and trace the transmission of the torsional moment into the profile. An assessment was
not carried out. A solution might be a stiffening plate shop-welded between the top parts of the
flanges to ensure that the torsional moment can be transferred into the arch profile.
7.8 Hanger connections along the tie
Different conditions are found for the hanger connections along the lower chord. Their size is
not as much constricted as is the case with the connections along the arch, and significantly larger
connection plates can be used (Figure 7.30). Therefore, design checks are not expected to
become critical and were omitted in this work. In cases where design checks cannot be satisfied,
the dimension of the connection can be increased. However, this has to account for transverse
prestressing, which might intersect with the hanger connection detail. An end plate was chosen to
be 410 mm long. Since the transverse prestressing thread bars are spaced with s = 270 mm,
intersections are inevitable. But if either a hole is used at the end of the hanger, GNTHER ET AL.
97
[15], or as in this case the cut out is extended to the lower end of the vertical hanger gusset plate
(Figure 7.30), the thread bars can go through the opening.
60
610
420 200
40
Fig.7.30. Hanger connection along the lower chord
The hanger plate is equipped with shear studs which transfer the horizontal hanger forces into
the bridge deck. STEIMANN [37] showed that 8 shear studs with a diameter 22 are sufficient. Even
though the bridge calculated in this work gives smaller maximum hanger forces, the hangers are
more inclined than in STEIMANN [37] which leads to a higher horizontal component. Without a
design check, it was decided to use 8 shear studs with a diameter 25 instead of additional studs.
It ought to be mentioned that it might be possible to reduce or even eliminate the shear studs,
since hanger forces are also transferred by friction increased by prestressing.
7.9 Notes on the FEM analysis
Creating three dimensional models and determining stress results with the help of a finite
element analysis programme incorporates certain aspects which have to be considered in order to
be able to judge the result. The following information is supposed to give assistance and explain
possible derivations from analyses carried out under different conditions.
In NE/Nastran (Annex C, Section C.1) the interaction between elements is based on forces
exerted at the so-called grid points, where elements are connected together. The stiffness of the
structure, discretised at the grid points, is generated with data on material properties and geometry.
Both stiffness and forces are used to calculate displacements with which stresses and strains are
generated. It was decided to use tetrahedron elements with an additional mid-node.
The mesh size of the model has significant influence on results. For each model the default
mesh size suggested by NE/Nastran was adopted. According to ABS [1] it is advisable to create
smooth transitions without abrupt changes in mesh sizes. The meshes did not show discontinuities
in the form of abrupt changes in element size, so it was not necessary to manually remesh the
model. Few warnings occurred during the analyses indicating very steep face angles exceeding
80.
The mesh sizes varied depending on the model between 10 and 20 mm, which corresponds
approximately to the thickness of the gusset plates.
The colour gradation appears to be very coarse. The different colours in the FEM result
illustrations represent mean values and the given magnitudes constitute upper and lower
boundaries of each colour. For the evaluation, only the mean values were considered.
98
Section 8
Erection of the bridge using a temporary lower chord
8.1 Why a temporary lower chord
Many common arch bridges with vertical hangers possess lower steel chords consisting of
transverse and longitudinal beams. Together with the arches and the hangers they form a steel
skeleton strong enough to bear itself when lifted at the ends of the bridge. This gives considerable
advantages for erection because the steel skeleton can be mounted accurately on side spans or
construction sites besides the bridge location and then be displaced.
To exploit properly the structural behaviour of network arch bridges no permanent steel chord
should be used. It is normally best to design the bridge with a concrete tie. But then the heavy self-
weight impedes then a simple lift with ordinary hoisting devices.
The solution given by PER TVEIT is a temporary lower steel chord similar to the lower steel
chords mentioned above. Since the temporary lower chord is only subjected to loads during
erection of the bridge, it can be of a very light and efficient design. For the transversal beams the
fresh concrete of the bridge deck gives maximum strains; decisive loads for the longitudinal beams
are the self-weights of formwork, reinforcement and tendons. After erection the temporary lower
chord is removed and can be reused, with some modifications, to erect other network arches or to
serve for other structures.
Fig. 8.1. Temporary lower chord and formwork at arch root point
For the bridge, the object of this work, a temporary lower chord was designed and assessed for
all construction phases. It was refrained from listing every detail of the investigation in this section.
All results are found in Annex E.
99
Diploma Thesis Brunn & Schanack Section 8: Erection using a temporary lower chord
8.2 The design of the temporary lower chord and formwork
Formwork sheets
The top layer consists of 18 mm thick plywood formwork sheets, for example Dokaplex
formwork sheet, [9]. In transverse direction six sheets with a width of 1.5 metres, two with a width
of 1.25 metres and two with a width of 1 metre are needed to ensure that all edges of the formwork
sheets are supported. The deflections due to fresh concrete do not exceed 0.83 mm, and therefore
a virtually plane bottom surface of the bride deck can be provided.
Formwork beams
The plywood rests on longitudinal timber formwork beams with a height of 200 mm, for
example Doka formwork beam H 20 P, [8]. They have a length of 4.5 metres and span 3.5 metres,
which means they provide double cross section due to overlapping on a length of 1 metre at the
supports. Their mid-span deflection of 4.7 mm equivalents L/750 and is tolerable in respect of a
plane bottom surface, ENV 1992-2: 1996; 4.4.3.1 (105).
Formwork for cantilevers
The formwork for the cantilevers consists of vertical plywood
posts sawn to the required shape and horizontal formwork sheets.
The vertical posts are spaced at distances of 0.5 metre. Laths are
applied to ensure a transversely non-sliding constraint (Figure 8.2).
Transverse steel beam
The formwork beams are supported by transverse steel beams.
This is the member of the temporary lower chord subjected to largest
forces, mainly bending moments. An IPEa 550 profile proved to be
necessary by assessment. As part of the wind bracing it also
receives axial forces. To prevent lateral torsional buckling it is Fig. 8.2. Formwork for cantilever
restrained at the top flange by wooden cams nailed to the timber
formwork beams. The structural system is a simply supported beam with a span of 10.15 metres
and 1.45 metres cantilevers at both sides. The deflections of the transverse beam will be
countervailed by applying a wooden camber with the shape of the deflection line on top of the steel
beam. This camber accounts for deflections due to fresh concrete, maximum 54.4 mm, and ballast,
sleepers and rails. Furthermore, it is used to compensate for the 20 mm difference in the height of
the longitudinal steel beams and the timber formwork beams.
Diagonal members of wind bracing
Serving as diagonal members of the wind bracing three profiles L 120x10 are collocated
between two transverse steel beams as shown in Figure 8.3. They are bolted to the top flange of
the transverse beams.
Longitudinal steel beam 13570
10150
The transverse steel beams are
supported by longitudinal steel IPEa 550
beams in the arch planes. The

upper flanges of the transverse
HEB 220
L
3500
12
beams are attached with bolts to the

0 x1
50
0
47
lower flanges of the longitudinal

beam. The longitudinal beam is
subjected to axial forces and
3380
bending moments during the first
Fig. 8.3. Structural system of the temporary lower chord
100
construction phases. This load case is decisive for the assessment. The bending moments will
increase when casting the bridge deck. Since the longitudinal beams are disconnected from the
arch at this time, they will not receive axial forces and therefore additional capacity is released to
bear more bending. A HEB 220 profile was found to be sufficient for the bridge calculated in this
work. Eight beams with a length of 12.5 metres are needed for each side of the bridge. They are
connected by butt joints with bolted cover plates.
At the lower hanger ends the longitudinal beam is connected with bolts to the end plate of the
hanger connection detail. Between this plate and the top flange of the profile an 18 mm thick
wooden board is placed, to ensure that the lower surface of the hanger connection plate is in one
plane with the bottom surface of the bridge deck. Cut-outs are made in the formwork at these
points.
8.3 Special considerations on the end of the temporary lower chord
The geometry described in Section 8.2 is appropriate for almost the whole span of the bridge.
At the ends there are adaptations necessary due to the higher load of the end cross girder and the
transmission of forces between the arch root point and the longitudinal temporary steel beam.
The bottom surface of the bridge deck is inclined near the end cross girder. This causes a kink
in the longitudinal steel beam. The additional forces rising from this eccentricity are distributed to
the nearby hangers by a truss. The web of the transverse beam at the break point forms the
compression member. To obtain the desired horizontal bottom surface of the end cross girder the
formwork beams receive a wooden triangular camber on their top, see Figure 8.1. The number of
timber formwork beams is increased and the last transverse steel beam consists of an IPEv 550
profile to bear the higher loads.
s mo
oth b
ar d
= 60
mm
steel plate
420x200x40 mm
plywood plate
HEB 220 420x200x18 mm
l = 11530 mm HE B 220
l = 2850
m m
L 120x10
l = 4750 mm steel plate steel plate L 120x10 HEB 200
1380x200x15 mm 1300x200x15 mm l = 4250 l = 300 mm
mm
IPEa 550
IPEv 550
l = 13570 mm
l = 13570 mm
Fig. 8.4. Connection of temporary lower chord to arch root point, deflected state
8.4 Sequence of erecting the bridge
In order to explain the erection of the bridge more clearly, its construction is divided into 4 phases.
1. Mounting the steel skeleton and displacement
On already built foreshore bridges or a suitable site near the final bridge location the steel
skeleton is assembled. Mounting the steel skeleton starts at one end of the span and is mainly
done by one crane. The sequence should prevent erected parts from obstructing the access by the
101
crane. Therefore, the erection method depends to some extend on the width of the place where
erection is done. The stages of erection are as follows:
1a: Erecting the scaffolding on the inner sides of the planes of the arches
1b: Laying out the transverse steel beams
1c: Attaching the longitudinal beams to the transverse steel beams
1d: Connecting the diagonal wind bracing members to the transverse beams
1e: Putting the arch segments on the scaffolding
1f: Welding the arch splices and connecting the arch root point to the temporary
longitudinal beams
1g: Connecting the members of the main wind bracing to the arches
1h: Connecting of the hangers first to the top and then to bottom connection details
1i: Removing of the scaffolding
The steel skeleton is now ready to be lifted at the ends of the span and transported to the final
position. The total weight to be lifted is about 300 tons. Depending on the particular circumstances
the appropriate method out from a great variety should be chosen. Pontoons or floating cranes are
a possibility for river crossings; mobile cranes for an inland crossing.
2. Preparations for casting the bridge deck
Once the steel skeleton is in place formwork, reinforcement and the prestressing tendons can be
laid out.
2a: Applying wooden parts to form the camber on top of the transverse steel beams
2b: Completing the timber formwork beams with wooden cams and putting them into place
2c: Laying out the formwork sheets
2d: Attaching the temporary railings
Steps 2b to 2d should be performed in a certain sequence. Work should start from
both ends of the bridge simultaneously.
2d: Finishing the formwork under the cantilevers
2e: Putting into place the reinforcement and prestressing tendons
Due to the complicated crossing of the reinforcement and the prestressing tendons
in the area of the concrete edge beam it is necessary to complete the
reinforcement for the whole bridge. It can not be made separately for different
construction stages. For an example of the sequence see TVEIT [45], page 50.
2f: Applying formwork sheets with special cut-outs for the reinforcement between the end
cross girder and the main part of the bridge deck, this also applies to the edge beams
2g: Casting of the end cross girder and first several metres of the bridge deck until the free
parts of the two designated longitudinal tendons are straight across the span
2h: Curing of the concrete
2i: Partial prestressing of transverse tendons in end cross girder, three days after step 2g
At this construction state the temporary longitudinal steel beams take decisive
forces.
2j: Partial prestressing of the two mentioned longitudinal tendons with 650 kN each
3. Casting the edge beams
3a: Casting one of the edge beams with a footpath beginning at mid-span and proceeding
simultaneously towards both ends of the bridge
102
Since it is desirable to avoid hangers relaxing in fresh or semi-hardened concrete,

it was examined whether starting from either both ends or mid-span gives a better
result (See Annex E, Section E.8 of this work). For the bridge, the object of this
work, starting from mid-span is more appropriate. It depends on the hangers
resistance to relaxation.
3b: Curing of the concrete
3c: Casting the other edge beam with a footpath beginning at mid-span and proceeding
simultaneously towards both ends of the bridge
3d: Curing concrete
3e: Removal of formwork between edge beam/end cross girder and the main part of the
bridge deck
3f: Partial prestressing of all longitudinal tendons with 25% of the design value, three days
after step 3c, meanwhile disconnecting the temporary longitudinal steel beams from
the arches
The concrete edge beam now contributes to the structural behaviour of the span.
4. Casting the rest of the bridge deck and finishing the main work
Due to the length of the bridge and the time necessary to shape the inclined upper surface
properly, the casting of the deck was divided into three construction stages as shown in Figure 8.5.
Edge beams, End cross girder,
already cast already cast
Construction stage:
III I II I III
3.5 15.5 m 15.5 m 31 m 15.5 m 15.5 m 3.5

Fig. 8.5. Casting sequence of bridge deck in construction phase 4
The correct order and direction of casting were found by numerical analysis and prevent any
hanger relaxing. The possibilities starting from either both ends or mid-span were also examined,
but found to be inappropriate due to extensive relaxation of hangers.
4a: Casting the bridge deck in the construction stage I

4b: Casting the bridge deck in the construction stage II
4c: Casting the bridge deck in the construction stage III
4d: Removal of formwork underneath the cantilevers
4e: Full prestressing of transverse tendons in end cross girder three days after step 4c
4f: Full prestressing of all longitudinal tendons
4g: Full prestressing of the transverse tendons simultaneously at both bridge sides
4h: Attaching the railings to the cantilevers
Now the temporary lower chord can be removed and the holes of the bolted
connections to the hangers filled. More details about this can be found in
TVEIT [45], pp. 52-55.
4i: Finishing the expansion joints at each end of the bridge
4j: Casting the space behind the anchorages of the longitudinal tendons
4k: Prestressing the transverse tendon behind the anchorages of the longitudinal tendons,
three days after step 4j
103
Section 9
Summary
The structural members of network arches are mainly subjected to axial forces. Generally,
structures with this characteristic are considered as efficient.
The cross-section of the arch is designed for the axial force and the bending moments in the
wind portal frame. The application of a truss with diagonal struts instead of a single bending
resistant top member in the portal frame decreases the portal frame length and therefore the
transverse bending moments in the arch.
After having carried out the optimisation process of the hanger arrangement the authors found
their assumptions confirmed. A simple solution for an improved hanger arrangement is obtained if
all hangers cross the arch at the same angle. The upper hanger nodes are equidistant. At the ends
of the arch special considerations are necessary due to the disturbance range caused by the
clamping. Several investigations have been carried out to determine a favourable size of the cross
angle. For final conclusions these investigations have to be continued.
Common solutions for hanger connection details have to be adapted for network arches
because the more slender arch profile does not provide enough space to accommodate the
connection details. Especially the application of a hole at the hangers end leads to unfavourable
results due to the small gusset plate. Possibilities have been found of shaping a hanger connection
detail with a satisfying resistance against fatigue strains.
At a transverse span of the bridge deck of about 10 meters concrete ties without transverse
prestressing show slight economic advantages compared to those with transverse prestressing.
Advantages of the prestressed alternative are less deflection, smaller depth and the possibility of
avoiding cracking as a result of systematic compressive stresses.
The arch root calls for special attention while designing it. The stress range due to live load is
likely to exceed the allowed limits, because of the skew weld between the arch and the end plate
which takes nominal stresses and shear stresses from the large axial force in the arch. One
possible solution to improve this detail is enlarging the flanges of the arch profile and transferring
the forces partially to the horizontal plate above the bearings. The minimum distance of the
prestressing strand anchorages and the end cross girder require an enlargement of the concrete tie
at the arch root.
The drainage of the bridge also needs adapted solutions due to the extraordinarily slender
lower chord. It has been shown that various possible methods exist. The feasibility of the drainage
has to be considered already when designing the cross-section of the tie.
The investigation of the alternative stilt bearing showed its applicability. It might be a cheaper
solution where material is more expensive than labour hours. It could also lead to more economic
details at the ends of the arches. Otherwise pot bearings convince with a clear transfer of forces,
no restraining forces and easy handling.
The erection of a network arch can be performed using a temporary lower chord. The steel
skeleton consisting of the arches, hangers, wind bracing and temporary lower chord can be lifted at
the ends of the bridge and transported to the final position. Following the described sequence the
longitudinal prestressing can be utilised to release the temporary longitudinal steel beams from
axial forces. The casting sequence of the edge beams and the bridge deck is to determine for each
project, since relaxing of hangers in fresh or semi-hardened concrete can occur. The necessary
structural steel for the temporary lower chord is 15 % of the permanent structural steel.
Compared to other railway arch bridges the calculated network arch saves a significant amount
of steel.
As stated in the introduction, the authors hope that more bridge engineers will consider using
this type of bridge.
104
Section 10
Criticism and proposals for future work
While working on this thesis, some investigations had to be aborted to proceed with others.
Nevertheless, the authors continued thinking about the results and found mistakes and new ideas.
This section is supposed to help to evaluate the results, prevent repetition of the mistakes and
encourage further investigation of the ideas which seem important to the authors.
Optimisation of the hanger arrangement
According to the theory described in Section 6.6.2 an improved hanger arrangement is likely to
be found by determining the line of thrust in a network arch. The authors decided on the radial
arrangement because of its easy applicability. It is necessary to further investigate a favourable
hanger direction especially at the arch ends. Furthermore it is desirable to determine the direction
of the arch deflection relevant for the hanger forces.
The authors underestimated the influence of the stiffness ratio between arch and tie on the
structural behaviour. Tests with various stiffness ratios showed significant deviations of cross
angles giving small bending in the arch. An investigation of this parameter is definitely needed.
Combining the various demands of the attributes of an optimisation process requires a certain
weighting. The one assumed in Section 6 is based on an equal weighting of the considered
attributes. It is necessary to determine weighting factors taking into account the real ratio of the
various demands in the assessment of a bridge.
Concrete tie
For the two alternatives without transverse prestressing the required number of longitudinal
tendons was determined approximately. Since total costs differ only little, the price of the
longitudinal tendons may influence the results of the comparison. Exact calculations of the
necessary number of tendons are desirable.
Stilt bearing
For the assessment of the vertical bearing plate, formulas according to Hertz pressing were
used. Considering the small cross section provided underneath the contact surfaces, these
formulas give non-conservative results. Examination of more appropriate methods, maybe using
finite element analysis, is desirable.
FEM-analysis of connection details
Applying finite element software requires verification with the help of a patch test. This principle
was not respected. The consequences of this omission will be ascertained at this point by carrying
out a patch test.
As a basis for the verification the standard example hole in a plate was chosen. The stress
concentration factor in a plate around a hole with a diameter of one third of the plate width is
105
Diploma Thesis Brunn & Schanack Section 10: Criticism and proposals for future work
approximately 2.5, GIRKMANN [12], page 144. Further information on this standard example is given
in Annex D, Section D.3.4. The following dimensions of the plate were used:
Width W = 300 mm
Thickness T = 25 mm
Length L = 1000 mm
Diameter of the hole D = 100 mm (symmetrical to the plate)
In a first test (Figure 10.1), the plate is modelled and meshed in such a way that the conditions
correspond to the analyses of Section 7. The plate is constrained at the left short face and
subjected to 1000 kN at the opposite face, leading to a nominal stress at the net section of
o = 1000 kN / (25 200) = 200 N/mm2. The stress peak is found to be 441.7 N/mm2 which gives a
stress concentration factor SCF of 441.7 / 200 = 2.21.
441.7
413.6
385.4
357.3
329.2
301.0
272.9
244.8
216.7
188.5
160.4
132.3
104.2
76.02
47.89
19.77
-8.36
2
Fig. 10.1. Reference plate with hole, coarse mesh, maximum principal stress [N/mm ]
In a second test (Figure 10.2) a finer mesh was chosen leading to a stress peak of
462.4 N/mm2 and therefore a SCF of 462.4 / 200 = 2.31. Considering a possible range of a SCF
between 2.0 and 3.0, both results appear to vary significantly from the predicted value of 2.5 in
GIRKMANN [12].
462.4
433.2
404.0
374.8
345.6
316.5
287.3
258.1
228.9
199.8
170.6
141.4
112.2
83.04
53.86
24.69
-4.49
2
Fig. 10.2. Reference plate with hole, fine mesh, maximum principal stress [N/mm ]
Considering the reference value, it is apparent that the analyses of the hanger connections
underestimate the actual stresses. The chosen meshes appear not to be fine enough, since only
the test with the refined mesh showed a tendency towards the reference result.
It is supposed that the analyses executed are still meaningful in respect of comparison
between the different connection types. Locations of stress concentrations are found, but, their
actual magnitudes are expected to be higher.
106
List of figures
Front page
Photo composition of the bridge, the object of this work
Section 1
-
Section 2
-
Section 3
-
Section 4
Figure 4.1, Tied arch with vertical hangers, structural behaviour with partial loading (TVEIT [45], page 16) 6
Figure 4.2, Tied arch with one set of inclined hangers, structural behaviour with partial loading (TVEIT [45], 6
page 16)
Figure 4.3, Areas, stiffnesses and influence lines for the lower and upper chord of two tied arches 7
(TVEIT [45], page 14)
Section 5
Figure 5.1, Upper hanger connection detail 8
Figure 5.2, System lines of the wind bracing, half bridge 9
Figure 5.3, Cross section of the bridge at mid-span 9
Figure 5.4, Reinforcement of bridge deck with transverse prestressing 10
Figure 5.5, Cross section with reinforcement of alternative design 1 11
Figure 5.6, Cross section with reinforcement of alternative design 2 11
Figure 5.7, Comparison of costs for different bridge deck designs 13
Figure 5.8, Steel construction at arch root point 14
Figure 5.9, Anchorage of the longitudinal prestressing strands 15
Figure 5.10, Pot bearings 16
Figure 5.11, Position of the transverse tendons at the end of the bridge 17
Figure 5.12, Lower hanger connection with transverse tendons 17
Figure 5.13, Outer shape of the end of the bridge 18
Figure 5.14, Adapted design of the steel structure at the arch root point 19
Figure 5.15, Handrail 19
Figure 5.16, Drainage options 20
Figure 5.17, Arrangement of bearings 21
Figure 5.18, Pot bearing in situ 21
Figure 5.19, Pot bearings at the bridge in this work 22
Figure 5.20, Isometric view of stilt bearing and compression support 22
Figure 5.21, Elastomeric reinforced deformation bearing 23
Figure 5.22, Steel weight of railway bridges with ballast 25
Figure 5.23, Bridges used for comparison 25
Figure 5.24, Steel weight of the bridge, calculated in this work 26
Figure 5.25, Formwork material needed for the erection 26
Figure 5.26, Changes in the steel weight due to the application of the stilt bearings 27
Figure 5.27, Changes in the steel weight for alternative bridge deck, proposal 1 27
Figure 5.28, Changes in the steel weight for alternative bridge deck, proposal 2 28
Figure 5.29, Photo composition of the bridge calculated in this work 28
Figure 5.30, Front view 29
Figure 5.31, Side view 29
Section 6
Figure 6.1, One set of hangers 33
Figure 6.2, Node distance spaced by the help of an ellipse (schematic illustration) 33
Figure 6.3, Results for the maximum axial force of all hangers 34
107
Diploma Thesis Brunn & Schanack List of Figures
Figure 6.4, Results for the average axial force of all hangers 34
Figure 6.5, Results for the maximum variation of axial forces of all hangers 35
Figure 6.6, Results for the average variation of axial forces of all hangers 35
Figure 6.7, Number of relaxed hangers 35
Figure 6.8, Hanger arrangement with parameters ellip = 0.54 and range = 0.65, span: 100m, arch rise 36
17 m, number of hangers per arch plane: 44
Figure 6.9, Hanger arrangement with parameters ellip = 0.84 and range = 0.75, span: 100m, arch rise 36
17 m, number of hangers per arch plane: 44
Figure 6.10, Minimum curves within the data range 36
Figure 6.11, Hanger arrangement with parameters ellip = 0.3 and range = 0.3, span: 100m, arch rise 17 m, 37
number of hangers per arch plane: 44
Figure 6.12, Hanger arrangement with parameters ellip = 0.3 and range = 0.3, span: 100m, arch rise 17 m, 37
number of hangers per arch plane: 44
Figure 6.13, Results of a refined variation of parameter range, while ellip = 0.5 37
Figure 6.14, Definition of start angle and angle change 39
Figure 6.15, Results for maxN 40
Figure 6.16, Hanger arrangement with a start angle of 68 and constant angle change of 0.6 40
Figure 6.17, Results for N 41
Figure 6.18, Hanger arrangement with a start angle of 70 and a constant angle change of 1.4 41
Figure 6.19, Results for SummaxN 42
Figure 6.20, Hanger arrangement with a start angle of 66 and a constant angle change of 0.9 42
Figure 6.21, Results for aveN 43
Figure 6.22, Results for aveN 43
Figure 6.23, Results for SumaveN 44
Figure 6.24, Spoked wheel, span 100 m, rise of the arch 17 m, number of hangers: 44 45
Figure 6.25, Schematic illustration of the suggested hanger arrangement 46
Figure 6.26, Two Michell structures, found in BECKER [5], page 7 46
Figure 6.27, An optimal load bearing network of cables underneath a parabolic arch, McCullough 47
Figure 6.28, Two optimised trusses with circular upper chord, found on the Internet [28] 47
Figure 6.29, Efficient truss, ZALEWSKI ET AL. [51] 47
Figure 6.30, Cascade bridge of the Erie Railway, USA around 1845 PETRASCHKA [25] 48
Figure 6.31, 48
Figure 6.32, 48
Figure 6.33, 48
Figure 6.34, 49
Figure 6.35, 49
Figure 6.36, Derivation of the proposed hanger arrangement 49
Figure 6.37, 49
Figure 6.38, 49
Figure 6.39, 49
Figure 6.40, 50
Figure 6.41, 50
Figure 6.42, The hangers cross symmetrically the radii with the same angle, the cross angle defined as 51
being the variable is marked grey
Figure 6.43, Different hanger arrangement with cross angle of 0.835, 25 and 41 51
Figure 6.44, Results for the hanger forces 52
Figure 6.45, Results for bending moments in the arch about the horizontal axis 53
Figure 6.46, Sums of scaled values, result of improvement process 54
Figure 6.47, Results of the refined analysis 54
Figure 6.48, Comparison of the results 55
Figure 6.49, Bridges with spans of 75 m, 100 m, 150 m and 200 m 56
Figure 6.50, Diagram with scaled and added values for determination of improved results 57
Figure 6.51, Variation of the number of hangers, results scaled with their minimum to 1 and then added 58
Figure 6.52, Results for bending moments in the arch about the horizontal axis for different arch rises 59
Figure 6.53, Results for hanger forces for different arch rises 59
Figure 6.54, Influence of the ratio between live load and dead load on the hanger arrangement 61
Figure 6.55, Results for the calculation with loads similar to road bridge loads 62
Figure 6.56, Results from the calculation 63
Figure 6.57, Locating the convex deflection zone by the deflection line 64
Figure 6.58, Hanger arrangement used in RCK [27] for a Network Arch bridge with combined road and 65
railway traffic and four arch planes
Figure 6.59, Stiffening effect of pairs of crossing hangers 66
Figure 6.60, Range where the hangers cross the radii at an angle of 45 66
Figure 6.61, Summary in a scheme 69
108
Figure 6.62, Hanger arrangement found by the geometric description, labelling of the hangers 70
Figure 6.63, Maximal and minimal bending moments in the arch 71
Figure 6.64, Hanger forces 71
Figure 6.65, Maximal and minimal bending moments in the arch, adapted hanger arrangement 72
Figure 6.67, Maximal and minimal bending moments in the arch, final hanger arrangement 73
Section 7
Figure 7.1, Example of hanger connection 74
Figure 7.2, First mode shape of hanger web, f = 0.97 Hz 76
th
Figure 7.3, 40 mode shape of hanger web, f = 3.24 Hz 77
Figure 7.4, Investigated types of hanger connections 78
Figure 7.5, Suggested hanger connection 78
Figure 7.6, Hanger connection type 1 79
2
Figure 7.11, Hanger connection type 1: Maximum principal stress [N/mm ] 83
2
2
2
Figure 7.15, Location of weld joint for connection type 4 86
2
Figure 7.16, Hanger connection type 4: Minimum principal stress [N/mm ] 87
2
2
Figure 7.18, Hanger connection type 5: Minimum principal stress [N/mm ] 89
2
Figure 7.19, Maximum principal stress due to horizontal deflection [N/mm ], connection types 1, 2, 3 90
2
Figure 7.20, Maximum principal stress due to horizontal deflection [N/mm ], connection types 4, 5 91
2
Figure 7.21, Hanger connection suggested in GNTHER ET AL. [15], maximum principal stress [N/mm ] 92
Figure 7.22, Infinitely long plate with hole, subjected to uniformly distributed load 92
Figure 7.23, Smooth round corner in cut-out of hanger gusset plate 93
2
Figure 7.24, Suggested hanger connection: Maximum principal stress [N/mm ] 93
Figure 7.25, Dimensions of suggested hanger connection 94
Figure 7.26, Comparison between results from nominal stress and geometric stress ranges 95
Figure 7.27, Welded connection to arch profile 96
Figure 7.28, Bolted connection to arch profile 96
Figure 7.29, Bolted end-plate connection for connection types 4, 5 97
Figure 7.30, Hanger connection along the lower chord 98
Section 8
Figure 8.1, Temporary lower chord and formwork at arch root point 99
Figure 8.2, Formwork for cantilever 100
Figure 8.3, Structural system of the temporary lower chord 100
Figure 8.4, Connection of temporary lower chord to arch root point 101
Figure 8.5, Casting sequence of bridge deck in construction phase 4 103
Section 9
-
Section 10
2
Figure 10.1, Reference plate with hole, coarse mesh, maximum principal stress [N/mm ] 106
2
Figure 10.2, Reference plate with hole, fine mesh, maximum principal stress [N/mm ] 106
Annex A
Figure A.1, Load Model 71 and characteristic values for vertical loads, ENV 1991-3: 1995, Fig. 6.2 A-2
Figure A.2, Distribution of the wheel loads of the Load Model 71 onto the upper surface of the concrete A-2
slab
Figure A.3, Equivalent load qA1d, according to ENV 1991-3: 1995, 6.7.1.2 (2) P A-4
Figure A.4, Equivalent load qA2d, according to ENV 1991-3: 1995, 6.7.1.2 (3) P A-5
Figure A.5, Load combinations for assessment in ultimate limit state, and for fatigue design checks A-11
Figure A.6, Load combinations for assessment in Serviceability limit state, characteristic/rare A-12
combinations
Figure A.7, Load combinations for assessment in Serviceability limit state, infrequent and frequent A-13
combinations
109
Figure A.8, Load combinations for assessment in Serviceability limit state, quasi-permanent A-14
combinations, combinations regarding deformations and vibrations and additional load
combinations
Annex B
Figure B.1, Statical system and loading in transverse direction B-1
Figure B.2, Load model 71: Distribution in transverse direction B-3
Figure B.3, Load model 71: Distribution in plane B-3
Figure B.4, Determination of bending moment at quarter point due to train load B-4
Figure B.5, Determination of bending moment at mid-span due to train load B-4
Figure B.6, Concrete cover to thread bars B-5
Figure B.7, Nomenclature for prestressed concrete cross section B-7
Figure B.8, Transverse prestressing with relevant sections for preliminary design B-9
Figure B.9, Preliminary design of bridge deck in transverse direction B-10
Figure B.10, Cross section with numbered points assisting the calculation in Figure B.11 B-12
Figure B.11, Calculation of section properties of one half of the bridge deck B-12
Figure B.12, Preliminary design of bridge deck in longitudinal direction B-18
Annex C
Figure C.1, Foreshortened view of the 3D-FEM-model used for calculations C-1
Figure C.2, Partitioning of the mesh of the bridge deck C-2
Figure C.3, Connections and Couplings of the cable elements C-2
Figure C.4, Arrangement of prestressing tendons within the plane elements of the bridge deck C-3
Figure C.5, Definitions of the hanger node numbers C-3
Figure C.6, Geometry of the plane elements of the tie C-3
Figure C.7, Coordinates of hanger nodes C-3
Figure C.8, NE/Nastran mesh C-4
Figure C.9, Definition of axes in arch cross section C-5
Figure C.10, Fifth mode shape, equivalent initial bow imperfection eo,d ENV 1993-1-1: 1992 Figure 5.5.1 C-5
Figure C.11, Preliminary linear calculation of bending moments My in the arch C-6
Figure C.12, Preliminary linear calculation of bending moments My in the arch C-6
Figure C.13, Influence lines for N and My at the critical sections C-6
Figure C.14, First mode shape of the whole bridge C-7
Figure C.15, Maximum internal forces for collapse about the z-z axis of the arch (half of one arch, C-7
isometric projection)
Figure C.16, Internal forces for ultimate limit state assessment C-8
Figure C.17, Internal forces for fatigue assessment C-8
Figure C.18, Values for assessment to satisfy serviceability limits C-8
Figure C.19, Definition of the numbers of the hangers C-9
Figure C.20, Hanger forces C-9
Figure C.21, Influence line of hanger 4 C-10
Figure C.22, Angular rotation of upper hanger ends due to static wind loads C-11
Figure C.23, Eigenfrequencies of the first mode shapes of the hanger web C-11
Figure C.24, Mode shapes and eigenfrequency of the hanger web C-12
Figure C.25, Labelling of members and nodes, initial bow imperfections C-13
Figure C.26, Maximum and minimum forces in the wind bracing in ultimate limit state C-15
Figure C.27, Internal forces in ultimate limit state C-15
Figure C.28, Variation of internal forces for fatigue assessment C-16
Figure C.29, Internal forces in the wind bracing in serviceability limit state C-16
Figure C.30, Transverse bending moments my, [kNm/m], x = 4060 m C-17
Figure C.31, Longitudinal bending moments mx, [kNm/m], x = 4060 m C-18
Figure C.32, Transverse shear forces vy and axial forces ny [kNm/m], x = 51.5 m C-18
Figure C.33, Longitudinal shear forces vx [kNm/m], x = 4060 m C-19
Figure C.34, Summary of design-relevant internal forces for ultimate limit states C-19
2
Figure C.35, Non-frequent combination of actions: Concrete stresses y.top [N/mm ], x = 4060 m C-20
2
Figure C.36, Non-frequent combination of actions: Concrete stresses x.top [N/mm ], x = 4060 m C-20
2
Figure C.37, Frequent combination of actions: Concrete stresses y.bottom [N/mm ], x = 4060 m C-21
2
Figure C.38, Frequent combination of actions: Concrete stresses x.bottom [N/mm ], x = 4060 m C-21
Figure C.39, Summary of design-relevant concrete stresses for serviceability limit states C-22
Figure C.40, Stresses and internal forces for concrete fatigue assessment C-23
Figure C.41, Steel stresses for prestressing steel fatigue assessment C-23
Figure C.42, Maximum ULS internal forces, h = 610 mm C-24
Figure C.43, Maximum ULS internal forces, h = 470 mm C-24
Figure C.44, Calculation of stresses in reinforcement, structural depth h = 610 mm C-24
110
Figure C.45, Calculation of stresses in reinforcement, structural depth h = 470 mm C-24

Figure C.46, Internal forces for assessment of end cross girder in ultimate limit state C-25
Figure C.47, Stresses in end cross girder in Serviceability limit state, transverse direction C-25
Figure C.48, Pot bearings TF-10, TGa-10, TGe-10, MAURER SHNE GmbH & Co.KG [20] C-26
Figure C.49, Stilt bearing, arrangement of the slide directions as in Figure C.48 C-26
Figure C.50, Bearing forces in ultimate limit state, wind forces from the right side, design check relevant C-27
values are framed
Annex D
Figure D.1, Fatigue checks for axial stresses D-7
Figure D.2, Steel strength of different members D-11
Figure D.3, Design-relevant hanger forces D-11
Figure D.4, Hanger connection type 1 D-13
Figure D.5, Hanger connection type 1, alternative D-18
Figure D.10, Suggested hanger connection D-24
Figure D.11, Suggested hanger connection D-25
Figure D.12, SCF for connection types 1 to 3 and suggested type D-31
Figure D.13, SCF for connection types 4 and 5 D-32
Figure D.14, Fatigue assessment of connection types 1 to 3 and suggested type (geometric stress D-33
ranges)
Figure D.15, Fatigue assessment of connection type 4 (geometric stress ranges) D-33
Figure D.16, Fatigue assessment of connection type 5 (geometric stress ranges) D-33
Figure D.17, Joint III D-37
Figure D.18, CHS end plate connection D-39
Figure D.19, Sketch of node II D-42
Figure D.20, Punching shear critical perimeter D-51
Figure D.21, Serviceability limit state assessment D-52
2
Figure D.22, Concrete stress distribution just before cracking [N/mm ] D-53
2
2
Figure D.25, Geometric properties for calculation of steel weight D-59
Figure D.26, Calculation of steel weight D-59
Figure D.27, Cross section D-60
Figure D.28, Cross section D-66
Figure D.29, Calculation of deflection, alternative design proposal 1 D-75
Figure D.30, Calculation of deflection, alternative design proposal 2 D-76
Figure D.31, Geometric properties for calculation of steel weight D-78
Figure D.32, Reinforcement steel weight of design proposal 1 D-78
Figure D.33, Reinforcement steel weight of design proposal 2 D-78
Figure D.34, Load distribution on the stilt bearing D-87
Figure D.35, Geometry of vertical plate D-88
Figure D.36, Nomenclature: Determination of location with maximum deflection D-92
Annex E
Figure E.1, Structural system of formwork sheets, load is increased by partial safety factor E-1
Figure E.2, Schematic illustration of the overlapping timber formwork beams E-3
Figure E.3, Maximum bending moments during casting E-3
Figure E.4, Loads and constraints of the transverse steel beam E-4
Figure E.5, Axial forces in the wind bracing of the temporary lower chord E-4
Figure E.6, Deflections of the transverse temporary steel beam due to fresh concrete E-6
Figure E.7, Internal forces for construction state 1, full wind load and self-weight E-8
Figure E.8, Internal forces for construction state 2, full wind load and self-weight E-9
Figure E.9, Axial forces in hangers during casting of the longitudinal concrete edge beam E-10
Figure E.10, Casting sequence of bridge deck in construction phase 4 E-12
Annex F
Figure F.1, Comparison of influence lines for compression in the arch, (thin line - kviksound network F-2
arch / thick line - test of an arrangement according to a proposal by Brunn & Schanack)
Figure F.2, Comparison of influence lines for the force in the hangers I (thin line - kviksound network F-3
111
Figure F.3, Comparison of influence lines for the force in the hangers II (thin line - kviksound network F-4
Figure F.4, Comparison of influence lines for the bending moments in the arch I (thin line - kviksound F-5
network arch / thick line - test of an arrangement according to a proposal by Brunn &
Schanack) F-6
Figure F.5, Comparison of influence lines for the bending moments in the arch II (thin line - kviksound
network arch / thick line - test of an arrangement according to a proposal by Brunn & F-7
Schanack)
Figure F.6, Maximum load on the kviksound network arch in the ultimate limit state, the wheel loads are F-8
in the middle of the span
Figure F.7, Maximum load on the bridge with a test hanger arrangement according to the proposal of F-9
Brunn & Schanack in the ultimate limit state. The wheel loads are in the middle of the span
Figure F.8, Forces due to live loads on the left 54% of the kviksound network arch in the ultimate limit F-10
state. Half the weight of asphalt on the whole span is assumed. Three hangers relax
Figure F.9, Forces due to live loads on the left 54% of the bridge with a test of the hanger arrangement
proposed by Brunn & Schanack in the ultimate limit state. Half the weight of asphalt on the F-11
whole span is assumed. No hangers relax
Figure F.10, Composition of some relevant values from the comparison
112
List of standards
[S.1] ENV 1991-2-1: 1995 Basis of design and actions on structures Actions on structures, Densities, self-
weight and imposed loads
[S.2] ENV 1991-2-4: 1995 Basis of design and actions on structures - Actions on structures - Wind actions
[S.3] ENV 1991-3: 1995 Basis of design and actions on structures - Traffic loads on bridges
[S.4] ENV 1992-1-1: 1991 Design of concrete structures General rules and rules for buildings
[S.5] ENV 1992-2: 1996 Design of concrete structures Concrete Bridges
[S.6] ENV 1993-1-1: 1992 Design of steel structures General rules and rules for buildings
[S.7] ENV 1993-2: 1997 Design of steel structures Steel bridges
[S.8] DS 804 (B6) 09.2000 Vorschrift fr Eisenbahnbrcken und sonstige Ingenieurbauwerke (VEI)
113
List of references
[1] ABS, AMERICAN BUREAU OF SHIPPING, Guide for the Fatigue Assessment of Offshore Structures, Houston, Texas,
USA, April 2003
[2] ACI COMMITTEE 435, Control of Deflection in Concrete Structures, (Appendix B of ACI 435R-95), American
Concrete Institute, 2003
[3] ALMAR-NSS, A. (editor), Fatigue handbook Offshore steel structures, Tapir Publishers, Trondheim, Norway,
1985
[4] ARCELOR LONG COMMERCIAL S.A., Information on American section range, Laccolith - Cloche d'Or,
Luxembourg, 2003,
http://www.europrofil.lu/Download/PDF/PV/W.pdf
[5] BECKER W., Analogien: Bionikstrukturen (A): Wer leicht baut kann, auch schwer bauen, Institut fr Tragwerkslehre
und Ingenieurholzbau, Fakultt fr Architektur, TU-Wien, Austria, May 2003,
http://www.iti.tuwien.ac.at/download/files/lehre/arch/modul/TWS_BRc_Petraschka.pdf
[6] BRUGER P., Drehbare Kurzwellen-Vorhangantennen fr den Einsatz bis zu sehr hohen Windgeschwindigkeiten,
Stahlbau 66, Heft 3, pp. 174-176, Ernst & Sohn Verlag fr Architektur und technische Wissenschaften GmbH,
Berlin, Germany, 1997

[7] CORUS TUBES STRUCTURAL & CONVEYANCE BUSINESS, Celcius structural hollow sections, Corby, Northants, United
Kingdom, 2003,
http://www.corustubes.com/upload/FFCF8210-4EBB-11D5-80890040CA182248/pdf/... Hot_finished_SHS_v2.pdf
[8] DOKA UK FORMWORK TECHNOLOGIES LTD., Doka formwork beams H 16/H 20D List of items, Maidstone, Kent,
United Kingdom, 2003,
http://www.doka.com/imperia/md/content/doka/english/listofitems/29.pdf
[9] DOKA UK FORMWORK TECHNOLOGIES LTD., Doka formwork sheets List of items, Maidstone, Kent, United
Kingdom, 2003, http://www.doka.com/imperia/md/content/doka/english/listofitems/30.pdf
[10] DYWIDAG-SYSTEMS INTERNATIONAL, DYWIDAG Bonded Post-Tensioning Systems, Southam, Warwickshire,

United Kingdom, 2003,
www.dywidag-systems.com/docs/downloads.php?topic=brochures&content=b_europe
rd
[11] GHALI A., FAVRE A., ELBADRY M., Concrete Structures: Stresses and Deformations, 3 edition, Spon Press, London,
United Kingdom, 2002
th
[12] GIRKMANN K., Flchentragwerke Einfhrung in die Elastostatik der Scheiben, Platten, Schalen und Faltwerke, 6
edition, Springer-Verlag, Wien New York, 1978
[13] GRAF W., STRANSKY W., Optimierung Optimaler Entwurf von Tragwerken, Lehrstuhl fr Statik, Technische
Universitt Dresden, Germany, February 2000
[14] GRAE W., Einfhrung DIN-Fachbericht 101 Lasten und Einwirkungen auf Brcken einschlielich
Kombinationsregeln, G-M-G-Ingenieurpartnerschaft, Dresden, Germany, 2002,
http://www.gmg-dresden.de/publikationen/FB_101.pdf
[15] GNTHER G. H., HORTMANNS M., SCHWARZKOPF D., SEDLACEK G., BOHMANN D., Dauerhafte Ausfhrung von
Hngeranschlssen an sthlernen Bogenbrcken, Stahlbau 69, Heft 11, pp.894-908, Berlin, Germany, 2000
114
Diploma Thesis Brunn & Schanack List of references
[16] HARTMANN D., GALFFY M., WELLMANN A., Lebensdauerorientierter Entwurf der Hngestangen bei Bogenbrcken-
Modellierungen als Mehrebenenproblem, Lehrstuhl fr Ingenieurinformatik im Bauwesen, Ruhr-Universitt Bochum,
Contribution to the International Conference on the Applications of Computer Science and Mathematics in
Architecture and Civil Engineering, Weimar, Germany, 2003,
http://euklid.bauing.uni-weimar.de/papers/184/M_184.pdf
[17] HERZOG M., Stahlgewichte moderner Eisenbahn- und Straenbrcken, Der Stahlbau 9 (1975), pp. 280-282, 1975
[18] JAY A., Network arches memoire en vue de lobtention du titre dingenieur maitre, Hgskolen i Agder 4890,
Grimstad, Norway, 1998
[19] MAURER SHNE & CO. KG, Information on Maurer elastomeric deformation bearings, Munich, Germany, 2003,
http://www.maurer-soehne.com/ma15/kunden/maurer/maurer.nsf/7F09
[20] MAURER SHNE & CO. KG, Information on Maurer Pot bearings, Munich, Germany, 2003,
http://www.maurer-soehne.com/ma15/kunden/maurer/maurer.nsf/0170
[21] MCCULLOUGH J., Skyhook underneath An artist tangles with structural engineering, Department of Civil
Engineering, University of Strathclyde, Glasgow, Scotland, United Kingdom, 1996
http://www.strath.ac.uk/Departments/Civeng/staff/macleod/skyhook/
[22] NAKAI H., KITADA T., KUNIHIRO M., KITAZAWA M., HASINO F., Proposition of Methods for Checking the Ultimate
Strength of Arch Ribs in Steel Nielsen-Lohse Bridges, Stahlbau 64, Heft 5, pp. 129-137, Ernst & Sohn Verlag fr
Architektur und technische Wissenschaften GmbH, Berlin, Germany, 1995
[23] PEIL U., TELLJOHANN G., Dynamisches Verhalten hoher Bauwerke im bigen Wind, Stahlbau 66, Heft 3, pp. 99-
109, Ernst & Sohn Verlag fr Architektur und technische Wissenschaften GmbH, Berlin, Germany, 1997
[24] PETERSEN C., Stahlbau Grundlagen der Berechnung und bauliche Ausbildung von Stahlbauten, Friedr. Vieweg &
Sohn Verlagsgesellschaft mbH. Braunschweig, Germany, 1988
[25] PETRASCHKA K., Brckenbau (C), Institut fr Tragwerkslehre und Ingenieurholzbau, Fakultt fr Architektur, TU-
Wien, Austria, March 2003,
http://www.iti.tuwien.ac.at/download/files/lehre/arch/modul/TWS_BRc_Petraschka.pdf
th
[26] PUCHER A., Influence lines of elastic plates, 5 revised edition, Springer-Verlag, Wien New York, Austria, 1977
[27] RCK M., Entwurf einer kombinierten Strassen-Eisenbahn-Netzwerkbogenbrcke, Diploma thesis, Institut fr
Tragwerke und Baustoffe, Technische Universitt Dresden, Grimstad, Norway, 2003
[28] RODIN S. I., Optimisation of trusses 1 Genetic algorithm optimization in structural mechanics, Demonstration of
the software FrameGA, Department of material science and mechanics, Stullia, Scotland, United Kingdom, 1998
http://stullia.hypermart.net/economics/science/genetic_alg/gen_alg1.htm
th
[29] SCHNEIDER K.-J. (issuer), Bautabellen fr Ingenieure, 12 edition, Werner-Verlag, Dsseldorf, Germany, 1996
[30] SCHULTE H., BRUGER P., Drehbare Kurzwellen-Vorhangantenne auf der Funksendestelle Marlow, Stahlbau 69, Heft
10, pp. 781-787, Ernst & Sohn Verlag fr Architektur und technische Wissenschaften GmbH, Berlin, Germany,
2000
[31] SCHWARZ S., Sensitivittsanalyse und Optimierung bei nichtlinearem Strukturverhalten, Dissertation, Bericht Nr. 34
(2001), Institut fr Baustatik der Universitt Stuttgart, Germany, 2001
[32] SEDLACEK G., Praktische Anwendung der Windingenieurtechnik, Lehrstuhl fr Stahlbau, Rhein.-Westf. Technische
Hochschule Aachen, Germany, 2002,
http://www.stb.rwth-aachen.de/lehre/umdrucke/Wind_2002.pdf
[33] SEDLACEK G., Umdruck zur bung Stahlbau I Ermdung, Lehrstuhl fr Stahlbau, Rhein.-Westf. Technische
Hochschule Aachen, Germany, 2002
http://www.stb.rwth-aachen.de/lehre/umdrucke/ermuedung.pdf
[34] SEIDEL C., Grosser Beleg 2002 Projekt Havelberg II, Entwurfsplanung fr einen Ersatzneubau der Sandauer
Brcke in Havelberg, unpublished student project, Institut fr Tragwerke und Baustoffe, Technische Universitt
Dresden, Germany, 2002
115
Diploma Thesis Brunn & Schanack List of references
[35] SELLKE F., Brckenweb.de Brcke Nr. 230, 230A Groer Kolonnenweg, Archivnummer BAS 2002, Krefeld,
Germany
http://www.brueckenweb.de/Datenbank/Suche/brueckenblatt1.php3?brueckennummer=2002
[36] SPERLING D., Grenzzustnde der Gebrauchstauglichkeit Beschrnkung von Verformungen, Institut fr Baustoffe,
Massivbau und Brandschutz, Fachgebiet Massivbau, Technische Universitt Braunschweig, Germany 2002/2003,
http://www.ibmb.tu-bs.de/fgm/orga/script/durchbiegung.pdf
[37] STEIMANN U., Berechnung und Konstruktion einer sthlernen Eisenbahn-Stabbogenbrcke mit Netzwerkhngern,
Diploma thesis, Institut fr Tragwerke und Baustoffe, Technische Universitt Dresden, Grimstad, Norway, 2002
[38] TEICH S., WENDELIN S., Vergleichsrechnung einer Netzwerkbogenbrcke unter Einsatz des Europischen
Normenkonzeptes, Diploma thesis, Technische Universitt Dresden, Institut fr Tragwerke und Baustoffe,
Technische Universitt Dresden, Grimstad, Norway, 2001,
http://pchome.grm.hia.no/~pert/
nd
[39] TIMOSHENKO S., GOODIER J. N., Theory of Elasticity, 2 edition, McGraw-Hill, Kgakusha, Tokyo, 1951
[40] TRINITY RIVER CORRIDOR PROJECT, Trinity Bridges, City of Dallas, Texas, USA,
http://www.trinityrivercorridor.org/html/trinity_bridges.html
[41] TVEIT P., The design of network arches, The Structural Engineer Volume 44, No 7, pp. 247 - 259, London, United
Kingdom, July 1966
[42] TVEIT P., Report 7205 Network arches for railway bridges, Ren og Anvendt Mekanik, Danmarks
Ingenirakademi, Bygningsafdelingen, Aalborg, Denmark, August 1973
[43] TVEIT P., Considerations for Design of Network Arches, Journal of Structural Engineering, Vol. 113, No. 10, pp.
2189-2206, ASCE, ISSN 0733-9445/87/0010-21897 Paper No. 21892, October 1987
[44] TVEIT P., Wie konstruiert man die schlankste und leichteste Bogenbrcke der Welt Ein Vortrag ber
Netzwerkbgen. (How to design the most slender and lightest arch bridge in the world. A lecture on network
arches. In German), Publication series no: 2, Hgskolen I Agder University College, Publication series no: 2.,
ISSN 82-7117-301-4, Grimstad, Norway, 1996
[45] TVEIT P., The Network Arch An Extended Manuscript from 21 Lectures in 12 Countries, Internet edition,
Grimstad, Norway, May 2001
[46] TVEIT P., Optimal design of of Network Arches, Contribution to the IABSE Symposium in Melbourne, Australia,
September 2002
[47] TVEIT P., Preliminary design of network arch road bridges Examples with spans of 135 and 160 meters ,
Grimstad, Norway, March 2003
[48] TVEIT P., Revision of Erection of optimal Network Arches, Grimstad, Norway, May 2003
[49] VAN DER BURGH A.H.P., HARTONO, Rain-wind induced vibrations of a simple oscillator, Faculty of Information
Technology and Systems, Department of Applied Mathematical Analysis, Delft University of Technology, Delft, The
Netherlands, 2003,
http://ta.twi.tudelft.nl/TWA_Reports/cablis3new.pdf
[50] VERCH W., Umdruck zu den Vortrgen im Fach Entwurf und Gestaltung von Ingenieurbau-werken - Brckenbau,
Technische Universitt Dresden, Fakultt Bauingenieurwesen, Institut fr Tragwerke und Baustoffe, Dresden,
Germany, October 2001,
http://www.tu-dresden.de/biwitb/mbau/download/verch_all.pdf
[51] ZALEWSKI W., ALLEN E., IANO J., Shaping Structures: Statics CD ROM, John Wiley & Sons, Inc., New York, USA,
February 1998
http://shapingstructures.com/cdrom.html
116
Annexes
Photo composition of the bridge calculated in this work

Annex A
Load Assumptions
A.1. Permanent actions including self-weight
1a: Profile of the arches

American wide flange beams
W 360x410x900 : g1a = 9 kN/m
W 360x410x634 : g1a = 6.34 kN/m
1b: hanger
smooth surface steel bars
diameter: 60 mm, st = 78.5kN/m3 : g1b =0.222 kN/m
1c: lower cord slab
reinforced concrete
A = 4.96 m2, c = 25kN/m3 : g1c = 118.13 kN/m
1d: wind bracing
CHS 219.1 x 8, [7] : g1d = 0.417 kN/m
CHS 219.1 x 10, [7] : g1d = 0.515 kN/m
1e: sleepers, rails and ballast DIN-Fachbericht 101,
two tracks ( v 160 km/h ) Tabelle 1 (M.1), [14]
width of two tracks: 8.4 m : g1e = 105 kN/m
1f: cable trough
each side:
2 x U 350 (DIN 1026) : = 1.212 kN/m
1 x Bl 380x5 : = 0.15 kN/m
cables : = 1.138 kN/m
: g1f = 2.5 kN/m
1g: transversal prestress Annex B,
DYWIDAG threadbar 36D every 270 mm Figure B.9
each with : p1g = 989.1 kN
1h: longitudinal prestress Annex B,
6 x DYWIDAG Type 6827 each side of bridge Figure B.12
each with : p1h = 5467.5 kN
A-1
Diploma Thesis Brunn & Schanack Annex A: Load assumptions
A.2. Traffic loads
A.2.1 Actions due to railway operations
A.2.1.1 Vertical loads
21a: Load Model 71
Q vk = 250 kN 250 kN 250 kN 250 kN
q vk = 80 kN/m q vk = 80 kN/m
(1) 0.8m 1.6m 1.6m 1.6m 0.8m (1)
(1) no limitation
Fig. A.1. Load Model 71 and characteristic values for vertical loads, ENV 1991-3: 1995, Fig. 6.2
The factor is assumed to be 1.0.

According to ENV 1991:3 1995, 6.3.5.1 & 6.3.5.2 the wheel loads of the
Load Model 71 are distributed by the rails, sleepers and the ballast in
longitudinal direction to the upper surface of the concrete slab. (Figure A.2)
250/2 kN 3 x 250/2 kN 250/2 kN
250/4 kN 2 x 250/4 kN 2 x 250/4 kN 250/4 kN 80 kN/m

SO +- 0.00 m
-0.70 m
0.405m 0.41m 0.19m 0.19m

4:1
0.19m 0.19m 0.41m 0.405m
distributed Load Model UIC 71

304.9 kN/m
152.4 kN/m
152.4 kN/m
Fig. A.2. Distribution of the wheel loads of the Load Model 71 onto the upper surface of the
concrete slab
According to ENV 1991-3: 1995, 6.3.5.3 the actions are distributed

transversely to a total width of 2.77 meters (width of sleepers = 2.60 m,
thickness of ballast below sleepers = 0.34 m).
21b: Eccentric Load Model 71
An eccentricity is taken into account by distributing the total axle load by the ENV 1991-3: 1995,
ratio 1.25:1.00 on the two wheels. 6.3.1 (3)P
21c: Load Model unloaded train

qvk = 12.5 kN/m ENV 1991-3: 1995,
6.3.4
According to ENV 1991-3: 1995, 6.3.5.3 the actions are distributed
transversely to a total width of 2.77 meters (width of sleepers = 2.60 m,
thickness of ballast below sleepers = 0.34 m).
A-2
21d: non-public footpaths ENV 1991-3:1995,

2 6.3.6.1
On both sides of the bridge, 0.75 m wide qfk = 5 kN/m 0.75 m
qfk = 3.75 kN/m
A.2.1.2 Dynamic effects
(2,3): dynamic factors for load cases 22a, 22b and 22c
It was taken for granted that the first natural bending frequency of the bridge
loaded by permanent actions, n0, is within the limits of Figure 6.9 in ENV 1991-
3: 1995. In Section C.2 the relevant mode shape was found to be the second
natural mode shape of the bridge. Its eigenfrequency is:
n0 = 2.34 Hz
The criterion to fulfil is:
94.76 L-0.748 > n0 > 23.58 L-0.592 ; for 20m < L 100m
with L = 100m
3.02 Hz > 2.34 Hz >1.54 Hz
Within the limits of n0 there are no dynamic analysis, acceleration and
fatigue check at resonance required. The dynamic effects are considered by
using the dynamic factors with static analysis.
Assuming standard maintenance the factor 2 is chosen.
2.16
2 = + 0.73 ENV 1991-3: 1995,
L 0 .2
6.4.3.2 (1b)
for the arches: L = half span

ENV 1991-3: 1995,
with L = 50m 2 = 1.044
table 6.2 case 5.5
for the hangers: L = half span

with L = 50m 2 = 1.044
In ENV 1991-3: 1995, table 6.2 case 5.6, for suspension bars in
conjunction with stiffening girders a L of 4 times the longitudinal
spacing of the suspension bars is given. In this bridge there are no
stiffening girders, but the railway traffic loads are distributed by the
concrete tie. Additionally the spacing of the hangers is much smaller
than it is in the bridges that the Eurocode refers to. So the dynamic
effect on the hanger would be overestimated by the suggestion of
Eurocode. A reasonable L = half span was suggested by TVEIT.
for the concrete deck slab: L = twice span of deck slab ENV 1991-3: 1995,
with L = 19.9m 2 = 1.237 table 6.2 case 4.3
A.2.1.3 Horizontal forces characteristic values
21e: nosing force ENV 1991-3: 1995,

Acts horizontally at the top of the rails, perpendicularly to the centre-line of 6.5.2
the track and always combined with a vertical traffic load.
Qsk = 100 kN
A-3
21e: actions due to traction and braking ENV 1991-3: 1995,

6.5.3
These forces act at the top of the rails in longitudinal direction of the track
and will be combined with the corresponding vertical loads.
Traction force: Qlak = 33 kN/m La,b [m] 1000 kN
LM 71 La,b = 100 m Qlak = 1000 kN
Braking force: Qlbk = 20 kN/m La,b [m] 6000 kN
LM 71 La,b = 100 m Qlbk = 2000 kN
Most unfavourable forces: LM 71 Qla,bk,71 = 2000 kN

21g combined response of structure and track to variable actions
Assuming rail expansion devices at both sides of the bridge, there are no ENV 1991-3: 1995
actions due to thermal effects to be considered. 6.5.4.3 (3)
The requirement of considering longitudinal actions occurring as a result of ENV 1991-3: 1995
the vertical displacement of the bridge deck should be specified by the relevant 6.5.4.5
authority. It is taken for granted that this is not demanded.
All traction and braking forces are resisted by the fixed bearings because of ENV 1991-3: 1995
the assumed rail expansion devices provided at both ends of the deck. 6.5.4.4 (2)
Fbk = Qla,bk
21h: aerodynamic effects as a result of passing trains

The area of the structure subjected to a travelling wave of alternating
pressure and suction by the passing of rail traffic is very small. It is assumed
that actions caused by this effect are negligible.
A.2.2 Accidental actions
22a: derailment on bridges design situation I
centreline of track 0.52m 0.88m qA1d = 111.11 kN/m 2

l = 6.4 m
0.45m 0.45m
Fig. A.3. Equivalent load qA1d, according to ENV 1991-3: 1995, 6.7.1.2 (2) P
Since, in network arch bridges, structural elements above the levels of the DIN-Fachbericht 101,
rails could be damaged or destroyed by collision, lateral guideways and safety 3.4.8 [14]
catches shall be placed at a distance of 180 mm from the rail. So the lateral
evasion in derailment situation is limited to this value.
A-4
22b: derailment on bridges design situation II

centreline of track 0.88m qA2d = 177.78 kN/m 2
l = 20 m
0.45m
Fig. A.4. Equivalent load qA2d, according to ENV 1991-3: 1995, 6.7.1.2 (3) P
Since, in network arch bridges, structural elements above the levels of the DIN-Fachbericht 101,
rails could be damaged or destroyed by collision, lateral guideways and safety 3.4.8 [14]
catches shall be placed at a distance of 180 mm from the rail. So the lateral
evasion in a derailment situation is limited to this value.
22c: accidental action due to the severance of overhead line equipment
The catenary is connected to the lowest node of the wind bracing truss. The ENV 1991-3: 1995,
accidental force acts horizontally in a longitudinal direction with a design value 6.7.2
of twice 20 kN.
The most unfavourable sense of direction must be considered.
A.2.3 Assessment of traffic loads on railway bridges
The groups of loads are set up according to ENV 1991-3: 1995, 6.8 and
DIN - Fachbericht 101, 3.4.5 [14] separately for each factor 2.
gr 11: maximum vertical 1

= 1.044(LM 71)+Qla,bk+0.5Qsk for the arches
= 1.044(LM 71)+2000kN+50kN
= 1.044(LM 71)+Qla,bk+0.5Qsk for the hangers
= 1.044(LM 71)+2000kN+50kN
= 1.237(LM 71)+Qla,bk+0.5Qsk for the slab
= 1.237(LM 71)+2000kN+50kN
gr 12: maximum vertical 2
= 1.044(LM 71)+0.5Qla,bk+Qsk for the arches
= 1.044(LM 71)+1000kN+100kN
= 1.044(LM 71)+0.5Qla,bk+Qsk for the hangers
= 1.044(LM 71)+1000kN+100kN
= 1.237(LM 71)+0.5Qla,bk+Qsk for the slab
= 1.237(LM 71)+1000kN+100kN
gr 13: maximum longitudinal
matches gr 11
gr 14: maximum lateral
= 1.044(LM 71)+Qsk for the arches
= 1.044(LM 71)+100kN
= 1.044(LM 71)+Qsk for the hangers
= 1.044(LM 71)+100kN
A-5
= 1.237(LM 71)+Qsk for the slab

= 1.237(LM 71)+100kN
gr 21: maxium vertical 1 (two loaded tracks)
= 2(1.044(LM 71)+Qla,bk+0.5Qsk) for the arches
= 2(1.044(LM 71)+2000kN+50kN)
= 2(1.044(LM 71)+Qla,bk+0.5Qsk) for the hangers
= 2(1.044(LM 71)+2000kN+50kN)
= 2(1.237(LM 71)+Qla,bk+0.5Qsk) for the slab
= 2(1.237(LM 71)+2000kN+50kN)
gr 22: maximum vertical 2 (two loaded tracks)
= 2(1.044(LM 71)+0.5Qla,bk+Qsk) for the arches
= 2(1.044(LM 71)+ 1000kN+100kN)
= 2(1.044(LM 71)+0.5Qla,bk+Qsk) for the hangers
= 2(1.044(LM 71)+1000kN+100kN)
= 2(1.237(LM 71)+0.5Qla,bk+Qsk) for the slab
= 2(1.237(LM 71)+1000kN+100kN)
gr 23: maximum longitudinal (two loaded tracks)
matches gr 21
gr 24: maximum lateral (two loaded tracks)
= 2(1.044(LM 71)+Qsk) for the arches
= 2(1.044(LM 71)+100kN)
= 2(1.044(LM 71)+Qsk) for the hangers
= 2(1.044(LM 71)+100kN)
= 2(1.237(LM 71)+Qsk) for the slab
= 2(1.237(LM 71)+100kN)
A.2.4 Fatigue load models
The general design method, ENV 1991-3: 1995 F.2, is used for the fatigue
assessment.
For steel bridges the following condition must be satisfied:
c
Ff 2 71
Mf
The traffic loads to be considered for fatigue assessment are the following:
fatigue 1: =2LM 71= 1.044LM 71 for the arches
fatigue 2: = 2LM 71= 1.044LM 71 for the hangers
fatigue 3: = 2LM 71= 1.237LM 71 for the bridge deck
The load model has to be applied on one and/or both tracks.
A-6
A.3. Wind forces
A.3.1 Wind forces in y-direction (perpendicular to bridge ENV 1991-2-4: 1995,

centreline) 5.2
reference mean wind velocity pressure qref = /2v2ref,0

zone 1, altitude < 800 m = 0.625 kg/m3(27.6 m/s)2 ENV 1991-2-4: 1995,
qref = 476.1 kg/(ms2) Annex A.6
3a: lateral wind forces on the arches
average height: z = 30m
ENV 1991-2-4: 1995
force coefficient: cf = cf,0 = 2.01.0 = 2 10.6, 10.14
reference area: Aref = 0.448 m2/m
exposure coefficient: ce(z = 30m) = 2.6 (for terrain category III)
dynamic factor: cd (z = 30m) = 0.89
wind force: Fw1 = qrefcecdcfAref = 0.987 kN/m
attenuation of the wind force on the hind arch: Schneider [29], 6.1.5
= 0.1
a/d = 1
= 0.85
total lateral wind forces on the arches:

Fw = Fw1+ Fw1 = 1.826 kN/m
3b: lateral wind forces on the hangers ENV 1991-2-4: 1995

average height: z = 30m 8.1
roughness coefficient: cr (z = 30m) = 1.013 (for terrain category III) ENV 1991-2-4: 1995
topography coefficient: ct (z = 30m) = 1 (location: valley, no funnelling effects) 10.8.1
mean wind velocity: vm (z = 30m) = 27.96 m/s
b v m ( z = 30m)
Reynolds number. Re = = 7.5 10 4

cf,0 from figure 10.8.2 in ENV 1991-2-4: 1995
force coefficient: cf = cf,0 = 1.21.0 = 1.2

exposure coefficient: ce (z = 30m) = 2.6 (for terrain category III)
wind force: Fw1 = qrefcecdcfAref = 0.053 kN/m
attenuation of the wind force on the hind arch:

= 0.1 Schneider [29], 6.1.5
a/d = 1
= 0.85
A-7
total lateral wind forces on the hangers:

Fw = Fw1+ Fw1 = 0.098 kN/m
3c: lateral wind forces on the slab ENV 1991-2-4: 1995

10.11.2
average height: z = 20m
force coefficient: cf = cf,0 = 1.30.95 = 1.235 without traffic

force coefficient: cf = cf,0 = 1.750.775 = 1.313 with traffic
reference area: Aref = 1.5 m2/m without traffic

reference area: Aref = 4.9 m2/m with traffic
exposure coefficient: ce (z = 20m) = 2.25 (for terrain category III)

wind force: Fw = qrefcecdcfAref = 1.43 kN/m without traffic

Fw = qrefcecdcfAref = 6.134 kN/m with traffic
A.3.2 Crosswind forces in z-direction
3d: crosswind forces in z-direction ENV 1991-2-4: 1995

10.11.3
average height: z = 20 m
force coefficient: cf,z = 0.15 (d/b = 19.2)

exposure coefficient: ce (z = 20m) = 2.25
wind force in z-direction:Fw,z = qrefcecdcf,zAref = 1.645 kN/m
A.3.3 Longitudinal wind forces in x-direction
3e: longitudinal wind forces in x-direction ENV 1991-2-4: 1995

50 % of the wind forces in y-direction 10.11.4
total wind forces in y-direction: 585.26 kN without traffic

1055.66 kN with traffic
wind force in x-direction: Fw,x = 292.63 kN without traffic

527.63 kN with traffic
A.4. Actions due to temperature
4a: variation of the constant component of temperature ENV 1991-2-5: 1997,

6.1.3.3
The maximum variation causes longitudinal actions due to friction in the
bearings.
Mounting temperature: T0 = 10 C ENV 1991-2-5: 1997
Minimum of constant part of temperature: Te,min(Tmin=-24 C) = -17 C fig. 6.1, group 3
Maximum of constant part of temperature: Te,max(Tmax=37 C) = 40 C
A-8
Maximum negative variation: TN,neg = Te,min T0 = -27 K

Maximum positive variation: TN,pos = Te,max T0 = 30 K
Maximum variation: TN,max = 30 K
Enhancement of TN,max for the calculation of the horizontal displacement at the ENV 1991-2-5: 1997,
bearings. 6.1.3.3 (4)P & (5)P
TN,max = 40 K
Maximum longitudinal displacement: l = 1010-6 /K40 K100 m
l = 0.04 m
Forces due to friction in the 4 bearings: Qlfk = 4(Loadvertical) PTFE insert: in
Qlfk = 4(0.0510000kN) = 1280 kN PETERSEN [21], page
1157
4b: vertically linear variable component of temperature
The maximal negative temperature change (TM,neg) while the upper ENV 1991-2-5: 1997,
surface is colder than the lower one causes an additional vertical displacement 6.1.4.1
of the bridge.
This effect is considered for TM,neg = 1.0[-8 C] = 8 C
4c: horizontally linear variable component of temperature

This effect is considered for T = [5]C ENV 1991-2-5: 1997
6.1.4.2
4d: differences in the constant component of temperature of different
structural elements
A difference of 15 K between the slab and the arches with the hangers will be ENV 1991-2-5: 1997,
considered. 6.1.6
A.5. Combination of actions
The combinations are to be set up using the following formulas: ENV 1991-3: 1995
Annex G
Ultimate limit state:
1. Transient and persistent situations

j
G, j G k, j "+" P Pk + Q,1 Q k,1 "+" Q,i 0,i Q k,i
2. Accidental situations

j
GA, j G k, j "+" A d "+" P Pk + 1,1 Q k,1 "+" 2,i Q k,i ENV 1991-3: 1995
G2.1.2 (4)
It is assumed that the client demands the combination of LM71 with

the accidental actions.
For fatigue:
1. Minimal stress variation
G
j
k, j "+" P Pk + 2 LM71
2. Maximum stress variation

G
j
k, j "+" P Pk + twice 2 LM71
A-9
Serviceability limit state:

1. Characteristic/rare load combinations
G
j
k, j "+" Q k,1 " +" 0,i Q k,i
2. Infrequent load combinations

G
j
k, j "+" Pk + '1 Q k,1 "+" 1,i Q k,i
3. Frequent load combinations

G
j
k, j " +" Pk + 1,1 Q k,1 "+" 2,i Q k,i
4. Quasi-permanent load combinations

G
j
k, j "+" Pk + 2,i Q k,i
The load combinations used for the calculations are shown in figures A.5 to
A.8. Depending on the design checks, to obtain the most unfavourable load
position, the influence lines of structural elements were analysed.
The large number of load combinations is due to the requirements of
Eurocode. It is no problem, with present-day soft- and hardware, to calculate all
load combinations but the analysis of the results is very time-consuming. Since
for most design checks the same load combinations gave most unfavourable
results, it might be allowed to limit the number to the load combinations, found
to be decisive by experience.
A-10
Load combinations in ULTIMATE LIMIT STATE

Persistent and transient situations Accidental
Fatigue
load comb. Number traffic actions dominant wind forces dominant temp. situations
Description 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Selfweight 1a 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
arches - - - - - - - - - - - - - - - - -
Hangers 1b 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
hangers - - - - - - - - - - - - - - - - -
Bridge deck 1c 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
bridge deck - - - - - - - - - - - - - - - - -
Wind bracing 1d 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
wind bracing - - - - - - - - - - - - - - - - -
Ballast, rails, 1e 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
sleepers - - - - - - - - - - - - - - - - -
Selfweight 1f 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.00 1.00
cable trough - - - - - - - - - - - - - - - - -
Prestressing 1g 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
trans. tendons - - - - - - - - - - - - - - - - -
Prestressing 1h 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
long. tendons - - - - - - - - - - - - - - - - -
Load Model 71 21a 1.50 1.00 1.00 1.00
0.8 - - -
Eccentricity 21b 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.50 1.00 1.00 1.00 1.00 1.00
load model 71 - - - - - - - - - - 0.8 - - - - -
"unloaded train" 21c 1.50
1.0
Non-public 21d 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50
footpaths 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
Derailment, 22a 1.00
design situation I -
Derailment, 22b 1.00
design situation II -
Catenary rupture 22c 1.00
-
gr11 1.45 1.45 1.00
- 0.8 -
gr12 1.45 1.45
- 0.8
Groups of loads
gr14 1.45
-
gr21 1.45 1.45 1.00
- 0.8 -
gr22 1.45 1.45
- 0.8
gr24 1.45
-
Lateral wind on 3a 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50
arches 0.6 0.6 0.6 0.6 0.6 - - - - -
Lateral wind on 3b 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50
hangers 0.6 0.6 0.6 0.6 0.6 - - - - -
Lateral wind on 3c 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50 1.50
bridge deck 0.6 0.6 0.6 0.6 0.6 - - - - -
Longitudinal wind 3e 1.50
0.6
Fig. A.5. Load combinations for assessment in ultimate limit state, and for fatigue design checks
A-11
SERVICEABILITY LIMIT STATE - characteristic/rare load combinations

Load comb. number traffic actions dominant w ind forces dominant temp.
Description 18 19 20 21 22 23 24 25 26 27 28 29
Selfw eight 1a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
arches - - - - - - - - - - - -
Hangers 1b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
hangers - - - - - - - - - - - -
Bridge deck 1c 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
bridge deck - - - - - - - - - - - -
Wind bracing 1d 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
w ind bracing - - - - - - - - - - - -
Ballast, rails, 1e 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
sleepers - - - - - - - - - - - -
Selfw eight 1f 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
cable trough - - - - - - - - - - - -
Prestressing 1g 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
trans. tendons - - - - - - - - - - - -
Prestressing 1h 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
long. tendons - - - - - - - - - - - -
Load model 71 21a -
0.8
Eccentricity 21b 1.00 1.00 1.00 1.00 1.00 1.00 - - - - -
load model 71 - - - - - - 0.8 0.8 0.8 0.8 0.8
"unloaded train" 21c 1.00
-
Non-public 21d - - - - - - - - - -
footpaths 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
gr11 1.00 -
- 0.8
gr12 1.00 -
- 1.0
Groups of loads
gr14 1.00
-
gr21 1.00 -
- 0.8
gr22 1.00 -
- 1.0
gr24 1.00
-
Lateral w ind on 3a - - - - - 1.00 1.00 1.00 1.00 1.00
arches 0.6 0.6 0.6 0.6 0.6 - - - - -
Lateral w ind on 3b - - - - - 1.00 1.00 1.00 1.00 1.00
hangers 0.6 0.6 0.6 0.6 0.6 - - - - -
Lateral w ind on 3c - - - - - 1.00 1.00 1.00 1.00 1.00
bridge deck 0.6 0.6 0.6 0.6 0.6 - - - - -
Longitudinal w ind 3e -
0.6
Const. temp. 4a - - - - - - - - - - - 1.00
change 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 -
Fig. A.6. Load combinations for assessment in serviceability limit state, characteristic/rare
combinations
A-12
SERVICEABILITY LIMIT STATE

Infrequent load combinations frequent load combinations
traffic actions w ind forces traffic actions w ind forces
Load comb. number temp. temp.
dominant dominant dominant dominant
Description 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
Selfw eight 1a 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
arches - - - - - - - - - - - - - - - -
Hangers 1b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
hangers - - - - - - - - - - - - - - - -
Bridge deck 1c 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
bridge deck - - - - - - - - - - - - - - - -
Wind bracing 1d 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
w ind bracing - - - - - - - - - - - - - - - -
Ballast, rails, 1e 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
sleepers - - - - - - - - - - - - - - - -
Selfw eight 1f 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
cable trough - - - - - - - - - - - - - - - -
Prestressing 1g 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
trans. tendons - - - - - - - - - - - - - - - -
Prestressing 1h 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
long. tendons - - - - - - - - - - - - - - - -
Load model 71 21a -
0.8
Eccentricity 21b - - - - - - - - - - -
load model 71 1.0 1.0 1.0 1.0 0.8 0.6 0.8 0.8 0.8 0.6 0.6
Non-public 21d - - - - - - -
footpaths 0.5 0.5 0.5 0.5 0.5 0.5 0.5
gr11 - - -
1.0 0.8 0.8
Groups of loads
gr14 - -
1.0 0.8
gr21 - - -
1.0 0.6 0.6
gr24 - -
1.0 0.6
Lateral w ind on 3a - - - - - - - - - - - -
arches 0.5 0.5 0.5 0.6 0.6 0.6 0.0 0.0 0.0 0.5 0.5 0.5
Lateral w ind on 3b - - - - - - - - - - - -
hangers 0.5 0.5 0.5 0.6 0.6 0.6 0.0 0.0 0.0 0.5 0.5 0.5
Lateral w ind on 3c - - - - - - - - - - - -
bridge deck 0.5 0.5 0.5 0.6 0.6 0.6 0.0 0.0 0.0 0.5 0.5 0.5
Longitudinal 3e - -
w ind 0.6 0.5
Const. temp. 4a - - - - - - - 0.80 - - - - - - - -
change 0.6 0.6 0.6 0.6 0.6 0.6 0.6 - 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6
Fig. A.7. Load combinations for assessment in serviceability limit state, infrequent and frequent combinations
A-13
SERVICEABILITY LIMIT STATE Additional load

quasi-permanent load combinations regarding combinations
load combinations deformations and vibrations
rotation at the end of horizontal
deck tw ist the deck/maximum deflection of temp. effects w ind lift
Load comb. number vertical deflection the deck
Description 46 47 48 49 50 51
Selfw eight 1a 1.00 1.00 1.00 1.00 1.00 1.00
arches - - - - - -
Hangers 1b 1.00 1.00 1.00 1.00 1.00 1.00
hangers - - - - - -
Bridge deck 1c 1.00 1.00 1.00 1.00 1.00 1.00
bridge deck - - - - - -
Wind bracing 1d 1.00 1.00 1.00 1.00 1.00 1.00
w ind bracing - - - - - -
Ballast, rails, 1e 1.00 1.00 1.00 1.00 1.00 1.00
sleepers - - - - - -
Selfw eight 1f 1.00 1.00 1.00 1.00 1.00 1.00
cable trough - - - - - -
Prestressing 1g 1.00 1.00 1.00 1.00 1.00 1.00
trans. tendons - - - - - -
Prestressing 1h 1.00 1.00 1.00 1.00 1.00 1.00
long. tendons - - - - - -
Load model 71 21a - - - -
- 1.0 1.0 1.0
Eccentricity 21b - - -
load model 71 1.0 1.0 1.0
Non-public 21d one side one side
footpaths 0.5 0.5
Nosing force 21e - - -
0.8 0.8 0.8
Lateral w ind on 3a - - -
arches 0.5 0.5 0.5
Lateral w ind on 3b - - -
hangers 0.5 0.5 0.5
Lateral w ind on 3c - - -
Bridge deck 0.5 0.5 0.5
Crossw ind in 3d 1.00
z-direction -
Const. temp. 4a -
change 0.5
Vertically linear 4b -
temp. change 0.6
Horizontally linear 4c -
temp. change 0.6
Different
temperature 4d 1.00
in different
structural
elements -
Fig. A.8. Load combinations for assessment in serviceability limit state, quasi-permanent combinations,
combinations regarding deformations and vibrations and additional load combinations
A-14
Annex B
Preliminary design
B.1 Bridge deck
The preliminary design is be based on serviceability limit states, where

ENV 1992-2: 1996
category C is applicable to satisfy durability requirements (Annex D,
Table 4.118
Section D.5.2.2). Thus, the quasi-permanent combination of actions is relevant
for the decompression, where no concrete tensile stresses are allowed.
However, since the quasi-permanent combination does not include the Load
Model 71 it is desirable to carry out the preliminary design with loads of the
frequent load combination of actions applicable for the limit state of cracking. It
was decided for load combination 40 with the following components and partial
safety and combination factors (Annex A, Figure A.7):
- Self-weight: = 1.0
- Prestressing = 1.0
- Live load: 2 LM 71, = 0.6
- Constant temperature change: = 0.5
For the reason of approximation the constant temperature change is not
considered.
B.1.1 Transverse direction
Statical system and loading

0.79 1.375 3.0 1.4 3.0 1.375 0.79
0.875 8.4 0.875
load model 71
0.6 x
52 kN/m
ballast load
12.5 kN/m
3.75 kN/m dead load deck 3.75 kN/m

10.75 kN/m
0.79 10.15 0.79
Fig. B.1: Statical system and loading in transverse direction
B-1
Diploma Thesis - Brunn & Schanack Annex B: Preliminary design
Internal forces and moments
For the purpose of preliminary design it is desirable to determine concrete

stresses at the mid-span and at the quarter point. The respective internal forces
are calculated in the following.
Dynamic factor: 2 = 1.237 Annex A,

Section 2.1.2
10.15
1. Internal moments at x = m = 2.5375m (quarter point)
4
Bending moment due to dead load
kN kN
Dead load of footpath: 3.75 0.8m = 3
m2 m
kN kN
Dead load of deck between hangers: 10.75 2
10.15m = 110
m m
kN 110 kN kN
Reaction: 3 + = 58
m 2 m m
kN (2.5375m)
2
kN kN 0.8
m y = 58 2.5375m 3 + 2.5375 m 10.75 2
m m 2 m 2
kNm
m y = 104
m
Bending moment due to ballast load
kN (2.5375m 0.875m)
2
105 kN kNm
my = 2.5375m 12.5 2 = 115.9
2 m m 2 m
Bending moment due to LM 71
This bending moment results from the uniformly distributed load

80 kN/m. The bending moment resulting from the concentrated train load is
determined below (Figure B.4, Figure B.5).
0.6 80 kN (2.5375m 1.375m)

2
kN kNm
m y = 0.6 80 2.5375m 2
= 111
m 3 m 2 m
kNm kNm
m y = 2 111 = 137.3
m m
B-2
10.15
2. Internal moments at x = m = 5.075m (centre point)
2
Bending moment due to dead load
kN (5.075m )
2
kN kN 8
m y = 58 5.075m 3 + 5.075 m 10.75 2
m m 2 m 2
kNm
m y = 140
m
Bending moment due to ballast load
kN (5.075m 0.875m)
2
105 kN kNm
my = 5.075m 12.5 2 = 156.2
2 m m 2 m
Bending moment due to LM 71
kN kN
m y = 0.6 80 5.075m 0.6 80 (5.075m 1.375m 1.5m)
m m
m y = 138 kNm m
kNm kNm
m y = 2 138 = 171
m m
Distribution of concentrated train load of LM 71

For simplification in respect of the preliminary design, it is assumed that the
80 kN/m of the load model 71 are infinitely long without a gap. Hence, this load
is only distributed in transverse direction. The additional 76 kN/m within the 6.4
m range are distributed in transverse as well as longitudinal direction. This 2-
axial load distribution and the resulting bending moment is accounted for with
the help of the influence surface charts by PUCHER [26].
It was found by
TEICH & WENDELIN [38] that for the
transverse direction, results according to
PUCHER [26] are conservative compared
arch plane
to results determined with a 3D model.
Since the calculation of internal forces
2.875
will be carried out with the help of a 3D

0.6 x 250 kN
model in SOFiSTiK, this preliminary 0.6 x 80 kN/m
design is on the safe side.

2.200
7.170
4x
10.150
sleepers
2.200
4.40
0.6 x 250 kN
0.6 x 80 kN/m
2.60 2.60
4x
2.875
sleeper d = 0.2
4:1
ballast d = 0.34
2.77 2.77 6.4

7.17 arch plane
Fig. B.2. Load model 71: distribution in Fig. B.3. Load model 71: Distribution
transverse direction, dimensions in [m] in plane, dimensions in [m]
B-3
The bending moments in transverse direction are calculated at the quarter

point and mid-span in figures B.4 and B.5, respectively.
Quarter point: Evaluation with chart 3, PUCHER [26]

a [m] 6.4
Loaded area
b [m] 7.17
Width of the deck L [m] 10.15

Train load of LM 71 0.6 pTL [kN/m] 45.6
2
Distributed train load p [kN/m ] = 0.6 2 pTL / b 12.72
Ordinates of patches on 1/4 of loaded area: h [-]
0.5000 0.5000 0.5000 0.5200 0.5000
0.7000 0.7300 0.7500 0.8000 0.8200
1.0000 1.0000 1.0500 1.1000 1.2000
1.4500 1.4500 1.4500 1.5000 1.5000
1.7000 1.8000 1.8000 1.8000 1.8000
Sum of ordinates h [-] 27.9200
Area of a single patch A1 [-] = a / (10L) b/ (10L) 0.0045
Volume under 1/4 of loaded area V1/4 [-] = A1 h 0.1244

Bending moment my [kNm/m] my [kNm/m] = 4 p V1/4 L2 / (8 ) 25.9
Fig. B.4. Determination of bending moment at quarter point due to train load
Mid-span: Evaluation with chart 1, PUCHER [26]
a [m] 6.4
Loaded area
b [m] 7.17
Width of the deck L [m] 10.15

Train load of LM 71 per track 0.6 pTL [kN/m] 45.6
2
Distributed train load p [kN/m ] = 0.6 2 pTL / b 12.72
Ordinates of patches on 1/4 of loaded area: h [-]
1.20 1.25 1.30 1.40 1.45
1.80 1.90 2.00 2.00 1.90
2.50 2.70 2.80 2.80 2.60
2.80 2.90 3.80 3.50 3.10
6.00 5.20 4.50 3.90 3.40
Sum of ordinates h [-] 68.7000
Area of a single patch A1 [-] = a / (10L) b/ (10L) 0.0045
Volume under 1/4 of loaded area V1/4 [-] = A1 h 0.3060

2
Bending moment my [kNm/m] my [kNm/m] = 4 p V1/4 L / (8 ) 63.8
Fig. B.5. Determination of bending moment at mid-span due to train load
B-4
Materials
Concrete C80/60
ENV 1992-1-1 Table 3.1, 3.2
fck = 50 N/mm2 Characteristic cylinder compressive strength

2
fctm = 4.1 N/mm Mean value of the tensile strength
2
Ecm = 37000 N/mm Value of the secant modulus of elasticity
Reinforcing steel S 500
fyk = 500 N/mm2 Characteristic yield strength

Es = 200000 N/mm2 Modulus of elasticity DYWIDAG-SYSTEMS
INTERNATIONAL [10]
Prestressing DYWIDAG Bonded Post-Tensioning
thread bars 36D Systems
fp0.1k = 1080 N/mm2 Characteristic 0.1% proof-stress

2
fpk = 1230 N/mm Characteristic tensile strength
Py = fp0.1k Ap = 1099 kN Yield load
Pu = fpk Ap = 1252 kN Ultimate load
2
Ap = 1018 mm Nom. cross-sectional area
D = 36 mm Nominal diameter
k = 0.3 /m Wobble factor
= 0.5 Friction factor
min f = 92 mm Minimum centre distances
I.D. = 51 mm Internal diameter of sheathing
O.D. = 57 mm External diameter of sheathing
Rmin = 12.1 m Minimum radius for elastic bending
57.0
51.0
Concrete cover
36.0
28.5 7.5
Reinforcing steel: nom c = 40 mm

Prestressing thread bars: nom c = 67 mm Annex D,
103.0
Section D.5.1
67.0
40.0
reinforcement bars
concrete face
Fig. B.6. Concrete cover to thread

bars, dimensions in [mm]
B-5
Prestressed concrete section properties
zic cc d h
zcp
ci
zip zc
Ap
zp
Fig. B.7. Nomenclature for prestressed concrete cross section
The relevant geometric values for calculating stresses at the top and bottom
level of the prestressed concrete section as well as at the tendon level are
calculated as follows:
= Es / Ecm (Equ. B.1 ) Elastic modulus ratio
zc = h / 2 (Equ. B.2) Distance between centroid of concrete

section and bottom level
zcp = zc - zp (Equ. B.3) Distance between centroid of the

concrete section and the thread bar
zic = ( 1) Az zcp / Ai (Equ. B.4) Distance between centroid of concrete

section and centroid of composite
section
zip = zcp zic (Equ. B.5) Distance between the thread bar and
the centroid of the composite section
Ii = Ic + ( 1) Ap zcp zip (Equ. B.6) Second moment of area of the

composite section
Wit = Ii / (zc zci - h) (Equ. B.7 ) First moment of area of the composite
section at top level
Wib = Ii / (zc zci) (Equ. B.8) First moment of area of the composite
section at bottom level
Wip = Ii / zip (Equ. B.9) First moment of area of the composite

section at the prestressing level
B-6
Determination of prestressing force

ENV 1992-1-1: 1991
The calculations of prestressing forces and concrete stresses were carried Section 2.5.4.2
out by using a Microsoft Excel spreadsheet, Figure B.8. The formulas which
were used are given and explained in the following. Reference is made by
numbered equations.
The mean value of the prestressing force is given by:
Pm. = Po - P Pt(t = ) (Equ. B.10)

where:
Pm. is the mean value of the prestressing force at time t =

Po is the initial force at the active end of the thread bar immediately
after stressing
P is the loss due to friction
Pt(t = ) is the loss due to creep, shrinkage and relaxation at time t =
Thread bars are not subjected to nut draw-in.
Initial prestressing force
Maximum force applied to the thread bar immediately after stressing at the ENV 1992-1-1:1991
stressing end: Po 0.8 Ap fpk ( 0.9 Ap fp0.1k) Section 4.2.3.5.4 (P2)
Maximum prestressing force immediately after tensioning: ENV 1992-1-1: 1991

Po 0.75 Ap fpk ( 0.85 Ap fp0.1k) Section 4.2.3.5.4 (P3)
Loss of prestress
Loss of prestress in the thread bar due to friction ENV 1992-1-1: 1991
Section 4.2.3.5.5
(
P (x ) = Po 1 e ( +k x ) ) (Equ. B.11)
where:
is the coefficient of friction between thread bar and duct

is the sum of the angular displacements over a distance x
(irrespective of direction or sign)
k is an unintentional angular displacement (per unit length)
related to the profile of the thread bar
x horizontal distance from stressing end
Prestressing force after the effect of friction: P = Po - P (Equ. B.12)
B-7
Loss of prestress due to nut draw-in
Thread bars are especially suitable for short tendons, because they result in DYWIDAG-SYSTEMS
very low seating loss. The threadability of the anchor nut offers the advantage INTERNATIONAL [10]
of adjusting the prestress force in the tendon at any given time before grouting.
Time dependent losses due to creep, shrinkage and relaxation
Pt = p.c + s + r A p =
(
s. Es + pr + (t = .t o ) cg + cpo )A (Equ. B.13)
p
A p.m A 2
1+ 1 + c zcp (1 + 0.8 (t.t o ))
Ac Ic

where:
p.c+s+r is the variation of stress in the thread bars due to creep, shrinkage
and relaxation at location x, at time t =
S. is the estimated shrinkage strain at time t = taken from

ENV 1992-1-1: 1991 Table 3.4
= Es / Ecm Elastic modulus ratio
pr is the variation of stress in the thread bar at section x due to

relaxation, which is derived from ENV 1992-1-1: 1991 Figure 4.8,
for a ratio of Initial stress/Characteristic tensile stress p / fpk
calculated from: p = 0.85 pgo (Equ. B-14)
pgo is the initial stress in the thread bar due to prestress and
permanent action with:
Po P
pgo = (Equ. B.15)
Ap
The long term values of the relaxation losses are assumed to be
three times the relaxation losses after 1000 h:
pr. = 3 pr (Equ. B-16) ENV 1992-1-1: 1991
Section 4.2.3.4.1.
(t = ,to) is the creep coefficient according to ENV 1992-1-1: 1991 Table 3.3
with the notional size 2 Ac / u
where:
Ac is the cross-sectional area of concrete
u is the perimeter of that area exposed to environmental
conditions
cg is the stress in the concrete adjacent to the thread bars, due to

self-weight and any other permanent actions (ballast load)
m y.SW m y.B
cg = + (Equ. B.17)
Wp Wip
cpo is the initial stress in the concrete adjacent to the thread bars,
due to prestress
B-8
cpo =
(Po P ) (Po P ) zcp (Equ. B.18)
Ac f Wp f
Ap.m is the area of the prestressing bars at the level being considered
Ac is the area of the concrete section
Ic is the second moment of area of the concrete section
zcp is the distance between the centre of gravity of the concrete section
and the thread bar
According to the serviceability requirement for the limit state of cracking the
calculated minimum reinforcement is sufficient if the concrete stresses do not
exceed the mean tensile strength, which is 4.1 N/mm2 for C50/60 (Annex D.5,
Figure D.21: Requirements). This requirement will be the basis for the
preliminary design. Moreover, the concrete compressive stress shall be limited
to 22.5 N/mm2 at the time directly after prestressing (Annex D.5, Figure D.21:
Requirements).
It is desired to provide a margin covering unconsidered strains such as the
temperature change.
The spreadsheet in Figure B.9 provided the possibility to change the
distance between the thread bars. The concrete stress at the mid-span bottom
fibre appears to be decisive and a value of 0.83 N/mm2 (tension) led to a bar
distance of 270 mm.
It was necessary to split the calculation of friction losses at the quarter
span, since two adjacent thread bars have different geometries (Figure B.8).
Each thread bar is only stressed from one end. The dead ends lie at a higher
location than the stressed ends of the neighbouring bars.
For one and the same quarter point, the thread bars stressed from the left
experience different friction losses than the bar stressed from the right. Because
the prestressing is calculated for one metre width, the mean value of the friction
losses of two adjacent thread bars was calculated.
x
quarter point mid-span

x = 2.5375 m x = 5.075 m
plane
arch
Transverse tendons, DYWIDAG threadbar 36D, s = 270
Fig. B.8. Transverse prestressing with relevant sections for preliminary design
B-9
Material properties and factors

Compressive strength of C50/60 fck [N/mm2] 50
Tensile strength of C50/60 fctm [N/mm2] 4.1
Secant modulus of elasticity Ecm [N/mm2] 37000
2
Elastic modulus of steel Es [N/mm ] 200000
Elastic modulus ratio Equ. B.1 5.41
Cross-sectional area per thread ba [cm2] 10.18
Yield load fp0.1k A [kN] 1099
2
Yield stress fp0.1k [N/mm ] 1080
Ultimate load fpk A [kN] 1252
Ultimate stress fpk [N/mm2] 1230
Sheathing tube inner diameter I.D. [mm] 51
Sheathing tube outer diameter O.D. [mm] 57
Minimum radius for elastic bending Rmin [m] 12.1
Wobble factor k 0.0052
Friction factor 0.5
Prestressing
Allowable prestress force P0 [kN] = 0.9 fp0.1k A 989.1
P0 [kN] = 0.8 fpk A 1001.6
P0 [kN] = 0.85 fp0.1k A 934.15
P0 [kN] = 0.75 fpk A 939
Critical value P0 [kN] 934.15
Spacing of thread bars f [m] 0.27
Cross-sectional area of prestressing [m2/m] 0.0037704
Internal moments coordinate x 2.5375 5.075

Moment due to dead load my.DL [kNm/m] 104 140
Moment due to ballast my.B [kNm/m] 116 156.2
Moment due to LM 71 UDL load 80 kN/m my.LM71.80 [kNm/m] 137.3 171
Moment due to concentrated train load my.LM71.Train [kNm/m] 26 64
Moment due to complete LM 71 my.LM71 [kNm/m] 163.3 235
Section properties
Depth of the deck at distance x d [m] 0.3914 0.4300
Dstance: thread bar - bottom level zp [m] 0.1030 0.1030
2
Area of concrete section Ac [m /m] 0.3914 0.4300
2nd moment of area of concrete section Ic [m4] 0.0050 0.0066
1st m. of area at top of concrete section W t [m3] -0.0255 -0.0308
1st m. of area at bottom of concrete section W b [m3] 0.0255 0.0308
st
1 m. of area at thread bar level W p [m3] 0.0539 0.0592
Virtual area of composite section Ai [m2/m] 0.4080 0.4466
Dist.: centroid conc. sec. - bottom level zc [m] Equ. B.2 0.1957 0.2150
Dist: centroid of conc. sec. - thread bar zcp [m] Equ. B.3 0.0927 0.1120
Dist.: centroids conc. sec. - comp. sec. zci [m] Equ. B.4 0.0038 0.0042
Dist.: centroid comp. sec. - thread bar zip [m] Equ. B.5 0.0889 0.1078
nd
2 moment of area of composite section Ii [m4] Equ. B.6 0.0051 0.0068
st 3
1 m. of area at top of composite section W it [m ] Equ. B.7 -0.0257 -0.0311
1st m. of area, bottom of composite section W ib [m3] Equ. B.8 0.0267 0.0324
st
1 m. of area, thread bar level, composite W ip [m3] Equ. B.9 0.0577 0.0633
Fig. B.9. Preliminary design of bridge deck in transverse direction, continued below
B-10
Friction coordinate x: 2.5375 7.6125 5.075 5.075

Sum of angular displacements [] 7.363 12.235 7.363 12.235
P [kN] Equ. B.11 63.94 111.17 69.70 105.68
Prestress loss due to friction mean value P 87.55 87.69
P [kN] Equ. B.12 870.21 822.98 864.45 828.47
Prestressing force after friction mean value P 846.60 846.46
Prestress per metre width at time t = 0 [kN/m] 3135.55 3135.04

2
Thread bar stress [N/mm ] 831.63 831.49
Relaxation
Initial stress in thread bar due to P pgo [N/mm2] Equ. B.15 831.63 831.49
2
p [N/mm ] Equ. B.14 706.89 706.77
p / fpk [%] 56.46 56.45
ENV 1992-1-1:1991 Figure 4.8 relaxation [%] 1.50 1.50
2
Relaxation loss after 1000h pr [N/mm ] 10.60 10.60
Relaxation loss at time t= pr, [N/mm2] Equ. B.16 31.81 31.80
Creep, shrinkage, relaxation losses

Notional size [mm] = 430 u,28 = 1.50 shrinkage coefficient cs,u = 0.00028
Conc. stress adjacent to bars (DL) cg [N/mm2] Equ. B.17 3.94 4.83
2
Initial conc. stress adjacent to bars cpo [N/mm ] Equ. B.18 -13.05 -12.84
Prestress loss at time t= Pt [kN] Equ. B.13 -138.09 -130.72
Prestressing force after friction, relaxation, creep, shrinkage

Prestress in thread bar at time t= Pm,t [kN] Equ. B.10 708.50 715.74
Prestress per [m] width at time t= Pm,t [kN/m] 2624.09 2650.90
Thread bar stress [N/mm2] 695.98 703.09
Primary strain pm 0.003480 0.003515
Concrete stresses at frequent combination of actions

At time t = 0
Concrete stress due to prestressing
top [N/mm2] 3.37 4.10
bottom [N/mm2] -19.40 -18.68
Concrete stress due to dead load
top [N/mm2] -4.07 -4.54
bottom [N/mm2] 4.07 4.54
Sum of concrete stress
top [N/mm2] -0.70 -0.44
bottom [N/mm2] -15.32 -14.14
At time t =
Concrete stress due to prestressing
top [N/mm2] 2.82 3.47
2
bottom [N/mm ] -16.23 -15.80
Concrete stress due to dead load
top [N/mm2] -8.58 -9.56
2
bottom [N/mm ] 8.41 9.37
Concrete stress due to load model 71
top [N/mm2] -6.35 -7.55
2
bottom [N/mm ] 6.11 7.26
Sum of concrete stress
top [N/mm2] -12.10 -13.63
2
bottom [N/mm ] -1.72 0.83
Fig. B.9. Preliminary design of bridge deck in transverse direction
B-11
B.1.2 Longitudinal direction
The lower chord is mainly subjected to axial load. Bending moments are
secondary and will therefore be neglected in the preliminary design. However,
an additional prestress of 30% will be applied to compensate for the influence of
bending moments.
According to TVEIT [41], the axial force in the lower chord at a distance x
from one support due to a uniformly distributed load is to be calculated as
follows:
w (l x ) x 1
Ku = l w cot 2 v h (Equ. B.19)
2h 2
w uniformly distributed load per unit length of lane
l = 100 m length of the span
h = 17 m rise of the arch
vh = 60 average angle of the hangers crossing a vertical line
at a distance x from the support
The uniformly distributed load w is calculated further below.
Section properties
Z
CL
5.86
2.15
1.66
1=9 8
2 7 6
3
0.62
0.59
0.47
0.46
0.43
0.37
4 5
y
0.57
5.86
Fig. B.10. Cross section with numbered points assisting the calculation in Figure B.11
The numbering of the corner points in Figure B.10 corresponds to the first
column (i) in Fig. B.11.
Calculation of cross section properties

A B C AC D E = A (C2 - D)
i yi zi yi zi+1 - zi yi+1 yi + yi+1 zi + zi+1 zi zi+1
1 0 0.62 0.00 0 1.02 0 0.248 0
2 0 0.4 -0.228 0.57 0.86 -0.1961 0.184 -0.1266768
3 0.57 0.46 -0.2622 1.14 0.46 -0.1206 0 -0.05548152
4 0.57 0 0 6.43 0 0 0 0
5 5.86 0 2.5198 11.72 0.43 1.0835 0 0.46591102
6 5.86 0.43 1.2437 8.01 0.8 0.995 0.1591 0.59809533
7 2.15 0.37 0.6113 3.81 0.94 0.5746 0.2109 0.41122151
8 1.66 0.57 1.0292 1.66 1.19 1.2247 0.3534 1.09373084
9 0 0.62 0 0 1.24 0 0.3844 0
Ac = 0.5 A = 2.46 Sy = (A C) / 6 = 0.5935
eb = Sy / Ac = 0.242 Iy* = 1/12 E = 0.1989
units: [m], [m ], [m ] 2 4
Iy = Iy* - Ac eb2 = 0.0555
Fig. B.11. Calculation of section properties of one half of the bridge deck
B-11
B-12
The cross section properties calculated in Figure B.11. refer to one half of the
bridge deck.
Summary of cross section areas for one half of the bridge deck:
Ac = 2.46 m2
Iy = 0.0555 m4
The tendon geometry is symmetrical to the centroid of the concrete section.
Statical system and loading
The preliminary design is to determine the required prestressing force for one
edge beam. Therefore, only one half of the bridge loading is considered.
Dead load
Self-weight of prestressed concrete deck
Cross section area of one half of the concrete deck: A = 2.46 m2
kN kN
g deck = 25 3
2.46m 2 = 62
m m
Self-weight ballast, sleepers, rails

Annex A Load Assumptions
105 kN m
g track = = 52.5 kN m
2
Self-weight arch
For preliminary design purposes, one and the same profile will be assumed
for the whole arch length.
Arch profile: American wide flange W 360 x 410 x 990
Weight: 990 kg/m
Approximate arch length: 107m
g arch = 990 kg m 9.81m s 2 107m 100m = 10.4kN m
Total dead load
g k = g deck + g track + g arch

gk = (62 + 52.5 + 10.4 )kN m = 125 kN m
Total dead load without ballast load
gk.no ballast = (62 + 10.4 )kN m = 72.4 kN m
B-13
Live load
Load model 71
Uniformly distributed load:

q 71.UDL = 80 kN m
Concentrated train load:

Concentrated loads are converted into evenly distributed loads by
multiplying by 2 and dividing by the length of the influence line. This is because
the maximum ordinate of an influence line is usually nearly twice as big as the
average ordinate, TVEIT [47], page 5.
76 kN m 6.4m 2 kN
q 71.train = = 9.73
100m m
Total LM 71:
Multiplying by = 0.6 (frequent combination of actions)
q 71 = 0.6 (q 71.UDL + q 71.train ) = 53.8 kN m
Applying the dynamic factor:

q 71 = q 71 = 1.044 53.8 = 56.2 kN m
Total
w = gk + qk = (125 + 56.2)kN m = 181.2 kN m
Axial force in lower chord due to self-weight without ballast

kN
72.4 (100m 50m) 50m
m 1 kN
K u.DL = 17m 72.4 cot 2 60 = 5118kN
2 17m 2 m
Axial force in lower chord due to ballast load

kN
52.5 (100m 50m) 50m
m 1 kN
K u.B = 17m 52.5 cot 2 60 = 3711.5kN
2 17m 2 m
Axial force in lower chord due to live load

kN
56.2 (100m 50m) 50m
m 1 kN
K u.LL = 17m 56.2 cot 2 60 = 3973kN
2 17m 2 m
Axial force in lower chord due to total load

K u = K i = 12802kN
B-14
Materials
Concrete C50/60 ENV 1992-1-1

Table 3.1, 3.2
fck = 50 N/mm2 Characteristic cylinder compressive strength
fctm = 4.1 N/mm2 Mean value of the tensile strength
Ecm = 37000 N/mm2 Value of the secant modulus of elasticity
Reinforcing steel S 500
fyk = 500 N/mm2 Characteristic yield strength

Es = 200000 N/mm2 Modulus of elasticity
Prestressing strands DYWIDAG Bonded Post-Tensioning DYWIDAG-SYSTEMS

type 15mm (prEN 10138) Systems INTERNATIONAL [10]
fp0.1.k = 1500 N/mm2 Yield strength

fpk = 1770 N/mm2 Ultimate strength
Ap = 150 mm2 Cross-sectional area
Pu = fpk Ap = 265 kN Ultimate load
D = 15.70 mm Nominal diameter
Es = 195000 N/mm2 Modulus of elasticity
max.2.5% Relaxation after 1000h at 0.7 Pu
B-15
B-
Required number of strands
The axial forces in the tie result from loads acting on the bridge deck. These
loads are transferred through the hangers into the arch, resulting in a horizontal
thrust which is constrained by the tie. This leads to high axial forces in the tie.
Bending moments result from local effects of loads on the bridge deck
between constraints given by the hanger nodes. Furthermore, bending
moments occur due to deformation of the whole bridge structure. Strains due to
the axial forces are significantly higher than strains caused by the bending
moments.
For the preliminary design purpose the concrete section was intended to be
in compression after all prestress losses and due to axial loads, leaving a
margin for the not considered but minor bending moments in the tie. With the
concrete left in compression, the number of strands and tendons was
determined with the help of a Microsoft Excel spread sheet, Figure B.12.
STEIMANN [37] used tendons with 27 strands making 6 tendons per edge
beam necessary. It was decided to use the same tendon type:
Tendon type with 27 strands of type 15mm: Type 6827 with corrugated DYWIDAG-SYSTEMS
sheathing INTERNATIONAL [10]
I.D. = 110 mm Internal diameter of sheathing
O.D. = 118 mm External diameter of sheathing
Min f = 198 mm Minimum centre distances
k = 0.3/m Wobble angle
= 0.2 Friction factor
The equations used for the preliminary design for longitudinal direction
(Figure B.12) are basically the same as used for the transverse direction.
Differences are explained in the following.
Loss due to wedge draw-in
In contrast to the thread bars used in transverse direction, the prestressing

of tendons in longitudinal direction experiences losses due to wedge draw-in.
The slip at the stressing anchorage has to be taken to ln = 2 mm, according to
DYWIDAG SYSTEMS INTERNATIONAL [10]. Wedge draw-in has effects within the
influence length only. Since the effect of friction is largest at mid-span, this will
be the decisive concrete section. The influence length of the nut draw-in does
not reach the mid-span and therefore does not need to be considered.
Nevertheless, the equations are given in case it is desired to determine the
prestressing force at a different location.
B-16
Psl = Po (1 2 l e 1 ) (Equ. B.20) Prestress loss due to

wedge draw-in
ln E s A p
le = (Equ. B.21) Influence length
Po 1
ln nut draw-in in [mm]

1 average angle of deflection along the influence length le
of tendon behind the stressing end [rad/m]
coefficient of friction
Time dependent losses due to creep, shrinkage and relaxation
Creep and shrinkage coefficients
The notional size for the longitudinal direction differs from the value
obtained for the transverse direction. This has influence on the creep and
shrinkage coefficients, which are determined in the following.
2 Ac
Notional size =
u
The perimeter of the bridge deck section was measured to u = 24.7 m

2 A c 2 2.46mm 2
= = 200mm
u 24.7mm
(t = ,to) = 1.6 Creep coefficient for a loading age of 28 days

and 80% relative humidity
ENV 1992-1-1: 1991 Table 3.3
s. = -0.0003 Shrinkage strain at time t = taken from
ENV 1992-1-1: 1991 Table 3.4
The stress in the concrete adjacent to the tendons, due to self-weight and
ballast load, is calculated on the basis of the axial force only:
K u.DL + K u.B 5118kN + 3711.5kN N

cg = = 2
= 3.59 (Equ. B.22)
Ac 2460000mm mm 2
The initial stress in the concrete adjacent to the tendons, due to

prestressing, is calculated with the following:
(Po P )
cpo = (Equ. B.23)
Ac f
B-17
Material properties
Compressive strength of C50/60 fck [N/mm2] 50
Tensile strength of C50/60 fctm [N/mm2] 4.1
Secant modulus of elasticity Ecm [N/mm2] 37000
2
Elastic modulus of steel Es [N/mm ] 195000
Elastic modulus ratio Equ. B.1 5.27
2
Cross-sectional area per strand [mm ] 150
2
Yield strength fp0.1k [N/mm ] 1500
2
Ultimate strength fpk [N/mm ] 1770
Sheathing tube inner diameter I.D. [mm] 110
Sheathing tube outer diameter O.D. [mm] 118
Material properties min f [mm] 198
Wobble factor k 0.005235988
Friction factor 0.2
Nut draw-in ln [mm] 2
Number of strands per tendon n 27

Number of tendons per edge beam m 6
Prestressing area per tendon Atendon [mm2] 4050
Prestressing per tendon

Allowable prestressing force P0 [kN] = 0.9 fp0.1k 5467.5
P0 [kN] = 0.8 fpk 5734.8
P0 [kN] = 0.85 fp0.1k 5163.75
P0 [kN] = 0.75 fpk 5376.375
Critical value P0 [kN] 5163.75
Internal forces coordinate x 50

Axial force due to self-weight Ku.DL [kN] 5118
Axial force due to ballast load Ku.B [kN] 3711.5
Axial force due to live load Ku.LL [kN] 3973
Total axial force Ku [kN] 12802.5
Section properties
Concrete cross section area Ac [mm2] 2460000
Friction
Sum of angular displacements over dist. x [] 20
Prestress loss due to friction P [kN] Equ. B.11 593.86
Prestressing force after friction P [kN] Equ. B.12 4569.89
Prestress of edge beam at time t=0 P [kN] 27419.36

2
Tendon stress [N/mm ] 1128.37
Wedge draw-in
Average angle of deflection along le 1 [rad/m] 0.017
Influence length le [m] Equ. B.21 10.08
Not applicable at mid-span Pn [kN] Equ. B.20 0
Prestressing force after wedge draw-in Pn [kN] 4569.89
Fig. B.12. Preliminary design of bridge deck in longitudinal direction, continued below
B-18
Relaxation
Initial stress in thread bar due to prestress pgo [N/mm2] Equ. B.15 1128.37
p [N/mm2] Equ. B.14 959.11
p/fpk [%] 54.19
ENV 1992-1-1:1991 Figure 4.8 relaxation [%] 1
2
Relaxation loss after 1000h pr [N/mm ] 9.59
Relaxation loss after 1000h pr, [N/mm2] Equ. B.16 28.77
Creep, shrinkage, relaxation losses

Motional size [mm] = 200 cg [N/mm2] Equ. B.22 3.59
2
Creep coefficient u,28 = 1.6 cpo [N/mm ] Equ. B.23 -11.15
Shrinkage coefficient cs,u 0.0003 Ptt Equ. B.13 -599.67
Prestress in thread bar at time t = Pm,t [kN] Equ. B.10 3970.22

Prestress per edge beam at time t = Pm,t [kN] 23821.34
P (t=0) / P (t=00) 0.87
2
Tendon stress [N/mm ] 980.30
Primary strain pm 0.00502719
Concrete stresses at frequent combination of actions

At time t = 0
Concrete stress due to prestressing N/mm2 -11.15
2
Concrete stress due to dead load N/mm 2.08
2
Total N/mm -9.07
At time t =
Concrete stress due to prestressing N/mm2 -9.68
2
Concrete stress due to dead load N/mm 3.59
2
Concrete stress due to live load N/mm 1.62
2
Total N/mm -4.48
Fig. B.12. Preliminary design of bridge deck in longitudinal direction
The application of 6 tendons type-6827 (DYWIDAG [10]) per edge beam

leads to a concrete compressive stress of 4.48 N/mm2 after all losses at mid-
span (x = 50 m). The margin is to compensate for not considered bending
moments.
B-19
B.2 Arch
TVEIT [41] provides a formula for calculating an approximate axial force in

the arch. This equation will be used for the preliminary dimensioning of the arch
profile. Bending moments have been omitted, as they contribute only little to the
strain of the profile.
Axial force in the arch at a distance x from one support due to a uniformly
distributed load:
1 w (l x ) x 1
Ko = + h w cot 2 v h
cos(v o ) 2h 2
vo = 0 slope of the arch (max. arch force expected at mid-span)

w uniformly distributed load per unit length of lane
l = 100 m length of the span
h = 17 m rise of the arch
vh = 60 average angle of the hangers crossing a vertical line
at mid-span
The uniformly distributed load is calculated in the following.
Loading
Loads have partly been determined in Section B.1.2 of this Annex. An
additional live load on the non-public footpaths is required. The preliminary
design of the arch profile will be based on the ultimate limit state. Dead loads
and live loads are therefore multiplied by the partial safety factors of 1.35 and
1.5, respectively.
Dead load
Self-weight of the bridge deck, ballast, rails and sleepers:
Annex B,
gk = 125 kN m Section B.1.2:
g d = gk 1.35 = 168.8 kN m
Live load
q 71.UDL = 80 kN m Annex B,
q71.train = 9.73 kN m Section B.1.2.
2 = 1.044 Annex A,
q71.k = 2 (q71.UDL + q71.train ) = 1.044 (80 + 9.73 )kN m = 93.7 kN m Section 2.1.2
q footpath = 5 kN m 2 0.75m = 3.75 kN m

( )
q71.d = 1.5 q71 + q footpath = 1.5 (93.7 + 3.75 )kN m = 146.2 kN m
Total
w = g d + q d = (168.8 + 146.2)kN m = 315 kN m
B-20
1 w (l x ) x 1
Ko = + h w cot 2 v h
cos(v o ) 2h 2
315 kN m (100m 50m) 50m 1
Ko = + 17m 315 kN m cot 2 60
2 17m 2
K o = 24054kN
Arch profile
The bridge calculated in STEIMANN [37] uses an arch rise of 15 metres with
otherwise similar geometric properties and loading compared to the bridge
calculated in this work which has a rise of 17 metres. The higher rise leads to
smaller axial forces in the arch. It has therefore been decided to use a smaller
profile than that taken by STEIMANN [37] who uses an American Wide Flange W
360x410x990.
Chosen profile: American Wide Flange W 360x410x900 (Class 1)
Cross section area: A = 1149 cm2
ARCELOR LONG COMMERCIAL S.A. [4]
ENV 1993-2: 1997
Material: Steel S 460 ML Table: 3.1a
fy = 430 N/mm2
Resistance in ultimate limit state

ENV 1993-1-1: 1992
Section 5.4.4
NSd Nc.Rd
N
430
fy 2 mm 2 = 44915kN
Nc.Rd = Npl,Rd = A = 1149cm
M0 1. 1
Ko 24054kN
= = 0.54 < 1
Npl.Rd 44915kN
The margin in the check will compensate for bending moments as well as
weakening by fastener holes which are not considered in the preliminary
design.
Buckling of arch
The span and loading of the bridge calculated in this work corresponds to
the conditions in STEIMANN [37], where buckling appeared not to be critical. It
will therefore be omitted in this preliminary design.
B-21
B.3 Hangers
Loading
Dead load
Self-weight of the bridge deck, ballast, rails and sleepers:
gk = 125 kN m
There must be a distinction made between concentrated loads and

uniformly distributed loads. Even though live loads have been calculated in
Section B.1 and B.2, they will be listed here again.
Live load
Load model 71
Uniformly distributed load q 71.UDL = 80 kN m
Concentrated load Q 71 = 76 kN m 6.4m = 487kN
Live load on footpath q footpath = 5 kN m 2 0.75m = 3.75 kN m
It is desirable to obtained information about hanger forces in ultimate as

well as serviceability limit states. The characteristic loads are therefore
multiplied with the respective partial safety factors to determine design loads.
The dynamic factor will also be applied.
Dynamic factor: 2 = 1.044 (Annex A, Section 2.1.2)
Ultimate limit states:
g d.ult = gk 1.35 = 125 kN m 1.35 = 168.75 kN m

( )
q d.ult = q71.UDL 2 + q footpath 1.5 = (80 1.044 + 3.75 )kN m 1.5 = 130.9 kN m
Total uniformly distributed load: w d.ult = g d.ult + q d.ult = 300 kN m
Q d.ult = Q k 1.5 2 = 487kN 1.5 1.044 = 762.6kN
Serviceability limit states:

The loads for the serviceability limit states are equal to the characteristic
loads, since the partial safety factors are = 1.0. Only the dynamic factor has to
be applied.
g d.ser = gk 1.0 = 125 kN m 1 = 125 kN m

( )
q d.ser = q71.UDL 2 + q footpath 1.0 = (80 1.044 + 3.75 )kN m 1.0 = 87.27 kN m
Total uniformly distributed load: w d.ser = g d.ser + q d.ser = 212.27 kN m
Q d.ser = 2 Q 71 = 1.044 487kN = 508.43kN
B-22
Influence areas (reference bridge: Vienna 200B)
The influence areas of this bridge are calculated with the help of the
influence areas of the bridge design Vienna 200B, TVEIT [45]. Among all
network arches designed by PER TVEIT, Vienna 200B is most appropriate in Annex B,
respect of the ratio between dead load to live load. Section B.1.2
Influence areas of Vienna 200B (hangers 90-94):
+
Positive influence area: IA 200 B = 9.56m

Negative influence area: IA 200 B = 4.20m
+
Sum of influence area: IA 135 = IA 135 + IA 135 = 5.36m
+
Maximum value of influence line: I200 B = 0.38

Minimum value of influence line: I200 B = 0.1
The ratio between the spans of 200B and this bridge: k = 100m 200m = 0.5 .
Hence, the influence areas of this bridge with a span of 100m:
+ +
Positive influence area: IA 100 = k IA 135 = 4.78m

Negative influence area: IA 100 = k IA 135 = 2.10m
+
Sum of influence area: IA 100 = IA 100 + IA 100 = 2.68m
+
Maximum value of influence line: I100 = 0.38

Minimum value of influence line: I100 = 0.1
Approximate hanger force
Ultimate limit state
(
N d,max = (IA 100 g d.ult ) + IA 100
+
) ( +
q d.ult + I200 B Q d.ult )
Nd,max = (2.68m 168.75 kN m) + (4.78 130.91kN m) + (0.38 762.6kN) = 1368kN
Serviceability limit state
( ) (
Nk.max = (IA 100 gd.ser ) + IA +100 q d.ser + I200
+
B Q d.ser )
Nk,max = (2.68m 125 kN m) + (4.78m 87.27 kN m) + (0.38 508.43kN) = 945kN
(
Nk. min = (IA 100 g d.ser ) IA 100

) (
q d.ser + I200 B Q d.ser )
Nk,min = (2.68m 125 kN m) (2.10 87.27 kN m) (0.1 508.43kN) = 100.89kN
B-23
Materials
The hangers are smooth bars of S 460 ML steel with circular cross section. ENV 1993-2
fy = 430 N/mm2 Table: 3.1a
Determination of required cross-sectional area
The partial safety factor M0 is taken from ENV 1993-2: 1997 Section 5.1.1. ENV 1993-1-1: 1992
fy Section 5.4.3
Npl.Rd = A M0 = 1.0
M0
Minimum required area of hanger section

N 1.0 1368kN
A min = M0 d.max = 2
= 3182mm 2
fy 430 N mm
Minimum diameter of a circular hanger section:

4 A min 4 3182mm 2
D min = = = 64mm

The diameter is chosen to be D = 65 mm.
B-24
Annex C
FEM - Calculation
C.1 General
For the FEM-calculations two structural analysis software packages were

used. Most of the investigations were performed with SOFiSTiK AG, www.sofistik.com
Oberschleiheim-Germany. For the calculations of stress distribution in the
hanger connection details NE/Nastran for Windows from Norman Engineering www.nenastran.com
Inc. was used.
SOFiSTiK
The 3D-model used for calculations shows Figure C.1.
Fig. C.1. Foreshortened view of the 3D-FEM-model used for calculations
The bridge deck was modelled with plane elements. Their nodes were
aligned to the bottom plane of the tie. In that way it was possible to shape the
bridge deck and end cross girder like the real cross-sections by applying
different thickness to the plane elements. The cantilevers were connected by
couplings to the nodes of the bridge deck elements providing fixed connection
to the rigid body at the reference nodes. Figure C.2 shows the partitioning of the
bridge deck mesh.
C-1
Diploma Thesis Brunn & Schanack Annex C: FEM - calculation
Fig. C.2. Partitioning of the mesh of the bridge deck
It was decided to connect the hanger nodes with skew triangular elements.
This allowed square plane elements in the rest of the deck. Substitution of the
triangular mesh by skew quadrilateral partitioning would give better results for
the analysis, nevertheless. The square elements have a length of 0.6 metres.
The arches were modelled using beam elements with a length of about 0.5
metres. The truss members of the wind bracing were also beam elements with
hinges at the connections to the arch. In a test without hinges bending moments
of maximum 7 kNm occurred in the truss. They originate mainly in the torsional
moments in the arch due to the eccentric connection of the hangers and can be
ignored for the assessment.
The hangers were modelled using cable
elements that only sustain tension in case of
non-linear analysis. This has to be considered
when calculating influence lines. Since
analysis is carried out in linear fashion,
hangers will take compression forces, instead
of relax. This leads to increased internal forces
and is therefore on the safe side. The cable
elements were connected eccentrically to the
arch as shown in Figure C.3. At their
intersections the horizontal deflection
perpendicular to the arch plane was coupled.
In that way it was possible to calculate
deflections and mode shapes of the hanger
web.
Fig. C.3. Connections and couplings

of the cable elements
The SOFiSTiK module GEOS provides features to involve prestressing

tendons within the plane elements. Figure C.4 shows their arrangement. It is
possible to choose common tendons from reputable manufacturers. The
software itself calculates the loads to apply to the structure.
C-2
Fig. C.4. Arrangement of prestressing tendons within the plane elements of the bridge deck
Module ELSE was used for the analysis of the influence lines and module
ASE for second order analysis. Figures C.5 to C.7 give the geometric properties
of the final bridge design.
Ge om e try of hange r arrange m e nt

Node Coordinates of the hanger
number nodes of half span
100+n / low er hanger upper hanger
200+n nodes nodes
n x [m ] z [m ] x [m ] z [m ]
224
223
222
221
220
219
218
217
216
1 2.81 0 3.82 2.93

215
214
213
212
2 5.51 0 5.442 4.05

211
210
3 8.25 0 6.818 4.94

209
208
207
4 11.03 0 8.028 5.69

206
205
5 13.5 0 9.538 6.57

204
203
202
6 13.98 0 11.07 7.42

201
7 26.86 0 13.02 8.42 z

8 19.68 0 14.99 9.36 x
101
111
104
107
103
112
102
108
119
106
113
114
116
124
105
117
110
123
120
115
118
121
109
122
CL
9 22.22 0 16.99 10.2
Fig. C.5. Definitions of the hanger node numbers
10 22.3 0 19.02 11.1
11 25.23 0 21.06 11.86
12 27.68 0 23.13 12.59
13 29.29 0 25.21 13.26
hanger
14 30.73 0 27.31 13.88 CL

0.15
0.61
0.61
0.59
0.58
15 33.51 0 29.42 14.44 d [m]

0.43
0.42
0.41
0.40
0.39
0.39
0.38
16 35.03 0 31.55 14.95

17 36.33 0 33.69 15.4
18 39.21 0 35.84 15.8
19 39.7 0 38.01 16.14 0.57 0.30 0.60 0.60 0.60 0.60 0.60 0.60 0.60 0.60
20 42.18 0 40.17 16.42
0.22 [m]
21 44.13 0 42.35 16.65
Fig. C.6. Geometry of the plane elements of the tie
22 45.28 0 44.53 16.82
23 47.95 0 46.72 16.94
24 48.55 0 48.91 16.99
Fig. C.7. Coordinates of hanger nodes
C-3
NE/Nastran
The stress analysis used to evaluate the resistance against fatigue strains
of the hanger connection details was performed with NE/Nastran. The models
were shaped using SolidWorks 3D CAD Software. The length of the hanger www.solidworks.com
extending the connection detail was chosen to be 1 metre. The default settings
of NE/Nastrans auto-mesh feature were used for partitioning of the solid mesh,
which gave element sizes between 10 and 20 mm. Figure C.8 shows an
example of the meshes used. The static analysis was performed in linear
fashion. Constraints and both loads, axial force and horizontal deflection, were
applied to the corresponding surfaces of the 3D-model. Stress distributions and
values were found by the help of the Stress Wizard, which is a component of
NE/Nastran.
Fig. C.8. Example of solid mesh used for numeric analysis
In NE/Nastran the interaction between elements is based on forces exerted

at the so-called mesh grid points, where elements are connected together. The
stiffness of the structure, discretised at these grid points, is generated with data
on material properties and geometry. Both stiffness and forces are used to
calculate displacements, with which stresses and strains are generated. It was
decided to use tetrahedron instead of hexahedral elements, since the additional
effort required to generate high-quality hexahedral meshes cannot usually be
justified by the time required to do so. However, the tetrahedron elements were
equipped with additional mid-notes, leading to higher-order 10-noded elements.
C-4
C.2 Internal forces of the arch
The arches receive mainly axial compression forces and are therefore in
danger of collapse due to buckling. Additionally, there are in-plane bending
moments My due to the hanger forces and out-of-plane bending moments Mz
and torsional moments Mt due to horizontal forces (like wind) on hangers and
arches. Additionally the eccentricity of the hanger connections causes torsional
bending. The arches will be verified using second
z
order analysis to prove the buckling resistance.
For this purpose it is required to apply the initial
bow imperfection specified in ENV 1993-1-1: 1992, y
y
5.5.1.3 on the arch. The relevant buckling curve is
the first mode shape for each axis of the arch profile.
The mode shapes were determined by SOFISTIK
ASE-module. z
Fig. C.9. Defintion of axes in
arch cross-section
C.2.1 Ultimate limit state
C.2.1.1 Collapse about the weak axis y-y
The first mode shape for in-plane arch deflection is the second mode shape
of the whole bridge. But it is the symmetric vertical deflection and does not lead
to any buckling. The fifth mode shape of the whole bridge (Figure C.10)
represents the relevant buckling curve.
o,d
eo,d
Fig. C.10. Fifth mode shape, equivalent initial bow imperfection eo,d ENV 1993-1-1: 1992
Figure 5.5.1
Eurocode ENV 1993-1-1 does not consider the special buckling behaviour
of arches. Therefore, the equivalent initial bow imperfection cannot be
calculated according to ENV 1993-1-1: 1992 Figure 5.5.1. As the German
National Code, DIN 18800, includes clear statements about imperfections to be
applied to arches, these values are used for second order analysis.
l
eo,d= =0.25m DIN 18800 T.2,
400 Abschnitt 6, Tab. 23
The determination of the critical cross-sections of the arch was performed in
linear fashion. Thereto the envelopes of maximum My and minimum My were
calculated (Figures C.11, C.12). The first critical cross-section is found at the
top edge of the concrete tie at the clamping of the arch, where maximum My
occurs. Further upwards the maximum and minimum values of the bending
moment are distributed more or less uniformly along the arch. So it was only
necessary to calculate the internal forces for the lowest butt-welded splice,
C-5
since the axial force will be highest there. The influence lines of N and My for
these two critical sections are shown in Figure C.13. The design-relevant
internal forces can be found in Figure C.16.
109.2
109.6
84.4
107.6
Envelope of maximum My in the arch [kNm]
112.0
57.5
0
108.9
57 . 1
Design relevant sections: 49.3
97.9
102.
230.1
222.2
190.0
9
237.0
213.5
57.1
236.5
232.6
231.4
lowest end plate joint
197.1
281.2
49.9
215.4
195.9
216
280 7
2.3
clamping of the arch
231
316
.
.3
29
.4
.2
280
1 9 6 .3
.0
-49
21
28
3 .7
.7
.4
.5
267 .2
19
97
33
.7
8.4
66
.2
7
.2
0
51
2.2
27
2
.7
85
8.
C
L
3
Fig. C.11. Preliminary linear calculation of bending moments My in the arch
-101.4
-103.6
-127.5
-126.2
-103.6
-94.6
-87.5
-93.8
-132.8
-103.6
-143.9
-80.7
-100.9
Envelope of minimum My in the arch [kNm]

-89.7
-129.3
-125.6
0
-1 1
-106
-141 .1
-33.7
-38.4
2.3
-34.5
- 117
- 18
-70.6
-130
.6
-17
.8
-20.2
7 .8 .4
-39.9
.8
-17
-58.2
2.2
-50.6
5
-46
-15 -15
8.9
-5
.2
-13
.9
5.2
.3
0
51
8.3
27
2.7
C
L
Fig. C.12. Preliminary linear calculation of bending moments My in the arch
-1.12
My
0.0648
Influence lines at the root of the arch
-1.1
N
-0.0234
My
0.0442
Influence lines at the lowest end plate joint
0 20 40 60 80 l [m] 100
15.94 28.59 39.39 58.41 69.89
Fig. C.13. Influence lines for N and My at the critical sections
C-6
C.2.1.2 Collapse about the strong axis z-z
The first mode shape of the e o,d

whole bridge represents the CL
relevant buckling curve for out-of-
plane buckling (Figure C.14). As
in Section C.2.1.1 the equivalent
initial bow imperfection is taken
from DIN 18800.
l
eo,d= =0.2m
500
Fig. C.14. First mode shape of whole bridge
Maximum internal forces N and Mz occur at load combination 4

(Figure C.15). The design check will be performed for the section at the
clamping of the arch, where the maximum bending moment and the maximum
axial force are found. For all relevant forces see Figure C.16.
-27
Load combination 4 Mz [kNm]

-2
85
N x [kN]
80
6.3
-2
20
-2
79
73
.6
68
-2 49
78
.2
.3
-2 75
75 .4
-2 45 70.5
-2 778 .1 -168.3
-2 711 3.5 -185.3
7
-2 62 6.7
70 1. -161.2
-2
75 12.3 9
-2 0 -133.4
68 1.7
-27 48. -165.5
3 3 -193.2
-27 85.2
-26 290 -201.1
4 .1 -215.8
-27 84.9 -200.7
12
-27 3 . -100.7
12 8
-26 0 .5
544 -95.8
-27 .0
1
-27 83.6 378.4 -193.5
699 960.5
-27 .5
-28 275.8
057 819.0
-28 .
146 4
.8
-300.7
-1202.3
-1840.8
Fig. C.15. Maximum internal forces for collapse about the z-z axis of the arch (half of one arch,
isometric projection)
C.2.2 Fatigue stress spectra
The cross-sections to be checked are the butt-welded splices, the

connections to the wind truss beams, hanger connections and the root of the
arch. For the first three locations the biggest stress range occurs at the lowest
of each of these joints. The variation of internal forces due to load model 71 on
one and on both of the tracks, 1 and 1+2, can be found in Figure C.17.
C.2.3 Serviceability limit state
The serviceability limit states to be verified by numerical assessment are

limitations for stress, stress range and the maintenance of the specified
clearance gauges. The maximum nominal stress Ed,ser and stress range fre
C-7
to be limited occur at the root of the arch. The structural part in danger of
encroaching upon the railway traffic is the lowest member of wind truss. Its
horizontal and vertical deflections are x and z.
The appendant values can be found in Figure C.18.
C.2.4 Composition of the design relevant calculation results
Ultim ate lim it s tate

Collaps e about
axis y-y axis z-z
Se cond
Root of arch Firs t s plice Root of arch
s plice
Nx [kN] -28468.3 -27487.3 -26873.9 -28146.8
V y [kN] -198.3 135.4 128.7 -514.7
V z [kN] -511.5 -310.5 -278.3 -428.8
M y [kNm] 1115.8 210.8 195.9 841.8
M z [kNm] -809.9 -100.4 -164.5 -1840.8
M t [kNm] -87.5 -11.3 -18.3 -114.6
Fig. C.16. Internal forces for ultimate limit state assessment
Fatigue
Connection to wind truss member Lowest hanger connection
Forces due to Forces due to Forces due to Forces due to Forces due to Forces due to
dead load one LM 71 two LM 71 dead load one LM 71 two LM 71
Nx [kN] -10362 -15810.1 -17839.9 -10618.4 -16376.9 -18526.3
Vy [kN] -29.2 -65.2 -66.6 -24.4 -67.6 -85.5
Vz [kN] 134.7 134.9 149.9 -191.9 -259.9 -349.1
My [kNm] 30.8 5.10 13.2 -58.1 -59 -60.7
Mz [kNm] 86.6 202.1 209.3 -170.3 -265.9 -332.9
Mt [kNm] -6.2 -13.3 -15.1 -10.3 -22.1 -28.9
First splice Root of the arch
Forces due to Forces due to Forces due to Forces due to Forces due to Forces due to
dead load one LM 71 two LM 71 dead load one LM 71 two LM 71
Nx [kN] -10331.3 -15851.7 -17912.6 -10673.8 -16441.9 -18978.8
Vy [kN] -27.1 -28.6 -28.3 -24.4 -68.3 -86.5
Vz [kN] 89.0 28.5 30.3 105.2 75.2 97.7
My [kNm] 10.0 0.9 0.9 17.0 250.7 446.2
Mz [kNm] -43.1 -39.1 -40.8 3.6 354.9 468.1
Mt [kNm] -13.8 -12.1 -13.7 -2.9 -4.4 -6.0
Fig. C.17. Internal forces for fatigue assessment
Se rvice abiltiy lim it s tate

Root of the arch w indtrus s
Nom inal V ariation of
s tre s s force s Load com bination 12
Nx [kN] -19245.6 5758.5 horizontal def lection
V y [kN] -97.4 43.2 x 18.7 cm
V z [kN] 102.3 68.0 vertical def lection
M y [kNm] 54.7 95.6 y -4.0 cm
M z [kNm] 479.3 0.9
M t [kNm] -7.2 11.8
Fig. C.18. Values for assessment to satisfy serviceability limits
C-8
C.3 Hanger
C.3.1 Axial forces
The labelling of the hangers is defined in Figure C.19.
Set of "rising" hangers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 20 21 22 23 24
Set of "falling" hangers
48 4746 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25
Fig. C.19. Definition of the labelling of the hangers
To obtain the maximum strains on the hangers the influence lines of all
hangers were calculated and analysed by applying the load combinations for
ultimate, serviceability and fatigue limit states as described in Annex A. The
resulting maximum and minimum hanger forces are shown in Figure C.20.
Hanger forces [kN]

Ultimate limit state Fatigue Serviceability limit state
Number of
1+2 N, two
hanger, 1 N, one Rare / characteristic Frequent load
Maximum Minimum tracks
according to track loaded load combinations combinations
loaded
Figure C.19 Maximum Minimum Maximum Minimum
1 833.16 269.12 376.03 252.51 616.34 247.81 483.11 198.25
2 1002.06 375.82 417.49 246.83 741.29 346.06 593.05 276.85
3 1016.22 364.09 434.75 268.9 751.76 335.26 595.60 274.30
4 1061.85 413.33 432.35 245.11 785.52 380.6 628.42 304.50
5 1057.03 432.28 416.5 221.28 781.96 398.05 615.56 318.44
6 1020.65 457.37 375.52 170.08 755.04 421.15 604.77 336.92
7 932.3 381.23 367.38 195.22 689.68 351.04 541.38 280.83
8 973.93 314.44 439.66 295.34 720.48 289.54 586.38 231.63
9 924.91 318.55 404.24 258.73 684.22 293.33 547.87 234.66
10 925.75 348.19 385.04 226.96 684.84 320.62 547.87 257.68
11 902.22 334.39 378.55 226.59 667.43 307.91 533.94 246.33
12 893.61 336.98 371.09 218.12 661.06 310.29 528.85 248.23
13 881.19 346.54 356.43 199.55 651.87 319.09 521.50 255.27
14 886.15 350.01 357.43 199.02 655.54 322.29 534.32 257.83
15 813.45 313.66 333.19 191.02 601.76 288.82 481.41 231.06
16 865.82 339.91 350.61 196.71 640.5 312.99 512.40 250.39
17 724.45 296.07 285.59 151.88 535.92 272.63 428.74 228.57
18 829.49 302.97 351.01 213.19 613.63 278.98 490.90 223.18
19 839.22 331.52 338.47 188.43 620.83 305.27 496.66 244.22
20 836.03 336.31 333.15 181.12 618.47 309.68 494.78 265.32
21 941.83 493.63 298.8 123.65 696.73 454.54 557.38 363.63
22 970.73 558.19 275.03 112.45 718.11 513.99 574.49 411.19
23 994.5 752.93 161.05 89.5 735.69 693.31 588.55 553.82
24 661.81 438.15 149.11 86.23 489.58 50.78 391.66 38.65
Fig. C.20. Hanger forces
C-9
In ultimate limit state hanger, number 4 receives the maximum axial forces.
As an example the influence line of this hanger is presented in Figure C.21.
-0.11
0.14
0 12.19 20.69 33.91 x [m] 100
Fig. C.21. Influence line of hanger 4
C.3.2 Deflections perpendicular to the hanger axis
For the fatigue assessment of the hanger connection details, the horizontal
deflection of the hangers were examined. As the lower connections of the
hangers are cast in the concrete bridge deck, therefore it is assumed that
horizontal movements of the hangers are not transferred to the connection, but
absorbed by the concrete around it.
The upper connections consist of a steel plate that is welded to the arch
perpendicularly to the plane of the hanger web, so that it can move slightly
within the arch plane. Furthermore, all hangers are tied together at their
crossings. Because of that it can be assumed that movements of the hangers in
the plane of the arch are small and do not lead to important internal forces in
the hanger connection.
Consequently the deflections of interest are perpendicular to the hanger web.
C.3.2.1 Static deflection due to wind load
To obtain a first idea of the strains caused by horizontal deflections the

hanger web was loaded by the static wind forces according to Section A.3.1.
Due to gusts the centre of the web was loaded with 100% and the outer parts
with 60% of the full wind load, SCHULTE [27]. This gives larger horizontal
deflections of the web centre. Since the stiffness of the hangers depends on
their axial force, calculations were carried out for the fatigue load combination
two LM 71 giving maximum hanger forces and without live load giving
minimum hanger forces. In Figure C.22 the values of the angular rotations at
the upper end of the hangers are presented.
C-10
M inim um hange r force s M axim um hange r force s

Num be r of Hange r Y de fle ction Angular Y de fle ction Angular
hange r, le ngth until of firs t rotation of of firs t rotation of
according to firs t cros s ing uppe r hange r cros s ing uppe r hange r
figure C.19 cros s ing [m ] [m m ] e nd [m rad] [m m ] e nd [m rad]
1 1.125 0.3674 1.0000 0.0942 0.2564
2 1.3349 0.5715 0.4281 0.1833 0.3207
3 1.678 0.6321 0.3767 0.1903 0.3011
4 1.6848 1.1346 0.6734 0.3399 0.2996
5 1.6808 1.3460 0.8008 0.4004 0.2975
6 1.6772 1.5034 0.8964 0.4458 0.2965
7 1.6573 1.6373 0.9879 0.4869 0.2974
8 1.6651 1.7683 1.0620 0.5313 0.3005
9 1.6477 1.8381 1.1156 0.5569 0.3030
10 1.6697 1.9184 1.1489 0.5828 0.3038
11 1.6705 1.9767 1.1833 0.5989 0.3030
12 1.667 1.9644 1.1784 0.5958 0.3033
13 1.6668 1.9644 1.1785 0.5958 0.3033
14 1.6669 1.9767 1.1859 0.5989 0.3030
15 1.6619 1.9184 1.1543 0.5828 0.3038
16 1.6671 1.8381 1.1026 0.5569 0.3030
17 1.6581 1.7683 1.0665 0.5313 0.3005
18 1.6676 1.6373 0.9818 0.4869 0.2974
19 1.6702 1.5034 0.9001 0.4458 0.2965
20 1.6722 1.3460 0.8049 0.4004 0.2975
21 1.6782 1.1346 0.6761 0.3399 0.2996
22 1.408 0.6321 0.4489 0.1903 0.3011
23 1.4466 0.5715 0.3951 0.1833 0.3207
24 3.0948 0.3674 0.1187 0.0942 0.2564
Fig. C.22. Angular rotation of upper hanger ends due to static wind loads
C.3.2.2 Deflections due to dynamic wind excitation
For dynamic investigations mode eigen- mode eigen- mode eigen-

shape f requency shape f requency shape f requency
the mode shapes are of interest. number [Hz] number [Hz] number [Hz]
Searching for them, the 1 0.97 21 2.67 41 3.3
crossings of the hangers were 2 1.12 22 2.7 42 3.32
modelled by coupling the out-of- 3 1.27 23 2.72 43 3.35
4 1.42 24 2.77 44 3.37
hanger-web deflections of the
5 1.55 25 2.82 45 3.38
nodes. The first mode shapes 6 1.69 26 2.85 46 3.39
are shown in Figure C.24. Since 7 1.82 27 2.88 47 3.42
the amplitudes of the 8 1.84 28 2.92 48 3.44
9 1.96 29 2.95 49 3.45
oscillations are larger for less
10 1.97 30 2.97 50 3.47
loaded hangers and since it is 11 2.09 31 3.01 51 3.49
probable that wind excited 12 2.1 32 3.05 52 3.51
oscillations occur for long 13 2.22 33 3.06 53 3.54
14 2.22 34 3.07 54 3.55
periods of time without a train 15 2.34 35 3.14 55 3.58
passing over the bridge, the 16 2.36 36 3.14 56 3.59
mode shapes were calculated 17 2.46 37 3.17 57 3.61
18 2.48 38 3.2 58 3.63
without live load.
19 2.59 39 3.21 59 3.65
20 2.61 40 3.24 60 3.66
61 3.68
Fig.23. Eigenfrequencies of the first mode shapes of the
hanger web
C-11
1.
8.
f = 0.97 Hz f = 1.84 Hz
10.
2. f = 1.97 Hz
f = 1.12 Hz
12.
3.
f = 2.1 Hz
f = 1.27 Hz
15.
4.
f = 2.34 Hz
f = 1.42 Hz
5. 20.
f = 1.55 Hz f = 2.61 Hz
6.
40.
f = 1.69 Hz f = 3.24 Hz
7. 61.
f = 1.82 Hz f = 3.68 Hz
Fig. C.24. Mode shapes and eigenfrequency of the hanger web
The effects of dynamic excitation are described in Section 7.2.2. It is shown

there that it is not possible to determine the dynamic behaviour of the hanger
web only be means of the procedures provided by Eurocode. But it seems
improbable that the web would be excited by along-wind nor crosswind. That
means fatigue strains caused by wind are not existent. Nevertheless, while
determining the hanger connection details in Section 7 horizontal deflections of
the hangers were considered and the qualitative stress distribution caused by
them will influence the decision on the shape of the connection detail.
Further investigations should be carried out with the help of wind engineers.
C-12
C.4 Internal forces in the wind bracing
The labels of the eo1 = 0.167

m
eo2 = 0.04
33 m
members and the nodes of
16
VIII
the wind bracing can be seen
15
14
VII
in Figure C.25. The straight
13
12
struts consisting of members VI
11
2, 3 and 16 are CHS
10
V
219.1x10. The rest are CHS
9
219.1x8. IV
8
7
Because of the double III
6
symmetry of the truss only 4
5
II
2
half of one bridge side is
3
eo3
considered. 1
I
The wind bracing
members are loaded by self-
weight and restraining forces
from the arch due to wind
forces and buckling effects. A Fig. C.25. Labeling of members and nodes, initial bow
calculation of the secondary imperfections
bending moments at the connections showed that they are negligible. According
to TEICH & WENDELIN [38], page 46, as forces due to vortex shedding do not
influence the choice of the cross-section, they are therefore neglected.
For the analysis of the wind bracing an equivalent geometric imperfection
shall be applied to the arches. For this purpose the first mode shape, shown in
Figure C.14, is scaled so that its maximum deflection in the middle of the span
equals the initial bow imperfection eo1.
eo1 = kr L/500
1 1
with kr = 0. 2 + = 0.2 + = 0.837
nr 2
eo1 = 0.837 100m/500 = 0.167m
ENV 1993-1-1: 1992
As can be seen in Figure C.25, the nodes of the wind bracing lie in a
5.2.4.4 (5.3)
cylindrical sphere between the arches. That means the members of the truss do
not lie in a plane and their axial forces cause out-of-sphere movements of the
nodes. This has to be impeded by the straight members (2, 3, 6, 8, 10, 12, 14,
16) which therefore receive bending moments. In case the out-of-sphere
movements occur in compression members, the effect has to be treated as a
stability problem. This will be performed with second order analysis. The initial
bow imperfections to be applied are:
Wpl
( )
eo2 = 1.33 0.2 k
A
o2
9 .6 0 m
with = 0.21 ; buckling curve a
l
= A = A
1 E ENV 1993-1-1: 1992
iy 5.5.1.3
fy
with = 1 ; Euler case 2
l = 9.60 m
C-13
iy = 7.4 cm
E = 210000 N/mm2
fy = 355 N/mm2 ; S355
A = 1 ; Class 1 cross section
1 9.6m
= 1 = 1.698
210000N / mm 2
7.4cm
355N / mm 2
k = 0.23 ; M1 = 1.1 & buckling curve a
k = (1-k) + 2 k = (1-0.23) + 2 0.23 1.698 = 1.551
Wply = 438 cm3
A = 65.7 cm2
438cm 3
eo2 = 1.33 0.21 (1.698 0.2) 1.551 = 4.33 cm
65.7cm 2
And:
Wpl o3 e o3
( )
eo3 = 1.33 0.2 k
A
2.275 m 5.05 m 2.275 m
with = 0.21 ; buckling curve a
l
= A = A
1 E
iy
fy
l = 4.80 m
iy = 7.4 cm
E = 210000 N/mm2
fy = 355 N/mm2 ; S355
1 4.8m
= 1 = 0.846
210000N / mm 2
7.4cm
355N / mm 2
k = 0.23 ; M1 = 1.1 & buckling curve a
k = (1-k) + 2 k = (1-0.23) + 2 0.23 0.846 = 1.159
Wply = 438 cm3
A = 65.7 cm2
438cm 3
eo3 = 1.33 0.21 (0.846 0.2) 1.159 = 1.39 cm
65.7cm 2
These node deflections were applied to the 3D-FEM-model and a second

order analysis was performed.
The wind bracing members are loaded by self-weight and restraining forces
from the arch due to wind forces and buckling effects.
C-14
The maximum and minimum axial forces of all members and the bending
moments in the straight members (2, 3, 6, 8, 10, 12, 14, 16) are shown in
Figure C.27.
Minimum axial forces [kN]

9.6
- 21
87.9 -109.8
-44
-4
-1
10.9
-26.3
-1
-52.3
-66.83
32.8
1.7
-6
-8
5.
70
04
-48.7
6.
7.
2
- 23
.4
.9
6
3
0.4
Maximum axial forces [kN]

-61.1 105.5
7.6
157.7
22
48.6
9
2
.9
0.
.0
43.2
4.
.3
71.8
50.1
63
10
84
1
.2 -17
43
-1
47.9
51 3 .6
CL
Fig. C.26. Maximum and minimum forces in the wind bracing in ultimate limit state
Internal forces in the wind bracing in ultimate limit state

Number of max Nx min Nx Number max My min My max Mz min Mz
member [kN] [kN] of node [kNm] [kNm] [kNm] [kNm]
1 51.2 -441.7 I - - - -
2 105.5 87.9 II -34.4 1.6 -15.9 -1.8
3 -61.1 -109.8 III 9.2 6.9 5.6 4
4 -173.6 -219.6 IV 13.2 8.8 0 0
5 227.6 -230.4 V 14.2 9.3 0 0
6 157.7 -66.8 VI 15.1 9.7 0 0
7 100.9 -104.9 VII 15.7 10.1 0 0
8 71.8 -48.7 VIII -27.3 -2.4 0 0
9 84 -87.3
10 47.9 -52.3
11 63.9 -66.6
12 50.1 -26.3
13 43.3 -45.2
14 48.6 32.8
15 -114.2 -170.4
16 43.2 10.9
Fig. C.27. Internal forces in ultimate limit state
The critical members for compression force and buckling are number 1 and
15. Member 3 has to be designed for bending and compression force in node II.
C.4.2 Fatigue
The variation of the internal forces due to the fatigue load models, one and
two load models 71, are shown in Figure C.28. Critical for the assessment are
nodes I and II, which are connected to members 1 and 1, 2, 3, 4 and 5.
C-15
V ariation of inte rnal force s in w ind bracing due to fatigue load m ode ls
V ariation of axial V ariation of be nding V ariation of be nding
force [k N] m om e nt M y [k Nm ] m om e nt M z [k Nm ]
Num be r of Num be r of
one LM 71 tw o LM 71 one LM 71 tw o LM 71 one LM 71 tw o LM 71
m e m be r node
1 120.9 174.8 I - - - -
2 15.5 26.4 II 15.4 20.3 7.4 10.3
3 30.7 47.2 III 9.2 13.4 2.3 4.5
4 14.2 19.8 IV 2.1 3.5 0 0
5 136.5 154.8 V 1.6 2.8 0 0
6 58.4 70.2 VI 1.4 2.5 0 0
7 39.8 65.3 V II 2.5 3.9 0 0
8 35.6 61.7 V III 13.2 15.6 0 0
9 32.1 59.2
10 28.4 54.3
11 31.4 59.7
12 35.6 61.2
13 34.5 60.1
14 6.2 8
15 15.8 17.3
16 9.6 14.8
Fig. C.28. Variation of internal forces for fatigue assessment
For the limitation of the nominal stress and nominal stress variation the
internal forces from rare/characteristic and frequent load combinations are
needed. The values are given in Figure C.29.
Inte rnal force s in the w ind bracing in s e rvice ability lim it s tate
Fre que nt load
Characte ris tic/rare load com binations
com binations
Num be r of m in Nx Num be r of m ax M y m in M y My
m ax Nx [k N] Nx [k N]
m e m be r [k N] node [k Nm ] [k Nm ] [k Nm ]
1 -4.2 -238.5 I - - 139.8 -
2 83.5 56.4 II -23.7 -2.1 21.1 16.2
3 -47.8 -75.4 III 5.1 5.2 37.8 10.7
4 -143.6 -184.3 IV 7.4 6.1 15.8 2.8
5 170.3 -172.8 V 7.9 6.5 123.8 2.2
6 112.5 -12.4 VI 8.1 6.6 56.2 2
7 75.8 -69.3 V II 8.2 7.4 52.2 3.1
8 35.8 -45.2 V III -21.5 -1.9 49.4 12.5
9 38.4 -41.8 47.4
10 32.9 -38.4 43.4
11 37.4 -47.2 47.8
12 35.6 -12.6 48.9
13 34.5 -32.1 48.1
14 28.6 25.9 6.4
15 -145.6 -140.7 13.8
16 27.6 8.4 11.8
Fig. C.29. Internal forces in the wind bracing in serviceability limit state
C-16
C.5 Internal forces in the bridge deck
The results given in the following were obtained from analysis of the 3D bridge
model in SOFiSTiK. The computation did not account for creep, shrinkage and
relaxation effects; these losses were taken from the preliminary design (Annex B) and
the prestressing forces in the model were reduced respectively.
The decisive forces for the bridge deck in the ultimate limit states result from load
combination 4 (see Annex A.5). One load model 71 on each track is placed in the
centre of the bridge giving maximum bending moments for both longitudinal and
transverse directions. The maximum values do not occur at mid-span, but rather under
each track (Figure C.30, Figure C.31). Moreover, the bending moments per metre width
are only single-symmetric about the longitudinal bridge axis, which is due to the braking
force.
The results in the following figures are only shown for the longitudinal mid-range
from x = 40 m to 60 m, whereas the full width of the deck is illustrated. The included
footpaths are to be neglected here, as they are treated separately.
Illustrations of membrane forces were not included, since results vary only slightly.
The membrane forces occurring at locations of maximum bending moments were
directly read from the software output and are given below at the appropriate position.
628.7
569.8
522.3
474.8
427.3
379.8
332.4
284.9
237.4
189.9
142.4
95.0
47.5
0.0
-47.5
-95.0
-142.4
-178.4
Fig. C.30. Transverse bending moments my, [kNm/m], x = 4060 m
Design-relevant internal forces taken from Figure C.30 (a summary of all forces
relevant for assessment can be found in Figure C.34):
Maximum bending moment: my.max = 628.7 kNm/m
(x = 49.1 m, y = 3.28 m)
Respective axial force: ny = -3139 kN/m
Bending moment at smallest depth: my.1.36 = 230 kNm/m (y = 1.36 m)
C-17
422.7
358.4
325.9
293.3
260.7
228.1
195.5
162.9
130.3
97.8
65.2
32.6
0.0
-32.6
-65.2
-97.8
-130.3
-162.9
-195.5
-228.1
-229.1
Fig. C.31. Longitudinal bending moments mx, [kNm/m], x = 4060 m
Design-relevant internal forces taken from Figure C.31 (a summary of all forces
relevant for assessment can be found in Figure C.34):
Maximum bending moment in the slab: mx.max = 422.7 kNm/m
(x = 49.1 m, y = 2.69 m)
Respective axial force: nx = -2743 kN/m
Maximum bending moment in the edge beam: mx.max = 407.9 kNm/m
(x = 48.3 m)
Figure C.32 shows the internal shear forces in transverse direction at the section
where the maximum value was found. The maximum occurs at the hanger node which
cannot be relevant for the shear design check. The decisive value was taken at a
distance y = 1.5 d (d : effective depth of the member). For punching shear, the
vertical component of the maximum hanger force is critical.
shear forces [kN/m]
axial forces [kN/m]
Fig. C.32. Transverse shear forces vy and axial forces ny [kNm/m], x = 51.5 m
C-18
Design-relevant internal forces taken from Figure C.32 (a summary of all

forces relevant for assessment can be found in Figure C.34):
Critical shear force: vSd.y = 710.2 kN/m
(y = 1.5 d = 1.5 565 mm = 847.5 mm)
Maximum hanger force (No. 4): Nmax = 1061.9 kN
Vertical component with = 75.2: Nmax.v = 1027 kN
615
478
382
287
191
96
-96
-191
-287
-382
-178
-573
-669
-765
-860
-956
-1051
-1105
Fig. C.33. Longitudinal shear forces vx [kNm/m], x = 4060 m
Design-relevant internal forces taken from Figure C.33:

Critical shear force: vSd,x = 615.1 kN/m (taken at a distance
y = 1.5 d from the maximum value)
Summary
Summary of design-relevant internal forces
m [kNm/m] v [kN/m] n [kN/m]
transverse 628.7 -3139

Bending
longitudinal slab 422.7 -2743
edge beam 407.9 -1691
transverse 710.2 -3113

Shear
longitudinal 615.1 -2410
Punching 1027 [kN]
Fig. C.34. Summary of design-relevant internal forces for ultimate limit states
C-19
In the following figures (C.35 to C.38), only the mid-range between x = 40 m and
60 m is illustrated. The footpath is not shown. It was decided not to include illustrations
of all relevant stress distributions for each combination of actions. Instead, it was
decided to include one representative example for each direction and location (top and
bottom fibre). Figures C.35 and C.36 contain the maximum compressive stresses at the
top fibre occurring in the non-frequent combination of actions. Figures C.37 and C.38
show maximum tensile stresses at the bottom fibre in the frequent combination,
relevant for crack width assessment. A summary of the design relevant forces can be
found in Figure C.39.
1.8
0.0
-1.1
-2.2
-3.3
-4.4
-5.5
-6.6
-7.7
-8.8
-9.9
-11.0
-12.1
-13.2
-14.3
-15.4
-16.5
-17.6
-18.7
-19.8
-20.2
2
Fig. C.35. Non-frequent combination of actions: Concrete stresses y.top [N/mm ], x = 4060 m
-2.3
-3.3
-4.1
-4.9
-5.7
-6.5
-7.3
-8.1
-9.0
-9.8
-10.6
-11.4
-12.2
-13.0
-13.8
-14.7
-15.5
-16.3
-17.1
-17.9
-18.6
2
Fig. C.36. Non-frequent combination of actions: Concrete stresses x.top [N/mm ], x = 4060 m
C-20
0.67
0.00
-0.42
-0.83
-1.25
-1.67
-2.09
-2.50
-2.92
-3.34
-3.75
-4.17
-4.59
-5.01
-5.42
-5.84
-6.26
-6.68
-7.09
-7.51
-7.67
2
Fig. C.37. Frequent combination of actions: Concrete stresses y.bottom [N/mm ], x = 4060 m
-0.30
-0.84
-1.26
-1.67
-2.09
-2.51
-2.93
-3.35
-3.77
-4.19
-4.60
-5.02
-5.44
-5.86
-6.28
-6.70
-7.12
-7.53
-7.95
-8.27
2
Fig. C.38. Frequent combination of actions: Concrete stresses x.bottom [N/mm ], x = 4060 m
C-21
Summary
Summary of design relevant concrete stresses

2 Longitudinal x Transverse y
Concrete stresses [N/mm ]
min max min max
Top -18.6 -1.51 -20.2 -2.5
Non-frequent Location deck centre edge beam deck centre ede beam
Bottom -9.38 4.6 -7.13 6.9
Location deck centre deck centre edge beam deck centre
Top -14.1 -3.09 -13.8 -1.25
Frequent Location deck centre edge beam deck centre edge beam
Bottom -9.75 -0.3 -8.23 0.67
Location smallest depth deck centre edge beam deck centre
Top -12.3 -5.86 -9.38 -1.16
Qquasi-permanent Location near diaphragm edge beam near diaphragm edge beam
Bottom -11.4 -3.41 -11.8 -4.66
Location smallest depth near diaphragm smallest depth edge beam
Top -16.5 -3.7 -12.3 2
After Location near diaphragm near diapgragm near diaphragm deck centre
prestressing Bottom -17.1 -4.4 -16.8 -4.2
Location smallest depth near diaphragm deck centre edge beam
Fig. C.39. Summary of design-relevant concrete stresses for serviceability limit states
C-22
C.5.3 Fatigue
The following table lists concrete stresses and shear forces necessary for
the assessment of concrete in compression and shear.
Stre s s e s and inte rnal force s : Concre te fatigue

Fre que nt com bination of actions
Com pre s s ion: ENV 1992-2: 1996 Section 4.3.7.4 (101)

Maximum compressive stress [N/mm2 ] -14.1
Corresponding minimum [N/mm ] 2
-0.3
She ar: ENV 1992-2: 1996 Section 4.3.7.4 (103)

Maximum shear f orce [kN/m] 702
Corresponding minimum [kN/m] 556
Punching: ENV 1992-2: 1996 Section 4.3.7.4 (104)

V ertical component of maximum hanger f orce [kN] 630
Corresponding minimum [kN] 305
Fig. C.40. Stresses and internal forces for concrete fatigue assessment
In Figure C.41, stresses in the prestressing tendons are calculated based

on an uncracked concrete section. Calculations were performed with the help of
the known concrete stresses at the top and bottom fibre, as well as the
geometric properties of the prestressed concrete section. The steel stress
ranges result from different steel stresses due to different loading. For the
conversion between concrete stress and steel stress, the elastic modulus ratio
n = 10 was used assuming equal strains.
The assumption of an uncracked section is an approximation, especially in
the case of two loaded tracks.
Steel stresses: Fatigue of prestressing steel ENV

1992-2: 1996 Section 4.3.7.5 (102) non-
frequent combination of actions
Concrete stresses [N/mm ]

2 Top Tendon
Stress range
at centre of bridge deck level
Bottom
-4.4 s,71 = 1+2 75

Dead load -7.5
-8.45 1 = 2 42
-9 with
Load model 71 on one track -3.3
-1.5 n = Es/Ec = 10
-20.2 h = 0.43 m
Load model 71 on two tracks 0
6.9 zp = 0.103 m
Fig. C.41. Steel stresses for prestressing steel fatigue assessment
C-23
C.5.4 Alternatives without transverse prestressing

The existing 3D model in SOFiSTiK was changed by removing the
transverse prestressing thread bars and increasing the slab thickness between
the edge beams. The internal forces of the slab in transverse direction were
taken directly from the software output of the new model. The internal forces in
the longitudinal bridge direction were assumed unchanged.
The following tables summarise design-relevant internal forces for the two
design proposals. The structural depth refers to the mid-section, where the
maximum moments occur. As with the main design, decisive shear forces were
taken at a distance y = 1.5 d from the hanger connection.
Alternative design proposal 1: h = 610 mm Alte rnative de s ign propos al 2: h = 470 m m
mSdy [kNm/m] mid-span 1195.6 m Sdy [k Nm /m ] mid-span 1041.2
vSdy [kN/m] at hangers 566.0 vSdy [k N/m ] at hangers 547.0
at smallest depth 383.0 at smallest depth 305.0
max NSd [kN] hanger 4 1110.5 m ax NSd [k N] hanger 4 1073.0
Fig. C.42. Maximum ULS internal forces, h = 610 mm Fig. C.43. Maximum ULS internal
forces, h = 470 mm
Serviceability limit states / Fatigue assessment
The non-frequent combination of actions is relevant for both fatigue and

serviceability limit state assessments. The following tables calculate steel
stresses in the reinforcement, assuming a cracked concrete section.
Calculation of steel stress in reinforcement - design proposal 1 (h = 610 mm) SCHNEIDER [26], Tafel
Combination:
2
As [cm /m]: 54.54 d [mm]: 557.5 2a, page 5.130
Non-frequent mSdy [kNm/m] Sd z [mm] s [N/mm ]
2
one LM 71 598.1 0.058 535.2 204.9

two LM 71 814.1 0.079 528.5 282.4
DL only 420.6 0.041 541.3 142.5
2
s1 [N/mm ] 62.4
2
s.71=s1+2 [N/mm ] 140.0
Fig. C.44. Calculation of stresses in reinforcement, structural depth h = 610 mm
Calculation of steel stress in reinforcement - design proposal 1 (h = 470mm)

2
Combination: As [cm /m]: 70.12 d [mm]: 398
Non-frequent mSdy [kNm/m] Sd z [mm] s [N/mm ]
2
one LM 71 484.0 0.093 374.9 184.1

two LM 71 700.0 0.134 363.8 274.4
DL only 306.5 0.059 382.1 114.4
2
s1 [N/mm ] 69.7
2
s.71=s1+2 [N/mm ] 160.0
Fig. C.45. Calculation of stresses in reinforcement, structural depth h = 470 mm
C-24
C.6 Internal forces of the end cross girder
The maximum internal forces occurring for the ultimate limit state are shown
in Figure C.46.
.2
61
-10
.5
-92
.2 .0 3.6 785.0
19 75 17.1 .8 -94 -
1 5.0 0.4 -75
5 0.2 14 .8
16 167 66.5 60.4
16 1 1 6.3 9.3
14 11 .7 4.3 My [kNm/m]
4 6.2 77 2
-70 9 6 .5 2
.
-45 4056 6
- 73. .4 8 .1
-25 -2542 629. 84.4 25
-2 6.0 9 17
-26 -268 659. 50.6 3.6
-2 -26 -270 27.
8
8.5 4.6 0.5 0.1
-27 -268 9 3 0.7 52.4
.9 -25 -261 -40 -424 -53
1 8
-6 1 N [kN/m]
5.1 8 .1
-34 -94.9 1.0
7.0 34 551. 98.6 5.5 -1 8 5.0 .1 9 4
360 5 40
.2
263 162.4 7.6 7.2 -28 -441 -655. 11.7 62.
-6 1.9 -32
7 1 -38
Vy [kN/m]
0
1.27
2.54
5.7
624
3.81
5.075
6.34 x [m]
7.61
8.88
10.15
Fig. C.46. Internal forces for assessment of end cross girder in ultimate limit state
The serviceability limit state to be verified by numerical assessment is

limitations for stress. The critical stresses on the top and the bottom surface of
the end cross girder are presented in Figure C.47.
Non-fre que nt load Fre que nt load Quas i-pe rm ane nt load
com binations com binations com binations
top le ve l bottom le ve l top le ve l bottom le ve l top le ve l bottom le ve l
M ax s tre s s
-4.03 -2.4 -2.4 -2.84 -1.3 -5.2
[N/m m 2 ]
M in s tre s s
-8.5 -7.9 -7.82 -9.2 -5.3 -9.6
[N/m m 2 ]
Fig. C.47. Stresses in end cross girder in serviceability limit state, transverse direction
C-25
C.7 Actions on the bearings
Two types of bearings were examined (for details see Section 5.9):
1. Two pot bearings at each of the four supports (Figure C.48.)
2. Stilt bearing with compression support in the middle, suggestion by
P. Tveit (Figure C.49)
bridge deck
non-sliding
y-direction slide
x-direction slide
all direction slide
Fig. C.48. Pot bearings TF-10, TGa-10, TGe-10, MAURER SHNE GmbH & Co.KG [20]
Fig. C.49. Stilt bearing, arrangement of the slide directions as in Figure C.48
The bearings had to be assessed for maximum vertical and horizontal

forces, horizontal deflections and tilt/angular rotations. The minimum vertical
forces do not undershoot 0 kN, and therefore there is no danger of lift-off. The
internal forces were calculated in ultimate limit state.
It is unusual to have three supports at an end cross girder about 10 metres
wide as in the stilt bearing. Therefore Figure C.50 shows the bearing forces
with and without an additional compression support in the middle of the end
cross girder, not only for the assessment of the bearings but also for general
perceptions about a middle support.
C-26
V e rtical be aring force s Fv and be nding m om e nts about the longitudinal axis of the bridge M x in ultim ate
lim it s tate
Pot bearings w ithout support in the middle Stilt-bearing w ith compression support in the middle
Load combination Fv [k N] M x [k Nm ] Fv [k N] M x [k Nm ] Fv [k N] M x [k Nm ] Fv [k N] Fv [k N] M x [k Nm ]
number lef t right lef t middle right
1 -14115 1204.46 -10791 -1004.94 -13847 430.9 -640.5 -10421 -247.58
1 LM 71
traffic dominant
lef t 2 -14189 1071.7 -10801 -998.06 -13860 359.92 -560.5 -10472 -294.6
3 -14031 1168.34 -10974 -921 -13694 471.32 -538.5 -10656 -219.28
4 -16052 2081.96 -15742 -2344.64 -15330 628.8 -1337 -15029 -941.5
2 LM 71 5 -16031 2004.42 -15743 -2279.42 -15365 606.02 -1225 -15087 -934.46
6 -15848 2040.52 -15905 -2125.22 -15238 720.28 -1116 -15402 -864.12
1 LM 71 7 -13356 851.44 -10301 -877.7 -13039 216.06 -495.5 -10005 -267.34
wind dominant
lef t 8 -13322 796.52 -10307 -881.06 -13044 181.16 -455.5 -10033 -292.38
9 -14806 1565.16 -14261 -1949.74 -14226 374.56 -1053 -13692 -822.76
2 LM 71
10 -14789 1502.86 -14262 -1897.52 -14253 356.14 -963.5 -13738 -816.8
unloaded train 11 -9877 -92.94 -9168 -206.56 -9877 -92.94 0 -9479 -206.56
Fig. C.50. Bearing forces in ultimate limit state, wind forces from the right side, design check relevant values are framed
The idea of P.TVEIT is, as described in Section 5.9, that without live load the
distance between the middle support and the bridge deck is 3 mm. The FEM-
software used for the analysis does not provide such bearing conditions.
Therefore the forces for the supports below the arches at a deflection of 3 mm
cannot be obtained by numerical analysis. Since the analysis was performed in
linear fashion, it is possible to scale the maximum values of the calculation
without middle support:
Deflection due to load combination 4: 1 = -7.5 mm
Deflection due to dead load: 2 = 0 mm
Maximum vertical force due to dead load: Fv2 = -8753 kN
Maximum bending moment due to dead load: Mx2 = 13.85 kNm
3mm
Fv(1 = -3 mm) = (Fv1 - Fv2) + Fv2
1 2
3mm
Fv(1 = -3 mm) = ( 15742kN + 8753kN) 8753kN = 11548.6kN
7,5mm
3mm
Mx(1 = -3 mm) = M x1 M x 2 + M x2
1 2
3mm
Mx(1 = -3 mm) = ( 2345kNm 13.85kNm ) + 13.85kNm = 929.7kNm
7,5mm
The obtained values do not give decisive forces.
For the load combinations 1-10 the deflections of the end cross girder
without the middle support are all greater than 3 mm, and so the forces after
contact can be simulated by a settlement of the middle support of 3mm. This
was applied while calculating the forces of the stilt bearing shown in Figure
C.50.
The determination of the force acting on the middle support is difficult
because in addition to the static forces it receives impulsive loads. The dynamic
enlargement of this static force is 20% according to a proposal by P. TVEIT.
C-27
The maximum forces are:

1. For the pot bearings:
Fv = -15742 kN
FH = 5393.4 kN
Mx= -2344.6 kNm
2. For the stilt bearing:
left / right (see Figure C.50)
Fv = -15238 kN / -15087 kN Section A.4
FH = 4113.4 kN
Mx = 720 kNm / 934.5 kNm
3. For the compression support:
Fv = -1337 kN 1.2
The maximum horizontal deflections are:
1. Deflections due to changes of temperature; load case 4a

minl = -0.037 m
maxl = 0.04 m
2. Deflections due to creep and shrinkage
c,S,K= c,S,u +
(
u,28 cg + cp0 )
= 0.0003 +
(
1.6 3.456N / mm 2 11.6N / mm 2 )
Ec 2
37000N / mm
-4
Section B
c,S,K = -6.52 10
l = c,S,K 100m = -0.065m

3. Total
minl = -0.037-0.065 m = 0.102 m
maxl = 0.04 m = 0.04 m
The maximum angular rotation is:
1 = 3.49 mrad
C-28
Annex D
Assessment of the bridge
D.1 Materials, cross-sections, assessment parameters
Concrete
C50/60
ENV 1992-1-1: 1991
Characteristic compressive strength fck = 50 N/mm2
Table 3.1
Mean tensile strength fctm = 4.1 N/mm2
ENV 1992-1-1: 1991
2
Secant elastic modulus Ecm = 37 kN/mm Table 3.2
ENV 1992-1-1: 1991

Reinforcing steel 500 S Section 3.2
Yield strength: fy = 500 N/mm2 SCHNEIDER [29],
Elastic modulus: E = 200000 N/mm2 page 5.26
Prestressing steel
Longitudinal prestressing DYWIDAG Bonded Post-Tensioning System DYWIDAG-SYSTEMS
Strand type: 15 mm according to prEN 10138 INTERNATIONAL [10]
Tendon type: 6827
Steel strength: 1770/1500
Transversal prestressing DYWIDAG Bonded Post-Tensioning
Thread bars: Type 36D
Steel strength: 1230/1080
Structural steel
Steel grade S460ML

For structural elements: Arch profile, hanger, hanger connections ENV 1993-2: 1997
Table 3.1a
t 40mm Yield strength: fy = 460 N/mm2
Ultimate tensile strength: fu = 550 N/mm2
Elastic modulus: E = 210000 N/mm2
40 mm < t 100 mm Yield strength: fy = 430 N/mm2

Ultimate tensile strength: fu = 530 N/mm2 ENV 1993-2: 1997
Elastic modulus: E = 210000 N/mm2 Table 3.1b
Steel grade S355H

For structural elements: Wind bracing
t < 40 mm Yield strength: fy = 355 N/mm2
Ultimate strength: fu = 510 N/mm2
Elastic modulus: E = 210000 N/mm2
D-1
Diploma Thesis Brunn & Schanack Annex D: General
Arch profiles: American Wide Flange ARCELOR LONG

COMMERCIAL S.A. [4]
W 360x410x634 W 360x410x990
A [cm2] 808 1262
Iz [cm4] 274200 450200
Wel,z [cm3] 11570 16960
Wpl,z [cm3] 14220 21620
Iy [cm4] 98250 15330
Wel,y [cm3] 4634 6938
Wpl.y. [cm3] 7117 10710
474 531
77.1 319.8 77.1 106 319 106
188.1
188.2
y y
47.6
65.9
424
442
188.2
188.1
z z
American Wide Flange American Wide Flange

W 360 x 410 x 634 W 360 x 410 x 900
Wind bracing: Circular hollow sections
CORUS [7]
CHS 219.1x8.0 CHS 219.1x10.0
A [cm2] 53.1 65.7
I [cm4] 2960 3598
Wel [cm3] 270 328
Wpl [cm3] 357 438
CHS 219.7 x 8 CHS 219.7 x 10
Hanger
Smooth circular steel bar, D = 60 mm
D-2
Diploma Thesis Brunn & Schanack Annex D: General
Partial safety factors
Member assessment
Ultimate limit states
M0 = 1.0 ENV 1993-2: 1997
M1 = 1.1 Section 5.1.1
M2 = 1.25
Serviceability limit states
M,ser = 1.25
Connection assessment
Welds ENV 1993-2: 1997
Section 6.1
Mw = 1.25
Bolts
Mb = 1.25
Slip resistance, standard nominal clearance
ENV 1993-2 Section
Ms,ult = 1.25 6.4.7.1
Ms,serv = 1.1
Joints
ENV 1993-1-1: 1992
Mj = 1.1 Annex K.1 (8)
Fatigue assessment
Structural steel elements ENV 1993-2: 1997
Ff = 1.0 Section 9.3
Mf = 1.0 (for redundant structural elements: Hangers, wind bracing)
Mf = 1.15 (for key structural elements: Arch profile)
Reinforced/prestressed concrete elements ENV 1992-2: 1997
F = 1.0 Section 4.3.7.2
Sd = 1.0
c,fat = 1.5 for concrete
s,fat = 1.15 for reinforcement and prestressing steel
Data on fatigue assessment
The following was presumed and is applicable for all fatigue assessments in
Annex D of this work.
EC Mix L = 100 m 1 = 0.6 (ENV 1993-2 Table 9.5)
Traffic per year: 30 106 t/track 2 = 1.04 (ENV 1993-2 Table 9.6)
Design life: 100 years 3 = 1.0 (ENV 1993-2 Table 9.7)
n = 0.12
Data on reinforced/prestressed concrete members ENV 1992-1-1: 1991
The following was presumed and is applicable for Annex D, sections D.5, D.6. Table 4.1
Environmental exposure class: 2b ENV 1992-2: 1996

For decompression state and limit state of cracking: Category C Table 4.118
D-3
Diploma Thesis Brunn & Schanack Annex D: Assessment of the arch
D.2 Arch
D.2.1 Ultimate limit state assessment
D.2.1.1 Collapse about weak axis y-y
Design check at the root of the arch
Values of the section vide Section D.1

1. Shear resistance
ENV 1993-1-1: 1992,
VSd,y = -198.3 kN VSd,z = -511.5 kN 5.4.6
fy
Vpl,Rd = A v
3 M1
Av = A-2 b tf+(tw+2 r) tf Av = A-d tf

= 1149 cm2-2 44.2 cm 10.6 cm = 1149 cm2-28.9 cm 10.6 cm
+(6.59 cm+2 1.5 cm) 10.6 cm
= 313.6 cm2 = 842.7 cm2
430N / mm 2 430N / mm 2
Vpl,Rd = 313.6cm 2 Vpl,Rd = 842.7cm 2
3 1 .1 3 1 .1
= 7077.6 kN = 19019 kN
|VSd|= 198.3 kN < Vpl,Rd= 7077.6 kN |VSd|= 511.5 kN < Vpl,Rd= 19019 kN
Since the design values of the shear force VSd do not exceed 50% of the
ENV 1993-1-1: 1992,
design plastic shear resistance Vpl,Rd, no reduction need be made in the plastic
5.4.7 (2)
resistance moment.
2. Bending and axial force
Section C.2.4, Figure

NSd = -28468.3 kN
C-8
My,Sd = 1115.8 kNm
Mz,Sd = -809.9 kNm
A fy 1149cm 2 430N / mm 2
Npl,Rd = = = 44915.5 kN
M1 1.1
Wpl,y f y 10710cm3 430N / mm2
Mpl,y,Rd = = = 4186.6 kNm ENV 1993-1-1: 1992,
M1 1 .1 5.4.4
Wpl,z f y 21620cm3 430N / mm2
Mpl,z,Rd = = = 8451.5 kNm ENV 1993-1-1: 1992,
M1 1 .1 5.4.5.1
Criterion to be satisfied:
N Sd M y,Sd M z,Sd
+ + 1 ENV 1993-1-1: 1992,
Npl,Rd Mpl,y,Rd Mpl,z,Rd
5.4.8.1 (5.36)
28468.3kN 1115.8kNm 809.9kNm
+ + = 0.996 1
44915.5kN 4186.6kNm 8451.5kNm
D-4
Design check at the first splice
1. Shear resistance
ENV 1993-1-1: 1992,
5.4.6
VSd,y = 135.4 kN VSd,z = -310.5 kN
fy
Vpl,Rd = A v
3 M1

= 948.1 cm2-2 43.2 cm 8.89 cm = 948.1 cm2-29.02 cm 8.89 cm
+(5.56 cm+2 1.5 cm) 8.89 cm
= 256.1 cm2 = 690.1 cm2
430N / mm 2 430N / mm 2
Vpl,Rd = 256.1cm 2 Vpl,Rd = 690.1cm 2
3 1.1 3 1.1
= 5779.9 kN = 15574.9 kN
|VSd|= 135.4 kN < Vpl,Rd= 5779.9 kN |VSd|= 310.5 kN < Vpl,Rd= 15574.9 kN
ENV 1993-1-1: 1992,
design plastic shear resistance Vpl,Rd, no reduction need be made in the plastic 5.4.7 (2)
resistance moment.
NSd = -27487.3 kN Section C.2.4, Figure

My,Sd = 210.8 kNm C-8
Mz,Sd = -100.4 kNm
A ,net f y 771cm 2 430N / mm 2

Npl,Rd = = = 30139.1 kN
M1 1 .1
Wpl,y,net f y 6908cm3 430N / mm2 ENV 1993-1-1: 1992,
Mpl,y,Rd = = = 2700.4 kNm
M1 1 .1 5.4.4
Wpl,z,net f y 12138cm3 430N / mm2 ENV 1993-1-1: 1992,
Mpl,z,Rd = = = 4744.9 kNm
M1 1.1 5.4.5.1
N Sd M y,Sd M z,Sd
+ + 1
27487.3kN
210.8kNm 100.4kNm ENV 1993-1-1: 1992,
+ + = 1.01 1.03 5.4.8.1 (5.36)
30139 .1kN 2700.4kNm 4744.9kNm
An overstepping of the allowed values up to 3% is tolerated.
D-5
D.2.1.2 Collapse about the strong axis z-z
Design check at the root of the arch
1. Shear resistance ENV 1993-1-1: 1992,

5.4.6
VSd,y = -198.3 kN VSd,z = -511.5 kN
fy
Vpl,Rd = A v
3 M1

= 1149 cm2-2 44.2 cm 10.6 cm = 1149 cm2-28.9 cm 10.6 cm
+(6.59 cm+2 1.5 cm) 10.6 cm
= 313.6 cm2 = 842.7 cm2
430N / mm 2 430N / mm 2
Vpl,Rd = 313.6cm 2 Vpl,Rd = 842.7cm 2
3 1 .1 3 1 .1
= 7077.7 kN = 19019 kN
|VSd|= 514.7 kN < Vpl,Rd= 7077.7 kN |VSd|= 428.8 kN < Vpl,Rd= 19019 kN
Since the design values of the shear force VSd do not exceed 50% of the ENV 1993-1-1: 1992,
design plastic shear resistance Vpl,Rd, no reduction need be made in the plastic 5.4.7 (2)
resistance moment.

Section C.2.4, Figure
NSd = -28146.8 kN C-8
My,Sd = 841.8 kNm
Mz,Sd = -1840.8 kNm
A fy 1149cm 2 430N / mm 2
Npl,Rd = = = 44915.5 kN
M1 1.1
Wpl,y f y 10710cm3 430N / mm2
Mpl,y,Rd = = = 4186.6 kNm ENV 1993-1-1: 1992,
M1 1 .1 5.4.4
Wpl,z f y 21620cm3 430N / mm2
Mpl,z,Rd = = = 8451.5 kNm ENV 1993-1-1: 1992,
M1 1 .1 5.4.5.1
N Sd M y,Sd M z,Sd
+ + 1
28146 .8kN
841.8kNm 1480 .8kNm ENV 1993-1-1: 1992,
++ = 1.003 1.03
44915 .5kN 4186 .6kNm 8451 .5kNm 5.4.8.1 (5.36)
An overstepping of the allowed values up to 3% is tolerated.
D-6
D.2.2 Fatigue assessment
Axial stresses
The fatigue check has to be carried out for four critical cross sections. So,
first the necessary formulas are given and then the values will be presented in
Figure D.1234.
N min M y,min M z,min

p,min = + dead load
A ,net W el,y,net W el,z,net
Nmax,1 M y,max,1 M z,max,1
p,max,1 = + dead load and one LM 71
Nmax,1+ 2 M y,max,1+ 2 M z,max,1+ 2
p,max,1+ 2 = + dead load and two LM 71
1 = |p,max,1-p,min|
1+2 = p = |p,max,1+2-p,min|
E2 = 2 p
2 is already included in the stresses due to LM 71
ENV 1993-2: 1997,

= 1 2 3 4
Table 9.3
1 = 0.6 EC Mix, L=100m
Table 9.6
Table 9.7

5
1
5 ENV 1993-2: 1997,
4 = 5 0.12 + 0.88 1
+ 1
9.5.3 (8)
1+ 2
1+ 2
c for: ENV 1993-2: 1997
1. Connection to wind truss member c = 90 N/mm2 Table L.1 sheet 2
2. Hanger connection c = 80 N/mm2 Table L.1 sheet 2
3. Transverse splice between profiles c = 80 N/mm2 Table L.3 sheet 1
4. Root of arch at top edge of concrete tie c = 160 N/mm2 Table L.1 sheet 1
Mf = 1.15 ; key structural element
1. 2. 3. 4.
p,min [N/m m 2 ] -145.99 -160.20 -138.90 -141.20
p,max,1 [N/m m 2] -222.40 -220.90 -208.90 -289.70
p,max,1+2 [N/m m 2 ] -250.50 -276.50 -235.80 -334.80
1 [N/m m 2 ] 76.41 60.70 70.00 148.50
1+2 [N/m m 2 ] 104.51 116.30 96.90 193.60
4 0.79 0.71 0.78 0.81
0.49 0.44 0.49 0.51
E2 f [N/m m 2 ] 51.43 51.27 47.35 98.16
< < < <
c / Mf [N/m m 2 ] 78.3 69.6 69.6 139.13
Fig. D.1. Fatigue checks for axial stresses
D-7
Shear stresses
The maximum shear forces occurring in ultimate limit state do not exceed
2.5 % (see Section D.2.1.1) of the permitted limits, so the shear stresses are
considered to be not critical to fatigue strains. The fatigue check will be omitted.
Full penetration butt weld at arch root point
Force due to Force due to Force due to

FE2 = F1+2 F1
dead load LM 71 two LM 71
NSd [kN] -11208.3 -17548 -20219 9010.7 6339.7
Vy [kN] -20.18 -33.5 -35.8 15.62 13.32
Vz [kN] 67.3 84 139.7 72.4 16.7
My [kNm] 229.2 355.7 404.1 174.9 126.5
Mz [kNm] -47 -28.7 -129.7 82.7 18.3
MT [kNm] 3.613 -0.602 -0.803 4.416 4.215
c 106 159.5 ENV 1993-2: 1997,

Ff E2 Section 9
Mf A
E2 = 2 p
221
Section properties y
442
A = 2 (106mm 442mm ) = 93704mm 2
Wy = 2
(442mm )2 106mm = 6.903 10 6 mm 3 z
6
2
(106mm )3 442mm + 2 106mm 442mm 159.5mm + 106mm 2
12 2 = 1.627 10 7 mm 3
Wz =
106mm + 159.5mm
The influence of the shear forces and the torsional moment will be ignored, as
their contribution is very small and not decisive.
Stress variation due to two LM 71:

N1+ 2 M y1+ 2 M z1+ 2
p = 1+ 2 = + +
A Wy Wz
9010.7kN 174.9kNm 82.7kNm
p = 1+ 2 = 2
+ 6 3
+
93704mm 6.903 10 mm 1.627 10 7 mm 3
N N N N
p = 1+ 2 = 96.16 2
+ 25.34 2
+ 5.08 2
= 126.58
mm mm mm mm 2
Stress variation due to one LM 71:

N1 M y1 M z1
1 = + +
A Wy Wz
6340kN 126.5kNm 18.3kNm
1 = 2
+ 6 3
+
93704mm 6.903 10 mm 1.627 10 7 mm 3
N N N N
1 = 67.66 2
+ 18.3 2
+ 1.12 2
= 87.1
mm mm mm mm 2
D-8
1 = 0.6 EC Mix L = 100 m ENV 1993-2: 1997

2 = 1.04 Traffic per year:30 106 t/track Table 9.5
3 = 1.0 Design life: 100 years Table 9.6
Table 9.7
(
4 = 5 n + (1 n) a 5 + (1 a )
5
)
n = 0.12
1 87.1N mm 2
a= = = 0.688
1+ 2 126.58 N mm 2
[
4 = 5 0.12 + [1 0.12] 0.6885 + (1 0.688 ) = 0.76
5
]
= 1 2 3 4 = 0.6 1.04 1.0 0.76 = 0.47
E2 = 2 p
N N
E2 = 0.47 126.58 2
= 59.49
mm mm 2
Detail category: 80 (ENV 1993-2:1997 Table L.3/1
N
c = 80
mm 2
Mf = 1.15 (key structural element)

Ff = 1.0
N 80 N mm 2 N
1.0 59.49 = 69.6
mm 2 1.15 mm 2
D.2.3 Serviceability limit state assessment
D.2.3.1 Limitation of nominal stress for characteristic/rare

load combinations
The critical section is found at the root of the arch. The net cross section is
used for the design check, which is on the safe side.
Axial stress
N ser M y,ser M z,ser 21326 .2kN 446.2kNm 468.1kNm

Ed,ser = + = +
A ,net W el,y,net W el,z,net 1098.12cm 2 6775cm 3 14007cm 3
= -161.8 N/mm2
Shear stress
Vy,ser 86.5kN
Ed,ser = =
A vy 313.6cm 2
= -2.78N/mm2
D-9
The criterion to be satisfied is:

fy
( Ed,ser )2 + 3( Ed,ser )2 < ENV 1993-2: 1997,
M,ser
4.3 (1)
( 161.8N / mm ) 2 2
(
+ 3 2.78N / mm 2 )
2
= 161.9 < 430N / mm 2 =
430N / mm 2
1 .0
D.2.3.2 Limitation of nominal stress range for frequent load

combinations
The critical section is found at the root of the arch. The stress range will be
calculated from the variation of internal forces. The net cross section is used for
design check, which is on the safe side.
N ser M y,ser M z,ser 5758.5kN 95.6kNm 0.9kNm

Ed,ser = + + = 2
+ 3
+
A ,net W el,y,net W el,z,net 1098.12cm 6775cm 14007cm 3
= 66.6 N/mm2
fy
Ed,ser < 1.5 ENV 1993-2: 1997,
M,ser
4.3 (4)
430N / mm 2
Ed,ser = 66.6N / mm 2 < 645N / mm 2 = 1.5
1 .0
D.2.3.3 Limits for clearance gauges
The structural part nearest to the clearance gauge is the lowest member of
the wind truss. The shortest distance is 20 cm. The wind truss moves maximally
18.7 cm towards the railway traffic with a minimum upwards deflection of 4 cm
in ultimate limit state. This reduces the shortest distance to 10 cm. Since the
loads and the deflections are smaller for partial safety factors according to
serviceability limit state, there is no danger of structural parts encroaching upon
the clearance gauge.
D-10
Diploma Thesis Brunn & Schanack Annex D: Assessment of hanger and hanger connections
D.3 Hanger and hanger connections
D.3.1 General
Steel strength of different members
2 2
Member Criteria fy [N/mm ] fu [N/mm ]
Hangers 40mm < t 100mm 430 530
Connection plates t 40mm 460 550
Arch profile 40mm < t 100mm 430 530
Fig. D.2. Steel strength of different members
Design relevant internal forces (Critical: hanger No. 4)

Ultimate limit states
Design tensile force in the hanger NSd,ult [kN] 1061.9
Serviceability limit states
Rare/characteristic combination of actions NSd,ser [kN] 785.5
Fatigue assessment
Force range due to load model 71 on one track N1,Sd [kN] 295.3
Force range due to load model 71 on two tracks Np = N1+2,Sd [kN] 439.7
Fig. D.3. Design-relevant hanger forces
Eccentricity of connection detail to the arch profile

The connection detail is eccentric to the arch profile, which has to be taken
into account. The resulting bending moment contributes to a higher force
transmitted into one flange. In the following, a factor () is calculated, which is
then applied in the ultimate limit state and fatigue assessment.
Hanger type 4: Eccentricity of hanger to the arch profile: e = 3 cm
Bending moment (virtual hanger force = 1 [-]): M = 1 0.03 m = 0.03 m
Clearance between flanges: s = 31.98 cm
= M / s = 0.03 m / 0.3198 m = 0.094
For ultimate limit state: NSd = 1061.9 kN
NSd.ecc = NSd / 2 + NSd F = 1061.9 kN / 2 + 1061.9 kN 0.094 = 630.8 kN
For fatigue assessment: (Np = 440 kN)
Np.ecc.1 = Np / 2 + Np F = 440 kN / 2 + 440 kN 0.094 = 261 kN
Np.ecc.2 = 440 kN 261 kN = 179 kN
For determination of SCF (N = 1000 kN):
Necc.1 = 1000 kN / 2 + 1000 kN 0.094 = 594 kN
Necc.2 = 1000 kN 571 kN = 406 kN
Hanger type 5: Eccentricity of hanger connection type 5: e = 2 cm
Bending moment (virtual hanger force = 1 [-]): M = 1 0.02 m = 0.02 m
Width of hanger gusset plate: s = 28.0 cm
= M / s = 0.02 m / 0.28 m = 0.0714
For determination of SCF (N = 1000 kN):
Necc.1 = 1000 kN / 2 + 1000 kN 0.0714 = 571 kN
Necc.2 = 1000 kN 571 kN = 429 kN
For fatigue assessment (Np = 440 kN):
Np.ecc.1 = 440 kN / 2 + 440 kN 0.0714 = 251 kN
Np.ecc.2 = 440 kN 251 kN = 189 kN
D-11
D.3.2 Ultimate and serviceability limit state assessment
D.3.2.1 Hanger
Section properties
Cross section: circular

Diameter: D = 60 mm
Cross section area: A = 2827.43 mm2

ENV 1993-1-1: 1992
Criterion: NSd < N t.Rd Section 5.4.3
Design plastic resistance

A fy
N t.Rd = Npl.Rd =
M0
2827.43mm 2 430 N mm 2
Npl.Rd = = 1215.8kN
1 .0
NSd = 1061.9kN < Npl.Rd = 1215.8kN
fy ENV 1993-2: 1997

Ed.ser <
M.ser Section 4.3
Fser 785.5kN N
Ed.ser = = 2
= 277.82
A 2827.43mm mm 2
fy 430 N mm 2
= = 430 N mm 2
M.ser 1. 0
N fy N
Ed.ser = 277.82 2
< = 430
mm M,ser mm 2
D-12
D.3.2.2 Hanger connections
D.3.2.2.1 Connection type 1
Member assessment
cross section 1-1

1
double fillet
joint, a = 8 mm
III W 360x410x634
I II
B
1

strength bolts 220
d = 24 mm 75 70 75
single bevel 40 70 70 40
A e2 p2
III - detail category 36

60
e1
ENV 1993-2 Table L.5/1 154.9 70 94.9

Root failure in fillet weld
65
p1
227
277
II - detail category 80 t = 20 mm
60
ENV 1993-2 Table L.5/3 t = 10 mm
Fillet joint in shear
185
L3
p1
65
497
I - detail category 112
60
e1
ENV 1993-2 Table L.1/2
Preloaded high strength bolt
270
L2
220
B - detail category 36 119.9 70 69.8 60.1

ENV 1993-2 Table L.3/2 L4
319.8
Transverse butt weld
A - detail category 45 arch gusset plate
L1
80 60 80
Crack in hanger
hanger gusset plates (2)
Fig. D.4. Hanger connection type 1
Hanger gusset plate
Section properties
Number of plates: n=2

Width: w = 220 mm
Thickness: t = 10 mm
Total cross section area: A = 4400 mm2
Net cross section area: Anet = 2840 mm2 (3 holes with d0 = 26 mm)
Criterion: NSd < N t.Rd ENV 1993-1-1: 1992

Section 5.4.3
Design plastic resistance:
A fy
N t.Rd = Npl.Rd =
M0
D-13
4400mm 2 460 N mm 2
Npl.Rd = = 2024kN
1. 0
N Sd = 1061.9kN < N t.Rd = Npl.Rd = 2024kN
Design ultimate resistance of net cross-section at holes for fasteners

fu 550 N mm 2
N t.Rd = Nu.Rd = 0.9 A net = 0.9 2840mm 2 = 1124.6kN
M2 1.25
NSd = 1061.9kN < N t.Rd = Nu.Rd = 1124.6kN
fy
Ed.ser < ENV 1993-2: 1997
M.ser
Section 4.3
Fser 785.5kN N
Ed.ser = = = 204.6
A net 2840mm 2 mm 2
fy 430 N mm 2
= = 430 N mm 2
M.ser 1. 0
N fy N
Ed.ser = 204.6 2
< = 430
mm M.ser mm 2
Arch gusset plate
Section properties
Number of plates: n=1

Shear length: L3 = 185 mm
Thickness: t = 20 mm
Gross shear area: Av.gross = 2 185 mm 20 mm = 7400 mm2
Net shear area: Av = 5320 mm2 (2 holes with d0 = 26 mm)
The gusset plate is welded to the flanges as well as to the web. However,
the design checks consider only the welding to the flanges.

ENV 1993-1-1: 1992
Criterion: VSd < Vpl.Rd Section 5.4.6
Design force: VSd = NSd = 1061.9kN
Design plastic shear resistance:
A v fy 5320mm 2 460 N mm 2
Vpl.Rd = = = 1413kN
3 M0 3 1 .0
VSd = 1061.85kN < Vpl.Rd = 1232kN
D-14
Serviceability limit state ENV 1993-2: 1997

Section 4.3
fy
Ed.ser <
3 M.ser
VSd 785.5kN N
Ed.ser = = = 147.7
Av 5320mm 2 mm 2
fy 430 N mm 2
= = 248.6 N mm 2
3 M.ser 3 1 .0
N fy N
Ed.ser = 169 2
< = 248.6
mm 3 M.ser mm 2
Connection assessment
Bolt connection (I)
Type: Category C, slip-resistant at ultimate limit state

Properties
Bolt grade: 10.9

Yield strength: fyb = 900 N/mm2 Schneider [29],
Ultimate tensile strength: fub = 1000 N/mm2 Page 8.81
Bolt diameter: d = 24 mm
Hole diameter: do = d + 2mm = 26 mm (standard nominal clearance)
Bolt cross section area: A = 452 mm2
Stress area: As = 353 mm2
Number of bolts: n=6
Geometry
Distances measured from centres of fasteners: ENV 1993-1-1: 1992

Section 6.5.1
End distance in load direction: e1
Edge distance perpendicular to load direction: e2
Spacing between fasteners in load direction: p1
Spacing between rows of fasteners perpendicular to load direction: p2
Limits:
e1 = 60mm min e1 = 1.5 d0 = 39 mm ENV 1993-2: 1997
e2 = 40mm min e2 = 1.5 d0 = 39 mm Section 6.3
p1 = 65mm min p1 = 2.5 d0 = 65 mm
p2 = 70mm min p2 = 2.5 d0 = 65 mm
Maximum end and edge distances: members exposed to weather
e1 = 60 < e1.max = 40 mm + 4 t = 80 mm
(t: thickness of thinner outer connected part)
Maximum spacing in tension members
Outer row: p1.0 = 65 mm < p1.0.max = 14 t = 168 mm (200 mm)
Inner row: p1.j = 65 mm < p1.j.max = 24 t = 336 mm (400 mm)
D-15
Criteria
ENV 1993-1-1: 1992
Fv.Sd Fs.Rd slip resistance
Table 6.5.2
Fv.Sd Fb.Rd bearing resistance
Design shear force for ultimate limit state
N Sd 1061.9kN
Fv.Sd = = = 177kN
n 6
Bearing resistance
ENV 1993-1-1: 1992
2.5 fu d t Table 6.5.3
Fb.Rd =
Mb
e1 60mm
where is the smallest of: = = 0.77
3 do 3 26mm
p1 1 65mm 1
= = 0.58
3 do 4 3 26mm 4
fub 1000 N mm 2
= = 1.82
fu 550 N mm 2
or 1
= 0.58
Arch gusset plate:
2.5 0.58 550 N mm 2 24mm 20mm

Fb.Rd = = 306kN
1.25
Fv.Sd = 177kN < Fb.Rd = 306kN
Hanger gusset plate:
2.5 0.58 550 N mm 2 24mm 10mm 2

Fb.Rd = = 306kN
1.25
Fv.Sd = 177kN < Fb.Rd = 306kN
Slip resistance
ks n
Fs.Rd = Fp.Cd
Ms
ENV 1993-1-1
Slip factor: = 0.5 Class A surface: Section 6.5.8
ENV 1993-2: 1997 Section 6.4.7.1 (6)
ENV 1993-1-1: 1992 Section 6.5.8.3
ks = 1 standard nominal clearance
n=2 number of friction interfaces
Fp.Cd preloading force
Fp.Cd = 0.7 fub As = 0.7 1000 N/mm2 353 mm2 = 247.1 kN
1 2 0 .5
Fs.Rd = 247.1kN = 198kN
1.25
Fv.Sd = 177kN < Fs.Rd = 198kN
D-16
Weld joints
Weld between hanger and hanger gusset plate
Type: Full penetration butt weld (Single bevel joint)

Thickness of gusset plates: t = 10 mm
The design resistance of a full penetration butt weld shall be taken as equal ENV 1993-1-1: 1992
to the design resistance of the weaker of the parts joined. Section 6.6.6.1
The gusset plates have only been checked for resistance at the section
weakened by the bolts. Here, the resistance of the hanger gusset plates along
the weld joint and therefore the resistance of the full penetration butt welds will
be checked. Only the longitudinals welds are considered for force transmission.

ENV 1993-1-1: 1992
Criterion: VSd < Vpl.Rd Section 5.4.6
Design force:
VSd = NSd = 1061.9kN
Shear area:
A v = L 2 t 2 = 270mm 10mm 2 = 5400mm 2
A v fy 5400mm 2 460 N mm 2
Vpl.Rd = = = 1434kN
3 M0 3 1 .0
VSd = 1061.9kN < Vpl.Rd = 1434kN
Weld between arch gusset plate and arch profile (II)
Type: Double fillet joint

Throat thickness: a = 8 mm
The gusset plate is welded to the flanges as well as to the web. However,
the design check considers only the welding to the flanges.
Criterion: Fw.Ed Fw.Rd

Section D.3.1
Design weld force: VSd = NSd.ecc = 630.8 kN Figure D.3.3
Weld length L w = L 3 2 = 185mm 2 = 370mm
Design weld force per unit length:
VSd 630.8kN
Fw.Sd = = = 1705 N m m
Lw 370mm
Design resistance per unit length
Fw.Rd = f vw .d a
fu 3
Design shear strength f vw.d =
w Mw
D-17
Correlation factor w for fillet welds (fu 520 N/mm2): w = 1.0 ENV 1993-1-1: 1992
Section 6.6.5.3
530 N mm 2 N
f vw.d = = 244.8
3 1.0 1.25 mm 2 ENV 1993-2 : 1997
Table 6.1
Fw.Rd = 244.8 N mm 8mm = 1958 N mm
N N
Fw.Sd = 1705 2
< Fw.Rd = 1958
mm mm
Alternative: Arch gusset plate bolted to the arch profile (II)
cross section 1-1

1
II
I
B
1

strength bolts
d = 24 mm 8 slip resistant
220 high strength bolts
single bevel d = 22 mm
75 70 75
A
40 70 70 40 151.5
e2 p2 51.5 60 40
36
II - detail category 90
60
e1

58
One sided connection with

65
p1
227
preloaded high strength bolts

277
t = 12 mm
59
63 120
I - detail category 112 t = 10 mm
183
183
65

497
Double covered joint with

59
preloaded high strength bolts
B - detail category 36
270
L2
119.9 70 70 59.9
220
Transverse butt weld 319.8
A - detail category 45 arch gusset plates (2)

Crack in hanger L1
80 60 80
Fig. D.5. Hanger connection type 1, alternative
Type: Category C, slip-resistant at ultimate limit state

Properties
Bolt grade: 10.9

Yield strength: fyb = 900 N/mm2
Ultimate tensile strength: fub = 1000 N/mm2
Bolt diameter: d = 22 mm
Hole diameter: do = d + 2mm = 24 mm (standard nominal clearance)
Bolt cross section area: A = 380 mm2
Stress area: As = 303 mm2
Number of bolts: n=8
D-18
Geometry
Distances measured from centres of fasteners:
ENV 1993-1-1: 1992
End distance in load direction: e1 Section 6.5.1
Limits:
e1 = 40mm min e1 = 1.5 d0 = 36 mm ENV 1993-2: 1997
e2 = 36mm min e2 = 1.5 d0 = 36 mm Section 6.3
p1 = 60mm min p1 = 2.5 d0 = 60 mm
p2 non-existent Maximum distances are not critical.
Criteria
ENV 1993-1-1: 1992
Fv.Sd Fs.Rd slip resistance Table 6.5.2
Fv.Sd Fb.Rd bearing resistance
Section D.3.1
Design shear force for ultimate limit state: Nv.Sd = NSd.ecc = 630.8 kN
N v.Sd 630.8kN
Fv.Sd = = = 157.7kN
n 4
ENV 1993-1-1: 1992
Bearing resistance
Table 6.5.3
2.5 fu d t
Fb.Rd =
Mb
e1 40mm
3 do 3 24mm
p1 1 60mm 1
= = 0.58
3 do 4 3 24mm 4
fub 1000 N mm 2
= = 1.82
fu 550 N mm 2
or 1
= 0.56
2.5 0.56 550 N mm 2 22mm 12mm

Fb.Rd = = 162.6kN
1.25
Fv.Sd = 157.7kN < Fb.Rd = 162.6kN
ENV 1993-1-1: 1992
Slip resistance
Section 6.5.8
k n
Fs.Rd = s Fp,Cd
Ms
a) ENV 1993-2: 1997
Slip factor: = 0.5 Class A surface a)b) Section 6.4.7.1 (6)
ks = 1 standard nominal clearance b) ENV 1993-1-1: 1992

n=2 number of friction interfaces Section 6.5.8.3
Fp.Cd preloading force
Fp.Cd = 0.7 fub As = 0.7 1000 N/mm2 303 mm2 = 212.1 kN
1 2 0. 5
Fs.Rd = 212.1kN = 169.7kN
1.25
Fv.Sd = 157.7N < Fs.Rd = 169.7kN
D-19
D.3.2.2.2 Connection type 2
cross section 1-1
1
double fillet
joint, a = 8 mm
III W 360x410x634
I II
B
1
6 slip resistant high r=

strength bolts 34
d = 24 mm 2 .5
mm
Single bevel 220

A 75 70 75
III - detail category 36 40 70 70 40

ENV 1993-2 Table L.5/1 e2 p2
154.9 70 94.9
60
e1
p1
65

227
t = 20 mm
60
277
185
t = 10
L3
p1
65
497
60
e1
B - detail category 36
119.9 70 69.8 60.1
Transverse butt weld
270
L2
L4
220
A - detail category 90 319.8

arch gusset plate
Crack in hanger
L1
80 60 80

For member and connection design checks, please refer to connection type 1.
D-20
D.3.2.2.3 Hanger connection type 3
cross section 1-1
1
fillet joint,
a = 7 mm
III W 360x410x634
I II
B
1

r=
strength bolts
34
d=24mm 2.5
mm
Single bevel
both sides 220
A
75 70 75
40 70 70 40
III - detail category 36 e2 p2
154.9 70 94.9
e1
60
II - detail category 80 p1
65
ENV 1993-2 Table L.5/3 t = 10 mm
60
185
277
L3
227
t = 20
p1
65
60
e1
497

119.9 70 69.8 60.1
B - no suitable
detail provided in EC 3
270
319.8
L2
220
A - detail category 90 L4
Crack in hanger arch gusset plates (2)
80 60 80
L1
hanger gusset plate
For member and connection design checks, please refer to connection type 2.
In contrast to hanger connection types 1 and 2, this type uses only 1 gusset
plate welded to the hanger, whereas two gusset plates are welded to the arch.
However, the total cross sections are equal and the design checks are therefore
equivalent.
The net cross sectional area at the circular cut-out is larger than the net
cross sectional area at the bolts. An additional check is therefore not required.
D-21
Weld between arch gusset plate and arch profile (II)
In contrast to connection types 1 and 2, this type uses 2 gusset plates

welded to the arch profile. It is suggested using the following relation between Schneider [29],
weld leg and plate thickness: 2 a 0.7 min(t) = 0.7 10 mm = 7 mm. page 8.67
Type: Single fillet weld

Throat thickness: a = 7 mm
The gusset plate is welded to the flanges as well as the web. However, the
design check considers only the welding to the flanges.
Design shear force for ultimate limit state

Nv.Sd = NSd.ecc = 630.8 kN
Section D.3.1
Weld length: L w = L 3 2 = 185mm 2 = 370mm
Figure D.3.5
Design weld force per unit length
N v.Sd 630.8kN
Fw.Sd = = = 1705 N m m
Lw 370mm
Design resistance per unit length
Fw.Rd = f vw.d a
fu 3
w Mw
Correlation factor w for fillet welds (fu 520 N/mm2): w = 1.0 ENV 1993-1-1: 1992
Section 6.6.5.3
530 N mm 2 N
f vw.d = = 244.8
3 1.0 1.25 mm 2
ENV 1993-2: 1997
Fw.Rd = 244.8 N mm 7mm = 1713.6 N mm
Table 6.1
N N
Fw.Sd = 1705 2
< Fw.Rd = 1713.6
mm mm
D-22
cross section 1-1
1
W 360x410x634
II
double
39.6
fillet weld
1
145.1 85.1
A
r = 40 mm
single bevel
both sides
129
560.4
0.3
150
.8
L3
16
full penetration butt weld

r = 40 mm
r = 30 mm
462.5
422.5
90
L2
312.5
A - detail category 45
100
C - detail category 90 C
Transverse splice
60
180
Gusset plate
Criterion: VSd < Vpl.Rd
Design force: VSd = N Sd.ecc = 630.8kN Section D.3.1
Shear area: Av = 130 mm 20 mm = 2600 mm2

A v fy 2600mm 2 460 N mm 2
Vpl.Rd = = = 691kN
3 M0 3 1.0
VSd = 630.8kN < Vpl.Rd = 691kN
Weld joint to arch profile (II)
Design weld force: VSd = N Sd.ecc = 630.8kN Section D.3.1

Figure D.3.6
Weld length L w = L 3 2 = 150mm 2 = 300mm
Design weld force per unit length
D-23
VSd 630.8kN
Fw.Sd = = = 2103 N m m
Lw 300mm
Design resistance per unit length: Fw.Rd = f vw .d a
fu 3
w Mw
Correlation factor w for fillet welds (fu 520 N/mm2): w = 1.0
530 N mm 2 N ENV 1993-2: 1997

f vw.d = = 244.8
3 1.0 1.25 mm 2 Table 6.1
Weld leg:
2 a 0.7 min(t) = 0.7 20 mm = 14 mm
chosen: 9 mm
Fw.Rd = 244.8 N mm 9mm = 2203 N mm

N N
Fw.Sd = 2103 2
< Fw.Rd = 2203
mm mm
280
130 60 90

cross section 1-1
strength bolts
d = 22 mm
172
462
fillet joint
290
t = 12 mm
t = 12 mm
end plates
Fig. D.9. Hanger connection type 5 Fig. D.10. Suggested hanger connection
Figure D.9 shows connection type 5 as it was modelled in NE/Nastran. The

geometry varies only slightly from connection type 4, so that member design
checks are omitted.
The application of connection type 5 with a slightly modified geometry
(Figure D.10) was suggested by PER TVEIT, Section 7.7. Member design checks
were not carried out.
D-24
D.3.2.2.6 Suggested hanger connection
cross section 1-1
1
double fillet
joint, a = 7 mm
I II
1

strength bolts r=
d=24mm 34
2.5
mm
Single bevel
both sides 220
A
75 70 75
40 70 70 40
III - detail category 36 e2 p2
154.9 70 94.9
e1
60
p1
65
ENV 1993-2 Table L.5/3 t = 10 mm
60
185
L3
300
227
t = 20
p1
65
60
e1
520

119.9 70 69.8 60.1
B - no suitable
detail provided in EC 3 319.8
L2
293
A - detail category 90 L4
220
ENV 1993-2 Table L.4/1 arch gusset plates (2)

Crack in hanger
80 60 80
L1
hanger gusset plate
Fig. D.11. Suggested hanger connection
For member and connection assessment, refer to connection types 2 and 3.
D-25
c ENV 1993-2: 1997

Fatigue assessment: Ff E2
Mf Section 9.5
E2 = 2 p
Damage equivalence factor = 1 2 3 4
1 = 0.6 EC Mix L = 100 m ENV 1993-2 Table 9.5

2 = 1.04 Traffic per year: 30 106 t/track ENV 1993-2 Table 9.6
3 = 1.0 Design life: 100 years ENV 1993-2 Table 9.7
Fp.1 295.34kN
4 = 0.77 = = 0.67
Fp.1+ 2 439.7kN
max = 1.4 ENV 1993-2 Table 9.8
= 0.6 1.04 1.0 0.77 = 0.48
ENV 1993-2 Section

The damage equivalent impact factor 2 is already included in the design
9.5.3
forces.
D.3.3.1 Fatigue assessment based on nominal stress ranges

ENV 1993-2: 1997
Point (A) Table L.4/1
Detail category: 45 ENV 1993-1-1: 1992

c = 45 N/mm 2 Figure 9.6.1
Design force range: Np = 439.7 kN Section D.3.1

Cross section area of hanger: A = 2827.4 mm2 Figure D.3
Np 439.7kN N
p = = 2
= 155.5
A 2827mm mm 2
N N
E2 = 2 p = 0.48 155.5 2
= 74.65
mm mm 2
N
Ff E2 = 1.0 74.65
mm 2
N c 45 N
Ff E2 = 74.65 2
> = Check not fulfilled!
mm Mf 1.0 mm 2 A
74.65 N mm 2
Utilisation of fatigue strength: 100% = 166%
45 N mm 2
Point (B)
Transverse butt weld made from one side only.

ENV 1993-2: 1997
Detail category: 36
Table L.3/2
c = 36 N/mm2
D-26
Design force range: Np = 439.7 kN

Length of transverse butt weld: L1 = 60 mm Section D.3.1
Length of longitudinal butt welds: L2 = 270 mm Figure D.3
A = (L1 + 2 L 2 ) t 2 = (60mm + 2 270mm ) 10mm 2 = 12000mm 2

Np 439.7kN N
p = = 2
= 36.6
A 12000mm mm 2
N N B
E2 = 2 p = 0.48 36.6 2
= 17.6 2
mm mm
N
Ff E2 = 1.0 14.65
mm 2
N c 36 N
Ff E2 = 14.65 2
< =
mm Mf 1.0 mm 2
14.65 N mm 2
36 N mm 2
Longitudinal butt weld
The full penetration butt weld between hanger and gusset plate in shear is
categorized in detail category 100 (ENV 1993-2: 1997 Table L.5/3). The stress
distribution here can be assumed as equal to the transverse butt weld. Since
this check is fulfilled, it does not need to be assessed for the longitudinal weld.
Point (I)
Double covered joint with preloaded high strength bolts. ENV 1993-2: 1997
Detail category 112 Table L.1/2
c = 112 N/mm2 ENV 1993-1-1: 1992

Design force range: 439.7 kN Figure 9.6.1
Hanger connection plate

A net = (220mm 3 26mm ) 2 10mm = 2840mm 2
Np 439.7kN N
p = = 2
= 154.8
A net 2840mm mm 2
N N
E2 = 2 p = 0.48 154.8 2
= 74.3 ENV 1993-2: 1997
mm mm 2
Section 9.4, 9.5.1
N
Ff E2 = 1.0 74.3
mm 2
N c 112 N
Ff E2 = 74.3 2
< = I
mm Mf 1.0 mm 2
D-27
Point (II): welded
Double fillet joint ENV 1993-2: 1997

Table L.5/3
Detail category: 80
c = 80 N/mm2 ENV 1993-1-1: 1992

Figure 9.6.1
It is assumed that this shear loaded weld joining the gusset plate to the
flanges is the main load carrying weld. For the fatigue assessment, the nominal
stress range is therefore calculated with the respective weld area, neglecting
the web weld.
Design force range: Np = Np.ecc = 261.1 kN Section D.3.1
Weld length: L3 = 185 mm
Weld leg: a = 8 mm II
A v = L 3 t = 185mm 8 2mm = 2960mm 2

Np 261.1kN N
p = = = 88.2
Av 2960mm 2 mm 2
N N
E 2 = 2 p = 0.48 88.2 2
= 42.3
mm mm 2
N
Ff E2 = 1.0 42.3
mm 2
N c 80 N mm 2
Ff E2 = 42.3 < =
mm 2 Mf 1. 0
Point (III)
ENV 1993-2: 1997
Root failure in fillet joint: Detail category 36 Table L.5/1
The flange welds (II) are assumed to be load carrying only. An assessment of
point III is therefore omitted.
Point (II): bolted
One sided connection with preloaded high strength bolts. ENV 1993-2: 1997
Detail category 90 Table L.1/2
c = 90 N/mm2
Design force range: Np = Np.ecc = 261.1 kN Section D.3.1
Arch connection plate:

A net = (151.5mm 2 24mm ) 4 12mm = 4968mm 2 Figure D.3.4
Np 261.1kN N
p = = 2
= 53
A net 4968mm mm 2
II
N N
E2 = 2 p = 0.48 53 = 25
mm 2 mm 2
N
Ff E2 = 1.0 25
mm 2
N c 90 N
Ff E2 = 25 2
< =
mm Mf 1.0 mm 2
D-28
Point (A)
Detail category: 90 ENV 1993-2 : 1997

Table L.4/1
c = 90 N/mm2
Cross section area of hanger: A = 2827.4 mm2
Np 439.7kN N
p = = 2
= 155.5
A 2827mm mm 2
N N
E2 = 2 p = 0.48 155.5 2
= 74.65
mm mm 2
N
Ff E2 = 1.0 74.65
mm 2
A
N c 90 N
Ff E2 = 74.65 < =
mm 2 Mf 1.0 mm 2
74.65 N mm 2
90 N mm 2
Point (B)
See connection type 1, Section D.3.3.1.1.
Point (A)
See connection type 2
Point (B)
The detail does not correspond to any detail category given in Eurocode 3,
and so the nominal stress method cannot be used here. It is necessary to
obtain the local stress at the edge of the circular hole, which constitutes a stress
concentration.
Point (C)
Full penetration butt joint of hanger to widened end member

ENV 1993-2: 1997
Detail category: 80 Table L.3/1
c = 112 N/mm2
Cross section area of hanger (D = 60mm): A = 2827.4 mm2
Np 439.7kN N
p = = 2
= 155.5
A 2827mm mm 2
D-29
N N
E2 = 2 p = 0.48 155.5 2
= 74.65
mm mm 2
N
Ff E2 = 1.0 74.65
mm 2
N c 80 N
Ff E2 = 74.65 2
> =
mm Mf 1.0 mm 2
74.65 N mm 2 A
Utilisation of fatigue strength: 2
100% = 93%
80 N mm
Point (A)
Crack in hanger at joint to gusset plate ENV 1993-2: 1997

Detail category: 45 Table L.4/1
C
c = 45 N/mm2
Cross section area (D = 90 mm):

(90mm ) = 6361.7mm 2
2
A=
4
Np 439.7kN N
p = = 2
= 69.1
A 6361.7mm mm 2
N N
E2 = 2 p = 0.48 69.1 2
= 33.2
mm mm 2
N
Ff E2 = 1.0 14.65
mm 2
N c 45 N
Ff E2 = 33.2 2
< =
mm Mf 1.0 mm 2
33.2 N mm 2
45 N mm 2
Point (II)
ENV 1993-2: 1997
Longitudinal crack in the fillet weld
Table L.4/2
Detail category: 80
c = 80 N/mm2
II
Design force range: Np = Np.ecc = 261.1 kN
Cross section area: A = 150mm 2 8mm = 2400mm 2
Np 261.1kN N
p = = 2
= 109
A 2400mm mm 2
N N
E2 = 2 p = 0.48 109 2
= 52
mm mm 2
N
Ff E2 = 1.0 52
mm 2
D-30
N c 80 N
Ff E2 = 52 2
< =
mm Mf 1.0 mm 2
52 N mm2
80 N mm2
Connection type 5 varies only slightly from type 4.

For point (A) see connection type 3.
For point (II) see connection type 4.
D.3.3.2 Fatigue assessment based on geometric stress

ranges
Determination of stress concentration factors SCF
The stress concentration factors were determined by carrying out finite

element analysis with NE/Nastran. The geometric stress magnitudes SC were
divided by the nominal stress 0, which is the stress calculated with the net
cross section area. The numbers and results are listed in Figures D.12 and
D.13.3
De te rm ination of ge om e tric s tre s s conce ntration factors SCF
Location Type 1 Type 2 Type 3 s ugge s te d

type
A
Ne t are a A net [m m 2 ] 2827.40 2827.40 2827.40 2827.40
Nom inal s tre s s (F=1000 k N) 0 [N/m m 2 ] 353.68 353.68 353.68 353.68
Ge om e tric s tre s s SC [N/m m 2 ] 506.80 399.40 409.40 401.75
Ge om e tric SCF SCF = SC / 0 1.43 1.13 1.16 1.14
Ne t are a A net [m m 2 ] 4400.00 4400.00 3200.00 4400.00
Nom inal s tre s s (F=1000 k N) 0 [N/m m 2 ] 227.27 227.27 312.50 227.27
Ge om e tric s tre s s SC [N/m m ] 2 316.30 333.05 1217.00 305.35
Ge om e tric SCF SCF = SC / 0 1.39 1.47 3.89 1.34

Fig. D.12. SCF for connection types 1 to 3 and suggested type
D-31
Determination of geometric stress concentration factors SCF

Location Type 4 Type 5
C
Net area Anet [mm2] 2827.43 -
2
Nominal stress (F=1000 kN) o [N/mm ] 353.68 -
2
Geometric stress sc [N/mm ] 402.00 -
Geometric SCF SCF = SC / 0 1.14 -
A
Net area Anet [mm2] 6361.73 2827.43
2
Nominal stress (F=1000 kN) o [N/mm ] 157.20 353.68
2
Geometric stress sc [N/mm ] 284.00 447.80
Geometric SCF SCF = SC / 0 1.81 1.27
D1
Net area Anet [mm2] 2620.00 3600.00
2
Nominal stress1) o [N/mm ] 226.65 158.33
2
D2
Net area Anet [mm2] 3280.00 4000.00
1) 2
Nominal stress o [N/mm ] 123.84 107.50
2
IIa
Net area Anet [mm2] 3000.00 3440.00
1) 2
2
IIb
Net area Anet [mm2] 3000.00 3440.00
1) 2
2
Fig. D.13. SCF for connection types 4 and 5
1)
Nominal stress based on the following forces (due to eccentricity)
Type 4: For D1and IIb: F = 593.81 kN For D2 and IIa: F = 406.2 kN
Type 5: For D1, IIa, IIb: F = 570 kN For D2: F = 430 kN
Fatigue assessment
According to ENV 1993-1-1: 1992 Section 9.6.3, the fatigue strength curves for
fatigue assessments based on geometric stress ranges shall be for
- full penetration butt welds: Category 90 (assumed: weld profile and
permitted weld defects acceptance criteria satisfied)
- load carrying partial penetration butt welds and fillet welds: Category 36
The relevant SN curves are found in ENV 1993-1-1: 1992 Figure 9.6.1
The assessment is carried out in the following tables. Formulas are provided.
D-32
Fatigue assessment
Suggested
Location Type 1 Type 2 Type 3
type
A
2
Net area Anet [mm ] 2827.40 2827.40 2827.40 2827.40
Nominal stress range p.0 [kN] = Np / Anet 155.50 155.50 155.50 155.50
SCF 1.43 1.13 1.16 1.14
Geom. stress range p.SC [kN] = p SCF 222.36 175.71 180.38 177.27
2
Ff E2 = Ff 2 p [N/mm ] (T1) 106.95 84.29 86.40 84.78
> < < <
2
c / Mf [N/mm ] (T2) 90.00 90.00 90.00 90.00
(T1) / (T2) 100 % 119 94 96 94
B
2
Net area Anet [mm ] 4400.00 4400.00 3200.00 4400.00
Nominal stress range p.0 [kN] = Np / Anet 99.92 99.92 137.39 99.92
SCF 1.39 1.47 3.89 1.34
2
Ff E2 = Ff 2 p [N/mm ] (T1) 66.75 70.29 256.80 64.44
< < > <
2
c / Mf [N/mm ] (T2) 90.00 90.00 90.00 90.00
(T1) / (T2) 100 % 74 78 285 72
= 0.48 Ff and Mf = 1.0 Np [kN] = 439.66 2 incl. in Np
Fig. D.14. Fatigue assessment of connection types 1 to 3 and suggested type (geometric stress ranges)
Fatigue assessment: Connection type 4

Location C A D1 D2 IIa IIb
Force range Np [kN] 439.66 439.66 261.1 178.6 178.6 261.1
Net area [mm2] Anet [mm2] 2827.43 6361.73 2620.00 3280.00 3000.00 3000.00
Nom. stress range p [kN] = Np / Anet 155.50 69.11 99.66 54.45 59.53 87.03
Geometric SCF 1.14 1.81 2.90 3.26 4.85 2.96
Geom. stress range p.SC [kN] = p SCF 177.27 125.09 289.00 177.51 288.74 257.62
Ff E2 = Ff 2 p [N/mm2] (T1) 85.09 60.04 138.72 85.21 138.59 123.66
< < > < > >
2
c / Mf [N/mm ] (T2) 90.00 90.00 125.00 125.00 36.00 36.00
(T1) / (T2) 100% 95 67 111 68 385 343
= 0.48 Ff and Mf = 1.0 2 incl. in Np
Fig. D.15. Fatigue assessment of connection type 4 (geometric stress ranges)
Fatigue assessment: Connection type 5

Location A D1 D2 II
Force range Np [kN] 439.66 250.6 189.1 250.6
2 2
Net area [mm ] Anet [mm ] 2827.43 3600.00 4000.00 3440.00
Nom. stress range p [kN] = Np / Anet 155.50 69.61 47.28 72.85
Geometric SCF 1.27 2.10 2.36 3.53
2
Ff E2 = Ff 2 p [N/mm ] (T1) 94.79 70.17 53.55 123.44
> < < >
2
c / Mf [N/mm ] (T2) 90.00 125.00 125.00 36.00
(T1) / (T2) 100% 105 56 43 343
= 0.48 Ff and Mf = 1.0 2 incl. in Np
Fig. D.16. Fatigue assessment of connection type 5 (geometric stress ranges)
D-33
D.3.4 Patch test
The following refers to Section 10 and gives additional references on the
stress concentration in a plate around a hole.
The reference plate has the following dimensions:
Width W = 300 mm
Thickness T = 25 mm
Length L = 1000 mm
Diameter of the hole D = 100 mm (symmetrical to the plate)
These dimensions were used as input values in two reference calculations
described in the following:
A java applet provided by David Grieve (School of Engineering, University
of Plymouth)1 gives a SCF of 2.307 at a hole in a plate with above mentioned
dimensions. Regarding the used formula, reference is made to "Roark's
Formulas for Stress and Strain", 6th Ed. McGraw Hill, 1989 and "Stress
Concentration Factors", by R E Peterson, John Wiley and Sons, 1974.
A further source2 provides the following formula predicting the stress
concentration factor for the same problem:
SCF = 2 + (1 D W ) .
3
With the dimensions of the reference plate, the SCF gives
SCF = 2 +(1 100 300 )3 = 2.296 ,
which is virtually equal to the result given by Grieves applet and to the
result obtained with the finer mesh of the test plate. The exactness of the
theoretical SCF obviously depends on the approximation of the curve between
the two extremes SCF = 2.0 and SCF = 3.0. Since even the references found
show different results, is it likely that further approximations exist.
1
) http://www.tech.plym.ac.uk/sme/desnotes/Stressc.htm
2
) http://www.stacieglass.com/scf/indext.html
(Websites accessed in July, 2003)
D-34
Diploma Thesis Brunn & Schanack Annex D: Assessment of the wind bracing
D.4 Wind bracing

Fig. C.25
Buckling resistance of member 1
Values of the cross section see Section D.1
NSd = -441.7 kN ENV 1993-1-1: 1992,

5.5.1
The criterion to be fulfilled is: NSd Nb,Rd
fy
Nb,Rd = A A
M1
depends on
l
= A = A
1 E
iy
fy
l = 5.738 m
iy = 7.47 cm
E = 210000 N/mm2
fy = 355 N/mm2 ; S355
1 5.738m
= 1 = 1.004
210000N / mm 2
7.47cm
355N / mm 2
= 0.6656 ; buckling curve a ENV 1993-1-1: 1992
5.5.1 Table 5.5.2
A = 53.1 cm2
M1 = 1.1
355N / mm 2
Nb,Rd = 0.6656 1 53.1cm 2 = 1140.6kN
1. 1
|NSd| = 441.7 kN 1140.6 kN = Nb,Rd
Fig. C.25
Buckling resistance of member 15
Values of the cross section see Section D.1
NSd = -170.4 kN
The criterion to be fulfilled is: NSd Nb,Rd
fy
Nb,Rd = A A
M1
depends on
D-35
l
= A = A
1 E
iy
fy
l = 7.1364 m
iy = 7.47 cm
E = 210000 N/mm2
fy = 355 N/mm2 ; S355
1 7.1364m
= 1 = 1.249
210000N / mm 2
7.47cm
355N / mm 2
= 0.5002 ; buckling curve a
ENV 1993-1-1: 1992
A = 53.1 cm2 5.5.1 Table 5.5.2
M1 = 1.1
355N / mm 2
Nb,Rd = 0.5002 1 53.1cm 2 = 870.4kN
1 .1
|NSd| = 441.7 kN 870.4 kN = Nb,Rd
Flexural buckling of member 3 after second order analysis Fig. C.25
Values of the cross section vide Section D.1
1. Shear resistance ENV 1993-1-1: 1992,

5.4.6
VSd,y = -3.1 kN VSd,z = -14.1 kN
fy
Vpl,Rd = A v
3 M0
Av = 2A/ Av = 2A/
= 265.7 cm2/ = = 265.7 cm2/
= 41.83 cm2 = 41.83 cm2
355N / mm 2 355N / mm 2
Vpl,Rd = 41.83cm 2 Vpl,Rd = 41.8cm 2
3 1. 0 3 1.0
= 857.3 kN = 857.3 kN
|VSd|= 3.1 kN < Vpl,Rd= 857.3 kN |VSd|= 14.1 kN < Vpl,Rd= 857.3 kN
ENV 1993-1-1: 1992,

5.4.7 (2)
design plastic shear resistance Vpl,Rd, no reduction need be made in the plastic
resistance moment.
NSd = -109.8 kN Section C.4.1, Figure

My,Sd = -34.4 kNm C-234
Mz,Sd = -15.9 kNm
D-36
A fy 65.7cm 2 355N / mm 2 ENV 1993-1-1: 1992,

Npl,Rd = = = 2332.4 kN 5.4.4
M0 1 .0
Wpl,y f y 438cm3 355N / mm2 ENV 1993-1-1: 1992,
Mpl,y,Rd = = = 155.5 kNm 5.4.5.1
M0 1. 0
Wpl,z f y 438cm3 355N / mm2
Mpl,z,Rd = = = 155.5 kNm
M0 1. 0
ENV 1993-1-1: 1992,
N Sd M y,Sd M z,Sd
+ + 1 5.4.8.1 (5.36)
109.8kN 34.4kNm 15.9kNm
+ + = 0.37 1
2332.4kN 155.5kNm 155.5kNm
Design resistance of joint III (members 5, 6) ENV 1993-1-1: 1992,

Annex K.6
Joint III is assessed by member 6 profiles: CHS 219.1 x 8

determining the design resistance
of the members 5 (Figure D.17). Joint III
60
Section properties
Elements 5, 6: CHS 219.1x8
me
5
mb
er
di = 219.1 mm
mb
er
me
ti = 8 mm
5
A = 53.1 cm2 Fig. D.17. Joint III

Wel = 270 cm3
Angles: 1 = 2 = 60
Range of validity ENV 1993-1-1: 1992

Table K.6.1
1) 0.2 di / do = 1 1
2) 5 di / 2to = 13.7 25
3) 5 do / 2to = 13.7 25
4) 5 do / 2to = 13.7 20 for X joints
5) ov = q / p = 127.96 mm / 253.1 mm = 51% 25%
6) g = - 127.96 mm
Internal forces in the chord

NSd.6.max [kN] 157.7
NSd.6.min [kN] -66
NSd.5.max [kN] 227.6
NSd.5.min [kN] -230.4
My.Sd [kNm] 9.2
Mz.Sd [kNm] 5.6
D-37
Secondary moments ENV 1993-1-1: 1992

Annex K.4 (2)
Ratio between system length and depth of the chord member:
10.15 m / 0.2191 m = 46 > 12
Ratio between system length and depth of the brace member:
5 m / 0.2191 m = 23 < 12
Secondary moments considered with 7 kNm for both directions.
o is the maximum compressive stress in the chord at the joint due to axial force
and bending moment.
p is the value of o excluding the stress due to the horizontal components of

the forces in the braces at that joint
Critical: Chord in compression
NSd M y,Sd M z,Sd

o = + +
A W el W el
o =
66kN
+
(9.2 + 7)kNm + (5.6 + 7)kNm = 119.1 N
2
53.1cm 270cm 3 270cm 3 mm 2 ENV 1993-1-1: 1992
Figure K.3
Conservatively and for simplification, it is assumed that: p = o
fyo = fyi = 355 N/mm2
kp = 1 0.3 np (1 + np)
np = p / fyo = 119.1 N/mm2 / 355 N/mm2 = 0.36 (compression)
kp = 1 0.3 0.36 (1 + 0.36) = 0.85
is the ratio of the chord diameter to twice its wall thickness

do 219.1mm
= = = 13.7
2 to 2 8mm
g / to = -127.96 mm / 10 mm = -12.8
kg = 2.4
Resistance of bracings (members 5):

fyo t o2 1 .1
N1.Rd = [1.8 + 10.2 d1 do ] k p k g
sin 1 Mj
355 N mm 2 (8mm )
2
N1.Rd = [1.8 + 10.2 1] 0.85 2.4 1.1 = 640kN
sin(60) 1.1
sin 1
N 2.Rd = N1,Rd = 640kN < |NSd,5,min| = 230.4 kN
sin 2
D-38
Design resistance of joint I: End plate connection Fig. C.25
50 138 184
Internal forces t = 20 mm
42
NSd.min [kN] 51.2
177
NSd.max [kN] -441.7
49
Geometry
Angle [] 26 W 360 x 410 x 900
M 20, type C,
preloaded,
NSd.max. = sin(26) (-441.7 kN) = -193.6 kN slip-resistant
NSd.min. = sin(26) 51.2 kN = 22.44 kN
NSd.max.|| = cos(26) (-441.7 kN) = 396.7 kN
Section properties CS
H 219
.1 x
8
Elements 1: CHS 219.1x8

Fig. D.18. CHS end plate connection
End plate: t = 20 mm
Arch profile: W 360 x 410 x 900: tf = 106 mm
Weld design
Weld length: Conservatively and for simplification, the circular cross section is
taken instead of the elliptical one.
Lw = d = 219.1 mm = 688 mm
Design force per unit length:

NSd,max 441.7kN
Fw,Sd = = = 642 N m m
Lw 688mm
Design resistance per unit length: Fw.Rd = fvw.d a
fu 3
Design shear strength fvw .d =
w Mw
fu nominal ultimate tensile strength of weaker part joined

fu = 530 N/mm2
w appropriate correlation factor
w = 0.9
530 N mm 2 N
f vw,d = = 272
3 0.9 1.25 mm 2
Weld leg: a = 5 mm
Fw.Rd = 272 N mm 2 5mm = 1360 N mm

Fw.Sd = 642 N mm 2 < Fw.Rd = 1360 N mm 2
D-39
Bolt connection
Since the connection also takes tensile forces, the bolts must be preloaded
Preloaded, slip resistant, type C
Grade: 10.9
d = 20 mm, d0 = 22 mm
A = 314 mm2
Tensile stress area of bolt: As = 245 cm2
Number of bolts: 8
Design shear force per bolt for the ultimate limit state:
Fv,Sd = 396.7 kN / 8 = 49.6 kN
Ft,Sd = 22.4 kN / 8 = 2.8 kN
Geometry
End distance in load direction: e1

e1 = 50 mm min e1 = 1.5 d0 = 33 mm
e2 = 42 mm min e2 = 1.5 d0 = 33 mm
p1 = 184 mm min p1 = 2.5 d0 = 55 mm
p2 = 177 mm min p2 = 2.5 d0 = 55 mm
Maximum end and edge distances: members exposed to weather
50 40 mm + 4 t = 120 mm (t = 20: thickness of end plate)
Maximum spacing in compression members
Outer row: p1,0 = 177 mm 14 t = 280 mm (200 mm)
Bearing resistance
2 .5 f u d t
Fb,Rd =
Mb
e1 50mm
3 d o 3 22mm
p1 1 174mm 1
= = 2 .8
3 d o 4 3 22mm 4
fub 1000 N mm 2
= = 1.82
fu 550 N mm 2
= 0.76 < 1 ENV 1993-1-1: 1992,
Table 6.5.3
2.5 0.76 530 N mm 2 20mm 20mm
Fb,Rd = = 322kN > Fv.Sd = 49.6.1kN
1.25
Tensile resistance
0.9 fub A s 0.9 1000 N mm 2 245mm 2

Ft.Rd = = = 176.4kN > Ft.Sd = 2.8kN
Mb 1.25
D-40
Combined tension and shear

ENV 1993-1-1: 1992
(
k s n Fp.Cd 0.8 Ft.Sd ) Section 6.5.8.4
Fs.Rd =
Ms,ult
Slip factor: = 0.5 Class A surface: ENV 1993-2 Section 6.4.7.1 (6)
ENV 1993-1-1 Section 6.5.8.3
ks = 1 standard nominal clearance
n=1 number of friction interfaces
Fp,Cd preloading force
Fp,Cd = 0.7 fub As = 0.7 1000 N/mm2 245 mm2 = 171.5 kN
Fs.Rd =
(
1 1 0.5 171.5 N mm 2 0.8 2.8 N mm 2 )
= 67.7kN > Fv.Sd = 49.6kN
1.25
Node I
Internal forces
N1 [kN] 120.9
N1+2 [kN] 174.8
Aweld = d a = 219.1 mm 5 mm = 3441.6 mm2
N1 120.9kN N
1 = = = 35 one LM 71
A 3441 .6mm 2 mm 2
N1+ 2 174.8kN N
1+ 2 = p = = 2
= 50.8 two LM 71
A 3441 .6mm mm 2
c
Criterion: Ff E2 with E2 = 2 p
Mf
Damage equivalent factor = 1 2 3 4
1 = 0.6 EC Mix L = 100 m


5
1
5
4 = 5 0.12 + 0.88 1 + 1 = 0.76
1+ 2
1+ 2
= 0.6 1.04 1.0 0.76 = 0.47
N N
E2 = 0.47 50.8 2
= 23.88
mm mm 2
Detail category: 40 ENV 1993-2: 1997,

Table L.6 sheet 2
c 40 N mm 2
Ff E 2 = 1.0 23.88 =
Mf 1 .0
D-41
Node II
In the sketch in Figure D.19 the B

critical section for the check of the node II
fatigue resistance can be seen. In point
A the hollow section of member 1 is
fillet-welded to the other straight
member 2-3. This is to range in detail A
ENV 1993-2: 1997,
category 71. Only the stresses in Table L.6 sheet 2
member 1 have to be checked.
Fig. D.19. Sketch of node II
In point B the hollow sections of members 4 and 5 are fillet-welded to

member 2-3 with an overlap > 30 %. Since, to/ti = 1.25<1.4 according to ENV
1993-2: 1997, Table L.7 sheet 2 this connection has to be ranged in detail
category 56. The assessment has to be carried out for the members 2 and 3 as
well as for members 4 and 5.
1. Node II, critical point A
N x,1 120.9kN
1 = = 2
= 22.8N / mm 2 one LM 71
A 53.1cm
N x,1+ 2 174.8kN
1+2 = p = = 2
= 32.9N / mm 2 two LM 71
A 53.1cm
E2 = 2 p
= 1 2 3 4
1 = 0.6 EC Mix, L=100m
ENV 1993-2: 1997,
2 = 1.04 Traffic per year: 30 106 t/track 9.5.1

5
1
5
4 = 5 0.12 + 0.88 1
+ 1 = 0.766
1+ 2
1+ 2
The criterion is:
E2 c/Mf with Mf = 1.0 ; redundant structural element
E2 = 0.61.041.00.76632.9N/mm2 = 15.7N/mm2
c/Mf = 71N/mm2/1.0 = 71 N/mm2
15.7 N/mm2 < 71 N/mm2
2. Node II, critical point B, assessment of the braces, critical member

number 5
N x,1 136.5kN
1 = = = 25.7N / mm 2 one LM 71
A 53.1cm 2
N x,1+ 2 154.8kN
1+2 = p = = = 29.2N / mm 2 two LM 71
A 53.1cm 2
E2 = 2 p
D-42
= 1 2 3 4
1 = 0.6 EC Mix, L=100m

5
1
5
4 = 5 0.12 + 0.88 1
+ 1 = 0.898

1+ 2
1+ 2
The criterion is:
ENV 1993-2: 1997,
E2 c/Mf with Mf = 1.0 ; redundant structural element
9.5.1
E2 = 0.61.041.00.89829.2N/mm2 = 16.4N/mm2
c/Mf = 56N/mm2/1.0 = 56 N/mm2
16.4 N/mm2 < 56 N/mm2
2. Node II, critical point B, assessment of the chord, critical members 2-3
N x,1 M y,1 M z,1 30.7kN 15.4kNm 7.4kNm

1 = + + = 2
+ 3
+ = 74.2N / mm 2
A W el,y W el,z 65.7cm 328cm 328cm 3
one LM 71
N x,1 M y,1 M z,1 47.2kN 20.3kNm 10.3kNm
1+2 = + + = 2
+ 3
+ = 100.5N / mm 2
A W el,y W el,z 65.7cm 328cm 328cm 3
two LM 71
E2 = 2 p
= 1 2 3 4
1 = 0.6 EC Mix, L=100m

5
1
5
4 = 5 0.12 + 0.88 1
+ 1 = 0.793

1+ 2
1+ 2
The criterion is:
E2 c/Mf with Mf = 1.0 ; redundant structural element ENV 1993-2: 1997,
E2 = 0.61.041.00.793100.5N/mm2 = 49.7N/mm2 9.5.1
c/Mf = 56N/mm2/1.0 = 56 N/mm2
49.7 N/mm2 < 56 N/mm2
Shear stresses
The maximum shear forces occurring in ultimate limit state do not exceed
2 % (see Section D.4.1) of the permitted limits, so the shear stresses are
considered to be not critical to fatigue strains. The fatigue check will be omitted.
D-43
D.4.3.1 Limitation of nominal stress for rare load

combinations ENV 1993-2: 1997,
4.3
The critical cross section is found in members 2 and 3 at node II.
Axial stress
N ser M y,ser M z,ser 75.4kN 23.7kNm 2.1kNm

Ed,ser = + + =
A W el,y W el,z 65.7cm 2 328cm 3 328cm 3
= -90.13 N/mm2
Shear stress
Vz 14.1kN
Ed,ser = =
A vz 41.8cm 2
= -3.4N/mm2
fy
( Ed,ser )2 + 3( Ed,ser )2 <
M,ser
( 90.13N / mm ) 2 2
(
+ 3 3.4N / mm 2 )
2
= 90.3 < 355N / mm 2 =
355N / mm 2
1 .0
D.4.3.2 Limitation of nominal stress range for frequent load

ENV 1993-2: 1997,
combination 4.3 (4)
The critical cross section is found in member 1. The stress range will be
calculated from the variation of internal forces.
N ser 139.8kN
Ed,ser = = = 26.6N / mm 2
A 53.1cm 2
fy
Ed,ser < 1.5
M,ser
355N / mm 2
Ed,ser = 26.6N / mm 2 < 532.5N / mm 2 = 1.5
1 .0
D-44
Diploma Thesis Brunn & Schanack Annex D: Assessment of the bridge deck
D.5 The bridge deck
D.5.1 Concrete cover
Exposure class
The railway bridge is assumed to be exposed to a humid environment with
frost. The non-public footpath will not be subjected to de-icing agent.
Exposure class: 2b ENV 1992-1-1: 1991
Table 4.1
Minimum concrete cover
Requirements for minimum cover to prestressing sheets:

- 50 mm a) ENV 1992-2: 1996
- 35 mm with a reduction of 5 mm for concrete strength class C50/60 b) Section 4.1.3.3 a)
- Equal to the diameter of the duct b):
D = 57 mm for transverse prestressing ENV 1992-1-1: 1991
D = 118 mm for longitudinal prestressing Section 4.1.3.3 b)
Critical values: Transversal prestressing: min c = 57 mm

Longitudinal prestressing: min c = 118 mm
Requirements for minimum cover to reinforcement:
- 30 mm a)
- 25 mm with a reduction of 5 mm for concrete strength class C50/60 b)
Critical value: min c = 30 mm
Nominal concrete cover
Allowance for tolerance: h = 10 mm ENV 1992-1-1: 1991

Section 4.1.3.3
nom c = min c + h
Transversal prestressing: nom c = 67 mm
Longitudinal prestressing: nom c = 128 mm
Reinforcement: nom c = 40 mm
D.5.2 Main design Bridge deck with transverse prestressing
D.5.2.1 Ultimate limit state
D.5.2.1.1 Bending with longitudinal force ENV 1992-1-1: 1991

Section 4.3.1
Bridge deck in transverse direction
ENV 1992-2: 1996
The axial compression forces are virtually evenly distributed over the bridge
Section 4.3.1
length, which is due to the transverse prestressing. The critical point is where
the bending moment has its maximum.
x = 49.1 m
y = 3.28 m
D-45
Geometry
z p [mm] 103
For geometry see
Internal forces h [mm] 403.1 Annex B, Figure B.7.
mSd,y [kNm/m] 628.7 zs1 = h / 2 - zp
d [mm] 300.1
nSd,y [kN/m] -3139
z s 1 [mm] 98.55
Annex C
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,x nSd,y zs1

msds = 628.7 kNm/m (-3139 kN/m) 0.09855 m = msds = 938 kNm/m
mSds
Sds =
b d2 fcd
938 kNm m
Sds = = 0.31 > 0.206 compressive reinforcement
(0.3m)2 33.3 N mm 2
Required re-bar diameter: 20 Schneider [29],
d2 = 40 mm + (20 mm) / 2 = 50 mm pages 5.126, 5.130
d2 / d = 50 mm / 300 mm = 0.17
1 = 0.366 (1 and 2 are interpolated for d2 / d = 0.17)
2 = 0.155
fpd = 0.9 fpk / s = 0.9 1230 N/mm2 / 1.15 = 962.5 N/mm2

1
A s1 = (1 b d f cd + N Sd )
fpd
A s1 =
1
962.5 N mm 2
(
0.366 1000m 300mm 33.3 N mm 2 3139 1000 N m ) ENV 1992-1-1: 1991
Section 4.2.3.3.3 (6)
A s1 = 5.4cm 2 m
10, s = 14 cm (gives As1 = 5.61 cm2/m)
fcd
A s2 = 2 b d
f yd
33.3
A s2 = 0.155 100cm 30.01cm = 35.6cm 2
435
20, s = 9 cm (gives As2 = 34.91 cm2/m)
(34.91 < 35.6, but difference less than 3% permissible)
Check at smallest depth
x = 49.1 m
Geometry
y = 1.36 m
z p [mm] 134
Internal forces h [mm] 374
mSd,y [kNm/m] 230 d [mm] 240
nSd,y [kN/m] -3170 z s 1 [mm] 53
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,x nSd,y zs1

msds = 230 kNm/m (-3170 kN/m) 0.053 m
msds = 398 kNm/m
D-46
mSds
Sds =
b d2 fcd Schneider [29],
page 5.130
398 kNm m
Sds = = 0.2
(0.24m)2 33.3 N mm 2
= 0.2327
fpd = 0.9 fpk / s = 0.9 1230 N/mm2 / 1.15 = 962.5 N/mm2 ENV 1992-1-1: 1991
Section 4.2.3.3.3 (6)
1
As = ( b d fcd + NSd )
fpd
1 N
As = 0.2327 1000mm 240mm 33.3
2
3170 1000N
962.5 N mm mm 2
A s = 13.6 cm 2 m < 0
No additional reinforcement required.
Bridge deck in longitudinal direction
At the critical point, the maximum bending moment faces a moderate axial
compression force.
At point (x = 50.9; y = 2.69), the axial compression force is considerably
smaller, but acts together with a bending moment of mx 0, which is not
relevant.
x = 49.1 m Geometry
y = 2.69 m z s [mm] 45
mSd,x [kNm/m] 422.7 d [mm] 385

Annex C
nSd,x [kN/m] -2743 z s 1 [mm] 170
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,x nSd,x zs1

msds = 422.7 kNm/m (-2743 kN/m) 0.17 m
msds = 889 kNm/m
mSds
Sds = Schneider [29],
b d2 fcd page 5.130
889 kNm m
Sds = = 0.18
(0.385m)2 33.3 N mm 2
= 0.88 z = d = 385 mm 0.8 = 308 mm
= 0.3 x = d = 385 mm 0.3 = 115.5 mm
s1 = 0.00902
p = pm + s1
fsd = fyk / s ENV 1992-1-1: 1991

fsd = 500 N/mm2 / 1.15 = 435 N/mm2 Section 4.2.2.3.2 (5)
D-47
1 m Sds
As = + n Sd
fsd z
1 889 kNm m
As = 2
+ ( 2743kN / m) = 3.3 cm 2 m
435 N mm 308mm
ds = 10 mm, s = 20 mm, As = 3.93 cm2/m

Minimum reinforcement: As.min = 5.24 cm2/m > 3.93 cm2/m
Edge beam
The critical section is located between the two hanger connections at x = 48

m and x = 48.6 m, where the bending moment is maximal and the axial
compression force minimal.
Geometry
x = 48.3 m z s [mm] 45
mSd,x [kNm/m] 407.9 d [mm] 565
nSd,x [kN/m] -1691 z s 1 [mm] 260

Annex C
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,x nSd,x zs1

msds = 407.9 kNm/m (-1691 kN/m) 0.26 m
msds = 847.6 kNm/m
mSds Schneider [29],

Sds = 2
b d fcd page 5.130
847.6 kNm m
Sds = = 0.08
(0.565m)2 33.3 N mm 2
= 0.95 z = d = 565 mm 0.95 = 537 mm
= 0.13 x = d = 565 mm 0.13 = 73 mm
s1 = 0.02
fsd = fyk / s ENV 1992-1-1: 1991
fsd = 500 N/mm2 / 1.15 = 435 N/mm2 Section 4.2.2.3.2 (5)
1 mSds
As = + nSd
ftd z
1 847.6 kNm m cm 2
As = + ( 1691kN / m ) = 2 . 6 <0
435 N mm 2 537mm m
D-48
D.5.2.1.2 Shear
ENV 1992-1-1: 1991
Transverse direction Section 4.3.2
The relevant shear force is taken at a distance y = 1.5 d = 847.5 mm from ENV 1992-2: 1996
hanger 20 at x = 60.3 m. Section 4.3.2
Geometry
v Sd,y [kN/m] 710.2
z s [mm] 45
nSd,y [kN/m] -3113
d [mm] 565
Elements not requiring design shear reinforcement: VSd VRd1
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d
Rd = 0.48 N/mm2 Basic design shear for C50/60
1 = As1 / (bw d) 0.02
As1 = 5.24 cm2/m

d = 0.565 m
1 = (5.24 cm2/m) / (1 m 0.565 m) = 0.00093
ENV 1992-1-1: 1991

cp = nSd / Ac = 3113 kN / ( 610 mm 1m) = 5.1 N/mm2
Section 4.3.2.3
(Compression positive)
ENV 1992-1-1: 1991

k = 1.6 d = 1.6 0.575 = 1.025
Table 4.8
N N 1
VRd1 = 0.48 2
1.025 (1.2 + 40 0.00093 ) + 0.15 5.1 1m 0.575m
mm mm 2 m
VRd1 = 790 kN/m
VSd = 710.2 kN m < VRd1 = 790 kN m
No shear reinforcement required.
Minimum shear reinforcement:

Concrete C50/60, steel S500: w,min = 0.0013
ENV 1992-1-1: 1991
Section 5.4.2.2
1 = As1 / (bw d) 0.02
A sw = w b w s
A sw = 0.0013 100cm 100cm = 13cm 2 m
D-49
Longitudinal direction
The relevant shear force is taken at x = 60.3 m.
Geometry
v Sd,x [kN/m] 615.1 z s [mm] 45
nSd,x [kN/m] -2410 d [mm] 565
Elements not requiring design shear reinforcement: VSd VRd1
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d

ENV 1992-1-1: 1991
1 = As1 / (bw d) 0.02 Section 4.3.2.3
As1 = 5.24 cm2/m ENV 1992-1-1: 1991

d = 0.565 m Table 4.8
1 = (5.24 cm2/m) / (1 m 0.565 m) = 0.00093
cp = NSd / Ac = 2410 kN / ( 610 mm 1m) = 3.95 N/mm2

(compression positive)
k = 1.6 d = 1.6 0.575 = 1.025
N N 1
VRd1 = 0.48 2
1.025 (1.2 + 40 0.00093 ) + 0.15 3.95 1m 0.565m
mm mm 2 m
VRd1 = 678 kN/m
VSd = 615.1kN m < VRd1 = 678 kN m
Minimum shear reinforcement:

Concrete C50/60, steel S500: w,min = 0.0013
1 = As1 / (bw d) 0.02

A sw = w b w s
bw = 1.3 m (width of edge beam)

A sw = 0.0013 130cm 100cm = 17cm 2 m
The minimum shear reinforcement for both directions has not been added
together.
The greater value is applied: Asw = 17 cm2/m
12, 4-shear stirrups, spacing s = 20 cm (gives Asw = 22.6 cm2/m)
D-50
D.5.2.1.3 Punching
Punching shear is checked for with the maximal hanger force 1061.85 kN,
which is found in hanger number 4 with a slope of 75.2. The vertical ENV 1992-1-1: 1991
component of the hanger force is Nmax = 1027 kN. Section
ENV 4.3.4.5.1
1992-1-1: 1991
Internal force Section 4.3.4
NSd [kN] 1027
Geometry of bearing plate Geometry of concrete section

l [mm] 420 h [mm] 610
w [mm] 150 z s [mm] 45
f [mm] 120 d [mm] 565
No shear reinforcement: vSd vRd1 84.75 42
12
Shear per unit length: v Sd = sd
20
u
VSd total design shear force; VSd = 1027 kN
coefficient taking eccentricity of
loading into account; here = 1.0
u critical perimeter; u = 2w + l + d1.5 + 2 f Fig. D.20. Punching shear critical perimeter
u = (2 20 + 42 + 56.5 1.5 + 2 12 )cm = 372cm
1027kN 1.0
v Sd = = 276 N mm (Shear per unit length)
3720mm
vRd1 shear resistance per unit length

vRd1 = Rd k (1.2 + 40 1) d
Rd basic design shear strength;

Rd = 0.48 N/mm2
k k = 1.6 d = 1.6 0.575 = 1.025
cpo
1 = 1x 1y + 0.015
f yd
cpo concrete stress due to initial prestress;
transverse direction: Npd.y = 3135 kN/m
cpo.y = 3135 kN/m / 0.61 m = 5.1 N/mm2 Annex B:
longitudinal direction: cpo.x = 11.6 N/mm2 Preliminary Design
average value: cpo = (5.1 + 11.6)/2 N/mm2 = 8.35 N/mm2
fyd design yield stress of the reinforcement;
fyd = fy / 1.15 = 435 N/mm2
Asx = Asy = 5.24 cm2/m

1x = 1y = (5.24 cm2/m) / (1 m 0.565 m) = 0.00093
1 = 0.00093 + 8.35 N/mm2 / 435 N/mm2 = 0.02 > 0.015
vRd1 = 0.48 N/mm2 1.025 (1.2 + 40 0.015) 565 mm = 500 N/mm

vSd = 276 N/mm vRd1 = 500 N/mm
No punching shear reinforcement required.
D-51
D.5.2.2 Serviceability limit states
Category C is assumed for the classification of assessment conditions. ENV 1992-2: 1996
Table 4.118
The serviceability limit state assessment is mainly based on comparison of
both concrete and steel stresses with certain limits. All decisive concrete and
steel stresses are listed in Annex C, Figure C.39. In Figure D.21, relevant
values are compared with the serviceability limit state requirements.
State / comb. ENV 1992-2: Requirement Stress [N/mm 2]
1996 Section Longi- Trans-
tudinal verse
Time of 4.4.1.1 (102) Max. concrete compressive stress
prestressing 0.6 f c(t) or 0.45 f c(t)
0.45 f c(t) = -22.5 N/mm2 -17.1 -16.8
Decompression 4.4.2.1 (106) Concrete in compression -3.41 -4.66
(quasi- 4.4.1.1 (106) Stress in tendons 1150.5 N/mm (L)

2
980 -
permanent) after losses 0.65 f pk 800 N/mm2 (T) - 703
4.4.1.1 (103) Max. concrete compressive stress -18.6 -20.2
0.6 f ck = - 30 N/mm2
Non-frequent 4.4.1.1 (105) Limitation of tensile stress in 354 88.5
reinforcement to 0.8 f yk = 400 N/mm2 1) 1)
4.4.2.2.2 (101) c - 1 N/mm2, 4.6 6.9
otherw ise minimum reinforcement 2) 2)
Crack width 4.4.2.3 (101) if c f ctm = 4.1 N/mm2, then min. reinf. -0.3 0.67
(frequent) according to (4.4.2.2.2 (101)) sufficient
Notes:
1)
The tensile stress in the reinforcement is calculated further below in the text
2)
The minimum reinforcement is calculated further below in the text
Fig. D.21. Serviceability limit state assessment
Limitation of steel stress in reinforcement
The tensile stress in the reinforcement steel should be limited to SCHNEIDER [29]
0.8 fyk = 400 N/mm2 for the non-frequent combination of actions. The page 5.116
approximate and conservative calculation was carried out with the help of the
tension triangle in the uncracked stage. The concrete stresses were taken from
Figure C.39, Annex C.
In transverse direction, the prestressing steel lies within the tensile zone
and is therefore additionally considered to the reinforcement. In longitudinal
direction, only reinforcement lies in the tensile zone and is therefore considered
exclusively.
Transverse direction
Concrete stress at top level: y,top = -20.2 N/mm2

Concrete stress at bottom level: y,bottom = 6.9 N/mm2
D-52
0.43
Tensile zone: x' = 6.9 = 0.11m
(6.9 + 20.2)
Distance of thread bar from bottom level: zp = 103 mm < 110 mm
Thread bar lies within tensile zone.
Area of prestressing steel: Ap = 37.7 cm2/m
Area of reinforcement: As = 5.24 cm2/m
Force in the tension triangle: F = 0.5 0.11m 6.9 N/mm2 = 380 kN/m
Steel stress:
s = F / As = 380 kN/m / (5.24 +37.7) cm2/m = 88.5 N/mm2 < 400 N/mm2.
Longitudinal direction
Concrete stress at top level: y,top = -18.6 N/mm2

Concrete stress at bottom level: y,bottom = 4.6 N/mm2
0.43
Tensile zone: x' = 4.6 = 0.085m
(4.6 + 18.6)
Force in the tension triangle: F = 0.5 0.085m 4.6 N/mm2 = 195.5 kN/m
Area of reinforcement: As = 5.24 cm2/m
Steel stress: s = F / As = 185.5 kN/m / 5.24 cm2/m = 354 N/mm2 < 400 N/mm2.
Minimum reinforcement transverse direction
As = (0.8 kc k fctm Act) / s ENV 1992-2: 1996

Section 4.4.2.2.3
fctm tensile strength of the concrete: fctm = 4.1 N/mm2
Act area of concrete within tensile zone just before cracking;
concrete stress due to prestressing:
P = 2650.9 kN (Annex B, Figure B.9)
p = P / Ac = - 2650.9 kN / 0.43 m2 = - 6.16 N/mm2
Act = 0.0859 m2/m -10.26 -16.42
s maximum stress permitted in the reinforcement

immediately after cracking, depending on the re-
43 cm
8.59 cm
bar diameter; for a bar size of 12 mm in a

prestressed concrete section: s = 240 N/mm2
(ENV 1992-2: 1996 Table 4.120)
kc coefficient taking into account the nature of the -6.16 10.26 4.1
concrete stress due concrete stress due concrete stress just
stress distribution within the section immediately to axial prestressing to cracking moment before cracking
prior to cracking; kc = 0.4 (conservative)

Fig. D.22. Concrete stress distribution just
2
before cracking [N/mm ]
k 0.9 (interpolated value for h = 0.43 m)
As = (0.8 0.4 0.9 4.1 N/mm2 859 cm2/m) / 240 N/mm2 = 4.3 cm2/m
ds = 10 mm; s = 15 cm (gives As = 5.24 cm2/m)
D-53
Minimum reinforcement longitudinal direction
Concrete slab
ENV 1992-2: 1996
As = (0.8 kc k fctm Act) / s
Section 4.4.2.2.3
fctm tensile strength of the concrete: fctm = 4.1 N/mm2.
Axial force due to prestressing: P = 2400 kN (SOFiSTiK output)
p = P / Ac = -2400 kN / 0.43 m2 = - 5.6 N/mm2
Act = 0.0907 m2/m -9.7 -15.3
s maximum stress permitted in the reinforcement

immediately after cracking, depending on the re-
bar diameter; for a bar size of 12 mm in a
43 cm
9.07 cm
prestressed concrete section: s = 240 N/mm2
(ENV 1992-2: 1996 Table 4.120)
kc coefficient taking into account the nature of the
stress distribution within the section immediately -5.6 9.7 4.1
prior to cracking; kc = 0.4 (conservative) concrete stress due
to axial prestressing
concrete stress due concrete stress just
to cracking moment before cracking

Fig. D.23. Concrete stress distribution just
2
before cracking [N/mm ]
As = (0.8 0.4 0.9 4.1 N/mm2 907 cm2/m) / 240 N/mm2 = 4.5 cm2/m
10 mm; s = 15 cm (gives As = 5.24 cm2/m)
Edge beam
Axial force due to prestressing: Np = 3000 kN (SOFiSTiK output)
Concrete stress: p = Np / Ac = 3000 kN / 0.61 m2 = 5 Nmm2
Area of concrete within tensile zone just before cracking: Act = 1374 cm2/m
-8.5 -14.1
61 cm
13.74 cm
-5.0 8.5 4.1

concrete stress due concrete stress due concrete stress just
to axial prestressing to cracking moment before cracking
2
Fig. D.24. Concrete stress distribution just before cracking [N/mm ]
k = 0.78 (interpolated value for h = 0.61 m)

As = (0.8 0.4 0.78 4.1 N/mm2 1374 cm2/m) / 240 N/mm2 = 5.9 cm2/m
12 mm; s = 15 cm (gives As = 7.54 cm2/m)
The spacing of reinforcement steel bars shall not exceed 200 mm. This
criterion is fulfilled for all reinforcement layers.
D-54
D.5.2.3 Fatigue
Relevant combination of actions: Frequent
Member in compression
If the following criterion is fulfilled, sufficient resistance against fatigue for ENV 1992-2: 1996
concrete under compression is given. Section 4.3.7.4 (101)
c,max c,min
0.5 + 0.45 0 .9
f cd fcd
c.max maximum compressive stress (frequent combination of actions)
c.max = -14.1 N/mm2
c.min minimum compressive stress at the same point where c.max occurs
c.min = -0.3 N/mm2
fcd fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
14.1N mm 2 0.3 N mm 2
0.5 + 0.45 0 .9
33.3 N mm 2 33.3 N mm 2
0.42 0.5 0.9
Member in shear
If the following criterion is fulfilled, sufficient resistance against fatigue for

concrete under shear without shear reinforcement is given.
max min
for 0.5 + 0.45 min 0.9 0 ENV 1992-2: 1996
Rd1 Rd1 max Section 4.3.7.4 (103)
Maximum shear force (frequent combination of actions): vSd.y.max = 702 kN/m

Minimum shear force (frequent combination of actions): vSd,y,min = 556 kN/m
max maximum shear stress under the frequent combination of actions;

max = vSd.y.max / d = 702 kN/m / 0.565 m = 1.24 N/mm2
(transverse direction)
min minimum shear stress at the same section where max occurs;
min = vSd,y,min / d = 556 kN/m / 0.565 m = 0.98 N/mm2
Rd1 Rd1 = VRd1 / (1m d) Section D.5.2.2.1

VRd1 = 968 kN/m (transversal direction)
Rd1 = 968 kN/m / 0.565 m = 1.71 N/mm2
1.24 N mm 2 0.98 N mm 2
0.5 + 0.45 0 .9
1.71N mm 2 1.71N mm 2
0.73 0.76 0.9
D-55
Punching
If the following criterion is fulfilled, sufficient resistance against fatigue for

concrete subjected to punching shear is given.
max min ENV 1992-2: 1996

0.5 + 0.45 min 0.9 for 0 Section 4.3.7.4 (104)
Rd1 Rd1 max
= vSd / d
Rd1 = vRd1 / d
Vsd
v Sd =
u
Vertical component of maximum hanger force: VSd,max = 630 kN

Vertical component of minimum hanger force: VSd,min = 305 kN
eccentricity coefficient; = 1.0 Section D.5.2.3

u critical perimeter; u = 3720 mm
vRd1 shear resistance per unit length; vRd1 = 500 N/mm
Vsd,max 630kN 1.0 N

v Sd,max = = = 170 (Shear per unit length)
u 3720mm mm
Vsd,min 305kN 1.0 N
v Sd,min = = = 82
u 3720mm mm
max = vSd,max / d = 170 N/mm / 0.565 m = 0.3 N/mm2

min = vSd,min / d = 82 N/mm / 0.565 m = 0.15 N/mm2
Rd1 = vRd1 / d = 500 N/mm / 0.565 m = 0.88 N/mm2
0 .3 0.15
0.5 + 0.45 0 .9
0.88 0.88
0.34 0.58 0.9
D-56
Prestressing steel
Transverse direction
F Sd s,equ
Rsk N* ( )
s, fat
ENV 1992-2: 1996
F = 1.0
Section 4.3.7.2
Sd = 1.0
s,fat = 1.15
ENV 1992-2: 1996
Rsk(N*) = 120 N/mm2
Table 4.116
N* = 106, k1 = 3, k2 = 7
ENV 1992: 1996
s,equ = s s,71 Annex A 106.3
s,71 Stress variation due to load model 71 for the non-frequent

combination of actions including the dynamic factor
s,71 = 75 N/mm2
s = s,1 s,2 s,3 s,4 ENV 1992-2: 1996
s,1 s,1 = 0.70 Table A 106.2
Vol
s,2 s,2 = k 2 6
Vol = 30 106 t / (track year)
25 10
30 10 6
s,2 = 7 = 1.026
25 10 6
Nyears
s,3 s,3 = k 2 Nyears = 100
100
100
s,3 = 7 =1
100
n + (1 n) s1 + (1 n) s 2
k2 k2 k2
s,4 s,4 =
n = 0.12 (taken from ENV 1993-2: 1997 Section 9.5.3 (8))
1 2
s1 = ; s2 =
1+ 2 1+ 2
1, 2 Stress variation due to load model 71 on one track
1+2 Stress variation due to load model 71 on two tracks
1 = 2 = 42 N/mm2
1+2 = 75 N/mm2
42 N mm 2
s1 = s 2 = = 0.56
75 N mm 2
s,4 = 7 0.12 + (1 0.12) 0.56 7 + (1 0.12 ) 0.56 7 = 0.76
s = 0.7 1.026 1 0.76 = 0.55

s,equ = s s,71 = 0.55 75 N/mm2 = 41.3 N/mm2
N N 120 N mm 2 N
1.0 1.0 50.7 2
= 41.3 2
= 104.3
mm mm 1.15 mm 2
D-57
D.5.2.4 Footpath
Structural system
Cantilever with length L = 63 cm
Geometry
Reinforcing steel: ds = 20 mm
Concrete cover: nom c = 4 cm
Depth: h = 15 cm
Effective depth: d = 15 cm 4 cm 1 cm = 10 cm
Loading
Live load: p = 5 kN/m2
Dead load: g = 25 kN/m3 0.15 m = 3.75 kN/m2
q = 1.35 g + 1.5 p = 12.56 kN/m2
Internal forces
mSds = q L2/2 =12.56 kN/m2 (0.63 m)2 / 2 = 2.5 kNm/m
m Sds
Sds =
b d2 fcd
2.5 kNm m
Sds = = 0.0075 take Sds.min = 0.05
(0.10m)2 33.3 N mm 2
= 0.966 z = d = 100 mm 0.966 = 96.6 mm

fsd = fyk / s = 500 N/mm2 / 1.15 = 435 N/mm2
1 m 1 2.5 kNm m cm 2
As = Sds = 2
= 0 .6
f sd z 435 N mm 96.6mm m
Minimum reinforcement
ENV 1992-2: 1996
As = (0.8 kc k fctm Act) / s Section 4.4.2.2.3
Act = 0.15 / 2 m2/m = 0.075 m2/m.
s maximum stress permitted in the reinforcement immediately
after cracking, depending on the re-bar diameter;
s = 320 N/mm2 for re-bar 20 (ENV 1992-1-1: 1991 Table 4.11)
kc coefficient taking into account the nature of the stress distribution
within the section immediately prior to cracking;
kc = 0.4
k 1.0
As = (0.8 0.4 1.0 4.1 N/mm2 750 cm2/m) / 320 N/mm2 = 3.1 cm2/m
10, s = 15 cm (gives As = 5.24 cm2/m)
D-58
D.5.2.5 Summary of reinforcement
Edge beam
Top: 12, s = 15 cm
Bottom: 12, s = 15 cm
Bridge deck: Transverse

Top: 20, s = 9 cm
Bridge deck: Longitudinal:

Top: 10, s = 15 cm
Footpath
Edge stirrups: 10, s = 15 cm
Shear stirrups: 12, s = 20 cm (4-shear)
Calculation of reinforcement steel weight
Geometry [m]
Bridge length 100
Width on the top 11.85 The calculation of the
Width on the bottom 10.59 steel weight is based on
Width of edge beam 1.3 the geometric properties
Length of cantilever 0.6 listed in Figure D.25.
Length of stirrup [m] 2.5
Length of cantilever strirrups 1
Fig. D.25. Geometric properties for calculation of steel weight
Calculation of reinforcement steel weight
Nom. mass Weight

[mm] s [cm] Number Length [m]
[kg/m] [kg]
Top 20 2.47 9.00 1111.11 13166.67 32522
Bottom 10 0.62 14.00 714.29 7564.29 4667
Bridge deck: Longitudinal
Top 10 0.62 15.00 74.33 7433.33 4586
Bottom 10 0.62 15.00 61.93 6193.33 3821
Edge beam
Top 12 0.89 15.00 17.33 1733.33 1539
Bottom 12 0.89 15.00 17.33 1733.33 1539
Cantilever stirrups 10 0.62 15.00 1333.33 1333.33 823
Shear stirrups 12 0.89 20.00 2000.00 5000.00 4440
1)
Additional reinforcement accounting for overlapping Total 53938
1)
of re-bars and unconsidered reinforcement + 15% 62028
Fig. D.26. Calculation of steel weight
D-59
D.5.3 Alternative design proposals for bridge deck
D.5.3.1 Alternative design 1: Slab depth 610 mm
D.5.3.1.1 Ultimate limit states
Bending
ENV 1992-1-1 Section 4.3.1
ENV 1992-2 Section 4.3.1
557.5
The design checks are performed at mid-span, where the
610
decisive bending moment is found.
Geometry 12
25
52.5
z s [mm] 52.5
Internal forces
mSd,y [kNm/m] 1195.6 h [mm] 610
d [mm] 557.5
90
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2 Fig. D.27. Cross section
mSds
Sds =
b d2 fcd
SCHNEIDER [29],
1195.6 kNm m
Sds = = 0.115 page 5.130
(0.5775m)2 33.3 N mm 2
= 0.93 z = d = 557.5 mm 0.93 = 518 mm
1 m Sds
As =
Sd z
1 1195 .6 kNm m cm 2
As = = 53
435 N mm 2 518mm m
Principal reinforcement:
Re-bars: 25, s = 9 cm
As = 54.54 cm2/m
Secondary transverse reinforcement:
ENV 1992-1-1: 1991
As = As / 5 = 10.9 cm2/m Section 5.4.3.2.1(2)
Re-bars: 12, s = 10 cm (gives As = 11.31 cm2/m)
D-60
Shear
ENV 1992-1-1 Section 4.3.2

Check for maximum shear force
Geometry
Internal forces z s [mm] 52.5
v Sd,y [kN/m] 566 h [mm] 610
d [mm] 557.5
Criterion for elements not requiring design shear reinforcement: VSd VRd1
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d ENV 1992-1-1: 1991

Section 4.3.2.3
1 = As1 / (bw d) 0.02 ENV 1992-1-1: 1991

Table 4.8
As1 = 54.54 cm2/m
d = 0.5575 m
1 = (54.54 cm2/m) / (1 m 0.5575 m) = 0.01 < 0.02
cp = 0 N/mm2 (No prestressing)

k = 1.6 d = 1.6 0.5575 = 1.0425 > 1
VRd1 = 446.4 kN/m
VSd = 566 kN m > VRd1 = 446.4 kN m Check not fulfilled!
VRd2 = 0.5 fcd bw 0.9 d

ENV 1992-1-1: 1991:
Effectiveness factor: = 0.7 fck / 200 0.5
Section 4.3.2.3 (3)
= 0.7 50 / 200 = 0.45 < 0.45
= 0.5
VRd2 = 0.5 0.5 33.3 N/mm2 0.9 557.5 mm = 4177.1 kN/m
Vwd = VSd VRd1 = 566 kN/m 446.4 kN/m = 120 kN/m
Vwd
A sw =
0.9 d f ywd ENV 1992-1-1: 1991
Section 4.3.2.4.3
120kN
A sw =
0.9 557.5mm 435 N mm 2
A sw = 5.5 cm 2 m
Minimum shear reinforcement
Concrete C50/60, steel 500 S: w,min = 0.0013

w = Asw / (bw sin())
ENV 1992-1-1: 1991
Asw = w bw sin() = 0.0013 100 cm sin(90)
Section 5.4.2.2
Asw = 13 cm2 / m > 5.5 cm2/m
D-61
The shear reinforcement for the edge beam in the longitudinal bridge
direction requires a minimum of 13 cm2/m as well and is added to the shear
reinforcement in transverse direction.
ENV 1992-1-1: 1991
Total required shear reinforcement: Asw = 26 cm2/m
Section 5.4.2.2
Stirrups 12, s = 15 cm, four-shear (gives Asw = 30.13 cm2/m)
Maximal spacing
Longitudinal spacing
VSd = 566 kN/m < 1/5 VRd2 = 4177.1 kN/m / 5 = 835 kN/m
smax = 0.8 d = 0.8 557.5 mm = 446 mm (300 mm)
s = 150 mm < 300 mm
Transverse spacing
smax = d = 557.5 mm (800 mm) not critical
Check for shear force at smallest depth
Geometry
Internal forces z s [mm] 52.5
v Sd,y [kN/m] 383 h [mm] 554
d [mm] 501.5
ENV 1992-1-1: 1991
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d Section 4.3.2.3
ENV 1992-1-1: 1991
1 = As1 / (bw d) 0.02 Table 4.8
As1 = 54.54 cm2/m
d = 0.5015 m
1 = (54.54 cm2/m) / (1 m 0.5575 m) = 0.01 < 0.02
cp = 0 N/mm2
k = 1.6 d = 1.6 0.5015 = 1.1 > 1
N 1
VRd1 = 0.48 2
1.1 (1.2 + 40 0.01) 1m 0.5015m
mm m
VRd1 = 423.7 kN/m
VSd = 383 kN m < VRd1 = 446.4 kN m Check fulfilled!
D-62
Punching
ENV 1992-1-1 Section 4.3.4
Hanger force
NS d [kN] 1110.5

l [mm] 420 z s [mm] 52.5
w [mm] 150 h [mm] 610
f [mm] 120 d [mm] 557.5
Criterion for elements not requiring shear reinforcement: vSd vRd1
Vsd
v Sd =
u
VSd total design shear force; VSd = 1110.5 kN
coefficient taking eccentricity of loading into account; here = 1.0
u critical perimeter; u = 2w + l + d1.5 + 2 f
u = (2 20 + 42 + 55.75 1.5 + 2 12 )cm = 3687mm
1110 .5kN 1.0

v Sd = = 301N mm (shear per unit length)
3687mm

vRd1 = Rd k (1.2 + 40 1) d

Rd = 0.48 N/mm2
k k = 1.6 d = 1.6 0.5575 = 1.0425 > 1
1 = 1x 1y 0.015
Asx = 11.31 cm2/m
1x = (11.31 cm2/m) / (1 m 0.5575 m) = 0.002
Asx = 54.54 cm2/m
1y = (54.54 cm2/m) / (1 m 0.5575 m) = 0.01
1 = 0.002 0.01 = 0.0045 < 0.015
vRd1 = 0.48 N/mm2 1.0425 (1.2 + 40 0.0045) 557.5 mm = 385 N/mm

v Sd = 301N mm v Rd1 = 385 N mm
No additional shear reinforcement required.
D-63
D.5.3.1.2 Serviceability limit state
ENV 1992-2: 1996
Relevant combination of actions: Non-frequent Section 4.4.2.2.3

Act = 0.61 m2 /2 = 0.305 m2/m.
s maximum stress permitted in the reinforcement immediately after

cracking, depending on the re-bar diameter; s = 200 N/mm2 (for
re-bar 25 ENV 1992-1-1: 1991 Table 4.11)
kc coefficient taking the nature of the stress distribution within the
section immediately prior to cracking into account;
kc = 0.4
As = (0.8 0.4 0.78 4.1 N/mm2 3050 cm2/m) / 200 N/mm2 = 15.6 cm2/m
15.6 cm2/m < 54.54 cm2/m
The minimum reinforcement at the bottom of the slab is already covered,
but it still needs to be applied at the top.
Re-bars: 14, s = 10 cm (gives As = 15.39 cm2/m)
As = 15.39 cm2/m < required 15.6 cm2/m, but the excess is less than 3% and
therefore acceptable.
Limitation of steel stress
Relevant combination of actions: Non-frequent
Requirement: Steel stress to be limited to 0.8 fy = 400 N/mm2

Ms = 814.1 kNm/m
m Sds
Sds =
b d2 fcd
814.1kNm m
Sds = = 0.079
(0.5775m)2 33.3 N mm 2
= 0.95 z = d = 557.5 mm 0.995 = 528.5 mm
As = 54.54 cm2/m
Ms 814.1kNm / m N N
s = = = 282 < 400
z A s 528.5mm 54.54 cm m
2
mm 2
mm 2
Crack width ENV 1992-2: 1996

2
For a steel stress of 282 N/mm and re-bar diameter of 25, a maximum Table 4.121
re-bar spacing of 150 mm is allowed. The actual spacing measures 90 mm.
D-64
D.5.3.1.3 Fatigue
Reinforcing steel
F Sd s,equ
Rsk N* ( )
s, fat
ENV 1992-2: 1996
F = 1.0
Section 4.3.7.2
Sd = 1.0
s,fat = 1.15
Rsk(N*) = 195 N/mm2 ENV 1992-2: 1996

N* = 106, k1 =5, k2 = 9 Table 4.116
s,equ = s s,71 ENV 1992: 1996

Annex A 106.3
s,71 Stress variation due to load model 71 for the non-frequent
s,71 = 140 N/mm2
s = s,1 s,2 s,3 s,4

s,1 s,1 = 0.65 ENV 1992-2: 1996
Vol Table A 106.2
s,2 s,2 = k 2 Vol = 30 106 t / (track year)
25 10 6
30 10 6
s,2 = 7 = 1.026
25 10 6
Nyears
s,3 s,3 = k 2 Nyears = 100
100
100
s,3 = 7 =1
100
n + (1 n) s1 + (1 n) s 2
k2 k2 k2
s,4 s,4 =
1 2
s1 = ; s2 =
1+ 2 1+ 2
1 = 2 = 62.4 N/mm2
1+2 = 140 N/mm2
61.55 N mm 2
s1 = s 2 = = 0.44
139.2 N mm 2
s,4 = 7 0.12 + (1 0.12 ) 0.44 7 + (1 0.12) 0.44 7 = 0.79

s = 0.65 1.026 1 0.79 = 0.53
s,equ = s s,71 = 0.53 140 N/mm2 = 73.8 N/mm2
N N 195 N mm 2 N
1.0 1.0 73.8 = 73.8 = 169.6
mm 2 mm 2 1.15 mm 2
D-65
D.5.3.2 Alternative design proposal 2: Slab depth 470 mm
D.5.3.2.1 Ultimate limit state
Bending and longitudinal force

ENV 1992-1-1 Section 4.3.1
398
Geometry
470
Internal forces z s [mm] 72
14
mSd,y [kNm/m] 1041.2 h [mm] 470 25
d [mm] 398
72
Assumption: Re-bars 25 in two layers with secondary
transverse re-bars 14 in-between.
140
Fig. D.28. Cross section
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
mSds
Sds =
b d2 fcd
1041.2 kNm m
Sds = = 0 .2
(0.398m)2 33.3 N mm 2
= 0.86 z = d = 398 mm 0.86 = 342.3 mm
1 m
As = Sds
Sd z SCHNEIDER [29],
page 5.130
1 1041 .16 kNm m cm 2
As = = 70
435 N mm 2 342.3mm m
Principal reinforcement
Re-bars: 25 in two layers, s = 14 cm (gives As = 70.12 cm2/m)
Secondary transverse reinforcement:
As = As / 5 = 14 cm2/m ENV 1992-1-1: 1991

Re-bars: 14, s = 10 cm (gives As = 15.39 cm2/m) Section 5.4.3.2.1(2)
D-66
Shear
ENV 1992-1-1 Section 4.3.2
Check for maximum shear force

Geometry
v S d,y [kN/m] 547 h [mm] 470
d [mm] 398
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d
1 = As1 / (bw d) 0.02

ENV 1992-1-1: 1991
As1 = 70.12 cm2/m Section 4.3.2.3
d = 0.398 m
1 = (71.12 cm2/m) / (1 m 0.398 m) = 0.018 ENV 1992-1-1: 1991
Table 4.8
cp = 0 N/mm2 (No prestressing)
k = 1.6 d = 1.6 0.398 = 1.02 > 1

N 1
VRd1 = 0.48 2
1.02 (1.2 + 40 0.018 ) 1m 0.398m
mm m
VRd1 = 440 kN/m
VSd = 547 kN m > VRd1 = 440 kN m Check not fulfilled!
VRd2 = 0.5 fcd bw 0.9 d

Effectiveness factor: = 0.7 fck / 200 0.5
= 0.7 50 / 200 = 0.45 < 0.45
= 0.5
VRd2 = 0.5 0.5 33.3 N/mm2 0.9 398 mm = 2982 kN/m ENV 1992-1-1: 1991:
Section 4.3.2.3 (3)
Vwd = VSd VRd1 = 547 kN/m 440 kN/m = 107 kN/m
Vwd
A sw =
0.9 d f ywd
ENV 1992-1-1: 1991
107kN
A sw = Section 4.3.2.4.3
0.9 398mm 435 N mm 2
A sw = 6.9 cm 2 m
D-67
Minimum shear reinforcement

ENV 1992-1-1: 1991
Concrete C50/60, steel S500: w,min = 0.0013 Section 5.4.2.2
w = Asw / (bw sin())
Asw = w bw sin() = 0.0013 100 cm sin(90)
Asw = 13 cm2 / m > 6.9 cm2/m
The shear reinforcement for the edge beam in the longitudinal bridge
direction requires a minimum of 13 cm2/m, as well, and is added to the shear
reinforcement in transverse direction.
Total shear reinforcement: ENV 1992-1-1: 1991
Section 5.4.2.2
Asw = 26 cm2/m
Stirrups 12, s = 15 cm, four-shear (gives Asw = 30.13 cm2/m)
Maximum spacing
Longitudinal spacing
VSd = 741.5 kN m < 2 3 VRd2 = 2 3 2982 kN m = 1988 kN m

smax = 0.6 d = 0.6 398 mm = 238.8 mm (300 mm)
s = 150 mm < 238.8 mm
Transverse spacing
smax = d = 398 mm (800 mm) not critical
Check for shear force at smallest depth

Geometry
v Sd,y [kN/m] 305 h [mm] 414
d [mm] 342
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d ENV 1992-1-1: 1991

Section 4.3.2.3
1 = As1 / (bw d) 0.02 ENV 1992-1-1: 1991

Table 4.8
As1 = 54.54 cm2/m
d = 0.342 m
1 = (54.54 cm2/m) / (1 m 0.342 m) = 0.016 < 0.02
cp = 0 N/mm2
k = 1.6 d = 1.6 0.342 = 1.126 > 1
N 1
VRd1 = 0.48 2
1.26 (1.2 + 40 0.016 ) 1m 0.342m
mm m
VRd1 = 380 kN/m
VSd = 305 kN m < VRd1 = 380 kN m Check fulfilled!
D-68
Punching
ENV 1992-1-1: 1991 Section 4.3.4
Hanger force
NSd [kN] 1073

l [mm] 420 z s [mm] 72
w [mm] 150 h [mm] 470
f [mm] 120 d [mm] 398
Criterion for elements not requiring shear reinforcement: vSd vRd1

V
v Sd = sd
u
VSd total design shear force; VSd = 1073 kN

coefficient taking eccentricity of loading into account; here = 1.0
u critical perimeter; u = 2w + l + d1.5 + 2 f
u = (2 20 + 42 + 39.8 1.5 + 2 12)cm = 2936mm
1073kN 1.0
v Sd = = 365 N mm (Shear per unit length)
2936mm

vRd1 = Rd k (1.2 + 40 1) d

Rd = 0.48 N/mm2
k k = 1.6 d = 1.6 0.398 = 1.2 > 1
1 = 1x 1y 0.015
Asx = 15.39 cm2/m
1x = (15.39 cm2/m) / (1 m 0.398 m) = 0.0039
Asx = 70.12 cm2/m
1y = (70.12 cm2/m) / (1 m 0.398 m) = 0.0176
1 = 0.0039 0.0176 = 0.008 < 0.015
vRd1 = 0.48 N/mm2 1.2 (1.2 + 40 0.008) 398 mm = 348 N/mm

vRd2 = 1.6 vRd1 = 1.6 348 N/mm = 556.8 N/mm
vRd3 = vRd1 + Asw fyd sin() / u
vRd3 = 348 N/mm + (19.3 cm2/m 1.5 398mm) 435 N/mm2 / 2596 mm
vRd3 = 541 N/mm
v Sd = 365 N mm < v Rd3 = 541N mm
D-69
D.5.3.2.2 Serviceability limit state
ENV 1992-2: 1996 Section 4.4.2.2.3

Act = 0.47 m2/m / 2 = 0.235 m2/m.
s maximum stress permitted in the reinforcement immediately after

cracking, depending on the re-bar diameter; s = 200 N/mm2 for
re-bar 25 (ENV 1992-1-1: 1991 Table 4.11)
kc coefficient taking into account the nature of the stress distribution
within the section immediately prior to cracking;
kc = 0.4
As = (0.8 0.4 0.88 4.1 N/mm2 2350 cm2/m) / 200 N/mm2 = 13.6 cm2/m
13.6 cm2/m < 70.12 cm2/m
The minimum reinforcement at the bottom of the slab is already covered,
but it still needs to be applied at the top.
14, s = 10 cm (gives As = 15.39 cm2/m)
Limitation of steel stress

Requirement: Steel stress to be limited to 0.8 fy = 400 N/mm2
Steel stress to be limited to 0.8 fy = 400 (non-frequent)
Ms = 700 kNm/m
m Sds
Sds =
b d 2 fcd
700 kNm m
Sds = = 0.134
(0.398m)2 33.3 N mm 2
= 0.914 z = d = 398 mm 0.935 = 363.8 mm
As = 70.12 cm2/m
Ms 700kNm / m N N
s = = = 274 < 400
z A s 363.8mm 70.12 cm 2 m mm 2 mm 2
Crack width
ENV 1992-2: 1996

For a steel stress of 274 N/mm2 and re-bar diameter of 25, a maximum
Table 4.121
re-bar spacing of 150 mm is allowed. The actual spacing measures 140 mm.
D-70
D.5.3.2.3 Fatigue
The fatigue assessment is performed for the reinforcing steel. ENV 1992-2: 1992
Relevant combination of actions: Non-frequent Section 4.3.7
F Sd s,equ
Rsk N* ( )
s, fat
ENV 1992-2: 1996
F = 1.0
Section 4.3.7.2
Sd = 1.0
s,fat = 1.15
ENV 1992-2: 1996
Table 4.116
Rsk(N*) = 195 N/mm2
N* = 106, k1 =5, k2 = 9
ENV 1992: 1996
Annex A 106.3
s,equ = s s,71
s,71 Stress variation due to load model 71 for the characteristic/rare
s,71 = 160.4 N/mm2
s = s,1 s,2 s,3 s,4 ENV 1992-2: 1996

Table A 106.2
s,1 s,1 = 0.65
Vol
s,2 s,2 = k 2 Vol = 30 106 t / (track year)
25 10 6
30 10 6
s,2 = 7 = 1.026
25 10 6
Nyears
s,3 s,3 = k 2 Nyears = 100
100
100
s,3 = 7 =1
100
n + (1 n) s1 + (1 n) s 2
k2 k2 k2
s,4 s,4 =
1 2
s1 = ; s2 =
1+ 2 1+ 2
1 = 2 = 68.8 N/mm2
1+2 = 160.4 N/mm2
68.8 N mm 2
s1 = s 2 = = 0.43
160.4 N mm 2
s,4 = 7 0.12 + (1 0.12 ) 0.43 7 + (1 0.12) 0.43 7 = 0.79

s = 0.65 1.026 1 0.79 = 0.53
s,equ = s s,71 = 0.53 160.4 N/mm2 = 85 N/mm2
N N 195 N mm 2 N
1.0 1.0 85 = 85 = 169.6
mm 2 mm 2 1.15 mm 2
D-71
D.5.3.3 Calculation of deflection
The deflections of concrete structures should be calculated for the quasi-

permanent combination of actions taking creep, shrinkage and cracking into
account, ENV 1992-2: 1996 Section 4.4.3.1 (106). However, the quasi-
permanent combination does not include live load due to railway traffic
(Annex A, Figure A.8). According to ENV 1991-3: 1995 Section 6.8.1, vertical
deflections of railway bridges are to be calculated with one track loaded.
Therefore, the following includes permanent loads as well as live loads in
the form of one load model 71. Firstly, the vertical deflections are calculated due
to permanent loads with respect to creep, shrinkage and cracking. Secondly, an
additional vertical deflection due to live load is determined for the cracked
section.
The calculation is based on a simplified procedure. Deflections are
predicted by empirical interpolation between lower (at uncracked stage) and
upper bound (cracked stage). The member is assumed to have a constant
section in each state, and deflections are obtained by using the
moment - curvature relation (GHALI A., ET AL. [11], pages 303-317).
The equations SPERLING [36] listed in the following were used to determine
the results given in Figure D.5.10 and Figure D.5.11. Reference is made by
numbered equations. The deflections calculated here are only deflections of the
bridge deck slab. The total deflection of the bridge is mainly constituted by
deflections of the edge beams.
The procedure is also adopted in ENV 1992-1-1: 1991 Appendix 4.
Calculation of effective elastic modulus

The effect of creep is accounted for by using an effective elastic modulus,
which is a reduced elastic modulus at time t = . The creep coefficient is
taken from Annex C Preliminary Design.
E cm
E c.eff = (Equ. D.1) Effective elastic modulus
1+
Uncracked
In the uncracked stage the bending moment acting on the member does not
exceed the cracking moment. The curvature is calculated with the second
moment of area of the total concrete section.
b h3
II = (Equ. D.2) Second moment of area of
12
I
1 MSd
k' = = (Equ. D.3) Curvature
r E c,eff II
D-72
Cracked
At this stage, the bending moment acting on the member exceeds the
cracking moment, which means that the concrete tensile strength is exceeded.
The stress distribution in the compressive zone is assumed to be linear. The
flexural stiffness of the member is determined by combining the compatibility
condition (Bernoulli hypothesis) and equilibrium conditions, SPERLING [36],
page 15.
Es
e = (Equ. D.4) Elastic modulus ratio
E c,eff
As
= (Equ. D.5) Reinforcement
bd
x
= e + e (2 + e ) (Equ. D.6) Compression zone
d
b x3
+ e A s (d x )
2
III = (Equ. D.7) Second moment of area
3
II
1 MSd
k' ' = = (Equ. D.8) Curvature
r E c.eff III
Tension stiffening (Interpolation)

Even though the section is cracked, the concrete between cracks in the
tensile zone gives contribution to the rigidity of the member. This contribution is
also referred to tension stiffening. Therefore an effective or mean value of the
member stiffness must be considered in the calculation of the curvature. This is
done by interpolation between the lower and upper bound, i.e. interpolation
between the curvatures in the uncracked and cracked stage, respectively. The
interpolation coefficient is given by the cracking moment, the bending moment
acting on the section and coefficients 1 and 2. GHALI A., ET AL. [11], page 314.
1 accounts for the used reinforcement:
1 = 1.0 for high-bond reinforcing bars
1 = 0.5 for plain bars
2 represents the influence of the duration and repetition of loading:
2 = 1.0 for short-time loading
2 = 0.5 for long-time loading
II I
1 1 1
= + (1 ) (Equ. D.9) Interpolation
rm r r
2
M
= 1 1 2 cr
(Equ. D.10) Interpolation coefficient
Ms
h2 b
Mcr = fctm W = fctm (Equ. D.11) Cracking moment
6
Shrinkage effect
The effect of shrinkage is accounted for by calculating curvatures in the

uncracked and cracked stage and interpolating between both similarly to the
procedure mentioned above. The formula corresponds to the one given in
ENV 1992-1-1: 1991 Appendix 4.
D-73
Uncracked
I
1 E A s z Is
k 'sh = = cs s
(Equ. D.12) Curvature due to shrinkage
rsh E c.eff II
Cracked
II
1 E A s z IIs
k 'sh
'
= = cs s
(Equ. D.13) Curvature due to shrinkage
rsh E c.eff III
Interpolation
II I
1 1 1
k sh = = + (1 ) (Equ. D.14) Interpolation
r r
rsh.m sh sh
Total curvature
The total curvature due to the applied long-term dead load and shrinkage
including creep effects is determined by superposition, SPERLING [36], page 21.
1 1 1
k= = +
(Equ. D.15)
rtot rm rsh.m
Deformation
The total deformation due to long-term dead loads and shrinkage is
determined with the following formula:
1
= k L2 (Equ. D.16)
r
with
L span
k parameter of moment distribution;
k = 0.104 for a uniformly distributed load acting on a simply
supported beam
Deformation due to live load
The calculation due to the applied live load on the cracked reinforced
concrete deck is based on the Branson effective moment of inertia, ACI 435 [2].
I Mcr
4 4
M III
Ie = cr I + 1 (Equ. D.17) Effective moment of inertia
M M

1 M
m = = (Equ. D.18) Mean curvature
rLL E Ie
3 4 2
Equation D-16 was used here with the parameter k = (Equ. D.19)
48 (1 )
= a / L, where a is the smaller distance of the single train load from the
support of the simple beam, and L is the span of the bridge deck.
D-74
Alternative design proposal 1: Calculation of deflections

Geometry
Width [m] b [m] 1
Depth [m] h [m] 0.61
Effective depth [m] d [m] 0.5575
Internal forces
Bending moment due to permament load M [kNm/m] 303.2
Bending moment due to live load M [kNm/m] 541.9
Cracking moment Mcr [kNm/m] 254.27 Equ. D.11
Properties
2
Reinforcement elastic modulus Es [N/mm ] 200000
2
Concrete elastic modulus Ecm [N/mm ] 37000
Creep coefficient 1.9
Shrinkage coefficient sc. -0.0003
2
Concrete effective modulus Ec.eff [N/mm ] 12758.62 Equ. D.1
2
Concrete tensile strength fctm [N/mm ] 4.1
3
First moment of area W [m ] 0.0620
Modulus ratio e = Es / Ec.eff 15.6757 Equ. D.4
2
Reinforcement area As [cm /m] 54.54
0.009783 Equ. D.5
x/d 0.421299 Equ. D.6
Depth of compression zone x 0.235
Calculation of deflection due to permament loads and shrinkage
Uncracked
4
Second moment of area I' [m ] 0.018915 Equ. D.2
Curvature due to permanent loads k' 0.001256 Equ. D.3
Curvature due to shrinkage k'sh 0.000342 Equ. D.12
Cracked
4
Second moment of area I'' [m ] 0.013218 Equ. D.7
Curvature due to permanent loads k'' 0.001798 Equ. D.8
Curvature due to shrinkage k''sh 0.000626 Equ. D.13
Interpolation
Coefficient for high bond re-bars 1 1
Coefficient for long-time loading 2 0.5
Interpolation coefficient 0.6484
Total curvature due to permanent loads k 0.001607 Equ. D.9

Total curvature due to shrinkage ksh 0.000526 Equ. D.14
Curvature due to permanent load + shrinkage ktot 0.002134 Equ. D.15
Deflection due to permantent load + shrinkage [mm] 22.9 Equ. D.16
Calculation of deflection due to applied live load
4
Effective second moment of area Ie [m ] 0.013494 Equ. D.17
Curvature k 0.001085 Equ. D.18
Parameter (single load at y = 2.875 m) 0.077860 Equ. D.19
Deflection due to live load [mm] 7.5
Total deflection (only concrete bridge deck) total [mm] 30.4

Fig. D.29. Calculation of deflection, alternative design proposal 1
D-75
Alternative design proposal 2: Calculation of deflections

Geometry
Width [m] b [m] 1
Depth [m] h [m] 0.47
Effective depth [m] d [m] 0.398
Internal forces
Bending moment due to permament load M [kNm/m] 286.1
Bending moment due to live load M [kNm/m] 541.9
Cracking moment Mcr [kNm/m] 150.95 Equ. D.11
Properties
Reinforcement elastic modulus Es [N/mm2] 200000
Concrete elastic modulus Ecm [N/mm2] 37000
Creep coefficient 1.9
Shrinkage coefficient sc. -0.0003
Concrete effective modulus Ec.eff [N/mm2] 12758.62 Equ. D.1
Concrete tensile strength fctm [N/mm2] 4.1
First moment of area W [m3] 0.036817
Modulus ratio e = Es / Ec.eff 15.6757 Equ. D.4
2
Reinforcement area As [cm /m] 70.12
0.0176 Equ. D.5
x/d 0.5167 Equ. D.6
Depth of compression zone x 0.206
Calculation of deflection due to permament loads and shrinkage
Uncracked
Second moment of area I' [m4] 0.008652 Equ. D.2
Curvature due to permanent loads k' 0.002592 Equ. D.3
Curvature due to shrinkage k'sh 0.000621 Equ. D.12
Cracked
Second moment of area I'' [m4] 0.006966 Equ. D.7
Curvature due to permanent loads k'' 0.003219 Equ. D.8
Curvature due to shirnkage k''sh 0.000911 Equ. D.13
Interpolation
Coefficient for high bond re-bars 1 1
Coefficient for long-time loading 2 0.5
Interpolation coefficient 0.8608
Total curvature due to permanent loads k 0.003132 Equ. D.9

Total curvature due to shrinkage ksh 0.000870 Equ. D.14
Curvature due to permanent load + shrinkage ktot 0.004002 Equ. D.15
Deflection due to permantent load + shrinkage [mm] 42.9 Equ. D.16
Calculation of deflection due to applied live load
Effective second moment of area Ie [m4] 0.006976 Equ. D.17

Curvature k 0.002099 Equ. D.18
Parameter (single load at y = 2.875 m) 0.077860 Equ. D.19
Deflection [mm] 14.5
Total deflection (only concrete bridge deck) [mm] 57.4

Fig D.30. Calculation of deflection, alternative design proposal 2
D-76
D.5.3.4 Summary of reinforcement

Neglecting end-cross girder
Alternative design proposal 1

Top: 14, s = 10 cm

Top: 10, s = 15 cm
Edge beam
Top: 12, s = 15 cm
Footpath:
Shear stirrups: 12, s = 15 cm (four-shear)
Alternative design proposal 2

Top: 14, s = 10 cm
Bottom: 25, s = 14 cm (two layers)

Top: 10, s = 15 cm
Edge beam
Top: 12, s = 15 cm
Footpath:
Shear stirrups: 12, s = 15 cm (four-shear)
D-77
Summary of reinforcement: Bridge deck
Geometry [m]
The calculation of the
Bridge length 100.00
steel weight is based on
Width on the top 11.85
the geometric properties
Width on the bottom 10.59
listed in Figure D.31.
Width of edge beam 1.30
Length of cantilever 0.60
Length of stirrup 2.50
Length of cantilever strirrups 1.00
Fig. D.31. Geometric properties for calculation of steel weight
Alternative design proposal 1: h = 0.61 m at mid-span

Nom. mass s Number Length Weight
[mm] [kg/m] [cm] [m] [kg]
Top 14 1.21 10.00 1000.00 11850.00 14339
Bottom 25 3.85 9.00 1111.11 11766.67 45302

Top 10 0.62 15.00 74.33 7433.33 4586
Bottom 12 0.89 10.00 92.90 9290.00 8250
Edge beam
Top 12 0.89 15.00 17.33 1733.33 1539
Bottom 12 0.89 10.00 26.00 2600.00 2309
Shear stirrups 12 0.89 15.00 2666.67 6666.67 5920

1)
of re-bars and unconsidered reinforcement + 15% 1) 95527
Fig. D.32. Reinforcement steel weight of design proposal 1
Alternative design proposal 1: h = 0.47 m at mid-span

Nom. mass s Number Length Weight
[mm] [kg/m] [cm] [m] [kg]
Top 14 1.21 10.00 1000.00 11850.00 14339
Bottom 25 3.85 14.00 1428.57 15128.57 58245
2 layers
Top 10 0.62 15.00 74.33 7433.33 4586
Bottom 14 1.21 10.00 92.90 9290.00 11241
Edge beam
Top 12 0.89 15.00 17.33 1733.33 1539
Bottom 14 1.21 10.00 26.00 2600.00 3146
Shear stirrups 12 0.89 15.00 2666.67 6666.67 5920

1)
of re-bars and unconsidered reinforcement + 15% 1) 114814
Fig. D.33. Reinforcement steel weight of design proposal 2
D-78
Diploma Thesis Brunn & Schanack Annex D: Assessment of the end cross girder
D.6 End cross girder
D.6.1.1 Bending and longitudinal force
The two critical points to be checked for are the middle of the end cross girder
and the end above the bearings.
1. Middle of end cross girder

x=0m y = 5.075 m
Internal forces Geometry
z p [mm] 103
h [mm] 780
mSd,y [kNm/m] 167.8
d [mm] 677
nSd,y [kN/m] -2650.6
z s 1 [mm] 287
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,y nSd,y zs1

msds = 167.8 kNm/m (-2650.6 kN/m) 0.287 m
msds = 928.52 kNm/m
m Sds
Sds =
b d2 fcd
928.52 kNm m SCHNEIDER [29],
Sds = = 0.06 page 5.126
(0.677m)2 33.3 N mm 2
Sds,lim = 0.206 (for concrete class > C40/50)
lim = 0.35
= 0.96 z = d = 677 mm 0.96 = 649.9 mm

= 0.35 x = d = 677 mm 0.35 = 236.95 mm
s1 = 0.02
p = pm + s1
pm = 0.0034 (corresponds to Pm,t in Annex B: Preliminary design) SCHNEIDER [29]
page 5.128 / Tafel 1a
p = 0.0034 + 0.02 = 0.0234 > 0.01
fpd = 0.9 fpk / s

fpd = 0.9 1229.86 N/mm2 / 1.15 = 962.5 N/mm2
1 mSds
As = + nSd
fpd z
1 928.51kNm m
As = + ( 2650 .6kN / m) = 12.7cm2 / m < 0 ENV 1992-1-1: 1991,
962.5 N mm 2 649.9mm Section 4.2.3.3.3 (6)
D-79
2. End of end cross girder above bearings
X=0m
y=0m
Internal forces Geometry
z p [mm] 625.4
mSd,y [kNm/m] -1062.1 h [mm] 1027.3
nSd,y [kN/m] -7046.2 d [mm] 401.9
z s 1 [mm] -111.75
fcd = fck / s = 50 N/mm2 / 1.15 = 33.3 N/mm2
msds = mSd,y nSd,y zs1

msds = -1026.1 kNm/m (-7046.2 kN/m) (-0.11175) m
msds = -1813.5 kNm/m
msds = 1813.5 kNm/m ! For upper reinforcement !
m Sds SCHNEIDER [29],

Sds =
b d2 fcd page 5.126
1813 .5 kNm m
Sds = = 0.337
(0.4019m)2 33.3 N mm 2
Sds,lim = 0.206 (for concrete class > C40/50)
= 0.86 z = d = 401.9 mm 0.86 = 345.6 mm

= 0.35 x = d = 401.9 mm 0.35 = 140.7 mm
s1 = 0.0073
p = pm + s1
Schneider 12 p. 5.128
pm = 0.0033 Tafel 1a
p = 0.0033 + 0.0073 = 0.0106 0.01
fpd = 0.9 fpk / s

fpd = 0.9 1229.86 N/mm2 / 1.15 = 962.5 N/mm2
1 m
As = Sds + n Sd ENV 1992-1-1: 1991,
f pd z Section 4.2.3.3.3 (6)
1 1813.5 kNm m
As = 2
+ ( 7046.2kN / m) = 18cm2 / m < 0
962.5 N mm 345 . 6mm
D-80
D.6.1.2 Shear
The relevant shear force is taken at a distance 1360 mm from the inner
bearing which is the thinnest point of the end cross girder.
Internal forces
vSd,y [kN/m] 598.6
nSd,y [kN/m] -2573.6
Geometry
h [mm] 726.3
nom c [mm] 40
d [mm] 681.3
Elements not requiring design shear reinforcement ENV 1992-1-1: 1991

Section 4.3.2.3
VSd VRd1
VRd1 = [Rd k (1.2 + 40 1) + 0.15 cp] bw d
1 = As1 / (bw d) 0.02
As1 = 5.24 cm2/m minimum reinforcement

d = 0.6813 m
1 = (5.24 cm2/m) / (1 m 0.6813 m) = 0.00077
cp = NSd / Ac = 2573.6 kN/m / ( 726.3 mm 1m) = 3.54 N/mm2/m (compression

positive)
k = 1.6 d = 1.6 0.6813 = 0.9187
VRd1=[0.48N/mm20.9187(1.2 + 400.00077) + 0.153.54N/mm2]/m1m0.6813m

VRd1 = 731.5 kN/m
VSd = 598.6 kN/m < VRd1 = 731.5 kN/m
D-81
D.6.1.3 Punching at bearings

ENV 1992-1-1: 1991
Internal force (maximal bearing forces, on the safe side) Section 4.3.4
NSd [kN] 16051.2
Geometry of bearing plate Geometry of end cross girder

h [mm] 726.3
l [mm] 1436
nom c [mm] 40
w [mm] 1040
d [mm] 681.3
Vsd
v Sd =
u
VSd total design shear force; VSd = 16051.2 kN
ENV 1992-1-1: 1991
corner column, = 1.5
Figure 4.21 ()
u critical perimeter,
u = w + l + /2d1.5
u = 1040 mm + 1436 mm + /2 681.3 mm 1.5 = 4093.06 mm
16051.2kN 1.5
v Sd = = 5882.34 N mm
4093.06mm

vRd1 = Rd k (1.2 + 40 1) d
Rd basic design shear strength ENV 1992-1-1: 1991

Rd = 0.48 N/mm2 Table 4.8
k k = 1.6 d = 1.6 0.6813 = 0.9137

cpo
1 = 1x 1y + 0.015
f yd
cpo concrete stress due to initial prestress;
cpo = 5.1 N/mm2
fyd design yield stress of the reinforcement;
fyd = fy / 1.15 = 435 N/mm2
Asx = 5.24 cm2/m minimal reinforcement
1x = (5.24 cm2/m) / (1 m 0.6813 m) = 0.00077
1y = 1x = 0.00077
1 = 0.00077 + 5.1 N/mm2 / 435 N/mm2 = 0.0162 > 0.0125
vRd1 = 0.48 N/mm2 0.9137 (1.2 + 40 0.0125) 681.3 mm = 507.9 N/mm

vSd = 5882.34 N/mm > vRd1 = 507.9 N/mm
Since, vSd > vRd1 reinforcement is required so that vSd vRd3.
vSd = vRd3 = vRd1 + ASW fyd sin / u

= 90
fyd = 435 N/mm2
u = 4093.06 mm
(v v Rd1 ) u (5882.34N / mm 537.8N / mm ) 4093.06mm
ASW = Sd = = 502.89
f yd sin 435N / mm 2 1
Chosen reinforcement: 85 x two-shear stirrups 20, A = 533.8 cm2
D-82
For classification of assessment conditions category C is assumed. ENV 1992-2: 1996

4.4.0.3 Tab. 4.118
D.6.2.1 Limitation of stress
1. At time of prestressing
Concrete compressive stress should be limited to 0.6 fc(t). Prestressing will

ENV 1992-1-1: 1991
be performed at the time when the concrete has its full compressive strength.
Section 4.4.1.1 (102)
0.6 fc(t) = 0.6 fck = 0.6 50 N/mm2 = 30 N/mm2
If the compressive stress in concrete does not exceed 0.45 fc(t), creep non-
linearity does not need to be considered.
0.45 fc(t) = 0.45 fck = 0.45 50 N/mm2 = 22.5 N/mm2

The maximum compression stress in the concrete at time of prestressing is:
|c| = 17.8 N/mm2 < 22.5 N/mm2
2. Non-frequent combination of actions:
The concrete compressive stress should be limited to: ENV 1992-2: 1991
2
0.6 fck = 0.6 50 N/mm = 30 N/mm 2 Section 4.4.1.1 (103)
The maximum compression stress in the concrete is:
|c| = 28.4 N/mm2 < 30 N/mm2
The tensile stress in the reinforcement steel should be limited to:

ENV 1992-2: 1991
0.8 fyk = 0.8 500N/mm2 = 400 N/mm2 Section 4.4.1.1 (105)
The maximum tensile force in the punch reinforcement is:
v Sd,ser u 3921.56N / mm 4039.6mm
S = = = 296.8N / mm 2
AS 533.8cm 2
S = 296.8 N/mm2 < 0.8 fyk = 400 N/mm2
3. Quasi-permanent combination of actions:
After all losses, the stress in the transversal prestressing steel shall not exceed
the elastic range and should be limited to: ENV 1992-2: 1991
0.65 fpk = 0.65 1229.86 N/mm2 = 800 N/mm2 Section 4.4.1.1,
According to Annex B: Preliminary Design, the stress level is as follows: (104) & (106)
S = 675.6 N/mm2 < 800 N/mm2
D.6.2.2 Limit states of cracking
1. Non-frequent combination of actions:

ENV 1992-2: 1996
Where the stress in concrete is greater than -1 N/mm2, minimum reinforcement 4.4.2.2.2. (101)
is required.
The minimum compression stress in concrete is -2.4 N/mm2 (Figure C-23).
There is no minimum reinforcement required.
D-83
2. Quasi-permanent combination of actions ENV 1992-2: 1996

4.4.2.1 (106)
No tensile stress shall be allowed.
The minimum compression stress in the concrete is -1.3 N/mm2

(Figure C.23), and so no tensile stresses occur.
3. Frequent combination of actions
If the stresses in the concrete are limited to c fctm no extra minimum

reinforcement is required.
c = -2.4 N/mm2 fctm = 4.1 N/mm2
There is no minimum reinforcement required to avoid cracking in the cross

girder.
The necessary minimum reinforcement is then to be calculated according to
ENV 1992-2: 1996, 4.4.2.2.3.
In transverse direction:
0.8 k c k f ctm
S =
S
kc = 0.4
k = 0.664 (h = 0.78 m)
fctm = 4.1N/mm2
ENV 1992-2: 1996,
S = 260 N/mm2 ( 10 mm)
Table 4.120
0.8 0.4 0.664 4.1N / mm 2
S = 2
= 3.35 10 3
260N / mm
AS = S Act = 3.35 10-3 (0.78 m 4.1/(8.2+6.8))
AS = 7.1 cm2/m
10 s = 10 cm (AS = 7.85 cm2/m)
In longitudinal direction (absence of prestressing):

0.8 k c k f ctm
S =
S
kc = 0.4
k = 0.664 (h = 0.78 m)
fctm = 4.1N/mm2
ENV 1992-2: 1996,
S = 360 N/mm2 ( 10 mm)
Table 4.120
0.8 0.4 0.664 4.1N / mm 2
S = 2
= 2.42 10 3
360N / mm
AS = S Act = 2.42 10-3 (0.78 m 4.1/(8.2+6.8))
AS = 5.2 cm2/m
10 s = 15 cm (AS = 5.24 cm2/m)
D-84
Diploma Thesis Brunn & Schanack Annex D: Assessment of the handrails
D.7 Handrails
Loads
ENV 1991-2-1: 1995,
Horizontal load on top holm of handrail:
6.4
0.8 kN/m (Q = 1.5)
Geometry q = 0.8 kN/m
CHS 60.3 x 3.2
Handrail: CHS 60.3x3.2
h = 1100 mm
Intermediate holms: Wire rope, d = 10 mm
Posts: I 120
I120
Anchorage plate: 130x200x20 [mm]
Distance of posts: l=3m
Height of handrail: h = 1.1 m d = 10 mm
l w = 100 mm
Length of welds: lw = 100mm
Thickness of welds: aw = 4mm
Material
fy 275N / mm 2
FE 275 fyd = = = 250N / mm 2
M0 1 .1
Bending in handrail
Mz,Sd = 0.086 q Q l2 = 0.086 0.8 kN/m 1.5 (3 m)2 = 0.9288 kNm SCHNEIDER [29],
page 4.11
M z,Sd
1
W el,z f yd ENV 1993-1-1: 1992,
5.4.8.2
0.9288kNm
= 0.48 1
7.78cm 3 250N / mm 2
Bending in posts
My,Sd = l q Q h = 3 m 0.8 kN/m 1.5 1.1m = 3.96 kNm

M y,Sd ENV 1993-1-1: 1992,
1 5.4.8.2
W el,y f yd
3.96kNm
= 0.87 < 1
18.2cm 3 250N / mm 2
Normal stress in the welds
My,Sd = l q Q h = 3 m 0.8 kN/m 1.5 1.1m = 3.96 kNm

M y,Sd
1
W el,y f u / Mw ENV 1993-1-1: 1992,
3.96kNm Annex M (5)
= 0.95 < 1
13.33cm 390N / mm 2 / 1.25
3
D-85
Diploma Thesis Brunn & Schanack Annex D: Assessment of the drainage
D.8 Drainage
In the following, the dimensions of drainage pipes are determined based on

rainfall data for Dresden, Germany.
SCHNEIDER [29],
Rain discharge: Qr = sm rT (n ) A E
page 13.62
AE is the area which has to be drained. It is equal to the plane area of the bridge
deck:
A E = 100m 11.75m = 1175m 2 = 0.1175ha
rT(n) is the rail yield factor depending on the location (here: Dresden, Germany):
l
rT (n ) = r15 (1) = 102
s ha
sm is the discharge coefficient, which varies for different surfaces and

materials. Since the bridge deck surface is partly of plain concrete (footpaths)
and partly covered with ballast, the discharge coefficient of each surface has to
be considered with the respective area.
For railway tracks: rw = 0.4

For concrete surfaces: c = 0.9
Area of railway track: A rw = 100m 8.4m = 840m 2
Area of concrete surfaces: A c = 1175m 2 840m 2 = 335m 2
840m 2 0.4 + 335m 2 0.9
sm = = 0.54
1175m 2
l l
Qr = 0.54 102 0.1175ha = 6.47
s ha s
Selected drainage pipe: Nominal diameter: DN = 250 mm

The critical values for slope and flow velocity in respect of sedimentation are:
vcrit = 0.52 m/s SCHNEIDER [29],
Icrit = 1.63 o/oo page 13.70
For these critical values, the DN 250 drainage pipe with roughness coefficient SCHNEIDER [29],
kb = 0.75 has a discharge capacity of Q = 26.3 l/s > Qr = 6.47 l/s page 13.72
D-86
Diploma Thesis Brunn & Schanack Annex D: Assessment of the bearings
D.9 Bearings
D.9.1 Pot bearings

MAURER SHNE
At each support there are two pot bearings either TF-10, TGa-10 or TGe-10
GmbH & Co. KG [20]
with a centre distance of a = 0.87m. Each pair shares the vertical force FV and
the bending moment about the horizontal longitudinal axis MX. Additionally there
is a bending moment due to the eccentricity of the arch centre line to the neutral
axis of the two pot bearings, e = 0.062 m. The allowed vertical load for each pot
bearing is 10000kN.
FV M F e
maxFv = + x v
2 a a
16052kN 2082kNm 16052kN 0.062m

= + = 9275 .2kN <10000kN
2 0.87m 0.87m
The allowed horizontal loads are 10% of the vertical load limits 1000kN.
The maximum horizontal forces caused by the actions on the bridge do not
exceed this limit. Therefore no additional structural elements need to be applied
to bear the horizontal forces.
MAURER SHNE GmbH & Co. KG provides bearings with allowed
horizontal eccentricities of the vertical force by shifting up to 150 mm. The
present maximal horizontal movement is 102 mm and thus below that value.
D.9.2 Stilt bearing
The fixed bearing can consist of a plate with a cylindrical upper surface,
which is attached to the abutment. Its dimensions can be smaller than the
dimensions in the moveable bearing. In the following only the moveable bearing
is considered.
D.9.2.1 Vertical plate
Due to the bending moments in the arches and the end cross girder and the
torsional moment in the edge beams, the stilt bearing receives a not uniformly
distributed load. It is shown in Figure D.34.
Load case: wind loads acting on opposite side Load case: wind loads acting on the side,
of where the assessed bearing is located where the assessed bearing is located
M x = 720 kNm Mx = 935 kNm

Forces to be taken by the stilt bearing
Fv = -15238 kN Fv = -15087 kN
s s
1385 kN/m 1798 kN/m
Distribution of the bending moment

-1385 kN/m -1798 kN/m
Distribution of the vertical force

-9768 kN/m -9671 kN/m
Fv* s = -15238kN * 0.062m = -945 kNm Fv* s = -15087kN * 0.062m = -935 kNm
1817 kN/m 1798 kN/m Bending moment due to the
eccentricity of the vertical force
-1817 kN/m -1798 kN/m
- 1.56 m - - 1.56 m -
Resulting load on the stilt bearing
-10200 kN/m -9336 kN/m -9671 kN/m
Fig. D.34. Load distribution on the stilt bearing
D-87
Due to the lack of appropriate calculation models for Hertz pressing

between cylinder and plane with a linear distributed load, the maximum value
was assumed to form the value of an evenly distributed dummy load. The width
of the stilt bearing is b = 1.56 metres. The vertical force assumed for
assessment is:
F = 10200 kN/m1.56 m = 15912 kN
1. Hertz pressing
The vertical force is transferred through a contact surface between the

vertical plates above and below and the stilt plate. Therefore, it has to be
ensured that this contact surface does not exceed the area supplied by the
vertical plate. The consideration of the horizontal movements to be provided
has to be included into the calculation. neutral position
(adjustment
temperature 10)
Maximum horizontal displacement:
l3 l2 l1 l4
Temperature changes: a live load thermal thermal Creep, a e
expansion contraction shrinkage
l1 = 0.037 m (towards mid-span)
l2 = 0.04 m (towards abutments)
Horizontal movement of end of span due to live load after adjustment of the bearing
l3 = 0.015 m (towards abutments)
Creep and shrinkage
l4 = 0.065 m (towards mid-span)
Width of the contact surface:
Fr PETERSEN [24],
2 a = 2 1.52
b E page 1153
with F = 15912 kN
r = 0.4 m l arc t e
b = 1.56 m
E = 210000 N/mm2
r 1
15912kN 0.4m h
2 a = 2 1.52
1.56m 210000N / mm 2
= 13.4 mm
2a
Eccentricity caused by angular rotation of the bridge deck: Fig. D.35. Geometry of the vertical
e = r 1 = 0.4 m 3.49mrad = 1.4 mm (towards mid-span) plate
Considering the deviation between the real load distribution underneath the
contact surfaces and the calculated load according to Hertz an additional safety
bow length is added:
lsafe = 10 mm
Required bow length of the cylindrical surface:
larc l1/2+ l2/2+ l3/2+ l4/2+2 a+ 2 e + lsafe
18.5 mm + 20 mm + 7.5 mm + 32.5 mm + 13.4 mm+ 1.4 mm + 10 mm
103.3 mm
With the thickness of the vertical plate t = 110 mm the present bow length of the
cylindrical surface is:
larc = 110.1 mm > 103.3 mm
D-88
Furthermore the Hertz pressing has to be examined in the contact surface.

Allowed Hertz pressing:
r, d,H = 1.77 f u = 1.77 550N / mm 2 = 973.5N / mm 2 ; S 460 ML ENV 1337-4
Present Hertz pressing:

F E 15912kN 210000N / mm 2 PETERSEN [24],
d,H = 0.418 = 0.418 = 967.3N / mm 2
b r 1560mm 400mm page 1153
d,H = 967.3N / mm 2 < 973.5N / mm 2 = r,d,H
2. Stresses in vertical plate

lK = 2 r = 800 mm ENV 1993-1-1: 1992,
bt3
1560mm (110mm ) 3 5.5.1.1
Im in 12 = 12
imin = = = 31.75mm
A bt 1560mm 110mm
lK 800mm
i min 31.75mm
= = = = 0.36
1 E 210000 N / mm 2

fy 430N / mm 2
= 0.88 ENV 1993-1-1: 1992, Table 5.5.2
F 15912kN
S,d = = = 92.7N / mm 2
b t 1560mm 110mm
fy 0.88 430N / mm 2
R,d = = = 344N / mm 2
M1 1 .1
S,d = 92.7 N/mm2 < 344 N/mm2 = R,d
3. Stresses in the concrete below the load distribution plate
For the abutments concrete class C30/37 is used. The plate below the stilt
bearing has a thickness of 20 mm. The pressure is distributed like in
Figure D.34 with a maximum of -10200 kN/m. The allowed maximal
compression stresses in the concrete is:
f ck 30N / mm 2
f cd = = = 20N / mm 2
c 1 .5
The necessary width wc of the load distribution plate is:
10200kN / m
wc = = 510mm
20N / mm 2
Stiffening plates have to be arranged to support the load distribution into

this horizontal plate.
MAURER SHNE
GmbH & Co. KG [19]
D.9.2.2 Middle compression support
The additional bearing in the middle of the end cross girder receives only
vertical loads. It is not subjected to horizontal forces, bending moments or
horizontal deflections. The maximum static load is -1337 kN. The dynamic
D-89
Diploma Thesis Brunn & Schanack Annex D: Deformations
increase is 20 % according to a proposal of P. TVEIT. So the maximum vertical

load the bearing is to assess for is:
maxFv = -1337 kN 1.2 = 1604.4 kN
MAURER SHNE GmbH & Co. KG an elastomeric bearing type 1/2 1800 MAURER SHNE
was chosen. The allowed maximal vertical load is 1800 kN and therefore above GmbH & Co. KG [19]
the necessary value.
The concrete pressure below the bearing is:
max Fv 1604.4kN
c,d = = = 13.37N / mm 2
A bearing 0.3m 0.4m
c,d = 13.37N/mm2 < 20 N/mm2 = fcd
D.10 Deformations
The relevant load combinations for calculating deformations are to be found
in Annex A, Figure A.8. The checks for the limits of deflection and vibration shall ENV 1991-3: 1995,
be made with only one track loaded. 6.8.1 (6)P
Vertical acceleration of the deck

This check is only required for design speeds v > 220 km/h, or when the ENV 1991-3: 1995
natural frequency of the bridge with a span of 100 m is above 3.02 Hz or below G.3.1.2.1
1.54 Hz. The data for the bridge calculated in this work are as follows:
Design speed: 160 km/h
Natural frequency: 2.34 Hz
The check is therefore not required.
Deck twist
For a design speed of 160 km/h, the maximum twist over a length of 3 m
shall not exceed t = 3.0 mm.
Actual value of maximum twist: t = 1.4 mm /3m < 3.0 mm/3m ENV 1991-3: 1995
G.3.1.2.2
The check is fulfilled.
Rotation at the end of the deck

For a double track bridge, the maximum transition between the deck and
the embankment shall not exceed = 3.5 10-3 radians.
Actual value of maximum transition: = 3.48 10-3 radians < 3.5 10-3 radians
The check is fulfilled. ENV 1991-3: 1995
G.3.1.2.3
Horizontal deflection of the deck
For a design speed of 160 km/h, the radius of curvature shall not exceed
the value of 9500 m.
Actual value of maximum horizontal deflection: h = 19.6 mm
Thus, the radius of curvature measures
L2 (100m) 2
R= = = 63776 > 9500
8 h 8 19.6mm ENV 1991-3: 1995
G.3.1.2.4
The check is fulfilled.
D-90
Vertical deflection
Vertical deflection of main design ENV 1991-3: 1995

G.3.1.3
For a comfort level very good of a 100 m railway bridge with one span and
a design speed of 160 km/h, the maximum vertical deflection shall not exceed
L/300 = 100 / 300 = 333 mm.
Actual value of maximum vertical deflection due to dead load and live load:
v = 142.4mm < 333mm (Location: edge beam)
The deflection due to dead loads is usually compensated by a respective

camber. Therefore, the relevant value for deflection shall be the one due to live
load only:
LL = 66.8mm
Vertical deflection of alternative design proposals
Alternative design 1:
v = 166.8mm < 333mm
v = 185.1mm < 333mm
Calculation of total deflections of alternative design proposals
It is necessary to compare the deflection of the prestressed concrete bridge

deck with the respective deflections of the alternative design proposals. Since
the bridge is loaded asymmetrically, the edge beams experience different
deflection. The higher dead load of the alternative designs lead to higher edge
beam deflections, as well as to higher deflections of the concrete slab between
the edge beams.
Alternative design 1 (due to DL and LL):

Maximum deflection of one edge beam: eb.1 = 156.7 mm
Maximum deflection of second edge beam: eb.2 = 117.4 mm
Relative deflection of concrete slab between edge beams:
Due to dead load: slab.DL = 21.5 mm

Due to live load: slab,LL = 7.5 mm
Total: slab = 29 mm
Alternative design 2 (due to DL and LL):

Maximum deflection of the first edge beam: eb.1 = 146.1 mm
Maximum deflection of the second edge beam: eb.2 = 104.8 mm
Relative deflection of concrete slab between edge beams:
Due to dead load: slab.DL = 42.9 mm
D-91
Due to live load: slab,LL = 14.5 mm

Total: slab = 57.4 mm
Location of maximum relative deflection of concrete slab
The maximum deflection of the main design with transverse prestressing

occurs at one edge beam. It is very likely that the high relative deflection of the
unprestressed bridge deck shifts the location of maximum total displacement to
somewhere between the edge beams.
The following calculation is an approximation of the total deflection at the
location where the greatest relative displacement of the bridge slab occurs.
Location of maximum relative deflection due to

traffic load
The centre of one track has a distance of

P
a = 2.850 m from the support, which is the x
arch plane. The span of the deck is 10.15 m

a b
and the maximum deflection is expected at a L
distance x from the right support (Figure
D.36). Fig. D.36. Nomenclature: Determination
of location with maximum deflection
If = a / L, and
= x /L ,
then the deflection line of a simple supported
beam subjected to an eccentric single load P
is given to:
( )
= 1 2 2 P L3 (6 EI)
Schneider [29],
The derivation of the deflection line with = 2.875 / 10.15 = 0.28325 leads to 13th edition, page 4.3
the location for the maximum deflection:
= 0.554 ,
hence, the x = 0.554 L = 5.623 m
Location of maximum relative deflection due to dead load
The maximum deflection due to dead load certainly occurs at the deck mid-
span: x = 5.075 m
Location of maximum relative deflection due to LL and DL
The deflection due to dead load is about three times higher than to live load.
The maximum deflection lies therefore closer at mid-span and the location is
assumed to be at x = 5.26 m.
In the following, the deflections of the deck are calculated at x = 5.26 m, but
only the components due to the edge beam displacement:
D-92
To obtain an approximate total deflection at the location, where the relative

maximum deflection of the slab is largest, the following is required:
With the help of both maximum edge beam displacements, the theoretical
displacement at the location x = 5.26 m neglecting the deflection of the slab can
be calculated:
156.7mm 117.4mm
eb.5.623 = (5.26m) + 117.4mm = 137.8mm
10.15m
146.1mm 104.8mm
eb.5.623 = (5.62m) + 104.8mm = 127.7mm
10.15m
These values are superimposed with the in Annex D, Section D.5.3.3

calculated relative deflections of the concrete slab giving the total displacement
at that location:
5.623 = eb.5.623 + slab = 137.8 mm + 29 mm = 166.8 mm
5.623 = eb.5.623 + slab = 127.7 mm + 57.4 mm = 185.1 mm
Both values are larger than the respective maximum edge beam
displacements and constitute therefore the total maximum deflection of the
bridge.
D-93
Annex E
Investigations on the temporary lower chord
E.1 General
The maximum deflection of the formwork should be limited to:

L + 40 ENV 1992-2: 1996,
with L = span [m].
2000 4.4.3.1 (105)
This value can be reduced to L/300 if cracking during casting is limited.
Both values were considered as limits.
Construction states are transient situation and therefore to be calculated in
ultimate limit state applying the appendant partial safety factors.
Load assumptions
ENV 1991-2-1: 1995,
Self-weight: Fresh concrete and reinforcement: 26 kN/m3 4.2 tab. 4.1
(G = 1.35) Plywood formwork sheet d = 18mm: 0.23 kN/m3 DOKA [9]
Timber formwork beams: 0.051 kN/m
Timber for transverse camber: 0.18 kN/m TEICH & WENDELIN [38]
Transverse steel beams (IPEa 550): 0.921 kN/m
Longitudinal steel beams (HEB 220): 0.715 kN/m EURONORM 19-62
Live load: On the cantilevers of the temporary lower chord due to casting
(Q = 1.5) and prestressing equipment and workers operating it:
0.5 kN/m2
Between casting the bridge deck and transverse prestressing
there is no access to the bridge deck allowed, so no further live
load has to be considered.
Section A.3.1
Lateral wind on formwork: 1.43 kN/m
E.2 Formwork sheet
Plywood formwork sheets with a thickness of 18 mm are used. The

structural system is shown in Figure E.1.
0.75
5.3 21.5 20.8
13.1 15.1
[kN/m 2 ]
[cm] CL
100 45 35 35 35 35 39.5 50 50 50 50 50 50 50
100 150 124.5 150 150
Fig. E.1. Structural system of formwork sheets, load is increased by partial safety
factor
E-1
Diploma Thesis Brunn & Schanack Annex E: Investigations on the temporary lower chord
1. Underneath the cantilevers (as a simply supported beam for no live load on
left side):
L = 0.45 m
0.45m + 40
Allowed deflection: = = 20.2 mm
2000
= 0.45m/300 = 1.5 mm
Present deflection:
5 q l4 5 6.05kN / m 2 0.45m 4
(6.05 kN/m2) = = = 0.83 mm
384 EI 384 8000N / mm 2 486000mm 4 / m
0.83mm < 1.9 mm

2. Underneath the edge beams (continuous beam):
L = 0.35 m
0.35m + 40
2000
= 0.35m/300 = 1.2 mm
Present deflection:
1 q l4 1 21.5kN / m 2 0.35m 4
(21.5 kN/m2) = = = 0.2 mm
384 EI 384 8000N / mm 2 486000mm 4 / m
0.2 mm < 1.5 mm

3. Underneath the main bridge deck (continuous beam):
L = 0.5 m
0.5m + 40
2000
= 0.5m/300 = 1.66 mm
Present deflection:
1 q l4 1 15.1kN / m 2 0.5m 4
(15.1 kN/m2) = = = 0.6 mm
384 EI 384 8000N / mm 2 486000mm 4 / m
0.6 mm < 1.75 mm

For the assessment of the plywood only allowed deflections are given by
the manufacturer. They are all larger than the tolerances given by Eurocode.
E-2
E.3 Timber formwork beams
The formwork beams are collocated as shown in Figure E.2.
7.601 kN/m
3.5 m
1.0 m 1.0 m
Fig. E.2. Schematic illustration of the overlapping timber formwork beams
The bridge deck is cast starting at a certain point and proceeding forward.
This leads to partial loading of the continuous beam constituted by the formwork
beams. Decisive bending moments are obtained as shown in Figure E.3
Load [kN/m]
Bending moment My [kNm]
Fig. E.3. Maximum bending moments during casting
L = 3.5 m
3.5m + 40
2000
= 3.5m/300 = 11.7 mm
Present deflection (SOFISTIK):
(7.601 kN/m) = 4.7 mm < 10 mm
Sagging bending moment:
Allowed bending moment: M = 5.0 kNm
Present bending moment (SOFiSTiK):
M(7.601 kN/m) = 4.7 kNm < 5 kNm DOKA [8]
Hogging at the supports (double cross section):
Allowed bending moment: M = 10.0 kNm
Present bending moment (SOFiSTiK):
|M(7.601 kN/m)| = 9.92 kNm < 10 kNm
The formwork beams first loaded are simply supported beams. In these
spans, double amount of formwork beams has to be applied.
E-3
E.4 Transverse steel beams
The transverse steel beams receive decisive loads during prestressing of

the transverse tendons. Therefore they are designed only for this construction
state. The contribution of the reinforced concrete cross section to the structural
behaviour is not considered, so the transverse steel beams have to bear all
loads.
The loads and the structural system are shown in Figure E.4.
1.45 m 10.59 m 1.45 m
2.625 kN/m 2.625 kN/m

26.6 26.6
26.6 19.9 26.7 26.7 26.7 26.7 26.7 26.7 19.9 26.6 [kN]
7.3 17.6 23.8 23.9 26.7 26.7 26.7 26.7 26.7 26.7 23.9 23.8 17.6 7.3
5.0
1.67 m 10.15 m 1.67 m
0.395 m 0.395 m
1.0 m 0.45 4 x 0.35 14 x 0.5 m 4 x 0.35 0.45 1.0 m
Fig. E.4. Loads and constraints of the transverse steel beam
To determine the axial forces acting in the transverse steel beam the wind
loads were applied to a 3-D-model in SOFiSTiK. The results are shown in
Figure E.5.
Axial force [kN] in the

wind bracing due to
wind from the right
(bottom in this
picture)
Axial force [kN] in the

wind bracing due to
wind from the left (top
in this picture)
Fixed bearing is in
top left corner in
the picture
Fig. E.5. Axial forces in the wind bracing of the temporary lower chord
E-4
As shown in Figure E.5 the axial forces are far from evenly distributed along
the span. The asymmetry is caused by the different bearing conditions. It is
important to connect the fixed bearing with a diagonal wind bracing member.
This reduces the axial forces by about 30% in the members of the first field.
Since the first two transverse steel beams will consist of stronger profiles, the
maximum and minimum forces are taken from the third field from the bearing.
There are two load cases to differentiate:
1. All loads:
Maximum sagging bending moment: Mf,d = 688.8 kNm
maxFx,d = -79.2 kN
Maximum hogging bending moment: Ms,d = -8.8 kNm
maxFx,d = -110.9 kN
maxVz,d = 294.4 kN
2. All loads except live load on cantilevers:

maxFx,d = -79.2 kN
maxFx,d = -110.9 kN
maxVz,d = 294.4 kN
It is not desired that the cross section of the transverse steel beams
develop their plastic moment resistance. Therefore, in the following design
ENV 1993-1-1: 1992,
checks it is treated as a class 3 cross section. 5.5.2 (8)
The transverse beams are restrained at the top flanges by wooden cams
nailed to the timber formwork beams, which are spaced at 50 cm c/c. This can
be considered as fully restrained. Therefore, the beams do not need to be
checked for lateral-torsional buckling.
My,Sd = Mf,d = 688.8 kNm

NSd = maxFx,d = -79.2 kN
VSd = maxVz,d = 294.4 kN
Cross section: IPEa 550
A = 117 cm2
Wel,y = 2190 cm3
fy 355N / mm 2
fyd = = = 322.7N / mm 2
M0 1 .1 ENV 1993-1-1: 1992,
5.4.8.2
Bending and axial force at mid-span, absence of shear force

N Sd M y,Sd
+ 1
A f yd W el,y f yd
| 79.2kN | 688.8kNm
2 2
+ = 0.99 1
117cm 322.7N / mm 2190cm 3 322.7N / mm 2
E-5
Shear resistance at the support ENV 1993-1-1: 1992,

5.4.6
Av = A 2 b tf + (tw+2r) tf
= 117 cm2 2 21cm 1.57cm + (0.9cm + 2 2.4cm) 1.57cm
= 60 cm2
Vpl,Rd =
(
A v fy / 3 ) = 60cm (355N / mm
2 2
/ 3 ) = 1118.1kN
M0 1 .1
VSd = 294.4 kN < 1118.1 kN = Vpl,Rd
VSd 294.4kN
and = = 0.26
Vpl,Rd 1118.1kN
Since VSd does not exceed 50% of Vpl,Rd, no reduction needs to be made in ENV 1993-1-1: 1992,
the resistance moments. Therefore, the bending moment at mid-span is 5.4.7 (2)
decisive.
Deflections for the design of the camber
To find the shape of the bottom surface of the bridge deck, the formwork
receives a camber. This camber is influenced by the deflections due to the self-
weight of the concrete bridge deck and the railway equipment (ballast, sleepers
and rails). The deflections due to the fresh concrete are shown in Figure E.6.
The deflections due to self-weight after transverse prestressing and removal of
the formwork are taken from Annex D, Section D.10.
[mm]
1.67 m 10.15 m 1.67 m
0.395 m 0.395 m
1.0 m 0.45 4 x 0.35 14 x 0.5 m 4 x 0.35 0.45 1.0 m
Fig. E.6. Deflections of the transverse temporary steel beam due to fresh concrete
E.5 Wind bracing members
The relevant axial force is taken from Figure E.5.
Buckling resistance of compression members ENV1993-1-1: 1992,

5.5.1
Geometry: l = 4.88 m
Cross section: L 120 x 10 (DIN 1028)

A = 23.2 cm2
iy = iz = 3.67 cm
fy = 355 N/mm2 S 355
NSd = -160.5 kN
E-6
The criterion to fulfil is: NSd Nb,Rd
fy
Nb,Rd = A A
M1
depends on
l
= A = A
1 E
iy
fy
l = 4.88 m
iy = 3.67 cm
E = 210000 N/mm2
fy = 355 N/mm2 ; S355
1 4.88m
= 1 = 1.74
210000N / mm 2
3.67cm
355N / mm 2
= 0.2461 ; buckling curve c ENV 1993-1-1: 1992
A = 23.2 cm2 5.5.1 Table 5.5.2
M1 = 1.1
355N / mm 2
Nb,Rd = 0.2461 1 23.2cm 2 = 184.3kN
1. 1
|NSd| = 160.5 kN 184.3 kN = Nb,Rd
E-7
E.6 Longitudinal steel beams
For the assessment of the longitudinal beams of the temporary lower chord
several construction phases have to be examined to find the decisive internal
forces. Only reasonable wind loads are considered. For example no casting
would be performed during a storm.
E.6.1 Construction phase 1 Mounting the steel skeleton and

displacement
In this construction phase the steel skeleton is assembled and supported

only at the arch root point. The formwork is still not applied.
N [kN]
M y [kNm]
N [kN]
My [kNm]
N [kN]
Fig. E.7. Internal forces for construction state 1, full wind load and self-weight
As can be seen in Figure E.7 bending moments in the temporary lower

chord are small. Two hangers tend to relax while one is relaxed. This is not a
problem, because in this construction state no casting is performed.
E.6.2 Construction phase 2 Preparations for casting the

bridge deck
After the steel skeleton has been put in place, the first approx. 3.5 meters
from the ends of the bridge are cast. Then the formwork of the whole bridge is
put into place starting from both ends. After that the longitudinal and transverse
tendons and the reinforcement are laid out. At the point before partial
prestressing of the longitudinal tendons, internal forces are calculated and
shown in Figure E.8.
E-8
N [kN]
M y [kNm]
N [kN]
My [kNm]
N [kN]
Fig. E.8. Internal forces for construction state 2, full wind load and self-weight
E.6.3 Construction phase 3 - Casting of the concrete edge

beam
This phase of construction is performed in one day. It is assumed that there

is no heavy wind on this day so wind loads are reduced to 25%. Two
longitudinal tendons on each side are already prestressed with 650 kN each. It
is possible to start casting from mid-span or from both ends. Since internal
forces do not differ too much between the two methods, the decisive criterion is
avoiding relaxed hangers when their bottom ends are already in concrete. If
both methods show the same resistance against hanger relaxation, starting
from both ends should be prioritised because the ages of the concrete of end
cross girder and edge beam do not differ that much, in this case.
The two different methods were examined concerning hanger relaxation.
The result is shown in Figure E.9, which extends over two pages.
E-9
State before casting the longitudinal State before casting the longitudinal
egde beam, longitudinal tendons egde beam, longitudinal tendons
prestressed with 1300 kN prestressed with 1300 kN
Begin casting from mid-span, Begin casting from both ends of the
continuing simultaneously towards bridge, continuing simultaneously
both ends, one hanger relaxed towards the middle, one hanger relaxed
Progress 5.25 m, one hanger

relaxed Progress 7 m, one hanger relaxed
Progress 10.5 m, one hanger tends to

relax
relaxed, one tends to relax

Progress 14 m
relaxed
Progress 15.75 m
Progress 17.5 m

tends to relax Progress 21 m, one hanger tends to
relax

Progress 24.5 m, one hanger tends to
tends to relax
relax
Fig. E.9. Axial forces in hangers during casting of the longitudinal concrete edge beam
Continued on next page
E-10
Progress 26.25 m
Progress 28 m
Progress 29.75 m
Progress 31.5 m
Progress 33.25 m
Progress 35 m
Progress 36.75 m
Progress 38.5 m, one hanger relaxed
Progress 40.25 m Progress 42 m
Casting of edge beam finshed, no

Casting of edge beam finished, one
hanger with bottom end in concrete
hanger with bottom end in concrete
relaxed or tended to relax during the
relaxed, one tended to relax during
construction
construction
Fig. E.9. Axial forces in hangers during casting of the longitudinal concrete edge beam
The results in Figure E.9 show that starting from both ends leads to relaxed
hangers when their bottom ends are already cast. Therefore it was decided to
start at mid-span and then proceed simultaneously to both ends of the bridge.
Since every network arch bridge will have a different geometry, hanger
relaxation should be examined for every special project.
The partial longitudinal prestressing relieves the longitudinal steel beam of
the temporary lower chord in such a way that the internal forces during
E-11
construction and after finishing the concrete edge beam do not exceed the
values of the second construction phase (before partial prestressing of
longitudinal tendons). The maximum bending moment is still caused by the end
cross girder (73.4 kNm). The maximum bending moment apart from this goes
up to 35 kNm while the axial force is 1085 kN
E.6.4 Construction phase 4 - Casting the main part of the

bridge deck
This construction phase starts after another partial prestressing of the

longitudinal tendons. All of them are prestressed with 25 % of the design value.
From then on the concrete edge beam contributes to the bearing of all loads
which leads to small bending moments in the arch. The longitudinal beam of the
temporary lower chord is then disconnected from the arch root point, that it does
not receive axial forces from the arch any more. This means that the internal
forces occurring in this construction phase are not decisive for the assessment
of the longitudinal steel beam. Therefore, as in construction phase 3, avoiding
hanger relaxation should decide the sequence of casting.
The two possibilities mentioned for construction phase 3 do not give
satisfying results any more. Both lead to extensive hanger relaxation. Since the
bridge deck is not cast in one day, other sequences are possible. It was decided
on three construction segments and then examined in which order and direction
they should be cast. Figure E.10 shows the result.
Edge beams, End cross girder,
already cast already cast
Construction stage:
III I II I III
3.5 15.5 m 15.5 m 31 m 15.5 m 15.5 m 3.5

Fig. E.10. Casting sequence of bridge deck in construction phase 4
This construction method does not show any relaxed hangers at all. The
maximum internal forces in the temporary longitudinal steel beam are:
My = 135 kNm
Nx = 0 kN
Vz = 188 kN
E.6.5 Assessment of the longitudinal steel beam
1. Construction phase 2:
maxFx,d = 1392 kN
Maximum hogging bending moment: Ms,d = -44 kNm
maxFx,d = 1392 kN
maxVz,d = 106 kN
E-12
It is not desired that the cross section of the longitudinal steel beams
develop their plastic moment resistance. Therefore, in the following design
checks it is treated as a class 3 cross section.
My,Sd = Mf,d = 73.4 kNm

NSd = maxFx,d = 1392 kN
VSd = maxVz,d = 106 kN
Cross section: HEB 220

Anet = 83.3 cm2
Wel,y,net = 694 cm3
fy 355N / mm 2
fyd = = = 322.7N / mm 2
M0 1. 1
Bending and axial force at maximum sagging moment, absence of shear force ENV 1993-1-1: 1992,
5.4.8.2
N Sd M y,Sd
+ 1
A net f yd W el,y,net f yd
1392kN 73.4kNm
2 2
+ = 0.85 1
83.3cm 322.7N / mm 694cm 322.7N / mm 2
3
Shear resistance at the support ENV 1993-1-1: 1992,

5.4.6
Av = A 2 b tf + (tw+2r) tf
= 78.1 cm2 2 20cm 1.5cm + (0.9cm + 2 1.8cm) 1.5cm
= 24.85 cm2
Vpl,Rd =
(
A v fy / 3 ) = 24.85cm (355N / mm
2 2
/ 3 ) = 463.02kN
M0 1 .1
VSd = 106 kN < 463.02 kN = Vpl,Rd
VSd 106kN
and = = 0.23
Vpl,Rd 463.02kN
Since VSd does not exceed 50% of Vpl,Rd no reduction needs to be made in
ENV 1993-1-1: 1992,
the resistance moments. Therefore the bending moment at mid-span is 5.4.7 (2)
decisive.
2. Construction phase 4:
Maximum sagging bending moment: Mf,d = 135 kNm
maxFx,d = 0 kN
maxFx,d = 0 kN
maxVz,d = 188 kN
E-13
My,Sd = Mf,d = 135 kNm

NSd = maxFx,d = 0 kN
VSd = maxVz,d = 188 kN
Cross section: HEB 220

Anet = 83.3 cm2
Wel,y,net = 694 cm3
fy 355N / mm 2
fyd = = = 322.7N / mm 2
M0 1. 1
Shear resistance at the support
ENV 1993-1-1: 1992,
Av = A 2 b tf + (tw+2r) tf 5.4.6
= 78.1 cm2 2 20cm 1.5cm + (0.9cm + 2 1.8cm) 1.5cm
= 24.85 cm2
Vpl,Rd =
(
A v fy / 3 ) = 24.85cm (355N / mm
2 2
/ 3 ) = 463.02kN
M0 1 .1
VSd = 188 kN < 463.02 kN = Vpl,Rd
VSd 188kN
and = = 0.41
Vpl,Rd 463.02kN
Since VSd does not exceed 50% of Vpl,Rd no reduction needs to be made in
the resistance moments. Therefore the bending moment at mid-span is ENV 1993-1-1: 1992,
decisive. 5.4.7 (2)
Bending and axial force at maximum sagging moment, absence of shear force
ENV 1993-1-1: 1992,
5.4.8.2
N Sd M y,Sd
+ 1
A net f yd W el,y,net f yd
0kN 135kNm
2 2
+ = 0 .6 1
83.3cm 322.7N / mm 694cm 322.7N / mm 2
3
E-14
Annex F
Data from the optimisation process
F.1 Comparison of influence lines
As mentioned in Section 6.6.2 a comparison between the kviksound network arch TVEIT [39],
page 73, and a hanger arrangement according to the authors` proposal was drawn. The influence
lines for bending moments in the arch and axial force in the arch and hangers served for this
purpose. Many influence lines of the kviksound bridge were calculated by A. JAY in 1998, [49].
The same geometry, cross-sections and material properties as in her work were used for the
authors` calculations. As a difference, a hanger arrangement with equal cross angles between the
arch and the hangers was chosen. The cross angle, as defined in Section 6.6.3, was 28 which
complies with the cross angle of the hangers and the axis of symmetry in the hanger arrangement
of the kviksound network arch. It should be noted that a different cross-angle may lead to better
results for the proposed arrangement.
Knowing that the radial hanger arrangement is not appropriate for the arch ends where the
clamping causes a disturbance range, the authors shifted the first few upper hanger nodes slightly
along the arch.
In the following, the relevant data of the calculation conditions are listed:
All hangers: Circular bars, diameter = 40 mm, E = 210,000 N/mm2, = 78.5 kN/m3
Arch profile: UC 356x406x393, E = 210,000 N/mm2, = 78.5 kN/m3
Concrete tie: E = 30,000 N/mm2, = 30.4 kN/m3
Span: 135 m
Arch rise: 20.25 m
Distance between arch planes: 7.55 m
The concrete tie was modelled with a thickness of 520 mm in the plane of the arch and
200 mm in the middle with a linear transition between them. The total width is 7.55 meters.
The influence lines were calculated for loads on the tie directly in the arch plane. Only one
arch plane and half the tie were considered for analysis. Figures F.1 to F.5 show the
comparison.
Additionally two load cases were calculated. The comparison can be seen in figures F.6 to F.9.
The applied loads can be seen in those figures as well.
A composition of some relevant values is shown in Figure F.10.
F-1
Diploma Thesis Brunn & Schanack Annex F: Comparison of influence lines
-1,73
-1,69
-1,75
Test of proposal Brunn & Schanack

-1,74
kviksound network arch
Fig. F.1. Comparison of influence lines for compression in the arch, (thin line - kviksound network arch / thick line - test of
an arrangement according to a proposal by Brunn & Schanack)
F-2
0.14
0.21
0.22
0.23
0.24

0.3
0.37
Fig. F.2. Comparison of influence lines for the force in the hangers I (thin line - kviksound network arch / thick line - test of
F-3
0.19
0.22
0.23
0.22
0.26

0.39
Fig. F.3. Comparison of influence lines for the force in the hangers II (thin line - kviksound network arch / thick line - test of
F-4
-0.21
0.11
-0.15
0.15
0.18
0.17
0.18

0.22
Fig. F.4. Comparison of influence lines for the bending moments in the arch I (thin line - kviksound network arch / thick
line - test of an arrangement according to a proposal by Brunn & Schanack)
F-5
-0.19
0.13
0.18
0.18
0.16

0.28
Fig. F.5. Comparison of influence lines for the bending moments in the arch II (thin line - kviksound network arch / thick
line - test of an arrangement according to a proposal by Brunn & Schanack)
F-6
Fig. F.6. Maximum load on the kviksound network arch in the ultimate limit state, the wheel loads are in the middle of the
span
F-7
Deflection of lower chord [mm]

Bending moments in the arch [kNm]
Axial forces in hangers [kN]
Live load: 9.3 kN/m*1.3

kN]
ch [
e ar
in th
Dead load: 59.7 kN/m

ces
l fo r
Axia
282 kN
282 kN
282 kN
Fig. F.7. Maximum load on the bridge with a test hanger arrangement according to the proposal of Brunn & Schanack in the
ultimate limit state. The wheel loads are in the middle of the span
F-8
Fig. F.8. Forces due to live loads on the left 54% of the kviksound network arch in the ultimate limit state. Half the weight
of asphalt on the whole span is assumed. Three hangers relax
F-9
Deflection of lower chord [mm]

Bending moments in the arch [kNm]
Axial forces in hangers [kN]

kN]
ch [
e ar
in th
42.1 kN/m
ce s
l for
Axia
3 x 231.3 kN/m
51.3 kN/m
Fig. F.9. Forces due to live loads on the left 54% of the bridge with a test of the hanger arrangement proposed by Brunn &
Schanack in the ultimate limit state. Half the weight of asphalt on the whole span is assumed. No hangers relax.
F-10
As can be seen in figures F.1 to F.9, the proposed hanger arrangement gives better results for
all compared properties. The following table will help to estimate how much the difference is. The
corresponding values of the influence lines and the maximum values from both load cases are
shown.
Test of
proposed "Test" /
kviksound hanger " kviksound"
bridge arrangement [%] average
Axial force in arch
(Figure F.1) -1.84 -1.74 94.57
-1.75 -1.75 100.00
-1.68 -1.69 100.60
-1.78 -1.73 97.19 98.09
Axial force in hange rs
(Figure F.2) 0.54 0.37 68.52
0.39 0.3 76.92
0.37 0.24 64.86
0.32 0.23 71.88
0.26 0.22 84.62
0.23 0.21 91.30
(Figure F.3) 0.43 0.39 90.70
0.35 0.26 74.29
0.34 0.22 64.71
0.28 0.23 82.14
0.23 0.22 95.65
0.22 0.19 86.36 79.33
Be nding m om e nts in the arch
(Figure F.4) 0.21 0.22 104.76
0.23 0.18 78.26
0.29 0.17 58.62
0.26 0.18 69.23
0.23 0.15 65.22
-0.21 -0.21 100.00
(Figure F.5) 0.28 0.28 100.00
0.19 0.16 84.21
0.28 0.18 64.29
0.23 0.18 78.26
-0.22 -0.19 86.36 80.84
Load on w hole s pan
(Figures F.6 and F.7)
maximum axial f orce in arch [MN] -10.3 -9.7 94.17
maximum hanger f orce [kN] 414.6 365 88.04
maximum bending moment [kNm] 170 125 73.53 85.25
maximum def lection [mm] 275 242 88.00
Load in 54% of s pan
(Figures F.8 and F.9)
maximum axial f orce in arch [MN] -9.2 -7.5 81.52
maximum hanger f orce [kN] 412 318 77.18
maximum bending moment [kNm] 220 126 57.27 71.99
maximum def lection [mm] 239 182 76.15
relaxed hangers 3 0
Fig. F.10. Composition of some relevant values from the comparison
With the proposed hanger arrangement forces and deflections are, on an overall average,
82 % of what they are in the kviksound network arch. If the cross angle and the position of the
first several hangers of the test hanger arrangement is optimised according to Section 6, the
difference will be even bigger.
F-11
N [kN] N [kN]
400
900
1400
1900
2400
300
500
700
900
1100
1300
1500
1700
1900
0 1836 2788 38
0.03 1361 1795 1818 38
0.01 1789 2692 36
0.02 1485 2057 2078 2774 36 0.04 1141 1361 1646 36
distances
0.03 1402 1828 1942 2325 34

0.05 1102 1303 1422 36
0.04 1251 1848 34
0.05 1149 1610 34 0.06 944 1099 1314 32
Diploma Thesis Brunn & Schanack
0.06 1078 1538 32

0.07 944 1099 1314 32
0.07 1031 1179 1467 30
0.08 900 1080 1400 14 0.08 832 1321 5
956
0.09 860 990 1355 11
0.1 827 921 1308 6 0.09 721 889 1331 2
0.11 787 1319 4

0.1 686 889 1320 0
0.12 754 938 1321 2
0.13 730 938 1321 0 0.11 662 735 895 1315 0
range = 0.1
0.14 710 940 1321 1
range = 0.0
0.12 645 709 904 1307 0

0.15 695 945 1318 0
param e te r ellip
param e te r ellip
0.16 685 743 953 1326 0 0.13 635 692 918 1308 0
0.17 679 731 964 1319 0
average N
0.14 631 679 933 1306 0
m axim um N
0.18 677 730 977 1325 0
average N
0.19 631 689 921 1230 0

m axim um N
0.15 632 683 952 1294 0
average variatio n N
M axim um variatio n N
0.2 635 705 936 1235 0
0.21 640 722 1237 0 0.16 637 700 970 1288 0

952
0.22 648 741 969 1232 0 Results for hanger arrangement from Section 6.5.1; 44 hangers
Results for hanger arrangement from Section 6.5.1; 44 hangers
0.17 643 718 991 1283 0

0.23 656 765 987 1243 0
F.2 Results from the variation of the lower hanger nodes by the node
0.24 666 791 1006 1270 0 0.18 653 744 1012 1320 0
0.25 678 822 1025 1296 0
0.19 665 772 1033 1343 0

0.26 691 853 1045 1327 0
F-12
On the following 13 pages internal forces from the calculations explained in Section 6.5.1 are listed.
Annex F: Results from Section 6.5.1
N [kN] N [kN]
450
950
1450
1950
2450
480
680
880
1080
1280
1480
1680
1880
0 1385 1874 36
0 1973 36
0.01 1233 1563 1748 38
0.02 1125 1337 1618 1625 1943 38

0.01 1717 36 0.03 1242 1439 0
1038
0.04 1000 1150 1425 36

0.02 1483 2074 2092 36 0.05 957 1102 1402 34
0.06 883 997 1348 31
0.03 1319 1740 1880 36 0.07 827 899 1294 26

2244
0.08 781 884 1243 8
0.09 748 875 1228 0

0.04 1199 1492 1724 38
0.1 715 868 1216 4
1860
0.11 691 865 1223 2
range = 0.15
param e te r
0.05 631 907 1217 0
0.12 672 735 865 1223 2
range = 0.2
688
param e te r ellip
ellip
0.13 656 718 867 1225 2
0.06 633 702 918 1216 0 0.14 645 1223 0
700 871
averag e N
maximum N
0.15 637 688 877 1222 0

averag e variat ion N
0.07 637 715 929 1218 0 0.16 636 710 932 1219 0
M aximum variat ion N
0.17 641 727 947 1225 0
average N
m axim um N
0.26 642 1222 0 0.18 648 745 962 1235 0

731 942
0.19 655 766 977 1224 0
0.2 664 791 993 1226 0

0.27 643 735 945 1222 0
0.21 674 816 1010 1228 0
F-13
N [kN] N [kN]
550
650
750
850
950
1050
1150
1250
1350
550
650
750
850
950
1050
1150
1250
1350
0.2 774 886 8 0.1 1041 1160 1404 36
0.21 749 1235 6 0.11 996 1270 36
857 872 0.12 964 34
989 1171
0.22 730 828 867 1208 5 0.13 917 933 1074 32
0.23 711 801 863 1183 4 0.14 864 996 30
0.24 696 777 861 1179 4 0.15 821 936 12
0.16 785 886 883 1271 7
0.25 683 760 860 1185 0 0.17 754 874 1233 6
average N
m axim um N
0.26 671 744 860 1193 2 0.18 730 869 1221 34
0.27 660 733 860 1192 2 0.19 706 864 1210 4
0.2 689 763 862 1193 3
0.28 652 719 862 1191 2
0.21 673 742 861 1201 2
0.29 646 708 865 1196 2 0.22 660 727 862 1204 2
0.3 640 703 869 1199 2 0.23 650 711 865 1203 2
0.31 636 692 873 1210 2 0.24 643 701 869 1207 2
0.25 637 691 874 1208 0
average N
0.32 634 684 878 1200 0

m axim um N
0.26 633 678 881 1212 0

0.33 632 680 884 1197 0 0.27 631 680 888 1208 0
0.34 632 685 890 1196 0 0.28 631 683 897 1211 0
0.29 632 906 1212 0

0.35 633 689 897 1200 0 691

0.3 634 703 915 1213 0
0.36 634 696 904 1202 0
param e te r
0.31 637 717 925 1213 0
0.37 635 705 911 1205 0 0.32 641 731 936 1212 0
0.38 637 718 919 1212 0 0.33 646 745 948 1224 0
ellip
0.34 652 760 959 1215 0
range = 0.35
range = 0.25
0.39 640 731 927 1207 0

0.35 658 775 971 1218 0
0.4 644 742 936 1206 0 0.36 664 795 984 1228 0
param e te r ellip
0.41 648 750 940 1208 0 0.37 672 816 996 1225 0
0.42 651 765 950 1210 0 0.38 681 840 1010 1241 0
0.39 690 863 1023 1233 0
0.43 655 775 960 1212 0
0.4 699 884 1036 1236 0
0.44 661 793 970 1214 0 0.41 715 890 1040 1240 0
0.45 668 802 981 1217 0 0.42 723 900 1060 1250 0
0.46 670 820 990 1218 0 0.43 733 920 1075 1260 0
0.44 741 960 1090 1270 0
0.47 678 843 1000 1220 0 0.45 750 984 1105 1292 0
0.48 685 858 1010 1224 0 Results for hanger arrangement from Section 6.5.1; 48 hangers 0.46 760 990 1120 1300 0
0.49 692 870 1020 1228 0 0.47 770 1005 1140 0
0.5 700 889 1030 1233 0 0.48 780 1030 1160 0

0.49 790 1050 1165 0
0.51 710 900 1040 1237 0 0.5 804 1081 1173 0
0.52 718 915 1050 1241 0 0.51 810 1100 1180 0
0.53 725 935 1060 1245 0 0.52 820 1130 1195 0
0.53 835 1140 1210 0
0.54 730 950 1070 1248 0 0.54 845 1180 1215 0
737 969 1080 1251 858 1207 1238
0.55 0 0.55 0
F-14
N [kN] N [kN]
550
650
750
850
950
1050
1150
1250
1350
500
700
900
1100
1300
1500
0.2 774 886 1268 8 0.13 1091 1052 1320 1531 1700 32
0.21 749 857 872 1235 6 0.14 1023 1022 1228 1495 30
0.15 966 999 1143 1457 28
0.22 730 828 867 1208 5
0.16 918 1085 1415 26
0.23 711 801 863 1183 4 0.17 878 1028 1381 12
0.24 696 777 861 1179 4 0.18 844 975 1338 8
0.25 683 760 860 1185 0 0.19 814 945 1300 6
average N
maximum N
0.26 671 744 860 1193 2 0.2 727 867 1198 4
0.21 707 863 1182 4
0.27 660 733 860 1192 2
average variat ion N

0.22 691 861 1191 4

0.28 652 719 862 1191 2
0.23 677 860 1190 2

0.29 646 708 865 1196 2 0.24 664 861 1199 2
0.3 640 703 869 1199 2 0.25 654 720 862 1198 2
0.31 636 692 873 1210 2 0.26 646 707 865 1200 2
0.27 641 701 870 1205 2
0.32 634 684 878 1200 0
0.28 636 690 880 1202 0
average N
0.33 632 680 884 1197 0

maximum N
0.29 633 680 880 1200 0

0.34 632 685 890 1196 0 0.3 632 681 887 1202 0
0.35 633 689 897 1200 0 0.31 631 684 894 1205 0
average variat ion N
0.36 634 696 904 1202 0 0.32 633 691 902 1213 0
0.37 635 705 911 1205 0 0.33 634 700 911 1207 0
0.34 636 715 919 1210 0
0.38 637 718 919 1212 0
param e te r ellip
range = 0.35
param e te r ellip
range = 0.35
0.35 640 728 929 1219 0
0.39 640 731 927 1207 0 0.36 644 740 938 1209 0
0.4 644 742 936 1206 0 0.37 648 752 948 1211 0
0.41 648 750 940 1208 0 0.38 653 765 959 1221 0
0.42 651 765 950 1210 0 0.39 658 779 969 1220 2
0.4 665 794 980 1220 0
0.43 655 775 960 1212 0
0.41 670 800 990 1222 0
0.44 661 793 970 1214 0 0.42 677 820 1000 1226 0
0.45 668 802 981 1217 0 0.43 685 840 1015 1229 0
0.46 670 820 990 1218 0 0.44 692 860 1025 1230 0
0.47 678 843 1000 1220 0 0.45 703 894 1037 1231 0
Re s ults for hange r arrange m e nt from Se ction 6.5.1; 44 hange rs
Re s ults for hange r arrange m e nt from Se ction 6.5.1; 44 hange rs
0.46 710 890 1045 1240 0

0.48 685 858 1010 1224 0
0.47 720 920 1055 1250 0
0.49 692 870 1020 1228 0 0.48 730 940 1066 1260 0
0.5 700 889 1030 1233 0 0.49 740 960 1080 1270 0
0.51 710 900 1040 1237 0 0.5 750 990 1090 1280 0
0.52 718 915 1050 1241 0 0.51 760 1010 1100 1290 0
0.52 770 1020 1120 1300 0
0.53 725 935 1060 1245 0
0.53 780 1030 1130 1310 0
0.54 730 950 1070 1248 0 0.54 787 1040 1140 1330 0
737 969 1080 792 1054 1154 1343
0.55 1251 0 0.55 0
F-15
N [kN] N [kN]
550
650
750
850
950
1050
1150
1250
1350
1450
540
640
740
840
940
1040
1140
1240
1340
0.2 899 925 1080 32 0.2 831 981 13371440 17
0.21 866 911 1033 30 0.21 802 938 1309 10
0.22 836 900 997 1344 24 0.22 777 899 1276 8
0.23 810 958 1320 10 0.23 755 867 1254 6
0.24 787 922 1295 8 0.24 736 868 1223 0
0.25 766 891 1269 14 0.25 719 864 1193 4
0.26 747 862 871 1245 6 0.26 704 862 1178 4
0.27 731 839 867 1219 6 0.27 692 860 1175 4
average N
0.28 716 817 864 1201 4 0.28 680 760 860 1189 0
m axim um N
0.29 704 796 861 1184 4 0.29 669 746 859 1204 2
0.3 692 781 860 1174 4 0.3 660 736 860 1184 2
0.31 682 767 859 1174 4 0.31 670 724 861 1189 2
0.32 672 751 859 1178 2 0.32 647 712 864 1201 2
0.33 664 742 859 1180 2 0.33 642 705 867 1197 0
0.34 657 733 860 1180 2 0.34 638 698 870 1194 2
average N
0.35 651 722 862 1185 1 0.35 635 689 874 1190 2
m axim um N
0.36 646 712 864 1188 2 0.36 634 684 879 1193 0
0.37 642 707 867 1188 2 0.37 633 682 884 1198 0
range = 0.45
0.38 639 700 870 1189 2 0.38 632 684 890 1194 0
range = 0.4
param e te r ellip
0.39 636 692 873 1197 2 0.39 633 690 896 1196 0
param e te r ellip
0.4 634 688 877 1191 2 0.4 634 696 902 1200 0
0.41 633 685 878 1189 0 0.41 635 700 905 1200 0
0.42 633 687 880 1191 0 0.42 638 710 915 1202 0
0.43 634 691 885 1193 0 0.43 640 720 925 1204 0
0.44 634 693 888 1194 0 0.44 642 734 930 1205 0
0.45 634 700 907 1195 0 0.45 645 747 937 1206 0
0.46 635 704 907 1197 0 0.46 648 755 940 1208 0
0.47 637 713 913 1201 0 0.47 652 765 948 1210 0
0.48 639 723 919 1202 0 0.48 658 776 955 1213 0
0.49 640 733 924 1213 0 0.49 662 785 968 1215 0
0.5 644 744 931 1203 0 0.5 666 798 975 1218 0
0.51 647 751 937 1208 0 0.51 670 810 985 1219 0
0.52 650 759 944 1206 0 0.52 675 824 997 1221 0
0.53 652 767 949 1211 0 0.53 682 840 1005 1223 0
0.54 656 776 957 1218 0 0.54 686 855 1010 1225 0
660 780 960 1220 693 870 1017 1226
0.55 0 0.55 0
F-16
N [kN] N [kN]
500
600
700
800
900
1000
1100
1200
1300
500
600
700
800
900
1000
1100
1200
1300
0.46 647 717 863 1182 2 0.41 644 711865 1185 2
0.42 641 707867 1187 2

0.47 644 711 865 1183 2
0.43 638 700 870 1184 2
0.48 641 708 867 1185 2
0.44 636 694 873 1187 2
0.49 639 701 869 1184 2 0.45 635 690 876 1189 2
0.46 634 687 880 1189 0

averag e N
0.5 637 697 872 1186 2
maximum N
averag e N
maximum N
0.47 633 683 884 1186 0

0.51 635 693 874 1184 0


0.48 633 685 888 1187 0

0.52 635 690 877 1184 2 0.49 634 690 892 1192 0
0.53 634 688 880 1187 1 0.5 634 694 896 1187 0
0.51 635 699 901 1193 0

0.54 633 686 884 1188 0
range = 0.55
0.52 635 704 905 1194 0
range = 0.5
0.55 633 685 887 1188 0
param e te r ellip
0.53 650 711 910 1197 0
param e te r ellip
0.56 634 688 889 1189 0 0.54 638 719 915 1200 0
0.57 635 695 893 1190 0 0.55 640 729 920 1208 0
0.56 642 735 925 1207 0

0.58 635 699 894 1190 0
0.57 644 740 930 1207 0
0.59 636 700 896 1191 0
0.58 646 745 935 1206 0
0.6 636 705 907 1191 0 0.59 648 750 940 1206 0
0.6 651 760 945 1205 0

0.61 637 711 909 1192 0
0.61 654 772 951 1205 0
0.62 638 718 913 1196 0
0.62 657 779 956 1205 0
0.63 636 690 880 1202 0 0.63 660 786 962 1208 0
F-17
N [kN] N [kN]
620
720
820
920
1020
1120
1220
550
650
750
850
950
1050
1150
0.5 699 802 860 1184 4 0.41 713 819 862 1209 4
0.51 693 794 859 1180 4 0.42 705 808 861 1192 4
0.52 687 787 858 1172 4 0.43 697 797 859 1182 4
0.53 682 780 858 1163 4 0.44 690 788 859 1175 4
0.45 683 779 858 1164 4
0.54 677 772 858 1162 4 0.46 677 769 858 1165 4
0.55 673 764 857 1166 4 0.47 672 759 858 1165 3
0.56 0.48 666 750 858 1169 2
0.57 0.49 662 745 858 1173 2
0.58 0.5 658 740 859 1173 2
0.51 654 733 860 1176 2
0.59 0.52 650 726 861 1177 2
0.6 0.53 647 720 862 1179 2
0.61 652 731 860 1174 2 0.54 645 715 864 1177 2
0.62 650 726 861 1177 2 0.55 642 711 865 1184 2
0.63 647 721 862 1174 2 0.56 2
0.64 0.57 2
0.58 2
0.65 0.59 2
0.66 0.6 2
0.67 0.61 635 691 878 1191 2
0.68 0.62 634 689 880 1183 2
0.69 0.63 633 678 882 1180 0
0.7 699 872 1179 2 0.64 0
637 0.65 0
0.71 636 697 873 1184 2 0.66 0
0.72 636 695 875 1181 2 0.67 0
0.73 0.68 0
0.74 0.69 0
0.75 0.7 645 710 903 1189 0
average N
0.71 636 708 905 1192 0
m axim um N
0.76 0.72 637 908 1192 0
range = 0.6
712
0.77
range = 0.65
0.73
0.78 0.74
param e te r ellip
0.79 0.75
param e te r ellip
0.8 0.76
0.81 0.77
0.78
0.82 0.79
0.83 0.8
0.84 637 703 898 1182 0 0.81
0.85 636 702 900 1184 0 0.82
0.86 636 705 902 1182 0 0.83
0.84
0.87
average N
0.85
maximum N
0.88 0.86
0.89 0.87
0.9 0.88
average variat io n N
0.89
M aximum variat io n N
0.91
0.9
0.92 0.91
0.93 0.92
0.94 0.93
0.95 0.94
0.96 0.95
0.97 0.96 890
644 745 926 1197 0.97 919 1093
0.98 645 1200 0 0.98 900 987 0
748 928 1120
0.99 0 0.99 1150 0

646 752 930 1205 950 1000
1 0 1 0
F-18
N [kN] N [kN]
550
650
750
850
950
1050
1150
1250
1350
570
670
770
870
970
1070
1170
1270
0.5 865 910 1057 1381 29 0.4 878 1071 30
0.51 854 906 1043 1372 28 0.41 863 1052 29
0.52 843 902 1030 1363 21 0.42 849 1035 27
0.43 836 1018 15
0.53 834 898 1018 1357 12 0.44 823 895 1000 0
0.54 824 895 1005 1348 12 0.45 812 890 984 10
0.55 815 891 992 1342 10 0.46 801 886 967 10
0.56 0.47 791 883 950 8
0.57 0.48 780 880 935 1298 8
0.49 771 877 919 1294 8
0.58 0.5 763 874 905 1283 8
0.59 0.51 755 872 892 1270 6
0.6 0.52 747 870 879 1261 6
0.61 770 876 921 1295 8 0.53 740 867 868 1253 0
0.54 733 856 866 1246 6
0.62 764 874 910 1286 8 0.55 727 847 865 1231 6
0.63 758 873 900 1279 6 0.56 696 800 859 1182 4
0.64 0.57 692 794 858 1179 4
0.65 0.58 687 789 858 1172 4
0.66 0.59
0.67 0.6
0.61 696 800 859 1182 4
0.68 0.62 692 794 858 1179 4
0.69 0.63 687 789 858 1172 4
0.7 721 840 860 1222 0 0.64
0.71 719 838 863 1221 6 0.65
0.72 715 832 862 1215 4 0.66
0.73 0.67
0.68
0.74 0.69
0.75 0.7 664 750 858 1167 2
0.71 661 747 858 1185 2
range = 0.7
0.76
param e te r ellip
0.72 660 740 868 1190 2
range = 0.75
0.77 0.73
0.78 0.74
param e te r ellip
0.79 0.75
0.8 0.76
averag e N
0.81 0.77
maximum N
0.82 0.78
0.83 0.79
0.8
averag e variat io n N
0.84 680 780 857 1162 4 0.81

M aximum variat io n N
0.85 678 776 857 1163 4 0.82

0.86 677 776 857 1162 4 0.83
0.87 0.84 640 707 867 1177 2
0.88 0.85 639 706 868 1179 2
0.86 639 704 869 1181 2
0.89 0.87
0.9 0.88
0.91 0.89
Results for hanger arrangement from Section 6.5.1; 44 ha ngers
0.9
average N
0.92
maximum N
0.93 0.91
0.92
0.94 0.93
average variat io n N
0.95 0.94
0.96 M aximum variat io n N 0.95
0.96
0.97 656 739 859 1170 0.97 634 883 1183
0.98 2 689
737 859 1167 0.98 634 687 884 1181 1
0.99 654 0 0
735 860 1166 0.99 634 685 885 1184
1 653 2 1 0
F-19
N [kN] N [kN]
800
900
1000
1100
1200
1300
1400
600
700
800
900
1000
1100
1200
1300
1400
0.7 983 1309 1453 36 0.5 981 1032 1301 36
1036 0.51 974 1280 1446 36
1016
0.71 978 1025 1295 1451 36 0.52 968 1001 1259 1444 36
0.53 961 987 1240 1436 36
0.72 974 1017 1283 1449 36 0.54 956 974 1222 1433 36
0.73 0.55 950 962 1205 1431 34
0.56
0.74 0.57
0.58
0.75 0.59
0.76 0.6
0.61 899 924 1107 1400 32
0.77 0.62 890 919 1093 1397 0

0.78 0.63 881 917 1081 1396 31
0.64
0.79 0.65
0.66
0.8 0.67
0.81 0.68
0.69
0.82 0.7 832 897 1017 1355 12
0.71 825 895 1009 1350 12
0.83 0.72 815 890 1000 1340 10
0.84 932 937 1162 1421 34 0.73
0.74
0.85 911 930 1160 1420 0 0.75
0.76
param e te r ellip
range = 0.85
0.86 884 929 1158 1419 34
range = 0.8
0.77
param e te r ellip
0.87 0.78
0.79
0.88 0.8
0.81
0.89 0.82
0.9 0.83
0.84 764 874 912 1286 8
0.91 0.85 760 873 906 1287 7
0.86 757 872 900 1277 6
0.92
average N
0.87
maximum N
0.93 0.88
0.89
0.94 0.9
average variation N
0.91
Maximum variation N
0.95
Results for hanger arrangement from Section 6.5.1; 44 ha ngers
0.92
0.96 0.93
0.94
average N
0.97 0.95
m axim um N
0.98 867 911 1065 1381 30 0.96

0.97 723 846 864 1229
0.99 864 910 1386 29 0.98 6
1060 721 842 863 1224

908 1056 0.99 6

1 860 1375 28 719 839 863 1221
1 6
F-20
Diploma Thesis Brunn & Schanack Annex F: Results from Section 6.5.1
N [kN] Results for hanger arrangement from Section 6.5.1; 44 hangers

range = 0.9 number of relaxed hangers
36
36
36
38
36
36
36
0
0
2090
maximum N
1963
1890
1940
1918
Maximum variation N
1700
1680
1650
average variation N
1603
1594
1599
1690
1748
average N
1734
1722
1675
1650
1640
1490
1204 1553
1198 1544
1192 1536
1479
1464
1451
1290
1314
1300
1290
1191
1184
1177
1090
1112
1105
1100
0.98 1060
0.99 1057
1 1054
890
0.7
0.71
0.72
0.73
0.74
0.75
0.76
0.77
0.78
0.79
0.8
0.81
0.82
0.83
0.84
0.85
0.86
0.87
0.88
0.89
0.9
0.91
0.92
0.93
0.94
0.95
0.96
0.97
param e te r ellip

N [kN]
range = 0.95 number of relaxed hangers

38
36
36
36
36
0
3600
3659
3625
3610
3100
3246
3225
3205
maximum N
Maximum variation N
2425
2408
2390
2600
2266
average variation N
2257
2251
average N
2424
2406
2400
2100
2228
2217
2205
1600
1657
1649
1640
1559
1553
1548
1100
0.84
0.85
0.86
0.87
0.88
0.89
0.9
0.91
0.92
0.93
0.94
0.95
0.96
0.97
0.98
0.99
param e te r ellip
F-21
F-22
1 1
0.98 0.98
0.96 0.96
0.94 0.94
0.92 0.92
0.9 0.9
0.88 0.88
0.86 0.86
0.84 0.84
0.82 0.82
0.8 0.8
0.78 0.78
0.76 0.76
0.74 0.74
0.72 0.72
855-860
0.7 0.7
0.68 0.68
0.66 0.66
0.64 0.64
1174-1182
860-865
0.62 0.62
0.6 0.6
0.58 0.58
0.56 0.56
865-870
1182-1190
0.54 0.54
0.52 0.52
0.5 0.5
0.48 0.48
870-875
1190-1198
0.46 0.46
AveN [kN]
MaxN [kN]
0.44 0.44
0.42 0.42
875-880
0.4 0.4
0.38 0.38
1198-1206
0.36 0.36
0.34 0.34
880-885
0.32 0.32
0.3 1206-1214 0.3
0.28 0.28
0.26 0.26
885-890
0.24 0.24
0.22 0.22
0.2 0.2

0.18 0.18
0.16 0.16
0.14 0.14
0.12 0.12
0.1 0.1
0.08 0.08
0.06 0.06
0.04 0.04
0.02 0.02
0 0
1
0
1
0
ellip
ellip
range
range
0,9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0,9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Annex F: Results from Section 6.5.1 Diploma Thesis Brunn & Schanack
N [kN] range
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0,9
1
1
0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.98
0.96
0.94
0.92
0.88
0.86
0.84
0.82
0.78
0.76
0.74
0.72
0.68
0.66
0.64
0.62
0.58
0.56
0.54
0.52
0.48
0.46
0.44
0.42
0.38
0.36
0.34
0.32
0.28
0.26
0.24
0.22
0.18
0.16
0.14
0.12
0.08
0.06
0.04
0.02
ellip
680-685 685-690 690-695 695-700 700-705 705-710
AveN [kN] range

0
0.15
0.3
0.45
0.6
0.75
0,9
1 0.96 0.92 0.88 0.84 0.8 0.76 0.72 0.68 0.64 0.6 0.56 0.52 0.48 0.44 0.4 0.36 0.32 0.28 0.24 0.2 0.16 0.12 0.08 0.04 0
ellip
630-635 635-640 640-645 645-650 650-655 655-660 660-665 665-670 670-675 675-680 680-685 685-690
F-23
F-24
1
0.98
0.96
0-1
0.94
0.92
1-2
0.9
0.88
2-3
0.86
0.84
3-4
0.82
0.8
4-5
0.78
0.76
5-6
0.74
0.72
6-7
0.7
0.68
7-8
0.66
0.64
8-9
0.62
0.6
9-10
0.58
0.56
0.54
10-11
0.52
0.5
11-12
0.48
0.46
0.44
12-13
0.42
0.4
0.38
13-14
Number of relaxed hangers
0.36
0.34
14-15
0.32
0.3
0.28
15-16
0.26
0.24
0.22
16-17
0.2
0.18
17-18
0.16
0.14
0.12
18-19
0.1
ra n g e
0.08
19-20
0.06
0.04
0.02
0
1
0
ellip
range
0,9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Annex F: Results from Section 6.5.1 Diploma Thesis Brunn & Schanack
F.3 Results from the variation of the lower hanger nodes by the slope of the
hangers
Results for hanger arrangemants from Section 6.5.2, start angle 50, number of hangers 44, no variable
angle change
[kN]
0 0 3 3 0 0 2 2 2 2 number of relaxed hangers

1700.00
1611.5
1601.4
1593.5
1588.9
1581.6
1578.1
1579.5
1574.5
1575.8
1500.00
1570.0
1300.00
1125.2
1096.8
1070.6
1046.2
1028.1
1010.0
988.5
1100.00
972.4
954.9
937.3
1086.3
1045.8
1008.6
978.5
900.00
954.3
930.0
912.4
maximum N
892.8
878.3
863.8
maximum variation N
700.00
768.4
average variation N
750.8
734.6
719.2
709.6
700.0
690.8
683.8
685.2
682.4
average N
500.00
0
3
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
1
2
4
6
8
1
2
3
4
5
.3
.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
-0
-0
angle change
[kN]
3 2 2 1 0 0 0 0 0 0 0 0 number of relaxed hangers
1500.00
1515.0
1496.5
1490.0
1488.5
1482.6
1481.6
1481.3
1482.0
1482.5
1480.8
1480.0
1300.00
1101.0
1075.2
1035.0
1031.2
1010.0
995.9
1100.00
981.1
968.5
957.3
947.7
932.9
1010.5
971.7
900.00
930.0
922.5
900.0
873.4
maximum N
863.2
851.9
846.5
692.2 843.7
843.0
maximum variation N
700.00
746.5
average variation N
729.9
710.0
705.0
695.0
689.2
683.8
682.5
682.5
684.2
average N
500.00
0
3
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
1
2
4
6
8
1
2
3
4
5
.3
.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
-0
-0
F-25
[kN] [kN]
500
700
900
1100
1300
1500
1700
-0
.3
500.00
600.00
700.00
800.00
900.00
1000.00
1100.00
1200.00
1300.00
1400.00
-0 -0 1500.00
.1 .3
-0 692.2 859.6 935.0 1450.8
0 705.6 837.4 941.3 1398.6 .1
? ?
0. 697.9 828.0 946.1 1386.3 0 683.8 842.3 941.0 1430.3
1
0. 690.2 818.5 951.0 1373.9
0.
1 682.5 824.9 947.0 1409.8
2
682.5 823.7 954.6 1398.9
? ?
0.
0.
3 686.3 817.5 958.9 1360.4 2

0. 684.3 818.6 967.9 1352.6
0.
3 684.2 826.1 964.4 1388.9
4
689.2 828.5 974.2 1378.8
? ?
0.
0.
5 683.7 822.5 978.3 1343.8 4
0. 685.0 831.5 990.2 1337.2
0.
5 695.0 834.1 985.5 1370.6
6
0. 689.3 846.0 1004.81332.7

0.
6 705.0 849.7 999.5 1365.9
7 0. 710.0 866.2 1015.1 1368.8
? ? ?
0. 698.1 861.0 1025.5 1458.6 7
8 0. 729.9 881.0 1030.0 1365.0
0. 709.3 882.0 1049.2 1614.0 8
9 0. 746.5 897.0 1045.0 1362.0
729.4 896.3 1087.9 9
1
6 ? 0 ? ? ? ? ? ? ? ? ?
1 903.0 1060.0 1360.0

1.
1 760.4 1051.1 1141.8
1.
797.2 1168.4 1
? 6 10 12
1.
2 1.
1. 2
3 1.
1. 3
4 1.
1. 4
5 1.
1. 5
6
angle change
angle change
1.
1. 6
7 1.
1. 7
8 1.
1. 8
9 1.
9
2 2
2.
1 2.
1
2.
2 2.
2
2.
4 2.
4
2.
6 2.
6
2.
8 2.
8
3 3
3. 3.
1 1
3. 3.
2 2
average N
average N
3. 3.
maximum N
maximum N
3 3
3. 3.
4 4
3. 3.
5 5
F-26
average variation N
average variation N
maximum variation N
maximum variation N
[kN] [kN]
500
600
700
800
900
1000
1100
1200
1300
1400
1500
500
600
700
800
900
1000
1100
1200
1300
1400
1500
-0 -0
.3 .3
-0 -0
.1 .1
0 0
?
0.
1 713.7 836.5 944.5 1338.5 0.
1 704.5 830.1 944.4 1362.8
0 ?
0.
2 704.3 823.5 948.2 1323.0 0.
2 695.8 818.0 948.5 1351.2
? ?
0.
3 695.0 810.4 951.9 1307.5 0.
3 689.8 812.2 954.6 1336.4

0.
4 692.2 809.4 957.8 1300.0 0.
4 686.2 811.2 962.6 1324.8
? ? ?
? ?
0.
5 689.3 808.4 963.8 1292.5 0. 684.1 812.5 971.9 1314.2
5
0.
6 683.6 806.4 975.6 1277.4 0.
6 684.1 816.5 983.1 1297.4
0.
7 684.3 810.8 986.2 1268.5 0.
7 686.0 826.7 994.8 1298.9
? ? ?
? ? ?
0.
8 686.1 821.6 999.4 1256.5 0. 690.6 842.3 1009.2 1293.9
8
0.
9 691.8 838.6 1013.6 1255.6 0.
9 696.9 860.2 1025.4 1292.9
? ?
1 698.2 856.8 1030.2 1246.1 693.0 850.0 1020.0 1292.0

1
? ? ?
1.
1 706.5 879.9 1047.6 1249.8 1. 690.6 880.0 1050.0 1293.9
1
? ?
1.
2 718.2 900.9 1068.7 1243.8 1. 730.0 928.8 1085.9 1285.9
2
1.
3 732.4 925.2 1091.2 1245.5 1. 746.0 965.0 1110.8 1281.8
3
? ? ?
? ?
1.
4 748.5 961.3 1116.6 1233.8 1. 764.8 1005.3 1138.4 1283.2
4
1.
5 767.8 986.4 1144.4 1269.8 1. 787.5 1046.3 1169.3 1303.2
5
? ?
? ?
1.
6 791.3 1041.6 1176.2 1325.7 1. 812.1 1088.4 1203.0 1355.5
angle change
6
angle change
1. 1.
7 7
1.
8 1.
8
1. 1.
9 9
2 2
2.
1 2.
1
2.
2 2.
2
2.
4 2.
4
2.
6 2.
6
2.
8 2.
8
3 3
3.
1 3.
1
3.
2 3.
2
average N
3.
maximum N
average N
3 3.
maximum N
3
3.
4 3.
4
3.
5 3.
5
F-27
average variation N
maximum variation N
average variation N
maximum variation N
[kN] [kN]
550
650
750
850
950
1050
1150
1250
1350
1450
550
650
750
850
950
1050
1150
1250
1350
1450
-0
.3
-0
-0 -0
.1
0 749.5 882.2 946.4 1313.4
7 2
0
0.
1 735.3 859.6 946.1 1296.8 723.9 846.8 945.0 1317.7
0.
2 722.7 838.2 947.1 1280.5 712.5 829.5 946.9 1303.2
0 0
0 0
0.
3 711.4 822.5 949.1 1265.4 702.3 815.4 949.8 1289.3

0.
4 701.4 807.9 952.1 1252.8 694.3 803.1 954.3 1278.4
0 0
0 0
0.
5 693.7 796.8 956.8 1241.3 690.0 800.8 961.8 1265.3
0.
6 691.4 796.2 963.0 1231.8 685.7 798.5 969.2 1252.1
0.
7 689.1 795.6 969.2 1222.3 684.5 800.4 979.2 1234.7
0 0 0
0 0 0
0.
8 684.5 794.4 981.7 1203.2 684.6 805.4 989.8 1227.7
0.
9 685.1 801.3 992.8 1193.1 687.7 818.6 1002.9 1219.6
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9
0 0
0 0
688.5 814.1 1006.1 1184.8 692.8 834.6 1017.5 1218.8

1
1.
1 693.7 830.5 1021.0 1179.1 706.6 865.2 1045.2 1211.3
0 0
0 ?
1.
2 700.2 849.7 1037.3 1174.5 713.4 880.5 1059.1 1207.5
1.
3 709.6 871.5 1055.8 1182.3 720.3 895.8 1073.0 1203.7
0 0
? ?
1.
4 722.1 890.6 1077.0 1227.6 734.5 921.6 1096.0 1230.3
0
1.
5 740.0 910.0 1097.0 1280.0 750.6 943.7 1121.0 1268.5
? ?
1.
6 771.2 996.9 1149.7 1317.5
angle change
angle change
1.
7
1.
8 1 1 1 1 1 1 1 1 1
1.
9
1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
2
2.
1
2.
2
2.
4
2.
6
2 2 2 2 2
2.
8
2 . 1 .2 .4 .6 . 8
3
3.
1
3.
2
average N
average N
3.
maximum N
maximum N
3
3.
4
3 3 3 3 3
3.
5
F-28
3 .1 .2 . 3 . 4 .5
average variation N
average variation N

maximum variation N
maximum variation N
[kN] angle change
6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 number of relaxed hangers
1276.7
1269.3
1260.7
1350
1244.7
1232.5
1228.4
1217.8
1205.4
1196.1
1188.6
1186.7
1168.0
1156.4
1156.2
1250
1149.8
1080.5
1059.2
1150
1040.3
1024.3
1009.1
995.6
984.2
971.7
1050
965.4
959.2
954.1
950.9
948.2
948.6
949.0
950
850
884.7
873.9
855.1
734.9 852.5
845.4
maximum N
721.6 831.0
826.2
710.3 815.6
809.7
750
700.4 800.6
685.5 796.1
693.2 790.2
691.0 789.8
688.9 789.5
684.5 788.7
maximum variation N
748.1
average variation N
723.7
710.7
650
701.0
694.5
689.2
average N
550
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9 1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
1 1 1 1 1 1 1 1 1
2 . 1 .2 .4 .6 . 8
2 2 2 2 2
3 .1 .2 . 3 . 4 .5
3 3 3 3 3
-0 -0
angle change
[kN]
? ? ? ? ? ? ? ? ? ? ? ? ? number of relaxed hangers

1277.5
1350
1238.8
1208.3
1197.7
1190.1
1191.5
1181.3
1168.2
1163.4
1161.3
1250
1147.5
1131.8
1124.1
1062.7
1150
1043.4
1026.7
1012.0
998.2
986.8
976.6
968.2
1050
961.3
955.8
951.0
950.7
953.1
950
850
859.7
841.0
maximum N
731.8 835.3
720.5 824.1
821.7
709.3 808.6
689.8 805.3
750
699.6 793.1
686.0 791.0
maximum variation N
692.4 783.0
683.4 783.0
687.5 780.6
685.5 780.7
average variation N
712.3
650
701.9
695.1
average N
550
0
3
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
1
2
4
6
8
1
2
3
4
5
.3
.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
-0
-0
F-29
[kN] angle change
? ? ? ? ? ? ? ? ? ? ? ? ? number of relaxed hangers
1400
1304.8
1262.0
1229.5
1300
1190.1
1190.2
1174.4
1160.6
1161.4
1146.2
1129.4
1126.4
1125.6
1111.1
1200
1065.1
1045.7
1029.0
1014.1
1100
1000.4
988.4
978.5
969.9
963.1
957.7
954.4
951.0
952.1
1000
853.6
835.3
836.7
900
817.1
816.9
801.6
800.9
785.6
786.8
777.4
777.6
774.7
774.9
maximum N
800
maximum variation N
average variation N
700
731.8
719.4
average N
713.5
708.2
702.7
698.8
695.5
692.3
689.3
687.9
686.4
685.3
684.4
600
0
3
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
1
2
4
6
8
1
2
3
4
5
.3
.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
-0
-0
angle change
[kN]
12 6 0 0 0 0 0 0 0 ? ? ? ? ? ? number of relaxed hangers

1400
1299.5
1315.2
1264.6
1228.7
1219.8
1300
1195.7
1171.7
1160.4
1155.1
1134.6
1128.1
1125.4
1109.9
1200
1096.8
1095.2
1048.3
1031.1
1016.0
1100
1002.5
990.2
980.1
971.6
964.8
960.3
959.1
959.1
955.3
952.2
953.5
952.0
1000
849.0
831.3
827.3
900
927.6
812.0
810.1
873.9
796.4
794.6
782.4
778.1
773.2
771.1
769.1
768.7
maximum N
800
maximum variation N
797.1
average variation N
765.3
700
744.2
730.5
average N
718.2
707.1
703.5
697.8
695.7
691.1
690.7
687.5
686.8
685.0
684.4
600
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9 1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
1 1 1 1 1 1 1 1 1
2 . 1 .2 .4 .6 . 8
2 2 2 2 2
3 .1 .2 . 3 . 4 .5
3 3 3 3 3
-0 -0
F-30
[kN] angle change
12 0 0 0 0 0 0 0 ? number of relaxed hangers
1450
1336.9
1424.3
1301.3
1263.1
1350
1223.5
1192.7
1159.6
1155.3
1250
1137.4
1129.4
1121.6
1100.0
1099.1
1088.6
1083.7
1050.5
1150
1032.8
1017.8
1004.1
991.7
981.5
973.6
969.6
965.8
1050
960.3
956.6
865.1 953.9
840.9 952.9
952.9
954.0
950
825.4
819.8
892.1
800.0
801.6
791.6
787.6
778.3
771.4
767.7
765.1
850 764.3
763.4 maximum N
750 maximum variation N
788.4
758.1
742.7
average variation N
729.1
719.0
705.9
704.4
650
697.3
695.7
691.3
691.0
688.0
687.1
684.7
684.5
average N
550
0
3
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
1
2
4
6
8
1
2
3
4
5
.3
.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
-0
-0
angle change
[kN]
4 0 0 0 0 0 0 ? ? ? ? ? ? ? ? number of relaxed hangers

1450
1325.5
1359.6
1267.2
1291.3
1350
1226.4
1189.4
1195.0
1158.7
1250
1136.1
1133.9
1122.5
1113.9
1094.8
955.0 1079.1
954.3 1062.9
954.1 1061.2
1052.0
1150
1035.6
1019.2
1005.6
993.2
982.6
973.8
966.8
1050
958.0
956.8
955.5
957.9
954.6
950
848.3
824.9
819.6
806.6
803.2
789.2
786.8
772.8
774.1
900.0
850
762.3
773.6 883.2
758.9
756.8
757.6
756.4 856.8
maximum N
750 maximum variation N
790.0
754.7
average variation N
739.6
726.2
714.1
705.2
650
703.4
698.0
690.6
690.9
686.5
687.3
684.2
684.6
average N
550
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9 1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
1 1 1 1 1 1 1 1 1
2 . 1 .2 .4 .6 . 8
2 2 2 2 2
3 .1 .2 . 3 . 4 .5
3 3 3 3 3
-0 -0
F-31
[kN] [kN]
550
650
750
850
950
1050
1150
1250
1350
1450
550
650
750
850
950
1050
1150
1250
1350
1450
-0
.3
-0
-0 -0
.1
0 981.2 998.1
861.4 1167.2
14
0.
1
0.
2
0. 832.0 953.0 970.0 1125.0
3

0.
4 830.7 946.3 971.9 1091.9 810.1 923.6 965.9 1104.2
18 8 3
0.
5 819.0
930.3 968.5 1082.7 788.9 894.0 960.7 1087.9
0.
6 810.0 918.2 966.1 1075.3 769.7 866.4 957.0 1073.4
0.
7 796.9 900.3 963.0 1061.5 752.9 839.7 954.8 1055.6
12 10 8 4 0
0 0 0
0.
8 786.6 885.4 960.4 1058.1 737.9 817.8 953.9 1059.2
0.
9 776.9 871.4 958.4 1059.3 724.6 800.0 954.2 1088.3
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9
0 0
1 739.5 818.2 958.2 1101.3 712.1 782.2 955.4 1116.2
0 2 ?
1.
1 702.2 765.0 958.0 1143.3 695.6 759.0 962.2 1149.5
? ?
1.
2 694.7 752.9 962.5 1167.3 690.0 752.7 968.1 1171.5
1.
3 689.3 746.5 968.4 1197.8 686.0 751.5 975.2 1195.1
? ? ?
? ?
1.
4 680.0 740.0 974.0 1240.0 683.9 752.2 983.8 1222.5
1.
5 683.6 747.4 984.3 1244.8 684.7 758.6 994.8 1265.9
1277.5
? ?
1296.7
? ?
1.
6 684.8 755.3 995.3 687.5 769.9 1007.3
angle change
angle change
1.
7 687.3 766.0 1007.8 1309.1 691.2 782.0 1020.7 1318.5
? ?
1.
8 691.3 778.0 1021.3 1343.8 697.0 801.0 1036.1 1358.3
? ? ?
697.7 796.1 1037.0
?
1 1 1 1 1 1 1 1 1
1.
9 1376.7 705.9 813.7 1053.7
1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
1386.6
2 706.5 809.3 1054.8
1409.9
2.
1 718.8 839.0 1075.9
1446.2
? ? ?
2.
2 731.1 868.7 1097.0
1482.5
2.
4
2.
6
2.
2 2 2 2 2
8
2 . 1 .2 .4 .6 . 8
3
3.
1
3.
2
average N
3.
maximum N
average N
3
maximum N
3.
4
3.
3 3 3 3 3
F-32
3 .1 .2 . 3 . 4 .5
average variation N

average variation N
maximum variation N
maximum variation N
[kN] [kN]
550
750
950
1150
1350
1550
1750
550
750
950
1150
1350
1550
-0
.3
-0
-0 -0
.1
15
0 13
03 1612.9 1859.0
1267.3 .4 989.2
0.
1
94 1018.8
1214.5 .8 1504.3
0.
2 884.5 1131.8
1411.3
0. 1169.81302.1 1693.9 856.1 980.2 979.6 1111.1
3 1336.3 1557.3

average N
maximum N
0.
4 1131.9 1222.1 830.8 946.3 971.9 1091.9
1441.3
3 16 14 12 0
0.
5 1152.9 1343.1 1264.5 808.7 915.9 966.1 1075.0
0.
6 1092.6 1258.7 1212.3 786.7 885.4 960.4 1058.1
average variation N
maximum variation N
0.
7 1039.7 1184.6 1209.7 767.7 857.7 956.9 1071.5
0 0 0
0.
8 993.2 1121.51211.7 751.0 831.2 954.6 1066.2
0.
9 951.8 1067.4 1219.2 731.0 803.2 955.0 1104.0
.3 . 1 0 0 .1 0 .2 0 .3 0 .4 0 . 5 0 . 6 0 . 7 0 .8 0 .9
0 0
915.4 1025.5 1223.4 721.0 789.1 955.2 1122.8

1
1.
1 884.4 989.7 1218.3 711.0 775.1 955.3 1141.7
34 34 34 34 36 36 34 30 26 20 4 4
0 0
1.
2 853.4 953.9 975.8 1213.2 700.9 761.0 958.2 1172.4
1.
3 827.4 922.9 967.7 1211.9 693.8 750.8 962.8 1199.5
0 0
4 2
1.
4 804.0 960.9
897.5 1236.5 688.6 746.5 968.8 1227.4
1.
5 782.8 955.4
877.3 1255.8 684.9 741.4 975.9 1241.2
1279.9 1271.7
0 0
2 2
1.
6 763.9 856.2951.5 683.6 742.1 985.0
angle change
angle change
1.
7 747.5 836.5948.9 1303.8 684.9 752.1 996.2 1306.0
0 ?
2 2
1.
8 732.2 824.2 947.4 1323.2 687.7 762.0 1008.7 1338.3
1 1 1 1 1 1 1 1 1
1.
9 718.3 815.9 946.9 1351.9 691.6 775.0 1022.3
1 . 1 .2 .3 . 4 . 5 .6 .7 . 8 . 9
1368.7
? ?
2 2
706.7 809.3 947.8 1370.2 698.3 791.3 1038.2
2 1400.3
2.
1 696.2 799.3 950.5 1390.8 707.1 805.6 1056.1
1433.2
? ?
2 2
2.
2 690.7 797.1 954.6 1410.4 718.0 834.0 1075.8
1459.1
2. 685.5
4 795.6 959.9 1431.3 732.1 866.5 1098.6
1474.0 1494.2
2. 682.4
6 797.5 976.2 766.9 922.1 1151.2
1565.1
806.8 997.9 1517.1
2 2 2 2 2
2. 684.8
8
2 . 1 .2 .4 .6 . 8
825.7 1025.8 1543.2

2 2 number
3 694.8
2 2 of
3.
1 711.6 873.7 1062.1 1579.8
3.
2 722.9 914.7 1083.5 1601.3
734.8 949.3 1107.2 1621.0
average N
3.
maximum N
3
4 relaxed
3.
4 734.8 949.3 1107.2 1621.0
744.8 979.3 1132.1 1647.0
3 3 3 3 3
3.
5
F-33
3 .1 .2 . 3 . 4 .5
average variation N
4 4
749.4 980.2 1157.5 1683.0

4 4 4hangers
? ? number of relaxed hangers
maximum variation N
[kN] [kN]
550
750
950
1150
1350
1550
1750
-0
.3
550.00
750.00
950.00
1150.00
1350.00
1550.00
1750.00
-0
.1
-0 -0
0 18
0. 59
1 .0
0.
2 1267.3 1503.4
16
0.
3 1214.5 93
1859.0 1394.8 15 .9
1267.3

average N
1503.4 1169.8 57
maximum N
0.
4 1302.1 14 .3
average N
maximum N
1214.5 1693.9 1394.8 0.
5 1131.9 1222.1 1
41
34 .3
1169.8 1302.1 1557.3 0.
6 1099.8 1152.91 3.
1
average variation N
25
maximum variation N
1131.9 1222.1 1441.3 0.
7 1072.3 1092.61
average variation N
8.
7
maximum variation N
18
1099.8 1152.9 1343.1 0.
8 1039.7 1048.9 4.
11
2 6
1072.3 1092.6 1258.7 0.
9 993.2 1028.9
1 0 1 .5
1039.7 1048.91184.6 951.8 1012.1 67
1
. 3 . 1 0 0 . 1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7 0 .8 0 .9
.4
993.2 1121.5 1.
1 915.4 997.8
951.8 1067.4 1.
2 884.4 9 986.8
2
915.4 1025.5 1.
3 853.4 8 2.9 975.8
97
884.4 989.7 1.
4 827.4 . 967.7
88 5
853.4 953.9 1.
5 804.0 0. 960.9
87 0
34 34 34 34 36 36 34 30 26 20 4 4 4
827.4 922.9 1. 780.3 7 954.0 1282.0
6
8 .0
angle change
angle change
804.0 897.5 960.9 1.

7 775.0 58 953.0 1290.0
.9
783.3 888.3 960.0 1285.7 1.
8 762.1 950.2 1303.5
83
770.3 872.1 958.0 1294.0 1.
9 745.4 9 947.7 1326.6
8 2 .3
1 1 1 1 1 1 1 1 1
742.5 848.5 951.2 1343.7 2 730.2 7. 946.2 1357.3
1 .1 .2 .3 .4 . 5 . 6 . 7 . 8 . 9
82 2
728.6 836.6 947.8 1368.5 2.
1 716.5 0. 945.6 1369.8
80 5
714.7 824.8 944.3 1393.3 2.
2 696.6 6. 949.6 1409.7
80 9
703.8 817.8 945.6 1409.4 2.
4 689.9 4. 953.8 1433.7
80 9
695.2 814.8 948.6 1428.5 2.
6 682.6 3. 966.8 1469.8
81 3
684.2 810.2 958.3 1463.9 2.
8 682.5 0. 985.6 1509.1
82 2
2 2 2 2 2
681.7 815.1 975.0 1504.7 3 689.4 4. 1010.8 1540.2
2 .1 .2 .4 .6 .8
858
702.6 1042.9 1571.0
34 34 34 34 36 36 34 30 26 20 4 4 4 2 4 4 2 2 2 2 2 2 2 2number
688.3 832.8 1000.2 1535.7

2 2 0 6 6 2 2 2 2 number
3. 9.
1 892
2 0 of
694.9 850.4 1025.5 1566.6 3.

2 711.6 0. 1062.1 1589.1
922
702.1 875.3 1042.5 1578.7 3.
3 721.6 4. 1083.3 1608.8
955
710.6 907.0 1061.7 1602.8 3.
4 731.4 5. 1106.3 1628.3
2 4 of4 relaxed
4 relaxed
96 9
4 4 4
719.7 932.4 1082.8 1614.9 3.

5 738.5 6. 1130.4 1651.8
96 3
4
3 3 3 3 3
727.3 958.1 1105.1 1632.9 736.6 1. 1153.6 1686.9
F-34
3 .1 .2 .3 .4 .5
4 4
729.6 952.8 1127.6 1652.3

hangers
4 4 4hangers
[kN] [kN]
550.00
750.00
950.00
1150.00
1350.00
1550.00
550.00
750.00
950.00
1150.00
1350.00
1550.00
1750.00
-0
.3
-0
3
685.2 892.8 972.4 .1
0
891.6 1574.5
16
992.1 1030.5 1133.1 0.
1
linear angle change []

0.
868.9 2
15
984.6 1001.0 1112.2

0.
3 1859.0
average N
maximum N
847.9 974.1 1085.3 0. 1267.3 1503.4
13
978.2 4
0.
5 1214.5 1394.8 1693.9
828.5 972.7 1081.5
10
949.4 0. 1169.8 1302.1 1557.3
6
average variation N
maximum variation N
0.
7 1131.9 1222.1 1441.3
4
810.4 925.0 968.4 1082.6
0.
8 1099.8 1152.9 1343.1
average N
maximum N
0
793.6 900.8 964.6 1091.8 0.
9 1092.6 1072.3 1258.7
1 1039.7 1184.6
0
777.9 878.7961.8 1102.2 1. 993.2 1121.5
average variation N
1
maximum variation N
1. 951.8 1067.4
0
763.4 859.5 960.0 1147.1 2
1.
3 915.4 1025.5
0
750.2 839.9 958.8 1169.3 1.
4 884.4 989.7
34 34 34 34 36 36 34 30 26 20 4 4
1.
5 853.4 953.9
0
738.2 822.9 958.6 1202.9
angle change
1.
6 827.4 922.9 967.7
angle change 1
0
727.1 807.5 959.1 1231.9 1.
7 804.0 897.5 960.9
1.
8 793.9 910.7 954.4 1328.3
0
716.9 793.9 960.4 1270.2 1. 773.5 887.9 949.4 1351.5
9
755.3 867.2 945.7 1371.2
0
707.5 782.2 962.1 1290.7 2
2.
1 738.9 855.7 943.2 1390.6
0
699.2 771.4 965.0 1326.5 2.
2 724.0 844.9 941.7 1411.9
2.
4 709.7 837.2 941.6 1435.8
692.2 765.4 968.9 1349.4 2.
6 693.0 830.1946.2 1467.9
683.3 827.8 957.0 1509.8
0
686.5 760.3 973.5 1384.2 2.
8
680.7 832.9 973.3 1536.8
0 number
4 2 6 4 4 4 4 4 2 number
3
0
681.9 757.1 979.2 1417.8

2 4 of
3.
1 684.8 844.7 995.8 1562.9
3.
2 688.2 859.1 1009.0 1569.2
0
678.6 758.2 985.9 1458.5 3.

3 693.9 880.6 1024.3 1580.3
of relaxed
4 relaxed
676.2 758.2 993.1 1480.3

3.
4 699.8 902.9 1041.2 1594.9
3.
5 719.7 932.4 1082.8 1614.9
F-35
Results for hanger arrangemants from Section 6.5.2, start angle 83, number of hangers 44, constant
675.0 761.1 1001.5

1511.6
4 4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.2
727.3 958.1 1105.1 1632.9
hangers
4 4 4hangers
[kN] [kNm]
200
400
600
800
1000
1200
1400
1600
1800
2000
1000.00
1500.00
2000.00
2500.00
3000.00
3500.00
4000.00
4500.00
5000.00
5500.00
661.8 881.1
306.8 762.7 1032.3 1986.6
1048.1
1047.0 1321.7 1930.1
cross angle []
cross angle []
model
0.83 2
1065.2 1342.3 1473.9 1821.1
4938.3
0.83 2 4
1067.8 1341.4 1488.7 1722.1
4
3398.3
1069.2 1341.8 1471.9 1624.0
3302.0
5324.5
6
1065.3 1340.5 1452.7 1525.2
6
3089.5
5201.3
1056.5 1344.4 1436.5 1425.9
2907.0
4939.4
1325.8
average N
8
1043.3 1345.2
maximum N
2724.3
4676.5
1026.2 1345.1 1229.4
2541.5
4415.1
1133.8
1350.2
2356.1
4150.2
1012.9 996.8 1040.5
2169.0
3880.1
928.2 1365.4
average variation N
981.8 1346.3
1994.4
3601.1
893.1 1382.6
maximum variation N
967.6 841.6
1826.0
3319.3
860.6 1325.6 1401.6 751.5
1657.6
3036.8
829.3953.2 1249.2 1414.9 659.0
799.5 945.6 1208.5 1445.5 532.4 849.8
784.7 941.8 1188.2 1460.8 469.2 753.5
769.8 938.0 1167.8 1476.1 405.9 657.3
755.1 930.6 1121.9 1483.8 326.2 569.7
259.6 477.0 880.1
743.8 928.6 1125.2 1496.6 216.7 404.3 828.3
729.5 925.2 1107.4 1512.3 191.3 362.1 667.5 794.0
720.0 924.7 1093.1 1527.2 175.1 335.1 593.8 720.7
713.7 926.1 1096.3 1542.8 165.5 321.0 574.5 685.1
691.3 919.4 1068.9 1559.2 158.4 308.8 552.8 665.7
667.9 912.5 1035.0 1567.8 152.8 297.6 531.1 649.7
647.1 907.4 1002.5 1577.8 148.4 287.6 509.7 639.4
628.0 903.4 968.6 1591.4 143.9 275.1 489.2 624.2
609.0 899.6 935.1 1583.9 139.8 263.3 468.0 611.0
136.0 251.7 452.2 597.7
591.7 897.3 900.0 1580.2 132.1 241.8 434.6 583.4
576.8 872.0 896.5 1585.9 128.8 232.0 416.0 569.5
564.7 846.5 897.1 1591.0 125.5 223.6 398.2 552.5
554.3 820.6 899.3 1585.5 122.3 215.6 387.2 532.6
546.2 799.7 904.0 1578.6 119.3 208.1 373.6 523.1
539.5 794.6 910.0 1574.4 116.4 201.0 359.3 517.4
533.7 787.8 917.1 1570.1 113.9 194.7 344.6 513.3
529.5 782.2 925.4 1569.9 112.0 189.2 334.6 510.6
527.2 1567.3 110.0184.1 325.3 504.1
775.9 935.6 108.3179.8 316.7 497.0
526.6 771.2 946.7 1583.2 107.0176.0 310.8 489.6
527.1 772.1 958.7 1606.6 106.5173.2 301.6 479.7
528.8 786.2 971.9 1620.1 106.1170.9 301.2 471.0
531.7 797.1 986.5 1633.3 107.7170.4 296.2 457.8
535.5 801.8 1002.7 1647.4 109.0170.4 286.9 450.2
540.2 807.2 1020.5 1660.2 110.1169.7 280.9 437.3
546.3 809.4 1040.2 1664.0 112.3 170.5 280.3 428.5
553.7 806.7 1062.3 1665.6 113.9173.1 281.5 418.5
562.7 803.8 1086.7 1667.2 117.7175.6 276.5 431.3
M
average M
0 24 35 38 38 38 38 40 40 40 38 38 36 36 33 33 33 32 30 30 28 26 22 8 6 4 4 4 4 4 4 4 2 2 2 2 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0
maximum M
123.5 180.6 278.1 446.9

0 24 35 38 38 38 38 40 40 40 38 38 36 36 33 33 33 32 30 30 28 26 22 8 6 4 4 4 4 4 4 4 2 2 2 2 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0
573.0 799.2 1113.3 1674.0 131.1184.7 284.9 468.9

8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Res ults for hanger arrangm ent from Section 6.6.3, s pan 100m , f 17m , num ber of hangers 44
F.4 Results from variation of the lower hanger nodes using the advanced
585.8 793.8 1142.7 1688.9

Results for hanger arrangment from Section 6.6.3, span 100m, f 17m, number of hangers 44
On the following two pages internal forces from the calculation explained Section 6.6.3 are listed.
601.4 794.2 1176.4 1736.3

maximum variation

0.00
F-36
621.7 802.5 1216.2 1831.4
average variation M
650.7 848.0 1264.2 1995.2 [kN]
10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
100.00
200.00
300.00
400.00
500.00
600.00
700.00
800.00
900.00
[kNm] [kN]
500
700
900
1100
1300
1500
1700
38
.2
100.00
150.00
200.00
250.00
300.00
350.00
400.00
450.00
500.00
38
.2 5
0
5 35 527.6 775.4 962.0 1609.0

.
0
35
.5 109.8 183.0 323.1 507.4 38 5
0
527.9 779.3 965.2 1612.8
cross angle []
cross angle []
38 .7
0
.7 109.1 181.9 321.3 502.9 5
5
0
3 528.3 782.7 968.5 1616.6
0
39 108.8 180.8 318.7 499.3 39 9
0
39 .2 528.8 786.2 971.9 1620.1
0
.2 108.3 179.8 316.7 497.0 5
5
0
529.4 789.7 975.5 1624.1

average N
39
0
maximum N
39 108.0 178.8 315.2 495.9 .5
average M
maximum M
.5
0
39
39 .7 530.1 793.0 979.0 1626.2
0
.7 107.5 177.8 314.2 493.5 5
5
0
530.9 795.3 982.7 1630.1

0
average variation N
107.3 176.9 312.8 491.2 40
maximum variation N
40
average variation M
40
0
maximum variation M
40 .2 531.7 797.1 986.5 1633.3
0
.2 107.0 176.0 310.8 489.6 5
5
0
40 532.6 798.4 990.3 1638.1
0
40
.5 106.7 175.2 308.6 486.1 .5
0
40
40 533.4 799.5 994.1 1639.3
0
106.5 174.5 306.2 484.6 .7
.7 5
5
0
534.5 800.5 998.6 1648.3
0
41 106.7 173.9 303.8 491.0 41
41
0
41 535.5 801.8 1002.7 1647.4
0
106.5 173.2 301.6 479.7 .2
.2 5
5
0
41 536.6 803.0 1007.0 1656.3
0
41
.5 106.2 172.5 300.9 476.4 .5
41
0
41 537.7 804.5 1011.4 1654.7
0
106.2 171.9 300.5 474.1 .7
.7
5 5
0
538.9 806.2 1015.8 1652.6
0
42 106.1 171.4 300.7 471.0 42
0
42
42 540.2 807.2 1020.5 1660.2
0
.2 106.1 170.9 301.2 471.0 .2
5 5
0
541.6 808.1 1025.2 1657.6
0
42
42
.5 106.7 170.8 300.7 466.7 .
42 5
0
42 543.0 808.8 1030.0 1661.6
0
.7 106.9 170.7 299.7 463.2 .7
5 5
0
544.7 809.5 1035.0 1665.8
0
43 107.3 170.7 298.2 460.4 43
43
0
43
0
.2 107.7 170.4 296.2 457.8 .2 546.3 809.4 1040.2 1664.0
5 5
0
0
43 107.9 170.1 294.1 454.7
43
.5 548.0 808.8 1045.6 1664.9
.5 43
43 549.8 808.2 1051.0 1666.0
.7 108.2 170.2 292.5 451.3 .7
5
5
44 108.5 170.3 290.0 447.0 44 551.6 807.4 1056.4 1664.8
44 44
.2 109.0 170.4 286.9 450.2 .2 553.7 806.7 1062.3 1665.6
5 5
44 109.2 170.4 285.1 448.3 44
.5 555.9 806.7 1068.2 1667.1
.5
44 44
.7 109.4 170.1 283.8 441.5 .7 558.1 807.1 1074.1 1665.6
5 5
45 109.4 169.6 282.8 439.0 45 560.4 805.5 1080.3 1665.7
45 45
.2 110.1 169.7 280.9 437.3 .2 562.7 803.8 1086.7 1667.2
5 5
F-37
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
110.8 169.8 278.9 434.0 565.0 801.9 1093.2 1667.7

[kNm] [kN]
350
550
750
950
1150
1350
1550
0.00
50.00
100.00
150.00
200.00
250.00
300.00
350.00
400.00
450.00
500.00
598.0 848.8 1066.1 1573.5
117.7 227.9 385.6 437.6
cross angle []
cross angle []
598.3 846.4 1080.1 1592.7
38 39
114.3 221.1 370.6 423.9
38 39
599.6 842.1 1095.5 1609.1
40
111.5 214.6 359.4 418.3
40
15 hangers
602.1 839.4 1112.5 1620.7
41
15 hangers
109.1 208.7 350.1 407.4
41
606.1 837.8 1131.2 1633.0
42
106.9 203.2 340.8 397.8
42
561.8 799.5 1003.6 1552.0
38
102.4 198.0 342.2 435.4
38
561.7 799.8 1016.5 1571.8
39
99.8 193.0 331.1 428.6
39
563.0 802.0 1031.0 1586.2
40
97.5 188.2 320.9 422.6
40
16 hangers
16 hangers
566.0 805.8 1047.4 1596.2
41
95.2 183.9 311.7 414.9
41
570.4 808.7 1065.5 1610.5
42
93.4 180.3 305.0 406.3
42
530.4 752.7 949.0 1535.3
38
98.9 178.1 298.8 436.1
38
531.1 750.0 961.7 1542.3
39
96.6 173.8 289.3 430.2
39
532.9 747.1 975.9 1555.7
40
94.0 169.8 281.9 424.4
40
17 hangers
17 hangers
535.7 750.7 991.0 1567.6
41
91.6 166.4 273.8 412.5
41
539.5 759.5 1007.9 1579.7
89.7 163.2 267.2 406.6 42
42
522.0 742.9 966.5 1556.6
38
86.0 153.5 262.6 415.1
38
F.5 Results from the variation of the number of hangers and span
513.3 734.5 945.8 1545.1

39
84.2 148.6 260.4 419.3
39
Res ults from Section 6.7.1, num ber of hangers : 15, 16, 17, 18, s pan=75m , f=12.75m
504.5 726.2 925.1 1533.5

40
82.3 143.7 258.1 423.5
40
average M
average N
maximum M
maximum N
18 hangers
18 hangers
Results from Section 6.7.1, number of hangers: 15, 16, 17, 18, span=75m, f=12.75m
507.1 727.2 939.5 1536.7

41
80.9 140.9 256.0 413.4
41
F-38
average variation M
average variation N
79.1 138.5 251.6 403.4 510.6 725.9 955.6 1550.8

42
42
maximum variation M
maximum variation N
[kN] [kNm]
350
550
750
950
1150
1350
1550
1750
50.00
100.00
150.00
200.00
250.00
300.00
350.00
400.00
450.00
577.2 846.4 1048.0 1642.2 500.00
122.1 210.9 385.7 501.6
cross angle []
cross angle []
579.0 843.8 1062.5 1658.8
38 39
117.8 205.4 374.3 492.0
38 39
581.7 845.1 1078.3 1675.9
40
114.1 200.4 364.0 488.2
40
20 hangers
585.3 857.3 1095.5 1682.8
41
20 hangers
110.3 195.9 355.0 473.3
41
590.4 869.9 1114.8 1691.5
42
106.4 191.8 350.0 460.4
42
551.5 819.8 1001.3 1621.1
38
96.6 179.9 358.2 501.2
38
553.0 823.6 1015.2 1638.0
39
94.7 176.1 346.3 498.5
39
555.5 828.2 1030.1 1650.7
40
93.3 173.2 335.2 485.9
40
21 hangers
558.9 828.5 1046.7 1662.8
41
21 hangers
92.2 170.8 328.3 475.9
41
563.8 828.2 1065.4 1674.0
42
91.3 168.6 319.0 463.1
42
505.1 752.5 919.6 1583.2
38
93.8 161.3 301.4 505.4
38
507.2 754.9 932.4 1602.6
39
91.9 157.7 293.5 499.9
39
509.9 753.0 946.4 1617.2
40
90.0 154.3 284.3 491.6
40
23 hangers
23 hangers
513.9 758.5 961.9 1629.0
41
41 88.5 151.5 277.5 484.1
518.7 768.6 979.3 1641.0
42
86.8 148.7 274.3 472.7
42
484.8 726.2 883.4 1567.9
38
93.4 154.9 283.9 506.2
38
486.7 730.6 895.7 1582.8
39
91.1 151.0 275.2 500.2
39
Results from Section 6.7.1, number of hangers 20, 21, 23, 24, span 100m, f 17m
489.2 736.2 909.1 1597.3
40
89.3 147.8 268.1 493.0
40
average N
average M
maximum N
maximum M
24 hangers
24 hangers
493.0 742.8 924.3 1611.9
41
87.9 145.6 264.4 484.2
41
F-39
average variation N
average variation M
498.2 743.0 941.3 1626.1 86.4 144.0 258.1 476.1

42
42
maximum variation N
maximum variation M
[kNm] [kN]
350
550
750
950
1150
1350
1550
1750
20.00
120.00
220.00
320.00
420.00
520.00
620.00
579.1 904.1 1067.8 1869.9
135.0 226.8 454.2 679.6
cross angle []
cross angle []
583.8 908.4 1084.2 1862.2
38 39
132.0 221.4 441.9 650.0
38 39
588.3 903.1 1101.8 1850.9
40
128.9 215.8 428.2 621.9
40
29 hangers
594.4 919.8 1121.1 1850.8
41
29 hangers
127.2 211.4 418.8 615.8
41
602.1 934.6 1143.1 1873.0
42
125.3 208.2 411.9 607.7
42
542.8 839.3 1002.5 1804.3
38
123.0 204.8 401.3 661.4
38
547.0 858.2 1017.9 1801.3
39
119.2 200.0 392.3 633.3
39
552.3 868.1 1035.2 1798.3
40
116.3 196.3 384.7 623.9
40
31 hangers
31 hangers
557.9 874.9 1053.7 1815.2
41
114.7 192.6 379.4 623.1
41
565.2 878.1 1073.9 1837.9
42
112.6 189.8 373.6 614.2
42
512.4 802.4 945.5 1750.3
38
108.7 181.4 360.1 635.5
38
516.1 815.7 960.0 1752.8
39
107.4 177.4 353.7 630.7
39
106.3 175.0 347.5 629.2 40 521.3 828.9 976.6 1758.5
40
33 hangers
33 hangers
526.5 829.8 994.2 1788.4
41
106.8 174.9 338.0 627.0
41
532.1 824.1 1013.5 1820.0
42
106.6 174.6 333.1 623.5
42
485.6 770.9 895.7 1702.1
38
96.4 162.6 329.4 641.3
38
488.6 772.8 908.4 1714.1
39
94.7 158.4 322.2 636.9
39
Results from Section 6.7.1, number of hangers 29, 31, 33, 35, span 150m, f 25.5m
493.4 776.1 924.1 1737.4

40
94.0 157.8 315.0 633.8
40
Results from Section 6.7.1, number of hangers 29, 31, 33, 35, span 150m, f 25.5m
average N
average M
maximum N
maximum M
35 hangers
35 hangers
498.9 780.2 941.1 1768.8

41
96.0 159.5 306.8 633.4
41
F-40
average variation N
average variation M
99.4 162.8 301.1 635.8 504.2 789.1 959.4 1799.0
42
42
maximum variation N
maximum variation M
[kNm] [kN]
450
650
850
1050
1250
1450
1650
1850
2050
20.00
120.00
220.00
320.00
420.00
520.00
620.00
720.00
820.00
944.1
590.3 1086.9 2014.6
149.1 241.5 503.2 845.9 954.5
cross angle []
cross angle []
593.4 1103.2 2024.8
38 39
147.1 238.2 492.0 823.3
38 39
950.0
599.6 1122.2 2040.9
40
145.4 234.9 482.5 790.1
40
968.3
38 hangers
38 hangers
608.5 1143.8 2043.7
41
143.3 234.2 471.7 772.1
41
970.1
617.6 1167.3 2029.2
42
142.7 234.9 461.8 760.9
42
882.1
549.3 1013.2 1952.2
38
131.5 209.1 438.2 817.2
38
886.6
553.6 1028.4 1962.2
39
128.6 208.1 432.1 809.2
39
895.8
559.2 1045.3 1975.9
40
126.4 206.6 424.0 785.1
40
900.7
41 hangers
41 hangers
567.0 1065.7 1999.0
41
125.6 207.5 414.7 775.2
41
912.2
576.4 1087.8 1989.1
42
126.5 210.1 421.6 754.0
42
832.3
513.5 948.8 1886.4
38
119.0 190.2 388.9 776.8
38
835.4
517.6 963.4 1894.4
39
117.3 189.3 383.8 772.0
39
836.2
523.6 979.6 1915.2
40
117.5 189.8 379.0 766.2
40
846.4
44 hangers
44 hangers
530.0 997.5 1938.1
41
117.9 188.7 372.1 754.8
41
855.6
539.6 1019.0 1943.0
42
119.4 190.7 389.5 755.9
42
778.1
483.0 892.0 1832.7
38
107.3 177.7 348.7 777.6
38
793.8
486.5 906.3 1840.5
39
106.8 173.8 341.1 769.4
39
792.5
491.6 921.6 1866.6

40
105.6 171.7 336.8 761.8
40
average N
average M
799.1
maximum N
maximum M
47 hangers
47 hangers
497.7 938.4 1885.4

41
107.5 173.2 334.4 765.2
41
804.7
F-41
average variation N
average variation M
108.6 173.4 340.1 768.3 505.6 958.0 1913.9
42
42
maximum variation N
maximum variation M
Results from Section 6.7.1, span 100m, f 17m, cross angle 41

[kN]
0 0 0 0 0 0 0 2 2 2 2 2 4 4
2389.6
2156.7
2300
2080.9
maximum N
1987.7
average N
average variation N
1824.4
1759.6
1743.9
maximum variation N
1682.8
1647.4
1611.9
1800
1576.9
1550,8
1525.5
1523.5
1498.9
1476.0
1458.8
1347.7
1519.1
1208.6
1095.5
1300 1279.9
1002.7
1118.2
1163.5
924.3
1044.5
857.7
800.3
941.8
934
750.4
706.7
857.3
668.1
816.2
633.5
800 801.8
720.0
742.8
682.5
645.6
647.2
585.3
612.8 578.1
535.3
545.2523.3
493.0
457.3
426.0
399.1
375.6
354.9
336.3
300
20 24 28 32 36 40 44 48 52 56 60 64 68 72
number of hangers
Results from Section 6.7.1, span 100m, f 17m, cross angle 41

[kNm]
0 0 0 0 0 0 0 2 2 2 2 2 4 4
1352.2
1300.00
maximum M
1061.2
average M
1100.00 average variation M
943.9
maximum variation M
900.00
702.7
872.8
700.00
538.6
496.2
490.7
491.9
493.8
494.4
486.2
484.2
485.3
479.7
473.3
574.1
500.00
480.7
301.6
355.0
372.2
264.4
428.7
335.5
229.5
206.1
188.3
171.7
300.00
169.5
170.7
235.9
230.9
196.9
290.1
144.1
126.2
223.1
114.7
103.9
100.00
184.9
89.5
87.6
82.5
174.3
148.5
128.3
118.4
87.9
81.8
78.9
74.9
67.4
68.1
67.1
-100.00
20 24 28 32 36 40 44 48 52 56 60 64 68 72
number of hangers
F-42

Diplom Brunn Schanack

Hochgeladen von

Copyright:

Verfügbare Formate

Diplom Brunn Schanack

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Diplom Brunn Schanack

Hochgeladen von

Copyright:

Verfügbare Formate

TECHNISCHE

Berechnung einer zweigleisigen Eisenbahn-

Calculation of a double track railway network arch bridge

Benjamin Brunn Frank Schanack

Technische Universitt Dresden Dresden University of Technology

Grimstad, August 2003

Hiermit erklren die Autoren, dass die vorliegende Diplomarbeit vollstndig

Benjamin Brunn Frank Schanack

Grimstad, 18 August 2003

En este trabajo se presentan los resultados de las investigaciones realizadas sobre la

1 Task for the diploma thesis 1

4 What is a network arch? 6

6 Optimisation of the hanger arrangement 30

6.5 Preliminary investigations 33

7 Investigation of hanger connection details regarding fatigue strains 74

8 Erection of the bridge using a temporary lower chord 99

10 Criticism and proposals for future investigations 105

List of figures 107

List of standards 113

Annex A: Load assumptions A-1

Annex B: Preliminary design B-1

Annex C: FEM-calculations C-1

Annex D: Assessment of the bridge D-1

Annex E: Investigations on the temporary lower chord E-1

Annex F: Data from optimisation process F-1

Task for the diploma thesis

4. The following simplification to the first three points can be assumed:

Bridges are considered to be masterstrokes of engineering. Light, wide spanning structures

What is a network arch?

Network arch proposal 1980. 2

200 m Half lower chord A = 0.24 m 2 I = 0.27 m4

Bridge at Straubing built 1977.

Influence lines for bending

The bridge design

5.1 The arches

5.2 The hangers and hanger connections

solutions found to be best in Section 7 were 34

5.3 The wind bracing

members have a system length of 8.60 metres. For all

5.4 The bridge deck

5.4.1 Main design

Moveable supports 335 3900

Schematic illustration of post-tensioning

Fig. 5.3. Cross section of the bridge at mid-span

12, s = 20 cm Transverse tendons, DYWIDAG

5.4.2 Alternative design proposals (without transverse prestressing)

Longitudinal tendons, 25, s = 9 cm

Fig. 5.5. Cross section with reinforcement of alternative design 1

Longitudinal tendons, 25, s = 14 cm in two layers

Fig. 5.6. Cross section with reinforcement of alternative design 2

Effects on the edge beams

Effects on other structural elements

Comparison of costs for different bridge deck designs

Nonetheless, durability reasons might necessitate prestressing in order to reduce or even

5.5 The end cross girder

5.6 Constructive design of the arch root point

5.6.1 Anchorage of the arch

The occurring shear forces in the arch are small and

pressure around the steel construction of the arch root and