This action might not be possible to undo. Are you sure you want to continue?

**service d'Études techniques des routes et autoroutes
**

october 2006

Technical guide

Footbridges

Assessment of vibrational behaviour of footbridges under pedestrian loading

5 q q

1

§

¦

%

©

©

u

§

¦

£

¥

¤

¡

£

¢

¤

¥

¦

¥

¤

¦

¤

¦

¢

©

(

!

$

¢

T

¢

u

!

¤

¦

¦

§

¦

%

©

©

¦

§

¦

£

¥

¤

¡

£

¢

¤

¥

¦

`

¥

¤

¦

¤

¦

¥

¦

'

"

¢

¡

5

©

(

¤

¥

U

2

§

¦

¤

¥

¦

¤

¢

¤

¥

¤

¦

¢

¥

¥

¡

(

¦

#

¤

¦

¤

¦

¤

¥

¥

£

¤

¥

¦

`

¥

¦

¤

¦

¤

¦

¢

©

(

!

¢

¡

¤

¥

¦

©

¢

U

¦

4

5

¢

¥

¤

(

£

¢

¡

¤

¥

¢

¥

¢

©

a

¢

¤

¦

¤

¥

¥

¡

(

¡

¤

¥

¦

¢

§

¢

¤

¥

£

§

¢

2

#

T

¢

¢

¤

£

¤

¦

¢

£

¤

¢

¢

¤

£

¤

¥

¤

¥

¦

¤

¢

¢

©

¤

¥

¦

¢

©

0

£

§

¦

¢

¦

§

¥

T

W

¡

(

¡

¦

©

¢

¤

(

£

¤

¥

¢

¥

¡

¥

2

£

¢

X

©

¤

¦

¢

£

¤

¢

¥

¢

©

a

¢

¥

¢

¤

¦

¡

£

a

¢

#

§

¦

§

(

¢

¦

'

"

¢

¡

t

¤

¢

©

§

¢

%

¢

¨

¢

§

T

¦

¤

¥

¦

(

"

¤

¦

#

§

£

!

¤

¢

¦

©

¢

¢

¡

¢

U

¥

§

¤

¢

¦

©

¢

¨

Y

1

g

h

s

0

¢

£

¤

¦

g

¢

¢

§

T

¦

¥

%

¦

h

¢

¥

s

$

U

2

¢

h

#

¦

2

¢

¦

e

¡

£

¤

¢

g

¤

¦

1

g

g

i

¢

£

¤

¦

g

¢

'

¢

g

(

¦

¢

'

$

#

¤

¦

©

r

U

2

¢

h

§

¥

¦

¡

£

¤

¢

g

¦

¡

£

(

¤

¥

¦

`

¥

¤

¦

¢

¡

Y

¦

¢

©

#

¦

¤

¥

¢

£

¤

£

#

¦

2

Y

¤

¥

¦

`

¥

¤

¦

§

¦

¤

¥

¢

©

¤

¦

¢

¥

¤

¦

©

£

1

!

!

r

i

¤

¢

¢

©

¥

(

)

!

&

§

¢

¢

(

)

¥

¤

¡

£

¢

¢

(

'

&

¢

¤

¢

r

$

¢

£

¥

q

¤

¥

¢

¢

¤

¥

¤

!

§

¦

¤

¥

¢

U

e

£

¥

§

T

(

p

¢

¡

¦

§

(

£

¥

¦

©

¤

¥

1

"

i

¤

¢

¢

©

¥

(

)

!

&

§

¢

¢

(

)

¥

¤

¡

£

¢

¢

¢

(

)

¥

¥

¤

¢

¥

£

"

(

¦

¢

'

$

U

2

¢

h

§

¦

£

¥

¤

¡

£

¢

¤

¦

£

¥

¥

¤

¢

¥

£

"

Y

¤

¥

¦

`

¥

¤

¦

¡

£

¦

¢

¢

Y

¢

£

¥

%

¢

W

§

¦

¤

¥

¦

¢

©

V

¤

¥

£

¦

¥

#

¥

¤

(

£

¦

¡

£

¤

¢

g

¢

¡

¢

#

¦

§

©

¢

¡

§

§

¦

¢

¦

¥

£

¦

¦

'

"

¢

¡

F

P

7

D

B

8

S

A

B

C

F

S

P

f

B

G

R

C

b

C

B

A

@

9

8

7

6

W

5

#

¥

¤

(

£

§

¦

£

¥

¤

¡

£

¢

¢

¡

¤

¥

¦

¤

¥

¤

¦

¤

¥

¤

¥

¦

¢

¡

¢

(

T

¥

¤

£

W

Y

¢

¦

2

¢

¥

(

§

¦

£

¥

¤

¡

£

¢

¡

(

¡

¢

£

¥

£

¦

©

¤

¦

¢

¢

§

2

¤

U

©

§

¢

%

¢

W

Y

§

(

¢

£

¥

¥

¤

¢

¥

£

¢

¡

¤

¦

¢

©

§

¦

£

¥

¤

¡

£

¢

¢

¦

(

§

¦

%

¢

W

Y

¢

£

¤

¢

¥

¢

©

a

¢

¢

¤

¦

¡

£

a

¢

¢

¡

¢

W

Y

¦

¢

©

¤

¦

¢

¦

¤

¦

£

¦

¤

£

¢

¥

©

¤

¦

¢

¤

2

£

¢

X

©

¢

¢

¤

¢

¡

¤

(

£

£

¦

¤

¥

¢

U

¦

V

¦

'

"

¢

¡

¤

¢

©

§

¢

%

¢

¢

§

T

¦

¤

¥

¦

(

#

¥

§

¦

¢

¤

¥

©

¤

¦

§

¢

¥

¢

¡

¤

¥

¢

(

£

(

¤

¥

¢

¢

¤

¥

¤

¢

¤

¦

¦

¢

¡

¤

#

¢

T

¢

§

¦

£

¢

¦

§

#

¢

%

¦

¤

¥

¡

¥

e

A

B

C

8

7

A

d

G

8

A

C

A

D

8

7

A

d

G

B

8

Q

E

c

G

B

F

b

C

B

A

@

9

8

7

6

W

5

£

¢

X

©

a

¢

§

©

£

¤

¢

¥

¢

©

a

¢

¤

¦

¤

¥

T

W

Y

¤

¥

¦

`

¥

¦

¤

¦

¤

¦

¢

©

(

!

U

2

¢

¦

¢

¡

¤

¥

¡

¥

2

¦

§

(

£

¥

¦

©

¤

¥

¢

¢

¤

¥

£

¥

§

T

(

©

¢

¡

¦

2

(

©

W

Y

¦

¢

¡

¢

¦

¤

¦

U

2

¢

¤

¦

¢

¡

¢

£

¤

¢

¢

¡

£

¢

¡

¢

¢

¤

¦

¦

(

W

Y

¢

¥

£

¥

§

©

£

¥

§

T

(

©

¢

¡

¢

¦

(

§

¦

%

¢

¤

¦

#

§

©

©

¦

¢

¤

¥

¢

W

Y

¤

¢

¢

¦

¤

¦

¢

¦

¢

¡

¡

¥

2

©

§

¢

¡

W

Y

#

¥

§

¦

(

)

£

¢

X

©

¢

¡

¢

%

©

¥

V

¢

¤

¥

U

2

¢

¤

¢

¡

¢

¥

¥

£

¥

¥

£

¢

©

¢

¡

¢

¥

(

§

¤

¦

¢

¥

§

¡

¢

¤

¥

¦

¤

¥

¡

¥

2

1

¢

¥

¥

¤

(

£

¤

¦

T

(

¤

¦

¢

¥

¥

£

¢

¦

§

¢

¥

¤

(

£

$

¢

¥

¥

¤

(

£

§

¦

£

§

¤

¦

¢

¥

£

¤

¢

¦

¢

¦

"

¢

¥

§

©

©

(

¦

'

"

¢

¡

B

8

S

R

G

Q

P

I

H

E

F

8

7

A

D

A

B

G

F

F

E

D

C

B

A

@

9

8

7

6

5

¢

(

£

(

¤

¥

¢

¢

¤

¥

¤

¢

¢

¡

¤

¦

¤

¥

¦

©

¤

¦

¢

¡

¦

¢

¡

¥

¢

¥

¥

%

¥

£

¦

§

¢

¥

4

5

¢

(

£

(

¦

¤

4

¤

¦

©

¤

¦

#

¥

¤

¥

3

¢

¡

¤

¥

¡

¥

2

¤

¢

¦

©

¢

§

¦

£

¥

¤

¡

£

¢

¦

¥

1

¦

'

"

0

¢

(

(

¦

¢

¢

(

¢

¢

(

)

¥

¤

¡

£

¢

¢

(

'

&

¢

£

¥

%

¢

"

$

#

¢

¦

"

¦

¤

¦

¤

¥

¢

¢

¤

¥

¤

!

¢

¥

¤

¦

¦

©

¤

¦

¤

¢

¦

©

¢

¨

§

¦

£

¥

¤

¡

£

¢

¢

¡

Technical guide

Footbridges

Assessment of vibrational behaviour of footbridges under pedestrian loading

Published by the Sétra, realized within a Sétra/AFGC (French association of civil engineering) working group This document is the translation of the work "Passerelles piétonnes - Evaluation du comportement vibratoire sous l’action des piétons", published in under the reference 0611

Association Française de Génie Civil Régie par la loi du 1er juillet 1901 28 rue des Saints-Pères - 75007 Paris - France téléphone : (33) 01 44 58 24 70 - télécopie : (33) 01 44 58 24 79 mél : afgc@enpc.fr – internet : http://www.afgc.asso.fr

....................................... 54 4.................. 31 2............................................................ 43 DYNAMIC CALCULATION APPLIED TO FOOTBRIDGES .....6 x d w x w g h g w e 3............................................................................................ 10 1..2 WALKING ................................................................................ 7 1........................................................................................................ 31 2....................................3 PARAMETERS THAT AFFECT DIMENSIONING: FREQUENCY... 25 IMPROVEMENT OF DYNAMIC BEHAVIOUR ..................................... 40 2... COMFORT CRITERION...... 54 4.......................2 1.................. 37 STRUCTURAL CHECKS UNDER DYNAMIC LOADS ....... 76 DISCRETISATION OF THE CONTINUOUS SYSTEMS ....................1 1................................................1 EXAMPLES OF ITEMS FOR A FOOTBRIDGE DYNAMIC DESIGN SPECIFICATION ............ ETC............. ............................3 y x v h l x y g g w 1.. 27 1..........................................................................................................................................................2 EXAMPLES OF ITEMS TO BE INSERTED IN THE PARTICULAR WORKS TECHNICAL SPECIFICATION FOR A FOOTBRIDGE .......3 1............................................................. 82 y x d d 2...................................... COMFORT THRESHOLD...... 32 STAGE 4 IF NECESSARY: CALCULATION WITH DYNAMIC LOAD CASES................................Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 1 / 127 Contents f m k k m y x y x y w j y d d x y w x w x x e v x w e y v g g v x y e i e y x w y y x w v 1..................2006 x v w y y x x y x ....... 44 4................................................................................ 54 DYNAMIC TESTS OR TESTS ON FOOTBRIDGES .........................4 v h l x y g g w A SIMPLE OSCILLATOR .............. 34 STAGE 5: MODIFICATION OF THE PROJECT OR OF THE FOOTBRIDGE.1 STRUCTURE AND FOOTBRIDGE DYNAMICS .......2 STAGE 2: CHOICE OF COMFORT LEVEL BY THE OWNER ..................................................................1 STAGE 1: DETERMINATION OF FOOTBRIDGE CLASS .2 PEDESTRIAN LOADING .................................................................... 58 LINEAR SYSTEMS AT N DOF ............Footbridges .....................................1 2.......................2 y w y x h g w e PRACTICAL DESIGN METHODS ....3 STAGE 3: DETERMINATION OF FREQUENCIES AND OF THE NEED TO PERFORM DYNAMIC LOAD CASE CALCULATIONS OR NOT ..........................5 2.................................... 67 CONTINUOUS ELASTIC SYSTEMS ......1 3.................. 91 © Sétra ... .................... 87 RUNNING ...........................4 d w y w x v w y x 2........4 2........................

....................................................................... 124 REGULATIONS .3 6.................................... 103 RIBBED SLAB: NOISY-LE-GRAND FOOTBRIDGE ..... 104 BOW-STRING ARCH: MONTIGNY-LÈS-CORMEILLES FOOTBRIDGE ............................2 6...............5 4......... 126 © Sétra ............................... 97 TUNED LIQUID DAMPERS .................. 94 VISCOUS DAMPERS .....2 4........... 126 ADDITIONAL BIBLIOGRAPHY ............................................................................. 126 ARTICLES ON EITHER VIBRATING OR INSTRUMENTED FOOTBRIDGES ................................................................................................................................1 6......... 101 4... 124 CALCULATION METHODS .............................................5 6..............4 3............ 109 5.......................4 6..................................................................................... 103 STEEL BOX-GIRDER: STADE DE FRANCE FOOTBRIDGE ............................................................1 4.................................................................... 124 BEHAVIOUR ANALYSIS .............................................8 GENERALLY .............2006 n v y x g v w h l x y g g w ............................3 4.......... 106 CABLE-STAYED CONSTRUCTION: PAS-DU-LAC FOOTBRIDGE AT ST QUENTIN ........2 3............................................................ 124 DAMPERS ...8 f x y x w d d w d g v w l h k l x y g g w k WARREN-TYPE LATERAL BEAMS: CAVAILLON FOOTBRIDGE ....................... ...............3 3... 95 TUNED MASS DAMPERS (TMD) .......................7 6................ 106 STEEL ARCH: SOLFERINO FOOTBRIDGE ............................................................4 4.....................................................................................................................7 4......6 4.... 100 COMPARATIVE TABLE ...............................................................Footbridges ................................ 108 MIXED CONSTRUCTION BEAM: MONT-SAINT-MARTIN FOOTBRIDGE ...................................5 f x d g v w l h l x y g g w VISCO-ELASTIC DAMPERS ......2 e g w x d x h l y y w EXAMPLES OF COMPLETE CALCULATIONS OF FOOTBRIDGES ............... 126 SPECIFIC FOOTBRIDGES ..................................................... 119 6......1 5.... 110 SENSITIVITY STUDY OF TYPICAL FOOTBRIDGES.........................................1 3............................................... 105 SUSPENDED CONSTRUCTION: FOOTBRIDGE OVER THE AISNE AT SOISSONS ...................................Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 2 / 127 3.........6 6...............

Footbridges . w(t): displacements (m) x(t): position of a simple oscillator referenced to its balance position (m) : reduced pulsation : logarithmic decrement th i : i critical damping ratio of an oscillator with n degrees of freedom (no unit) : density (kg/m3) [ i]: ith eigen vector ( ): argument of Hr.phase angle between entrance and exit (rad) =2 f 0 : natural pulsation of a simple oscillator (rad/s): th i : i natural pulsation of an oscillator with n degrees of freedom (rad/s): i =2 : pulsation (rad/s) fi © Sétra . e . Main notations A( ): dynamic amplification [C]: damping matrix Ci: damping No. i in a system with n degrees of freedom (kg) S: cross-sectional area of a beam (m2) [X]: vector of the degrees of freedom Ci: generalised damping of mode i f : natural frequency of a simple oscillator (Hz) fi: ith natural frequency of an oscillator with n degrees of freedom (Hz) f: frequency Ki: generalised stiffness of mode i Mi: generalised mass of mode i n: number of pedestrians [p(t)]: modal participation vector [q(t)]: modal variable vector t: variable designating time (s) u(t). i in a system with n degrees of freedom (N/(m/s)) E: Young's modulus (N/m2) F(t): dynamic excitation (N) [F(t)]: dynamic load vector F0: amplitude of a harmonic force (N) [F0 ]: amplitude vector of a harmonic force (N) Hr.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 3 / 127 1. v(t). i in a system with n degrees of freedom (N/m) L: length of a beam (m) [M]: mass matrix Mi: mass No. e ( ): transfer function input e and the response r I: inertia of a beam (m4) J: torsional inertia of a beam (m4) [K]: stiffness matrix Ki: stiffness No.2006 .

These tests were funded by the Direction des Routes (Highway Directorate) and managed by Sétra. de l'Aménagement du territoire.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 4 / 127 2. as well as structural response to loading. An analytical methodology and recommendations are also proposed to guide the designer of a new footbridge when considering the resulting dynamic effects. des Transports.recommendations for the drafting of design and construction documents.LERMES) (DREIF and then Sétra) (Setec-TPI) (RFR and then ADP) (LCPC) (DREIF) (SNCF) (Sodeteg and then Arcadis-EEGSimecsol) (OTUA and then MIO) (University of Lyon 1 .a presentation of the practical methods for calculation of natural frequencies and modes. This document covers the following topics: . The methodology is based on the footbridge classification concept (as a function of traffic level) and on the required comfort level and relies on interpretation of results obtained from tests performed on the Solférino footbridge and on an experimental platform.2006 .a description of the dynamic phenomena specific to footbridges and identification of the parameters which have an impact on the dimensioning of such structures. these phenomena do not have adverse effects on structure. These guidelines were drafted by: Valérie BONIFACE Vu BUI Philippe BRESSOLETTE Pascal CHARLES Xavier CESPEDES François CONSIGNY Christian CREMONA Claire DELAVAUD Luc DIELEMAN Thierry DUCLOS Wasoodev HOORPAH Eric JACQUELIN Pierre MAITRE Raphael MENARD Serge MONTENS Philippe VION Actions exerted by pedestrians on footbridges may result in vibrational phenomena. Introduction These guidelines were prepared within the framework of the Sétra/AFGC working group on "Dynamic behaviour of footbridges". du Tourisme et de la Mer" and specifically by the DREIF (Division des Ouvrages d’Art et des Tunnels et Laboratoire Régional de l’Est Parisien). with the support of the scientific and technical network set up by the French "Ministère de l'Equipement.Footbridges . . . (RFR) (Sétra) (CUST . In general. led by Pascal Charles (Ile-de-France Regional Facilities Directorate and then Sétra) and Wasoodev Hoorpah (OTUA and then MIO). These guidelines round up the state of current knowledge on dynamic behaviour of footbridges under pedestrian loading. .a methodology for the dynamic analysis of footbridges based on a classification according to the traffic level.L2M) (Socotec) (OTH) (Systra) (Sétphenomenaa) © Sétra . although the user may feel some discomfort.

Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 5 / 127 Supplementary theoretical (reminder of structural dynamics. examples of recent footbridges. pedestrian load modelling) and practical (damping systems. typical calculations) data are also provided in the guideline appendices. © Sétra .Footbridges .2006 .

in this case the structure will be more slender and slim-line.Footbridges . Dynamic behaviour of footbridges under wind loads is not covered in this document.e. Two footbridges located in the centre of Paris and London. French regulations do not provide any indication concerning dynamic and vibrational phenomena of footbridges whereas European rules reveal shortcomings that have been highlighted in recent works. thus the structure will either be sturdy and probably unsightly. Maximum comfort does not tolerate any footbridge vibration. as concerns structural strength. this resulted in the need for thorough investigations and studies of their behaviour under pedestrian loading. © Sétra . These studies involved in situ testing and confirmed the existence of a phenomenon which had been observed before but that remained unfamiliar to the scientific and technical community.2006 . Minimum comfort allows moderate and controlled footbridge vibrations. These guidelines concern the normal use of footbridges as concerns the comfort criterion. parades. or slender but fitted with dampers thus making the structure more expensive and requiring sophisticated maintenance. balls. Dynamic analysis methodology is related to footbridge classification based on traffic level. it was considered useful to issue guidelines summarising the dynamic problems affecting footbridges. inaugurations etc. The part played by footbridge Owners is vital: they select the footbridge comfort criterion and this directly affects structural design. demonstrations. These guidelines are not meant to guarantee comfort when the footbridge is the theatre of exceptional events: marathon races. This means that footbridges in an urban environment are not treated in the same way as footbridges in open country. This phenomenon referred to as "forced synchronisation" or "lockin" causes the high amplitude transverse vibrations experienced by these footbridges. and take vandalism into consideration. In order to provide designers with the necessary information and means to avoid reproducing such incidents.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 6 / 127 Foreword This document is based on current scientific and technical knowledge acquired both in France and abroad. more aesthetically designed in general and possibly fitted with dampers. i. This document is a guide: any provisions and arrangements proposed are to be considered as advisory recommendations without any compulsory content. were closed shortly after inauguration: they showed hampering transverse oscillations when carrying a crowd of people.

This oscillator is supposed to move only by translation in a single direction and therefore has only one degree of freedom (herein noted "dof") defined by position x(t) of its mass.1).1. . As a consequence they more frequently require a thorough dynamic analysis. 3. The study of a basic model referred to as simple oscillator illustrates the dynamic analysis principles and highlights the part played by the different structural parameters involved in the process. directly useful to designers are mentioned here. dynamic loads are time-related and may be grouped in four categories: .pulsing loads corresponding to very brief loads.2 .periodically recurrent loads integrally repeated at regular time intervals referred to as periods.Footbridges . On the other hand.1 .harmonic or purely sinusoidal loads. intensity. It also presents the various tests carried out and the studies leading to the recommendations described in the next chapter.2006 .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 7 / 127 3. 3. impacted by an external force F(t) (figure 1. . footbridges follow this general trend and they are currently designed and built with higher sensitivity to strain.1. One of the main features of the dynamic loading of pedestrians is its low intensity.Simple oscillator The simple oscillator consists of mass m. Applied to very stiff and massive structures this load could hardly make them vibrate significantly. Generally speaking pedestrians loads are time-variant and may be classified in the "periodic load" category. Footbridge dynamics This chapter draws up an inventory of current knowledge on the dynamic phenomena affecting footbridges. technical and technological developments lead to ever more slender and flexible structures. . Detailed calculations are provided in Appendix 1.random loads showing arbitrary variations in time. direction…. Only the main results. However. © Sétra .Structure and footbridge dynamics 3. Appendix 1 to these guidelines gives a more detailed presentation of these results and deals with their general application to complex models. connected to a support by a linear spring of stiffness k and a linear damper of viscosity c. aesthetic.1 .General By definition static loads are constant or barely time-variant (quasi-static) loads.

unit is therefore expressed in kilograms. In c 2 km practice. The resonance phenomenon is particularly clear when the simple oscillator is excited by a harmonic (or sinusoidal) under the form F0 sin ( t ). the critical damping ratio can only be assessed. has a value that is always less than 1. by definition. welding).2006 . fo being the natural frequency (Hz). prestressed concrete). Damping has various origins: it depends on materials (steel.Footbridges . If.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 8 / 127 Figure 1. timber) whether the concrete is cracked (reinforced concrete. its S. : critical damping ratio (dimensionless) or critical damping percentage. Some of these curves are provided in figure 1. the steel jointing method (bolting. Since m 0 k m is a mass.1: Simple oscillator The dynamic parameters specific to this oscillator are the following: = 2 fo: natural pulsation (rad/s). It may be represented by a set of curves parameterised by .2 for a few specific values of the critical damping ratio. It should be noted that until experimental tests have been carried out.I. Dynamic amplification is obtained as a function of and . concrete. These curves show a peak for the value of R 1 2 2 characterising the resonance and therefore corresponding to the resonance pulsation © Sétra . its static response obtained with a constant force equal to F0 is: F F 0 0 /m x statique 2 k 0 the dynamic response may be amplified by a factor A ( ) and is equal to: xmax where 0 xstatique A ( ) is the reduced (or relative) pulsation and A( ) (1 2 1 )2 4 2 2 is the dynamic amplification.

Resonance amplification being directly related to damping.Complex systems The study of real structures. The new item with regard to the simple oscillator is the natural vibration mode defined by the pair constituted by the frequency and a vibratory shape ( i. may be considered as the study of a set of n simple oscillators.Footbridges . each one describing a characteristic vibration of the system. In this case the response is 1 2 2 and to the resonance frequency f R R 0 2 higher (or even much higher) than the static response. with a damping force proportional to speed). However. when modelling the system. all precautionary measures required for model analysis applications.2006 . Thus. we may consider that resonance occurs for = 0 and that amplification equals: A( R 1) 1 2 Figure 1. this is the assumption selected by most footbridge designers and engineers.3 . © Sétra . 3. which are generally continuous and complex systems with an important number of degrees of freedom. load F(t) = F0 sin ( 1 t) can generate displacements or stresses very much lower than load F(t) = (F0/10) sin ( 2 t) which however has an amplitude 10 times weaker. It should be noted that resonance does not occur for = 0 but for R . if this second load has a frequency much closer to the resonance frequency of the structure. Admitting that structural damping is weak in practice. it is therefore necessary to estimate this parameter correctly in order to obtain appropriate dynamic dimensioning. Computation of natural vibration modes is relatively intricate but designers nowadays have excellent software packages to obtain them. The approximation methods allowing such a conversion are detailed in Appendix 1.2: Resonance curves Dimensioning of the structures based on dynamic loading cannot be made using only the maximum intensity of the load impact.1. provided that they take. It should be noted that the simple oscillator study relies on the hypothesis of linear damping (viscous. i ) of the system. Practical use of natural vibration modes is dealt with in chapter 3. which is one damping type among others. for example.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 9 / 127 R .

but the experimental measurements performed show that it is periodic. The periodic function F(t).1 provides the estimated frequency values. the load of a pedestrian moving at constant speed v can therefore be represented as the product of a time component F(t) by a space component (x – vt). : walking frequency. : i-th harmonic amplitude.4 G0. depending on authors). physiological characteristics and apparel. Noting that x is the pedestrian position in relation to the footbridge centreline. 2 © Sétra . Designation Walking Running Specific features Continuous contact with the ground Discontinuous contact Frequency range (Hz) 1. that is n = 3. around 1. the coefficient values of the Fourier decomposition of F(t) are the following (limited to the first three terms. etc. that is a constant part increased by an infinite sum of harmonic forces.9 Hz. In any case.1 Conventionally. around 2 Hz (fm = 2 Hz) for vertical action. : first harmonic amplitude.1 . that is: P ( x. characterised by a fundamental parameter: frequency.20 Hz standard deviation (from 0.Footbridges . = 3 /2. may therefore be resolved into a Fourier series. t ) Several parameters may also affect and modify this load (gait. the coefficients of the higher of the terms being less than 0. 3. At mean frequency.1 G0): G1 = 0. has been studied rather thoroughly (see Appendix 2) and is translated as a point force exerted on the support.22.Pedestrian loading 3. being the Dirac operator. ground roughness.).5 F (t ) ( x vt ) Table 1.1 G0. Table 1. whether walking or running. the main conclusions resulting from the simple oscillator study can be generalised to complex systems. that is the number of steps per second. frequency may be described by a gaussian distribution with 2 Hz average and about 0. : phase angle of the i-th harmonic in relation to the first one.Effects of pedestrian walking Pedestrian loading. for normal walking (unhampered). n F (t ) i 2 G0 G0 G1 Gi fm i G1 sin 2 f m t Gi sin(2 if m t i ) with n : static force (pedestrian weight for the vertical component). The mean value of 700 N may be taken for G0.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 10 / 127 It should be emphasized that in some cases the problem can even be solved using a single simple oscillator.6 to 2. Recent studies and conclusions drawn from recent testing have revealed even lower mean frequencies. The sum of all unitary contributions of the terms of this sum returns the total effect of the periodic action.1. as a function of time and pedestrian position.4 2 to 3.2006 .2 .175 to 0. weight of one pedestrian. : number of harmonics taken into account.2.8 Hz . G2 = G3 0.

4 G0 sin(2 f m t ) Transverse horizontal component of one-pedestrian load: fm Fht (t ) 0. based on pedestrian behaviour studies. the transverse load frequency is equal to half the frequency of the vertical and longitudinal load.3 . for one same walk. in addition. footbridges are submitted to the simultaneous actions of several persons and this makes the corresponding dynamic much more complicated. One can merely set out reasonable and simplifying hypotheses. this behaviour can hardly be submitted to software processing. this load case does not cover exceptional events such as a marathon race which must be studied separately. This load case.2 G0 sin( 2 f m t ) It should be noted that.2. the pedestrian will modify his natural and normal gait in several ways. which may be very dimensioning in nature. should not be systematically retained.Effects of pedestrian running Appendix 2 presents the effects of pedestrian running. frequency. It is therefore very difficult to fully simulate the actual action of the crowd. Moreover. 3.05 G0 sin 2 t 2 Longitudinal horizontal component of one-pedestrian load: Fh l (t ) 0. possibly weighted by a minus factor. among others and facing a situation different to the one he expects. on the other. Very often. the crossing duration of joggers on the footbridge is relatively short and does not leave much time for the resonance phenomenon to settle. Moreover.2 .Footbridges .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 11 / 127 By resolving the force into three components. according to the number of persons present on the bridge. running effects shall not be considered in these guidelines. on the one hand. the transverse load period is two times higher than the vertical and longitudinal load and therefore the frequency is two times lower.2. this annoys the other pedestrians over a very short period. 3. As a result. the problem induced by intelligent human behaviour is such that. and possibly with the footbridge. Accordingly. pedestrians will generate loads which are more or less synchronous with each other.Random effects of several pedestrians and crowd In practice. speed) and. and then assume that the crowd effect is obtained by multiplying the elementary effect of one pedestrian. that is. This is due to the fact that the load period is equal to the time between the two consecutive steps for vertical and longitudinal load since these two steps exert a force in the same direction whereas this duration corresponds to two straight and consecutive right footsteps or to two consecutive left footsteps in the case of transverse load since the left and right footsteps exert loads in opposite directions. the following values of such components may be selected for dimensioning (in practice limited to the first harmonic): Vertical component of one-pedestrian load: Fv (t ) G0 0. Various ideas exist as concern crowd effects and they antedate the Solférino and Millenium incidents. each pedestrian has its own characteristics (weight. a «vertical» component and two horizontal components (one in the «longitudinal» direction of the displacement and one perpendicular to the transverse or lateral displacement). In fact. These concepts are presented in the © Sétra . Added to these.2006 . there are the initial phase shifts between pedestrians due to the different moments when each individual enters the footbridge.

Even the above-addressed N model has some deficiencies. The model finally selected consists in handling pedestrians' moves at random frequencies and phases. It is better to suggest simple elements to be improved on subsequently. © Sétra . It rapidly appears that knowledge of crowd behaviour is limited and this makes the availability of practical dimensioning means all the more urgent.34 – 0. This result can be demonstrated by considering a crowd with the individuals all at the same frequency with a random phase distribution. or in phase and at the natural frequency of the footbridge will produce the same effect as random pedestrians. but comprises a deficiency since it works on the assumption that all pedestrians are moving at the same frequency.09 log(np) for Np > 10 1 = 0. This result takes into consideration the phase shift between pedestrians. without any particular synchronisation) which enter a bridge at a rate of arrival (expressed in persons/second) the average dynamic response at a given point of the footbridge submitted to this pedestrian flow is obtained by multiplying the effect of one single pedestrian by a factor k T . [24]) propose a sparse crowd loading model on the first term of a Fourier representation the coefficient 1 of which depends on the number Np of persons present on the platform (for a 2 Hz walking frequency): for Np <10 1 = 0. using measurements made on instrumented platforms where small pedestrian groups move. Until recently dynamic dimensioning of footbridges was mainly based on the theoretical model loading case with one single pedestrian completed by rather crude requirements concerning footbridge stiffness and natural frequency floor values.25 Unfortunately. In fact. such requirements are very insufficient and in particular they do not cover the main problems raised by the use of footbridges in urban areas which are subject to the action of more or less dense pedestrian groups and crowds.2006 . Ebrahimpour & al. Several researchers have studied the forces and moments initiated by a group of persons. this means that n pedestrians present on a footbridge are equivalent to N all of them being synchronised. on a footbridge presenting different modes and in assessing each time the equivalent number of pedestrians which when evenly distributed on the footbridge. (Ref. Therefore several crowd load cases have been developed using probability calculations and statistical processing to deepen the random crowd issue. this model does not cover the cumulative random effects. rather than remaining in the current knowledge void. T being the time taken by a pedestrian to cross the footbridge (which can also be expressed by T = L/v where L represents the footbridge length and v the pedestrian speed). For a large number of independent pedestrians (that is. Practically. due to their different entrance time.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 12 / 127 following paragraphs together with a more comprehensive statistical study which will be used as a basis for the loadings recommended in these guidelines. this product T presents the number of N pedestrians present on the bridge at a given time. Obviously.Footbridges .

5m/s speed) is noted and the equivalent number of pedestrians which would be perfectly synchronised is calculated. The method used is explained in figure 1.2006 .175 Hz standard deviation.3. the maximum acceleration over a sufficiently long period (in this case the time required for a pedestrian to span the footbridge twice at a 1. a random phase and a normally centred around the natural frequency of the footbridge distributed random frequency f 2 and with 0.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 13 / 127 Several digital tests were performed to take into consideration the statistical effect. © Sétra .Footbridges . are selected. For each test including N pedestrians and for each pedestrian.

fixed damping and a fixed number of mode antinodes. with passerelle). +p -p +p Amax essai Amax (Neq) Amax test == Amax (Néq) donne N yields Neq éq Figure 1. d'écart-type 0.5m/s sur la footbridge passerelle random frequencies (see the distribution law opposite) and fréquences aléatoires totally random phases (voir loi de distribution ci-contre) et densité de probabilité pour les fréquences : Loi normale . moyenne 2Hz .2006 Accélération maximale = Amax (Néq) Maximum acceleration = Amax (Néq) ¨ ¡ ¡ £ ¥ ¡ ¤ ¡ ¥ ¡ © ¤ ¥ ¡ § ¥ ¨ ¥ ¨ ¡ ¥ ¥ ¥ § ¥ £ ¥ £ ¤ ¤ £ ¡ § ¦ £ © ¦ £ ¡ £ ¥ £ ¨ £ £ § ¤ ¥ § ¦ ¤ ¡ ¡ ¥ ¦ ¢ ¡ ¨ ¤ £ § ¢ ¦ ¨ § ¢ £ ¥ ¡ ¢ ¤ £ ¤ ¢ ¡ ¡ P =PNéqNéq / Longueur totale x cos( 2 xfcos( 2 ) f0 t + = / Total length x 280N 280N 0 t + ) w | w { w z w y w x w w v w w ~ v w } v w | v w { v w z v w y v w x v q o q o r r o p o p p p o o phases complètement aléatoires o o q o q o q o u p u p t p t p s p s p o o o o o o Random pedestrians circulating at speed v = 1. then the characteristic value. à la même the same frequency (that fréquence (celle de la of the footbridge). such as 95% of the samples © Sétra . au signe du mode. dont the loading sign related to the le signe du chargement est lié mode sign.5m/s on the Piétons aléatoires circulant à la vitesse V = 1. et en phase.3: Calculation methodology for the equivalent number of pedestrians Neq These tests are repeated 500 times with a fixed number of pedestrians.175 Hz . and in phase.Footbridges . at Piétons répartis.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 14 / 127 Déformée modale Mode shape Maximum acceleration = Amaxmax essai Accélération maximale = A test Distributed pedestrians.

This concept is explained in figure 1. he directly provokes the resonance phenomenon and since all pedestrians © Sétra .4 . are the Solférino footbridge and the Millenium footbridge which were submitted to thorough in-situ tests. Once again. 95 percentile or 95% fractile).Footbridges . it is possible to infer a law for the equivalent number of pedestrians. Chapter 3 explains how this loading is taken into consideration. these tests confirm that the phenomenon is clearly explained by the pedestrian response as he modifies his walking pace when he perceives the transverse motion of the footbridge and it begins to disturb him. will gradually coordinate at common frequency (that of the footbridge) and enters in phase with the footbridge motion.2006 . 3. We just have to distribute the Neq pedestrians on the bridge. the number of pedestrians. with frequencies randomly distributed around an average value and with random phase shifts.2. to apply to these pedestrians a force the amplitude sign is the same as the mode shape sign and to consider this force as the natural frequency of the structure and to calculate the maximum acceleration obtained at the corresponding resonance.8 N where N is the number of pedestrians present on the footbridge (density surface area) and the critical damping ratio. known cases of crowd lock-in have been limited to transverse footbridge vibrations. The most recent two cases. To compensate his incipient unbalance. Thus.4: Nombre Number of samples de tirages 95% 5% NumberNombre of pedestrianspiétons de Figure 1.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 15 / 127 give a value lower than this characteristic value (95% characteristic value. this law is the closest to the performed test results.Lock-in of a pedestrian crowd Lock-in expresses the phenomenon by which a pedestrian crowd. The following two laws are retained: Sparse or dense crowd: random phases and frequencies with Gaussian law distribution: N eq 10. Very dense crowd: random phases and all pedestrians at the same frequency: N eq This model is considerably simplified in the calculations. he instinctively follows the footbridge motion frequency. So far. the number of mode antinodes. now famous.85 N . 1.4: 95% fractile concept By varying damping.

Using a large footbridge with a 5.Footbridges .25m x 134m main span which can be submitted to a very dense crowd (up to about 2 persons/m²). pedestrian load may be understood as a negative damping. the critical number can then be written as: 8 m1 f 1 N K K is the proportionality factor. The model proposed for the Millenium footbridge study is as follows: the force exerted by a pedestrian (in N) is assumed to be related to footbridge velocity. © Sétra . [30]) observed that application of the above factor gave an under-estimation of about N 1 to 10 times of the actually observed lateral vibration amplitude. the problem is further amplified and theoretically the whole crowd may become synchronised. these authors propose a 0. For this structure. Seen this way. We can then note that a low damping. This is the phenomenon we call "lock-in". a low mass. on the one hand. This is how the convenient notion of critical number appears: this is the number of pedestrians beyond which their cumulative negative damping force becomes higher than the inherent damping of the footbridge. It will be noted that the value of factor K cannot a priori be generalised to any structure and therefore its use increases the criterion application uncertainty.2N multiplication factor to represent any loading which would equate that of a crowd of N persons. The consequence of lock-in is an increase of this negative damping force. when the footbridge movement is such that the pedestrians can no longer put their best foot forward. Only the main conclusions of this study are mentioned here.01 m).t where K is a proportionality factor (in Ns/m) and V the footbridge velocity at the point x in question and at time t. or a low frequency is translated by a small critical number and therefore a higher lock-in risk. F1 pedestrian KV x. actual synchronisation is much weaker and. and a detailed presentation is provided below. the situation would then be similar to that of an unstable oscillator: a small disturbance may generate indefinitely-amplifying movements… For the particular case of a sinusoidal horizontal vibration mode (the maximum amplitude of this mode being normalized to 1. Consequently. (Ref. the negative damping force induced by a pedestrian is directly deducted from it. the measurement magnitude obtained. Arup's team issued a very detailed article on the results obtained following this study and tests performed on the Millenium footbridge. They formed the hypothesis of synchronisation of a crowd walking in synchrony with the transverse mode frequency of their footbridge to explain the phenomenon and were thus able to prove. on the other. Fortunately. considering the maximum unit displacement) and assuming an uneven distribution of the pedestrians. induced by the participation of a higher number of pedestrians. Assuming a viscous damping of the footbridge. they have to stop walking and the phenomenon can no longer evolve. f1 representing the first transverse natural frequency and m1 the generalised mass in this mode. by retaining only the first term of the Fourier decomposition for the pedestrian-induced load. to increase the critical number it will be necessary to act on these three parameters.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 16 / 127 undergo it. (Ref.2006 . Fujino & al. with a value of 300 Ns/m in the case of the Millenium footbridge. allowing them to retrieve the magnitude of the effectively measured displacements (0. in this case. [38]).

2). for an individual pedestrian the amplitude of the pedestrian force remains constant.5 shows that.Footbridges . but these last oscillations should not be considered as they represent the end of the test. some tests were carried out on a reduced footbridge model by recreating. To maintain this continuity. The principle consists in placing a 7-metre long and 2-metre wide slab on 4 flexible blades moving laterally and installing access and exit ramps as well as a loop to maintain walking continuity (Photos 1. and in any case. In a second step. around 50 N. Figure 1. using a dimensional analysis.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 17 / 127 To quantify the horizontal load of a pedestrian and the pedestrian lock-in effects under lateral motion. © Sétra . lower than 100N.1 and 1.2006 . this number being clearly higher than the number of pedestrians present on the footbridge at a given time. a large number of pedestrians is of course needed on the loop. whatever the speed amplitude. the conditions prevailing on a relatively simple design of footbridge (one single horizontal mode).2: Description of the model By recreating the instantaneous force from the displacements measured (previously filtered to attenuate the effect of high frequencies) ( F (t ) mx(t ) cx(t ) kx t ). in a first step.1 and 1. CROSS-SECTION VIEW (with blade 70 cm) Scale 1/25 Detail A Detail B Detail C Flexible blade (thickness 8 mm) Detail D Attachment axis to the slab Detail E HEB Photographs 1. we observe that the force amplitude increases up to 150 N.

(N) (m/s²) Figure 1.75 m/s²) and the «efficient» force (instantaneous force multiplied by the speed sign.6 and 1. on the same figure. The following graphs (figures 1.6: Acceleration (m/s²) and efficient force (N) with 6 random pedestrians on the footbridge © Sétra . scale on the LH side expressed in N). the accelerations in time (pink curve and scale on the RH side expressed in m/s².53 Hz – pedestrians following each other Speed (m/s) F pedestrians (N) Time Exciting force (N) Figure 1. with the first harmonic of this signal being around 35 N.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 18 / 127 Comparison of F(t) and v(t) Test of 01-10-03 – 1 pedestrian at 0.1 to 0. averaged on a period.7) represent.2006 .5: Force and speed at forced resonant rate We have a peak value which does not exceed 100 N and is rather around 50N on average.Footbridges . variation from 0. which is therefore positive when energy is injected into the system and negative in the opposite case) for a group of pedestrians (blue curve.

0.7: Acceleration (m/s²) and efficient force (N) with 10 random pedestrians on the footbridge We observe that. even when the test crowd was made to walk at a frequency that had given rise to resonance. however. This would © Sétra . that is. In the inverse case. .The lock-in phenomenon effectively occurred for the first mode of lateral swinging for which the double of the frequency is located within the range of normal walking frequency of pedestrians. there is only some little synchronisation (maximum value of 100-150 N i.6m/s²).e. on the other hand. but this is quite sufficient to generate very uncomfortable vibrations (>0. .The concept of a critical number of pedestrians is entirely relative: it is certain that below a certain threshold lock-in cannot occur. However. beyond a threshold that has been proven various specific conditions can prevent it from occurring.Footbridges . The strong vertical movements disturb and upset the pedestrians' walk and do not seem to favour maintaining it at the resonance frequency selected for the tests.Lock-in appears to initiate and develop more easily from an initial pedestrian walking frequency for which half the value is lower than the horizontal swinging risk natural frequency of the structure. when the walking crowd has a faster initial pace several tests have effectively shown that it did not occur. it does not seem to occur for modes of torsion that simultaneously present vertical and horizontal movements.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 19 / 127 (N) (m/s²) Random rate acceleration: there will be as many forces participating in the movement as the forces opposed to it Beginning of synchronisation: There is a majority of forces the action of which participates in movement amplification. Several test campaigns were carried out over several years following the closing of the Solferino footbridge to traffic. Figure 1. The main conclusions to be drawn from the Solferino footbridge tests are the following: .2 to 0.On the other hand. then they were needed to check the efficiency of the adopted measures and.2006 . from the beginning these tests were intended to identify the issues and develop corrective measures.15 m/s² (straight line between the random rate zone and the incipient synchronisation zone). This threshold is around 0. finally. . cumulation is achieved in a better way.3 times the effect of 10 pedestrians). from a given value. the force exerted by the pedestrians is clearly more efficient and there is some incipient synchronisation. High horizontal acceleration levels are then noted and it seems their effects have been masked by the vertical acceleration. to draw lessons useful for the scientific and technical community.

For this reason.8: 1A Solferino footbridge random test: a crowd is made to circulate endlessly on the footbridge with the number of pedestrians being progressively increased (69 – 138 – 207).5m/s²). occur with the same equivalent © Sétra . which is more practical for defining a verification criterion. it is clear when the largest group of pedestrians is near the centre of the footbridge (sag summit). and from 0. regularly distributed on the structure. The concept of rate change critical threshold (shift from a random rate to a partly synchronised rate) becomes perceptible. The following graphs (figures 1. The equivalent number of pedestrians can be deduced from the instantaneous modal force.Correlation rate (%) Figure 1. which is certain. ratio between the equivalent number of pedestrians and the number of pedestrians present on the footbridge. but it could just as well be explained by an acceleration value felt by the pedestrian. Clearly lock-in occurs beyond a particular threshold. and in parallel the correlation or synchronisation rate.3m/s² then 0. The various "loops" correspond to the fact that the pedestrians are not regularly distributed on the footbridge. and this remains to be confirmed. It can also be seen that the three acceleration rise sags. that occur at increasing levels of acceleration (0.4m/s² and finally 0.12m/s². and both in phase and at the same frequency apparently inject an identical amount of energy per period into the system. The evolution of acceleration over time is shown (in green). It is the number of pedestrians who. with synchronisation reaching 30-35% when the acceleration amplitudes are already high (0.13) present a summary of the tests carried out on the Solferino footbridge.45m/s²). This threshold may be explained in terms of sufficient number of pedestrians on the footbridge (conclusion adopted by Arup's team).15m/s².2006 . they are concentrated in groups. In the test shown in Figure 1. Solferino footbridge: Test 1 A: Random crowd/Walking in circles with increasing numbers of pedestrians Acceleration (m/s) Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 20 / 127 - be worth studying in depth but it is already possible to explain that the pedestrian walking fairly rapidly feels the effects of horizontal acceleration not only differently. it becomes partly synchronised. behaviour is completely random. it can be seen that below 0.8. or rather at the ends of the footbridge (sag trough).Footbridges . but also less noticeably.8 to 1.

halted twice when the group of pedestrians reaches the end of the footbridge.Correlation rate (%) Figure 1.20 m/s². This time. the number of pedestrians has been increased more progressively.2006 .15 – 0. Solferino footbridge: Test 1 C: Random crowd/Walking in circles with increasing numbers of pedestrians Acceleration rate (m/s) © Sétra . In the following test. Solferino footbridge: Time 1 B: Random crowd/Walking in circles with increasing numbers of pedestrians Acceleration (m/s) Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.9.9: 1B Solferino footbridge random test: a crowd is made to circulate endlessly on the footbridge with the number of pedestrians being more progressively increased (115 138 161 92 – 184 – 202).Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 21 / 127 number of pedestrians each time. This clearly shows the fact that there is an acceleration rise.Footbridges . shown in Figure 1. the rate change threshold seems to be situated around 0. Maximum synchronisation rate does not exceed 30%.

10.5m/s²) and.11: 2A1 Solferino footbridge random test In the test shown in Figure 1.Correlation rate (%) Figure 1. the pedestrians are more grouped together and walk from one edge of the footbridge to the other.Footbridges . then more obvious synchronisation reaching up to 35%-40%. shown in Figure 1. that the level of vibration is higher (0. that the crowd is fairly compact and this favours synchronisation phenomenon among the pedestrians. however.10: 1C Solferino footbridge random test Test 1C. leads us to the same conclusions: change in threshold between 0.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 22 / 127 Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.11.Correlation rate (%) Figure 1.15m/s². © Sétra . The rise and subsequent fall in equivalent number of pedestrians better expresses the movement of pedestrians. on the other. This is higher than previously. it should be pointed out.10 and 0.2006 . and their crossing from an area without displacement (near the edges) and another with a lot of displacement (around midspan). on the one hand. Synchronisation rate rises to about 60%.9m/s² instead of 0. Solferino footbridge: Test 2 A1: Random crowd/Walking grouped together in a straight line 229p Acceleration (m/s) Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.

Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 23 / 127 Solferino footbridge: Test 2 A2: Random crowd/Walking grouped together in a straight line 160p Acceleration (m/s) Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.12: 2A2 Solferino footbridge random test In the test shown in Figure 1. synchronisation rate reaches 50%. there are only 160 people still walking slowly from one end of the footbridge to the other.12. 1B. Solferino footbridge: Test 2 B: Random crowd/Rapid walking in straight line 160p Acceleration (m/s) Correlation rate (%) Time (s) …… Acceleration (m/s²) -----.Correlation rate (%) © Sétra .Correlation rate (%) Figure 1.Footbridges .2006 . This time. and 1C. Accelerations are comparable with those obtained in tests 1A.

modelling and quantitative differences nevertheless lead to rather different damper dimensioning. It will then be possible to use the equivalent random pedestrian loadings mentioned above and this will lead to synchronisation rates to the order of 5 to 10%. although there was a compact crowd of 160 people. Below this threshold.15 m/s². the value 0. © Sétra .10 m/s² shall be noted. Acceleration thus systematically becomes uncomfortable. Consequently. Synchronisation rate can rise to more than 60% once this threshold has been exceeded. For dimensioning purposes. As soon as the amplitude of the movements becomes perceptible. Several models are available (force as a function of speed. Once this threshold has been exceeded accelerations rise considerably but remain limited. or even higher.13) is identical to the previous one except that. characterised by critical acceleration or a critical number of pedestrians. behaviour of pedestrians may be qualified as random. the pedestrians were walking fast. This clearly shows that where pedestrian walking frequency is too far from the natural frequency. high crowd synchronisation rate) but they all lead to accelerations rather in excess of the generally accepted comfort thresholds. the 0. Synchronisation rates in the order of 30 to 50% are reached when the test is stopped.10 to over 0. still random though after this they are no longer so.13: 2B Solferino footbridge random test This last test (figure 1.Footbridges .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 24 / 127 Figure 1. These tests show that there is apparently a rate change threshold in relation to random rate at around 0. this time. acceleration goes from 0.2006 . Pedestrian synchronisation phenomenon was not observed.10-0.10 m/s² rate change threshold becomes a threshold not to be exceeded. Thus. Acceleration actually corresponds to what the pedestrians feel whereas a critical number of pedestrians depends on the way in which the said pedestrians are organised and positioned on the footbridge. Passage from a random rate on a fixed support to a synchronised rate on a mobile support occurs when a particular threshold is exceeded. Critical acceleration may be interpreted as the acceleration produced by the critical number of pedestrians.60 m/s² relatively suddenly. the way in which the pedestrians are organised depends on the level of traffic on the footbridge (see below). Although it is true that the main principles and the variations in behaviour observed on the two footbridges concur. It should be noted that the concept of a critical number of pedestrians and the concept of critical acceleration may be linked. synchronisation does not occur. This value can rise to 60%. The various studies put forward conclusions that appear to differ but that actually concur on several points. It is therefore this critical acceleration threshold that will be discussed in these guidelines. The concept of critical acceleration seems more relevant that that of a critical number of pedestrians. crowd behaviour is no longer random and a type of synchronisation develops. when the crowd is compact.

[59] ) 1.7 Hz – 2. However.6 – 2.65 – 2. since they differ from one individual to another. Although risk frequency ranges are fairly well known and clearly defined.5 Hz – 2. However. Between 0 and 5 Hz <5 Hz <5 Hz 1. in most cases.1 . Owing to the fact that they are amplified. Where it is impossible to avoid resonance. 3.3 Hz 1. [28] ) CEB 209 Bulletin Bachmann ( Ref. even a footbridge of simple design will have an infinity of natural vibration modes. it seems advisable to further lower the lower boundary of the risk frequency range. Thus.3 . in construction practice.4 Hz Table 1 . it is necessary to consider pedestrian walking frequencies.3. comfort criterion. on the Millenium footbridge it was noticed that the lock-in phenomenon appeared even for a horizontal mode with a frequency considerably beneath that of the lower limit generally accepted so far for normal walking frequency.2006 . but are opposed in their horizontal action and this means transverse efforts apply at a frequency that is half that of the footsteps. Thus. drawn up for vertical vibrations: Eurocode 2 ( Ref. vital to the resonance phenomenon. for horizontal vibration modes.Footbridges . it is necessary to bear in mind a range of frequencies rather than a single one. [30] ) ISO/DIS standard 10137 ( Ref. [6] ) Regulations in Japan ( Ref. comfort threshold. it is not easy to avoid them without resorting to impractical rigidity or mass values. it is necessary to try to limit its adverse effects by © Sétra . [5] ) Appendix 2 of Eurocode 0 BS 5400 ( Ref. it is necessary to review the structural parameters. frequencies and critical damping ratios associated to it (see Appendix 1).Parameters that affect dimensioning: frequency.3 Hz 1. This leads to the concept of a range of risk frequencies to be avoided. The problems encountered on recent footbridges echo the well-known phenomenon of resonance that ensues from the matching of the exciting frequency of pedestrian footsteps and the natural frequency of a footbridge mode. where specified. represented by natural vibration modes (natural modes and natural frequencies) and the values of the structural critical damping ratio associated with each mode. In reality.35 Hz 1.5 Hz and 5 Hz.2 As concerns lateral vibrations. it is sufficient to study a few first modes. So. The first simple method for preventing the risk of resonance could consist in avoiding having one or several footbridge natural frequencies within the range of pedestrian walking frequency.6 Hz and 2. noticeable movements of the footbridge ensue and their consequence is a feeling of discomfort for the pedestrians that upsets their progress.4 Hz and. Along with this. between 2.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 25 / 127 3. the ranges described above are to be divided by two owing to the particular nature of walking: right and left foot are equivalent in their vertical action.Risk frequencies noted in the literature and in current regulations Compilation of the frequency range values given in various articles and regulations has given rise to the following table. etc. together with walking conditions and numerous other factors. [4] ) Eurocode 5 ( Ref.

The values are provided for vertical acceleration and are shown in Figure 1. the thresholds perceived by the pedestrians are assimilated with the measurable quantities generated by the footbridges and it is generally the peak acceleration value. a particular individual may react differently to the vibrations. 3. contradict each other with values that are either lower or higher (Ref. it is necessary to define appropriate load cases to apply to the structure in order to determine this acceleration. the same vibrations will not be perceived in the same way.3. so far. (Ref. the duration for which he is exposed to the vibrations affects what the pedestrian feels. This is followed by a second that can be related to various degrees of disturbance or discomfort (tolerable over a short period. Indeed. though highly desirable. Hence. 3. seated. Furthermore. such as Wheeler.Footbridges . [20]). there have only been a few suggestions relating to transverse movements. on the other. on the one hand. Obviously. that is retained. where a ceiling of 1m/s² for vertical acceleration has been given by Matsumoto and al. unacceptable). where projects are concerned.Comfort thresholds Before going any further. [18] and Ref. or even health problems. The first is a vibration perception threshold. in this way. depending on whether he is standing. knowledge in this field remains imprecise and insufficient. determining thresholds in relation to the comfort perceived by the pedestrian is a particularly difficult task. However. a third threshold may be determined in relation to the consequences the vibrations may entail: loss of balance. for a particular individual. noted acrit.Acceleration comfort criteria noted in the literature and regulations The literature and various regulations put forward various values for the critical acceleration. and Tilly et al. several thresholds may be defined. it is clear that in order to be used for footbridge dimensioning. furthermore. this concept is highly subjective. the comfort criterion to be respected is represented by this value that should not be exceeded. disturbing. called the critical acceleration acrit of the structure. obviously. In particular: from one individual to the next. Hence. it is necessary to consider the recommended values as orders of magnitude and it is more convenient to rely on parameters that are easy to calculate or measure. the concept of comfort should be specified. [16]). others.2 . moving or stationary.14 below: © Sétra . Clearly. Finally.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 26 / 127 acting on the remaining parameter: structural damping.2006 .3 .3. it will be necessary to have available criteria making it possible to determine the acceptable limits of the resonance. it is also well known that there is a difference between the vibrations of the structure and the vibrations actually perceived by the pedestrian. For instance. This verification is considered to be a service limit state.

4 .e.14: Vertical critical accelerations (in m/s²) as a function of the natural frequency for various regulations: some depend on the frequency of the structure.5 to 0. an acceleration threshold may be translated as a displacement threshold (which makes better sense for a designer) or even a speed threshold. 2 Acceleration frequency of 2 Hz movement Acceleration frequency of 2 Hz speed For instance. a speed of 0. a speed of 0.04m/s acceleration of 1m/s² corresponds to a displacement of 6.7mm.08 m/s but for a frequency of 1 Hz: acceleration of 0. It should be remembered that these values are mainly associated with the theoretical load of a single pedestrian. a crowd).Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 27 / 127 Figure 1. Eurocode 0 Appendix 2 proposes a horizontal critical acceleration of 0. With an existing footbridge.2 m/s² under normal use and 0.08m/s acceleration of 1m/s² corresponds to a displacement of 25. the text does not provide crowd loadings. there is apparently a consensus for a range of 0. Unfortunately.3 mm. for a frequency of 2 Hz: acceleration of 0.8 m/s².2mm.Improvement of dynamic behaviour What is to be done when footbridge acceleration does not respect comfort criteria? It is necessary to distinguish between a footbridge at the design stage and an existing one. others do not.2006 ¬ × Ë Ö ¬ ° Ê ß ª Ý É ° Þ « Ý Ç ª Ü Ò Ó Ñ ª Ü ¬ Ð Õ Ë Û Ô Ï È È Î Ú Ì Í Ê Ç Ù Ç Ø Ì Æ É Ó ¼ » º ¸ ´ ¹ ¸ µ · ¶ ² µ ´ ³ ² ± ª ª ª ª ª ª ª ª ¬ ª ¯ ® ¬ ª « « « « « « « « ° ° ° ª ª ª ª ª Ä Å Á Ã Ã ½ ½ Á Á ½ ½ ¾ ¾ À À ¿ Â Â ¿ ¿ . © Sétra . experience shows that it is generally cheaper to increase damping.4 m/s² for exceptional conditions (i.3mm. speeds and displacements are related.Footbridges . It is also useful to remember that since accelerations. it is logical to try to modify its natural frequency vibrations. it is also possible to try to modify its natural frequency vibrations. However.5m/s² corresponds to a displacement of 12. a speed of 0. In the case of a footbridge at the design stage. standard walking frequency.16 m/s 3.5m/s² corresponds to a displacement of 3. For lateral vibrations with a frequency of around 1 Hz. a speed of 0. then attempts should be made to increase structural damping. If it is not possible to modify them so that they are outside the resonance risk ranges in relation to excitation by the pedestrians. For vertical vibrations with a frequency of around 2 Hz.

CEB information bulletin No.0% 0.2% 0. Studies generally use critical damping coefficients ranging between 0. It should be noted that although the mass and rigidity of the various structure elements may be modelled with a reasonable degree of accuracy.6% 1.Increasing structural damping The critical damping ratio is not an inherent fact of a material. For instance.4.3 As concerns timber.4. practice indicates that an increase in stiffness is frequently accompanied by an increase in mass. The critical damping ratio also increases when vibration amplitude increases. damping properties are far more difficult to characterise. However. the AFPS 92 Guide on seismic protection of bridges (art.3) states the following: Material Welded steel Bolted steel Prestressed concrete Non-reinforced concrete Reinforced concrete Reinforced elastomer Table 1. which produces an inverse result. the difference between bolting and welding). 209 (Ref.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 28 / 127 3.0% Table 1. Eurocode 5 recommends values of 1% or 1.2. 4. Therefore the stiffness of the structure needs to be increased.5% 3.1% and 2.3% 0.5% 1.1 .2 . this is a difficult problem to solve.Modification of vibration natural frequencies A vibration natural frequency is always proportional to the square root of the stiffness and inversely proportional to the square root of the mass. It also depends on construction details that may dissipate energy to a greater or lesser extent (for instance. critical damping ratios are considerably higher and need to be SLS checked. 3. where steel is concerned. of mechanical joints.4 Critical damping ratio 2% 4% 2% 3% 5% 7% © Sétra .0% and it is best not to overestimate structural damping in order to avoid underdimensioning.Footbridges . [7] ). Most experimental results suggest that dissipation forces are to all practical intents and purposes independent of frequency but rather depend on movement amplitude. as with earthquakes. The general aim is to try to increase vibration frequency.8% 1.3% 0.2006 . Where vibration amplitude is high.5% depending on the presence. an important summary document dealing with the general problem of structure vibrations provides the following values for use in projects: Type of deck Reinforced concrete Prestressed concrete Metal Mixed Timber Critical damping ratio Minimum value Average value 0.4% 0. or otherwise.

5% 0.6 0.49 (horizontal) 0. or in the case of tension tie footbridges («stress ribbon»).6 span Critical damping ratio of the structure without dampers 0. it should be pointed out that an estimate of actual structural damping can only be achieved through measurements made on the finished structure. Appendix 3 describes the different types of dampers that can be used and describes the operating and dimensioning principle of a selection of dampers.5: Examples of the use of tuned dynamic dampers Note: In the case of the Millenium footbridge.22 (vertical) 0. through insertion of elastomer plates distributed between the prefabricated concrete slabs making up the deck.3% to 5.75 (horizontal) 2.2% to 0.6 2.2006 .2% 8% 1000 2500 to 0.5 (vertical) 1.2% 0.5% 0.4% 3.5% 2.075 1.6% to 0.5% 0.4% 1 0.95 (vertical) Name Mass Total effective kg mass % 2400 per 1. as well as ADAs.812 (horizontal) 1.6% to 0.6% 3.3% Critical damping ratio of the structure with dampers 4. viscous dampers were also installed to dampen horizontal movement. This said.7 0.8% 3% 2% 2% Passerelle du Stade de France (Football stadium footbridge) (Saint-Denis) Solferino footbridge (Paris) Ditto Ditto Millenium footbridge (London) Ditto 15000 4.94 (vertical) 2.Footbridges .8 (vertical) 0.6% 0. © Sétra .8% 10000 7600 2500 2. for instance through use of a wire-mesh structure.3% Structure frequency Hz France France Ditto Ditto England Ditto Japan USA South Korea Ditto Las Vegas (BellagioBally) Seonyu footbridge Ditto 0.4% 3. increased damping may be obtained from the design stage. The use of dampers is another effective solution for reducing vibrations.03 (vertical) Table 1.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 29 / 127 Finally. The table of examples below shows this is a tried and tested solution to the problem: Country Damper critical damping ratio 0.

as a function of the frequency value ranges. Obviously. the standard verifications (SLS and ULS) are to be carried out in compliance with the texts in force.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 30 / 127 4. The recommendations are the result of the inventory described in Appendices 1 to 4. The comfort level obtained can be qualified by the range comprising the values.Footbridges . very light footbridges may undergo vibrational phenomena. The methodology proposed makes it possible to limit risks of resonance of the structure caused by pedestrian footsteps. Footbridge dynamic analysis methodology For the purposes of Owners. or otherwise. The methodology is summarised in the organisation chart below.2006 . they are not covered by these guidelines. No calculation required Class IV Footbridge class Classes I to III Calculation of natural frequencies Traffic assessment Comfort judged sufficient without calculating negligible Resonance risk level sensitive Owner Dynamic load cases to be studied Max. when deciding on his approach. Nevertheless. resonance apart. the Owner needs to define the class of the footbridge as a function of the level of traffic it will undergo and determine a comfort requirement level to fulfil.1: Methodology organisation chart This chapter also includes the specific verifications to be carried out (SLS and ULS) to take into account the dynamic behaviour of footbridges under pedestrian loading (see 2. Thus. to determine structure natural frequency. prime contractors and designers. At source. on the other.6). these natural frequencies lead to the selection of one or several dynamic load cases. These load cases are defined to represent the various possible effects of pedestrian traffic. © Sétra . on the one hand. Footbridge class conditions the need. they are a summary of existing knowledge and present the choices made by the group. it must be remembered that. and of the work of the Sétra-AFGC group responsible for these guidelines. this chapter puts forward a methodology and recommendations for taking into account the dynamic effects caused by pedestrian traffic on footbridges. Where they are calculated. accelerations undergone by the structure Comfort conclusion Acceleration limits Comfort level Figure 2. Treatment of the load cases provides the acceleration values undergone by the structure.

1 . but do not become intolerable. It is for the Owner to determine the footbridge class as a function of the above information and taking into account the possible changes in traffic level over time. Minimum comfort: under loading configurations that seldom occur. in view of high media expectations. 4. it seems advisable to select at least Class III to ensure a minimum amount of risk control. elderly or disabled people). subjected to heavy traffic and that may occasionally be loaded throughout its bearing area. tourists. etc.2. Average comfort: Accelerations undergone by the structure are merely perceptible to the users.).Footbridges . It should be noted that the above information cannot form absolute criteria: the concept of comfort is highly subjective and a particular acceleration level will be experienced differently. Choice of comfort level is normally influenced by the population using the footbridge and by its level of importance. a lower class may be accepted in order to limit construction costs or to ensure greater freedom of architectural design. occasionally. It is possible to be more demanding on behalf of particularly sensitive users (schoolchildren. accelerations undergone by the structure are perceived by the users. Class III: footbridge for standard use. Indeed. that may occasionally be crossed by large groups of people but that will never be loaded throughout its bearing area.2006 . nearby presence of a rail or underground station) or that is frequently used by dense crowds (demonstrations. bearing in mind that the risk related to selecting a lower class shall be limited to the possibility that. these guidelines do not deal with comfort in premises either extensively or permanently occupied that some footbridges may undergo. © Sétra . comfort level is automatically considered sufficient. depending on the individual. and more tolerant in case of short footbridges (short transit times). His choice may also be influenced by other criteria that he decides to take into account. when the structure is subjected to a load where traffic and intensity exceed current values.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 31 / 127 4. Class IV footbridges are considered not to require any calculation to check dynamic behaviour. subjected to very heavy traffic. On the other hand. over and above their pedestrian function.Definition of the comfort level The Owner determines the comfort level to bestow on the footbridge.Stage 2: Choice of comfort level by the Owner 4.2 . some people may feel uncomfortable.Stage 1: determination of footbridge class Footbridge class makes it possible to determine the level of traffic it can bear: Class IV: seldom used footbridge.1 . built to link sparsely populated areas or to ensure continuity of the pedestrian footpath in motorway or express lane areas. Class II: urban footbridge linking up populated areas. Furthermore. In cases where the risk of resonance is considered negligible after calculating structure natural frequencies. For instance. a very light footbridge may present high accelerations without there necessarily being any resonance. Class I: urban footbridge linking up high pedestrian density areas (for instance. For very light footbridges. Maximum comfort: Accelerations undergone by the structure are practically imperceptible to the users. a higher class may be selected to increase the vibration prevention level.

1: Acceleration ranges (in m/s²) for vertical vibrations Acceleration ranges Range 1 Range 2 Range 3 Range 4 Min Mean 0 0. 2. the first 3 correspond to the maximum. using different dynamic load cases. for vertical and horizontal accelerations respectively.Footbridges .15 Max 0. it has been judged preferable to reason in terms of ranges rather than thresholds. 3 and 4. They are determined for 2 mass assumptions: empty footbridge and footbridge loaded throughout its bearing area.2. to the tune of one 700 N pedestrian per square metre (70 kg/m2). The ranges in which these frequencies are situated make it possible to assess the risk of resonance entailed by pedestrian traffic and. as a function of this. the dynamic load cases to study in order to verify the comfort criteria. Tables 2.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 32 / 127 4.1 0.Stage 3: Determination of frequencies and of the need to perform dynamic load case calculations or not For Class I to III footbridges.2 define 4 value ranges.1 and 2.3 0.3 .2: Acceleration ranges (in m/s²) for horizontal vibrations The acceleration is limited in any case to 0.Acceleration ranges associated with comfort levels The level of comfort achieved is assessed through reference to the acceleration undergone by the structure.2006 . transverse horizontal and longitudinal horizontal. In ascending order.8 Table 2. mean and minimum comfort levels described in the previous paragraph. Thus.10 m/s² to avoid "lock-in" effect 4. determined through calculation.5 Range 1 Range 2 Range 3 Range 4 Table 2. noted 1. Acceleration ranges 0 0.5 Max Mean Min 1 2. © Sétra . The 4th range corresponds to uncomfortable acceleration levels that are not acceptable.2 . it is not directly a question of the acceleration perceived by the users of the structure. These frequencies concern vibrations in each of 3 directions: vertical. it is necessary to determine the natural vibration frequency of the structure. Given the subjective nature of the comfort concept.

Table 2.3.7 2. å ç è æ ä ä ä ä ã ã ã ã â â â â á á á á à à à à Frequency 2.5 clearly defines the calculations to be performed in each case. : negligible risk of resonance.1 1.Footbridges .4: Frequency ranges (Hz) of the transverse horizontal vibrations 0 0. it is necessary to carry out dynamic structure calculations for all or part of a set of 3 load cases: Case 1: sparse and dense crowd Case 2: very dense crowd Case 3: complement for an evenly distributed crowd (2nd harmonic effect) Table 2. Table 2.Definition of the required dynamic calculations Depending on footbridge class and on the ranges within which its natural frequencies are situated.3. there are four frequency ranges.3 0.3 defines the frequency ranges for vertical vibrations and for longitudinal horizontal vibrations.1 .5 0 1 1. © Sétra .2006 .5 2.3: Frequency ranges (Hz) of the vertical and longitudinal vibrations Frequency 1.Frequency range classification In both vertical and horizontal directions.6 5 4. corresponding to a decreasing risk of resonance: : maximum risk of resonance.2 .4 concerns transverse horizontal vibrations.1 Range 1 Range 2 Range 3 Range 4 Table 2. : low risk of resonance for standard loading situations.3 Range 1 Range 2 Range 3 Range 4 Table 2.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 33 / 127 4. : medium risk of resonance.

and the way the constructions are to be modelised are given in the next chapter.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 34 / 127 Load cases to select for acceleration checks Natural frequency range Traffic Class 1 Nil Case1 Dense II I Case 2 Case 2 Case 3 Very dense Case No. in a simplified and practicable way.5 pedestrians/m2 0. If the previous stage concludes that dynamic calculations are needed. They have been constructed for each natural vibration mode. The number of equivalent pedestrians. under working conditions.4. This case is only to be considered for category III (sparse crowd) and II (dense crowd) footbridges. traditional SLS and ULS type checks.4 .Dynamic load cases The load cases defined hereafter have been set out to represent. The number of pedestrians involved is therefore: N = S d. including the dynamic load cases. in other words the number of pedestrians who. 3: Crowd complement (2nd harmonic) Case 1 Case 3 Nil 2 3 Sparse III Table 2. these calculations shall enable: checking the comfort level criteria in paragraph II. 4. The density d of the pedestrian crowd is to be considered according to the class of the footbridge: Class III II Density d of the crowd 0. the effects of fewer or more pedestrians on the footbridge.2 required by the Owner. Indications of the way these loads are to be taken into account and to be incorporated into structural calculation software.1 .8 pedestrians/m2 This crowd is considered to be uniformly distributed over the total area of the footbridge S.8 ( N)1/2 (see chapter 1). 2: Very dense crowd Case No.Stage 4 if necessary: calculation with dynamic load cases.2006 . would produce the same effects as random pedestrians. 1: Sparse and dense crowd Case No. the frequency of which has been identified within a range of risk of resonance.5: Verifications – load case under consideration: 4. The load that is to be taken into account is modified by a minus factor which makes allowance for the fact that the risk of resonance in a footbridge becomes less likely the further © Sétra .Footbridges . in frequency and in phase is: 10. being all at the same frequency and in phase. under the dynamic load cases as defined hereafter.

In the same way. and n the number of pedestrians on the footbridge (d x S).5 Hz – 1.2006 .3 Hz for the horizontal action.8 ( n)1/2.Footbridges . beyond 2. © Sétra . then divided by the loaded area S. if one is interested in the vertical and longitudinal modes. This factor falls to 0 when the footbridge frequency is less than 1 Hz for the vertical action and 0.6 0 0 Structure freq. G0. The vertical load case is applied for each vertical mode at risk.5 1. however.7 2.3 Figure 2.6 Hz for the vertical action and 1. 0.8 ( /n)1/2 d (140N) cos(2 flt) 10. Direction Vertical (v) Longitudinal (l) Transversal (t) Load per m² d (280N) cos( fvt) 10. and 0. See chapter 3 for further details. and the longitudinal load case for each longitudinal mode at risk. This component has no influence on acceleration.1 Hz for horizontal accelerations.1 1. Remember that the number of equivalent pedestrians was constructed so as to compare real pedestrians with fewer fictitious pedestrians having perfect resonance. Comment 4: These loads are to be applied until the maximum acceleration of the resonance is obtained. for vertical and longitudinal vibrations on the left.3 5 0.8 ( /n)1/2 d (35N) cos( fvt) 10.8 ( / n)1/2. however. for any random crowd. the second harmonic of pedestrian walking must be examined. represents the critical damping ratio (no unit). In this case. when passing through a node for example (see chapter III for further details).3 : Factor in the case of walking. Comment 3: The load cases above do not show the static part of the action of pedestrians. to be multiplied by the individual action of these equivalent pedestrians (F0 cos( t)) and by the minus factor Comment 2: It is very obvious that these load cases are not to be applied simultaneously.3 Hz for the horizontal action. and must be inverted each time the mode shape changes direction.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 35 / 127 away from the range 1.1 Hz for vertical accelerations.8 ( /n)1/2 ð ê ë ð ï î í ì ë ê é The loads are to be applied to the whole of footbridge. it shall be borne in mind that the mass of each of the pedestrians must be incorporated within the mass of the footbridge. to be applied for each direction of The table below summarises the load vibration.7 Hz – 2. 1 1 0 0 1 Structure freq. the number of equivalent pedestrians is calculated using the formula 10. adjusting on each occasion the frequency of the load to the natural frequency concerned.1 2. which is replaced by n/d (reminder n = S d). be selected to produce the maximum effect: the direction of application of the load must therefore be the same as the direction of the mode shape. and for lateral vibrations on the right. the factor cancels itself out. at any point. and the sign of the vibration amplitude must. 1. Comment 1: in order to obtain these values. which gives d 10.

1 2.0 (35N) fvt) fvt) fvt) 1. In this case. located. It is considered that the pedestrians are all at the same frequency and have random phases.5 Structure freq. . 0 1. The pedestrian crowd density to be considered is set at 1 pedestrian per m2. the number of pedestrians all in phase equivalent to the number of pedestrians in random phases (n) is 1. because of the uncertainty of the coincidence between the frequency of stresses created by the crowd and the natural frequency of the construction. The same comments apply as those for the previous paragraph: Direction Vertical (v) Longitudinal (l) Transversal (t) cos( cos( cos( Load per m2 1.3 according to the natural frequency of the mode under consideration. It is only to be taken into account for footbridges of categories I and II. 2. at double the frequency of the first harmonic. as for load case No. for vertical and longitudinal vibrations on the one hand.4 4. The density of the pedestrian crowd to be considered is 0. is defined by figure 2. 1 1 0 2. and transversal on the other. allowance is made for the random character of the frequencies and of the pedestrian phases. as for load case No. allowance is made for the random character of the pedestrian phases only.3 1. is given by figure 2.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 36 / 127 This load case is only to be taken into account for class I footbridges. Figure 2. because of the uncertainty of the coincidence between the frequency of stresses created by the crowd and the natural frequency of the construction.0 (280N) 1. but considers the second harmonic of the stresses caused by pedestrians walking.85 (1/n)1/2 This case is similar to cases 1 and 2. and transversal on the other.85 n (see chapter 1).4 according to the natural frequency of the mode under consideration. This crowd is considered to be uniformly distributed.8 pedestrians per m2 for category II. and 1.0 (140N) 1.2006 . 1. 7N transversally and 35N longitudinally.2 5 Structure freq. .4: Factor for the vertical vibrations on the left and the lateral vibrations on the right © Sétra .0 for category I. The second minus factor.85 (1/n)1/2 1.Footbridges .6 3. For category I footbridges.85 (1/n)1/2 1. This crowd is considered to be uniformly distributed over an area S as previously defined. on average. The second minus factor. The following table summarises the load to be applied per unit of area for each vibration direction.7 2. The individual force exerted by a pedestrian is reduced to 70N vertically. for vertical and longitudinal vibrations on the one hand. For category II footbridges.

4.2006 . when the frequency of the first mode is low.6% Steel 0. in comparison to the mechanical centre of that section. are outside the range of risk. In practice. it may be advantageous to reduce the frequencies so as to bring the first mode below the range at risk. the following formula approach can be used: EI m m mode material m EI m material m in which EIm is the contribution of the material m to the overall rigidity EI of the section. the determination of km. provided that the second mode remains above that range. so that the first mode. For traditional footbridges of hardly varying section. or steps to be taken if it concerns an existing footbridge (installation or not of dampers).1 . the critical damping ratio to be taken into account may be taken as the average of the ratios of critical damping of the various materials weighted by their respective contribution in the overall rigidity in the mode under consideration: m mode i material m km.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 37 / 127 4. the project is to be re-started if it concerns a new footbridge.i is the contribution of material m to the overall rigidity in mode i.Stage 5: Modification of the project or of the footbridge If the above calculations do not provide sufficient proof.3% Pre-stressed concrete 1% Mixed 0. This paragraph gives recommendations for reducing the dynamic effects on footbridges. but within the range of risk. Most of the time.Damping of the construction The dynamic calculations are made.2 . 4. in order to modify the natural frequencies of a construction significantly. These recommendations are given in decreasing importance.4% Timber 1% Table 2.i km. a way of increasing the natural frequencies is sought.5 . © Sétra .i material m in which km. In certain cases. and that of the second mode is sufficiently high.5.Footbridges . However.6: Critical damping ratio to be taken into account In the case of different constructions combining several materials. it is very often necessary to carry out extensive structural modifications so as to increase the stiffness of the construction.i rigidity is difficult.Modification of the natural frequencies The modification of the natural frequencies is the most sensible way of resolving vibration problems in a construction. and thus all the following modes. (such that EI m EI ) material m 4. taking into account the following structural damping: Type Critical damping ratio Reinforced concrete 1.

as the modulus only increases according to the cube root of the compressive strength. the frequency of vibration increases according to the square root of the tension of the cables divided by the linear density of the cables and of the deck. The torsional vibrations of the deck cause it to move vertically. other than by modifying the static diagram of the construction: creating recesses in the piers. it becomes more flexible and static deflection is increased. but the self-weight also increases. but it is not economical. In the case of a lattice deck. In the case of a deck with a solid-core mixed steel-concrete beam. the case of the vertical vibrations of a deck formed from a steel box girder. for example. A fan arrangement of stays is stiffer than a harp arrangement. an increase in the thickness of the bottom steel flange is more effective in increasing the frequency of vibration. whereas the section of the flanges (therefore the mass of the flanges) varies inversely to the depth. adding cable stays. In this case there is no solution for increasing the natural frequencies. in a great number of cases. This solution is effective. as the quantity of stays has to be increased. A deck formed by a box girder will thus have a higher frequency of vibration than a deck of the same thickness formed from rectangular beams.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 38 / 127 However. its stiffness can thus be increased without increasing its mass. Normal concrete can also be replaced by lightweight concrete in order to reduce the mass (with a slight reduction in stiffness) and thus increase the frequency of vibration. Taller pylons also lead to an increase in the stiffness without increasing the mass. by reducing the rigidity of the construction. etc. away from the longitudinal axis of the construction. without increasing its mass. to be increased. It is therefore advantageous to increase the depth in order to increase the frequency of vibration. The selfweight does not increase as quickly. the inertia increases in proportion to the thickness. The © Sétra . In the case of a suspended deck. in particular. this is not very satisfactory.2006 . Their deflection must. due to the mass of the concrete slab. Another traditional way of increasing stiffness without increasing the mass consists of replacing a rectangular section with an I-section. which reduces the overall effect. In the case of a concrete deck. for functional (height of passage under the footbridge) or architectural reasons. which does not increase. an increase in the strength of the concrete enables its modulus. with no offset. In the case of a cable-stayed deck. If the depth of the box girder can be increased. but in reduced proportions. It is sufficient to retain the thicknesses of the top and bottom flanges and to reduce the thicknesses of the webs in proportion to the increase in their depth. But.Footbridges . The vertical stiffness of a deck can also be increased by making the balustrades participate in the stiffness. the following can be quoted: Let us consider. and thus the stiffness of the deck. as it comprises a major part. it is not possible to increase the depth of the box girder. The values of the torsion vibration frequencies of the deck must also therefore be considered when the vertical vibrations felt by pedestrians are being studied. or not by enough. an increase in the sections of the stays generally allows the stiffness to be increased without increasing the mass by much. Of the ways of increasing the natural frequencies of the footbridge. and thus to an increase in the frequency of vibration. If the thicknesses of the flanges and of the webs of the box girder are increased. There is therefore no advantage in simply increasing the section of the cables. be reduced. In addition. the inertia varies according to the square of the depth.

There is therefore every advantage in designing a deck that is rigid in torsion. There are several ways of increasing the frequency of the torsion vibrations of a deck. At a given width.2 . 4. for example. over the Cher. rather than four S/2-section beams equally spaced under the deck. the deck of which is formed using two lateral beams. of course. The addition. However. [52]). elastomer panels may be positioned between the pre-cast concrete slabs from which the deck is formed in order to increase the damping. Another way of reducing accelerations is to use materials that are naturally damping. lateral cables can be used to stiffen the construction.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 39 / 127 frequency of torsional vibration is proportional to the square root of the stiffness in torsion and inversely proportional to the square root of the polar density of the deck. A box girder deck thus has greater torsional inertia than a deck formed with lateral beams. an attempt must be made to reduce accelerations (without directly affecting the damping). In the case of a cable-stayed bridge with lateral suspension. which will most usually be tuned mass dampers (these are the easiest to install: to work properly. to a deck formed from filler joists supported by lateral beams.3 . and not into two independent lateral pylons. concrete etc. Increased damping may be obtained.2006 . The frequency of horizontal vibration is proportional to the square root of the horizontal stiffness and inversely proportional to the square root of the mass of the deck.). In addition. by the use of a lattice balustrade. it shall be realised that. This is the case for the suspended footbridge at Tours. damping systems can be installed. But at a cost that is no less obvious. One obvious way of increasing horizontal stiffness is to increase the width of the deck. viscous dampers often require the construction of complex devices to recreate major differential movement). One of these.Structural reduction of accelerations If it is not possible to increase the frequencies sufficiently. This has a direct effect on accelerations (it should be remembered that they are inversely proportional to the mass). or if the bridge is an old one. will also allow the torsional stiffness to be increased. of a bottom horizontal lattice wind-brace connecting the bottom flanges of the two beams. and the damping provided by the deck may make an appreciable contribution to the total damping. to which substantial modification cannot be made. anchoring the cable stays into the axial plane of the construction (on an axial pylon or to the top of an inverted Y or inverted V pylon). if this deck is connected to the structure. for the damping of these materials to be mobilised. 4.3 (Ref. or if the increase leads to a design that could make the project unviable. the mass of the construction can be increased by using a "heavy" deck (asphalt.5. if the previous solutions do not work. In the case of "stress ribbon" footbridges.Installation of dampers As a last resort.Footbridges . they must play a part in the overall rigidity. A tuned mass damper consists of a mass connected to the © Sétra . This is what was done for the footbridge for the Palais de Justice in Lyon. will allow the torsional frequency to be increased by a factor of close to 1. two S-section lateral beams can be used. is to increase the torsional inertia. The torsional inertia can be increased yet more by increasing the cross-sectional area of the box girder. To do this. the frequencies will not be reduced by too much. but by a smaller amount than a box girder.5. In the case of cable-stayed or suspended footbridges that are very narrow in relation to their span. one way of increasing horizontal stiffness consists of providing resistant elements on the edges of the deck: for example.

torsion. the corresponding forces must also be calculated to check whether they remain permissible. when accelerations have to be calculated. On a proposed footbridge. therefore. If the methodology does not require the calculation of accelerations. 4. this check need not be made. It is important to note that static and dynamic vibratory shapes are combined. because of the added weight (approximately 3 to 5% of the modal mass of the mode under consideration). etc. 4. that precautions must be taken when using calculation software. the effects of static loading associated with the dynamic case under consideration.ULS type checks specific to the dynamic behaviour As these consist of checking the strength of the construction. displacements. corresponding on the one hand to the permanent loads and on the other hand to the weight of the pedestrians set out identically. it may not be sufficient. as defined by the relevant regulations. preferably.3). they do need a minimum level of routine maintenance: Owners must be made aware of this.6 . using levels of stresses that are assumed to be close to the elastic limit of the materials (otherwise. and also the measurement of the natural frequencies. This device allows the vibrations in a construction to be reduced by a large amount in a given vibration mode. etc. longitudinal modes. Thus. 3% of guaranteed damping will be achieved: on very lightweight constructions (for which the ratio of the exciting force divided by the mass is high). specific service combinations for the dynamic loads of pedestrians should be built up. vertical.2 . there are disadvantages.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 40 / 127 construction using a spring. despite the apparently attractive character of these solutions (substantial increase in damping at low cost). the damper must be set (within about 2-3%) at a frequency of the construction that changes over time (deferred phenomena) or according to the number of pedestrians (modification of the mass). For complex footbridges.) resulting from one of the dynamic load cases 1 to 4.SLS type checks specific to the dynamic behaviour In addition to the usual checks to be made on service limit states. this solution will only work on an existing footbridge if it has sufficient spare design capacity. obviously.1 .Structural checks under dynamic loads 4. even though manufacturers claim that dampers have a very long life-span.6.) of risk. and thus doubles. This shall only be considered as a last resort. and.2006 .6. it may be very onerous to implement. there would be © Sétra . The reduction in effectiveness is appreciable. the designer may need to resize the construction. with a damper positioned in parallel. as. using a load of 700 N per pedestrian. which is the most typical case: as many dampers are needed as there are frequencies of risk. by applying the same methodology as for the determination of comfort. by combining: the effects (stresses. the addition of a damper degenerates. under the action of a periodic excitation of a frequency close to the natural frequency of this vibration mode of the construction (see Annexe 3 §3. which have many modes (bending. transversal. If tuned mass dampers are used. the natural frequency under consideration: this complicates the overall dynamic behaviour.Footbridges .

with the following modifications: the density of the crowd is taken as 1 pedestrian per m².Footbridges . the ultimate limit states set out here should be carried out in addition to the traditional ULSs imposed by the relevant regulations. In the transverse direction.2006 ¦ ¦ ¦ ¦ ¥ ¨ © ¥ ÿ þ ô ù ú ô ö ú ý ô ú û ô ø ø û £ £ ü ø ù ô ô û ù ö ú ô ô ú ö õ ø ú õ ø õ ù ¤ ö ¤ ¤ ¤ ô ø ô õ ô ô ÷ ú ¤ õ ø £ £ û ö ô ö ô û ö õ ú § ÷ ¢ õ ô ù ý ó ô ô ò û û ñ ó . there may be an advantage in looking to which acceleration the maximum dynamic force thus calculated corresponds. Experience has shown that there may be a problem of strength in particular cases only. it could occur under very exceptional conditions (rhythmic demonstrations. or can be more tolerant on the limits of the materials. and if the consequences on sizing are major (which should not be the case. as: The lateral action of a pedestrian is much weaker than the vertical action (35 N instead of 280 N). as it assumes the perfect coordination of a whole crowd of pedestrians. The consequences on sizing may be negligible because of the sizing of the footbridge for wind. However. etc. assuming that extreme wind situations and a large crowd on the footbridge are not combined. the individual effects of pedestrians are combined directly.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 41 / 127 no strength problem). whatever the category of the footbridge. for which the permanent loads can be weighted differently from the fundamental ULS. with no minus factor (nequivalent = n and = 1). The load case to be considered is similar to load case No.7 In the same way as for service limit states. the strength of the construction is not very likely to be a problem. the dynamic calculations must be made using the structural damping given in table 2. An accidental type combination should be formed in order to simulate a case of vandalism or an exceptionally large public demonstration on the footbridge. it could be given a reasonable value less than the acceleration due to gravity (between 0. processions.3. due to the small difference between sizing with and without dynamic loads). be a lack of strength if the construction is optimised and everything precisely dimensioned. there may. which is taken as 700 N/m². This load case is extremely pessimistic. If this acceleration is too great. under certain particular circumstances. beyond which it can be considered that walking becomes impossible. allowance must be made for the static load applied by the crowd. © Sétra .0 g). races.7 below: ¡ ù ý ó Table 2.) or under lock-in situations. apart from the permanent loads on the footbridge.2). Whatever. as it must not be forgotten that such a load case represents an accidental ULS. If such is the case. 1 previously defined (see § 2. The check is to be made as soon as a comfort check for a natural frequency within a range at risk has been made.5 g and 1.

it should be noted that.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 42 / 127 The transverse acceleration beyond which a pedestrian can no longer keep his balance is much smaller than the vertical acceleration.3 g.2006 . in these checks.Footbridges . © Sétra .1 g and 0. Finally. but a check should first be made. That is why this guide recommends Owners not to permit the holding of exceptional public demonstrations on pedestrian footbridges. but to take them into account in the sizing of the footbridge. It can. It is therefore fairly unlikely that an exceptional pedestrian crowd could cause a problem for the sizing of a pedestrian footbridge. be limited to a value of between 0. or of its strength. for example. as it may be difficult to prevent them from taking place throughout the lifespan of the footbridge. the individual safety of the pedestrians in the event of accelerations that are too great because of these exceptional public demonstrations has not been covered.

The number of degrees of freedom is thus considerably reduced and non-linked equations are obtained.1. the amplitude of that resonance. Chapter 3: Practical design methods 5. There is. One of the modes then has a much larger response than the others. The two latter points may be encountered in certain footbridge vibration problems (use of finite elements.Direct integration This method consists of integrating the equations of the dynamics directly. to determine "apparent frequencies" of normal vibrations in order to be able to set the dynamic loads producing the maximum effect. The power of this method is that the natural vibration modes are the natural vibration modes of the construction. There do exist. and only in this case. The second advantage is that they form a base. therefore. It is used very little in practice for vibrations of footbridges. as the phenomena affecting them are resonance phenomena. a resonance phenomenon is produced.2 . two major methods can be used: direct integration design and modal design. © Sétra . dampers. when the problem is not linear.Footbridges . may. methods that are close to modal methods. and that the actual solution is therefore a linear combination of these modes. the natural frequencies of the constructions must be known. when a construction is excited at one of its natural frequencies.Modal calculation The modal method is always carried out in two stages. construction with non-linear behaviour). however. thus easy to resolve. however. and it is often advantageous. in which the excitation is known (imposed earthquake accelerogram.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 43 / 127 5.Practical design methods For the dynamic design of continuous constructions and systems. 5. Finally. even so.1 . turn out to be necessary in one of the following cases: when the modal superposition method cannot be used with a reduced number of modes. then only one unknown in the problem. and the calculation of the actual response (if necessary) on the basis formed by these natural vibration modes. namely the determination of the natural vibration modes initially. 5.1. which means that. It is used more in earthquake-resistant design. with an imposed load.1 . A problem with several degrees of freedom is thus restricted to a problem with one degree of freedom. for example) The direct integration method. A knowledge of these resonance frequencies is crucial for the analysis of the dynamic problem.2006 . more expensive in analysis time than modal calculations. the third advantage is that. when the damping is not proportional or when it is concentrated (viscous dampers for example). to predict them. They therefore represent the most likely vibration modes of the construction.

GJ the torsional rigidity.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 44 / 127 5. for footbridges that are not very wide (low width in comparison with span) and rigid in torsion (closed profiles).). In practice.2006 . Mode Simple bending to n sags n2 n 2 Natural pulsation L2 n L fn fn n Natural frequency fn n 2L ES S n2 2 L2 EI S EI S v n ( x) sin Mode shape n x L sin sin Tension-compression to n sags n ES S u n ( x) ( x) n x L n x L Torsion to n sags n n L GJ Ir n 2L GJ Ir Table 3.1). However. Figure 3.Dynamic calculation applied to footbridges 5. a certain number of situations may be determined analytically. or when the sections are wide (on the right). It is also often possible to have orders of magnitude of natural frequencies. the torsion modes are at high frequencies. In this case. the calculation of the natural vibration modes is made by using software for complex constructions (which is achieved. (See figure 3. etc.2.2 . as are the tensioncompression modes. "plate and shell" type modelisations may turn out to be necessary. EI the bending rigidity. as soon as there are several beams. only an analysis of the bending modes is pertinent. ES the tension-compression rigidity. If the sections are flexible in torsion (case of open profiles). the torsion modes must be taken into account. an analytical calculation is possible (see table 3.1 .1: analytical modes for a simply-supported beam S is the linear density of the construction (including the permanent loads and the varying loads). In such a case. If the sections are wide (slabs or double beams. or even in both cases. even in the case of complex constructions. for example).1). For a simply-supported beam of constant characteristics.1: Torsion mode (on the left) and differential bending mode (on the right) to be taken into account when the profiles are open (on the left). in practice.Footbridges . I r is the moment of inertia of the construction.Calculation of natural frequencies and natural vibration modes In practice. © Sétra . allowance must also be made for the differential bending or transverse bending modes.

2006 .2 illustrate the bending and torsion modes encountered on a footbridge with a width one-fifth of the span. Mode 1: Bending 1 sag Mode 2: Bending 2 sags Mode 3: Torsion 1 sag Mode 4: Bending 3 sags Mode 5: Torsion 2 sags Mode 6: Bending 4 sags Mode 7: Torsion 3 sags Mode 8: Bending 5 sags © Sétra .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 45 / 127 The following figures 3.Footbridges .

[8]).Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 46 / 127 Mode 9: Torsion 4 sags Figure 3.2 Mode shape of a wide simply-supported footbridge For any beam. one factor depending on the material E and one I factor depending on the section . gives the values of the factor depending on the support conditions: n Table 3.2: Influence of the boundary conditions on the natural vibration modes © Sétra . taken from the CECM 89 rules (Ref.2006 . S The following table. continuous over supports. with different boundary conditions.Footbridges . the natural EI which can be broken down into one factor S pulsations are always in the form n n L2 depending on the shape of the beam n . also known as turning radius.

3 On strip footings or piles. If this is very high.3. provide rigidity which will have a direct effect on frequencies. At the base of the support. These points cover the dynamic calculation itself. 5. But this is not always the case. it will. a certain number of questions should first be asked on the calculation assumptions and the model being created. In general.Footbridges . Welds or bolted connections lead to different damping. either its horizontal and/or its vertical components may be taken into account. Figure 3. on piles or a strip footing. only its mass is relevant.2. or for the modelling of more complex constructions on particular calculation programs.2 Supports Particular attention shall be paid to supports.1. intermediate connections will have a direct effect on frequencies.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 47 / 127 Before carrying out any digital calculation of the dynamics of a footbridge. as the supports are often more rigid than the footbridge deck and provide more damping. if yes. The links between the various elements must be described and taken into account: swivel joints. all the usual static structure modelling assumptions and techniques should be carried out. etc. but not that of its supports. and damping. This concerns. and thinking about the modelling of links in steel constructions. In most typical cases.2006 . the relevance of the finite element grid. In certain cases. The dynamic behaviour of the structure of a footbridge deck is often taken into account. in addition.3.1.1 Structure of the deck Is the deck topping participating or not? If not. there is the foundation. a foundation has a certain rigidity ksol Then there is the pier. The modeliser must take particular care with the following aspects and questions before starting the digital modelling of the construction: 5. due to the friction between the ground and the construction. Obviously. its flexibility becomes important and it may affect the natural frequencies of the overall construction. this simplification is permissible. © Sétra . The following list is intended to set out the main points that have to be watched carefully to do the modelling successfully. whether for simple cases using frequency and acceleration calculation formulae.2. it also provides a high level of damping. for example. which will have a direct effect on the overall structural damping. recesses. This foundation has certain flexibility linked to that of the ground.

2. useful for the calculation of the moment of inertia. the support fittings must not be forgotten. the thickness of one layer of which is e. with a shear modulus G and with n layers. not to confuse rotation inertia. size a a. and torsion inertia. The theoretical model and simplified diagram may become as follows: Ground Pier and foundation Support fitting mapp ksupport fitting Kstructure Structure K. A complete model.1.2006 .4: Very high pier: The rigidity of this type of support. For a support fitting in reinforced elastomer. 5. there is not necessarily a single stiffness for each of the above elements.Footbridges . as the mass has a direct influence on the frequencies of the modes. for each mode. M M mfondati mpile kpier kground Figure 3. held at its base and free at its top may be evaluated as kpile = 3EI/h3 in which h is the height. either standing still or walking on the footbridge. Finally. the horizontal stiffness is: ksupport fitting = G a2 / n e All these intermediate elements may modify the natural frequencies of the footbridge and their associated damping. nor their effect on the overall rigidity of the construction and on overall damping.3. For very lightweight footbridges. © Sétra . the mass of the pedestrians. will then be necessary. shall be taken into account.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 48 / 127 Figure 3.5: "Simplified" diagram of the construction down to the foundations In the case of continuous systems. the frequency of the mode in question is located within a varying range between two extreme frequencies f1 and f2. useful for the calculation of rigidity in torsion. Thus. Care must be taken for example. often digital. I the inertia and E the elastic modulus. It is an important factor that must be properly dealt with by the software by distinguishing it from the static effects of self-weight. these two frequencies may be some distance apart. one calculated with the mass of the construction and of the pedestrians (f1) and the other with the mass of the construction only (f2).3 Making allowance for the mass Certain dynamic calculation software remains confused as to making allowance for mass. In addition.

). etc. pendulum effects.1. pre-tension in cable stays or ties. Annexe 3 gives the analytical solution of the dynamic amplification in accordance with the relationship between the exciting frequency and the initial frequency of the footbridge.2. be taken into account: existence of an initial constraint that influences the natural frequencies (pre-stressing in pre-stressed concrete. shall. etc.2. large displacements. it is a question of adding a mass.Footbridges . of no danger to a crowd of pedestrians) and global modes. when calculating accelerations.2. The analytical calculation may be made mode by mode if the frequencies of the footbridge are known. and also its modal masses and modal stiffness.3. vibration of a slab unit on a footbridge. there must be coherence between the load case carried out and the mass taken into account (and several calculations made if necessary).2006 .Practical calculation of the loading response á â © Sétra . 5. likely to modify the dynamic model of the footbridge. These dampers may be incorporated directly into the digital model if the model allows them to be modelised.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 49 / 127 They should. thrust from arches. or in the case of a footbridge of a design that would not allow the saving by using dampers to be made). in most cases. finally..4 Making allowance for dampers The dynamic calculation may be carried out with an allowance for tuned mass dampers (in the case of a recalculation of an existing footbridge that has to be upgraded by the addition of dampers.3. design software generally allows this to be done easily. In the same way. spring. 5. therefore.1. stiffness of the ground. This graph reveals the two natural vibration modes thus degenerated and their associated amplifications (thus the damping).2 . materials. Kfootbridge the generalised stiffness of the footbridge mode under consideration and Cfootbridge its generalised damping. both be calculated. In this case.5 Particular features The following points. it is possible to calculate the natural frequencies of the following system for each of the modes: Cfootbridge Cdamper Mdamper Kfootbridge Mfootbridge Kdamper Figure 3. non-linearity due to the cables or ties. As. characteristics of the materials to be taken into account (concrete. etc.6: Schematic representation of a Construction – Damper system Mfootbridge is the generalised mass of the mode of the footbridge under consideration.). 5. damper system. distinction between local modes (for example.

in other words in a steady state. with m j t j M j . there is then resonance of that mode. especially. The response qj of mode j is much greater than the others and the global response. is close to: n X t i i 1 ! ä qi t j ! â á qj t # " ä is obtained by The amplitude of the dynamic response qj(t).Footbridges . regularly spaced. the load must be such that the amplitude of the force is of the same sign as the mode shape. the transitory response is not interesting and only the F 1 2 j t j 0 mj j As specified in the previous chapter. © Sétra . which is the number of fictitious pedestrians that are all in phase.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 50 / 127 Remember the equation governing the behaviour of the mode i with modal damping (see annexe 1): t [ i ][ F t ] qi (t ) 2 i i qi (t ) i2 qi (t ) pi (t ) mi Each of these responses qi is calculated separately from the others. As a result. in t other words that: 2 j q j static mj j F0 ) by the dynamic amplification factor The response obtained is therefore: Displacement: q j t mj F0 1 1 2 j 2 j j t F0 cos j t transitory response Acceleration: q j t mj j 1 2 j j t cos t transitory response It can be seen that the displacement and the acceleration are out of phase by a quarter of the natural period in comparison with the excitation. at the natural frequency of the footbridge. The global displacement response can be written: Xt static j X qj t and in acceleration: X t qj t j amplitude of the permanent response Remember that the various calculations set out in this guide are based on the notion of number of equivalent pedestrians. then the global response is deducted by recomposing the modes.2006 ä á 1 . at the frequency of one of the modes (mode j for example). whose action is in the same direction as the mode shape at each point. and. multiplying the static response (that obtained as if the load was constant and equal to F0 . 2 j is interesting. The speed is in phase with the excitation. after a transitory period. at the maximum resonance. If the function F t is harmonic ( F t F0 sin t ).

This is because these programs. Several programs are available that take the dynamic calculation into account. in practice.Footbridges .7: Sign of the amplitude of the load in the case of a mode with several sags In the case of torsion modes with several sags. intended especially for seismic calculations. dynamic calculation software must be used to evaluate the natural frequencies. Although most of them allow a calculation of the natural vibration modes and natural frequencies. enough to calculate the natural © Sétra .2006 . are. and in blue the zones in which the amplitude of the load is negative. the amplitude of the force must be of the shape shown in figure 3. it is.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 51 / 127 In the case of modes with several sags in the longitudinal or transverse direction. that make calculations dynamically.7: Mode shape Figure 3. and also the accelerations obtained with the various loads. this means that the amplitude of the force must have the shape shown in figure 3.8 Figure 3. there are fewer that allow the temporal calculation of the accelerations under varying loads. in general. and because. In the general case. Noted in red are the zones in which the amplitude of the load is positive.8: Sign of the amplitude of the load in the case of a torsion mode with several sags. in such calculations.

In practice.2. the result does not change) 5. Indeed. The second member of the modal equation is then written mj The amplitude of the acceleration response. i The modal load is then formed. Care must be taken with the notion of generalised mass. for the mode j under consideration. However. together with the value of the natural frequency.1 Programs calculating only the natural vibration modes and modal characteristics.3. One can either define a load that varies over time in the form q cos t . In the case of pedestrian footbridges. With this type of program. at the F iVij 1 fj Vkj position of grid k: 2 j mj 1 2 j i M iVij2 i Vkj (value independent of the standard selected for the modes. is then written simply. j M .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 52 / 127 vibration modes. band looking at the dynamic response at the end © Sétra . F i has the same sign as Vij. But the modal mass has no physical reality in relation to the generalised mass of the mode that is a proper representation of the vibrating mass in the mode under consideration. having previously determined exactly the natural frequencies. independent). the results that they provide can be used to calculate the maximum accelerations. Its fj .2. It is written Fi F i cos t . the values of the displacements of the mode shape Vij. it is not necessary to use the tricks set out above. generalised mass. is a however. often given by the programs. if every Vij is multiplied by the same constant. in the way explained below: 5.3. these programs aimed at seismic design are not necessarily relevant. while the modal mass.2006 . by writing fj i FiVij in which Fi represents the value of the force on the grid i. Because of the loading mode of the beams. natural frequency. the projection of the load onto the mode.2. t 2 In practice. for each grid i of the model. modal damping (if possible). the modal information should be recovered for each natural vibration mode within the range at risk: vibratory shape at any point of the mode. the program must be asked to provide. at the resonance. the expression of which is t j M j value used typically in seismic calculations to evaluate the number of modes to be taken into account to represent properly the response to a seismic stress. With programs enabling temporal dynamic calculations. which is different from modal mass. The t generalised mass is the value m j which is used in the modal equation of the j M j dynamics and depends on the standardisation that has been selected (its final result is. it enables the natural vibration modes of low modal mass in relation to the mass of the construction to be eliminated. in the light of the selected standard.Footbridges . The generalised mass can then be formed by mj M iVi 2j in which Mi is the mass of the grid i. and the application of standardised response spectra s that avoids the need for any real dynamic calculation. value is given in chapter 2.2.2 Programs enabling temporal dynamic calculations.

= / © Sétra .3: modal values for a simply-supported beam subjected to a load on a footbridge depending on the mode with n sags. the characteristic values of the footbridge can be easily calculated. Then the displacement obtained at the end of a sufficiently long length of time is determined (in other 1 in order to give the 2 j words when the vibrations have smoothed out) and multiplied by acceleration at the resonance of the point under consideration. Type of value Natural pulsations n Literal expression n 2 2 EI S L2 fn v max M max V max Natural frequencies Maximum deflection Maximum moment Maximum shear force Maximum acceleration 1 1 4 FL4 2 n n 4 EI 5 1 1 4 FL2 3 2 n n2 1 1 4 FL 2 n n 2 n2 2 L2 EI S Acceleration max 1 4F 2 n S Table 3. The second method. On the other hand. if the exciting frequency is not exactly equal to the natural frequency 1 1 is replaced by the dynamic amplification A( ) 2 2 2 n (1 ) 4 2 2 n. the term (with n). F represents the amplitude of the force per unit length. which is a risk for the calculation. one must ensure that the frequency of the load is exactly equal to the natural frequency of the construction. easier to implement and carrying less risk.2006 .3.Footbridges . and then apply to it the same load as previously. It can be noted that. without the temporal section) which is easier to define. consists of selecting for the dynamic calculation only the mode under consideration (it must be possible to deactivate the other modes). the time at the end of which the amplitude of the accelerations becomes just about constant). 2 j For a footbridge with constant inertia on two supports. but assumed to be constant (therefore q alone.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 53 / 127 of a sufficiently long length of time (in practice. as shown in table 3.

very expensive. a series of tests can be programmed if the construction has had a number of vibration problems on occasions during its existence. but probably required. however.2 . when the design has imposed the use of dampers that have to be validated. and an improvement is needed. is a major. large damping assumptions. it will be necessary to have dynamic tests carried out (see next paragraph) so as to validate these assumptions or dampers. On an existing footbridge. The construction programme will have to make allowance for this. therefore. tests will not necessarily have to be carried out.Footbridges . which should only be considered in very particular cases.3 . minimum) In addition. The following shall. The Owner shall. Such an operation should be considered for a new footbridge when the design work has not succeeded in showing that the construction fully satisfies the previous criteria.2006 . or the testing of an existing footbridge. but not guaranteed. specify: The category of the footbridge (I to IV) The acceptable comfort level (maximum.Examples of items to be inserted in the particular works technical specification for a footbridge It is recommended that this guide be annexed to the technical "works" specification of a footbridge and to state that the working design must follow the recommendations in this guide. In the case where dampers are specified right from the start.Dynamic tests or tests on footbridges The carrying out of dynamic tests on a new footbridge. it is strongly recommended to measure the frequencies and damping of the modes of the completed footbridge before the final sizing of the dampers.Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 54 / 127 6. testing 6. if studies have shown that the dynamic behaviour can only be ensured by the use of dampers. minimum) 6. medium. If the dynamic behaviour can be assured without dampers. 6. and which should only be carried out by contractors who are skilled in this field. in order to adjust such sizing if applicable. and in accordance with the recommendations in this guide. or when the dynamic behaviour is assured using damping assumptions that are greater than those recommended in this guide. medium. or that there are probably. Chapter 4: Design and works specification. operation.1 . © Sétra . be stated once again: The category of the footbridge (I to IV) The acceptable comfort level (maximum.Examples of items for a footbridge dynamic design specification The specification shall state that the structural design of the footbridge shall be carried out in accordance with the recommendations in the SETRA / AFGC guide on the dynamic behaviour of pedestrian footbridges.

certain modes may be forgotten.Footbridges . In all cases. at programmable frequencies covering the range 0. The tests shall proceed in the following way: 6. at the natural frequencies of the footbridge. jumping. in circles on the footbridge.2 . If there are dampers. © Sétra . kneeling. beyond which the tests must be stopped. as they enable the dynamic behaviour and the validity of the dampers to be validated. 6. the tests shall be carried out with the dampers working and then. and not too intolerable in the case of exceptional loading tests (jumping. and. Crowd tests should be programmed. (accelerometers or motion measurers). installation of a visual monitoring system of the footbridge so as to be able to connect the dynamic measurements to the image of the footbridge. and also to the complexity of the management of such a crowd. horizontal and/or vertical excitation (depending on the footbridge). The tests may include the following elements: Random walking.5 Hz / 3 Hz minimum. positioned at the sags of the theoretical modes.).Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 55 / 127 The following paragraphs give recommendations for the successful completion of such a series of tests. etc. The number of pedestrians is to be determined according to the size of the footbridge. the following systems shall be provided: installation of dynamic sensors. This system must allow reliable and controlled. Dynamic analysis remains possible. Possible use of a mechanical exciter (of the out-of-balance type for example) so as to be able to characterise precisely the natural vibration modes. This system will have to be synchronised with the data acquisition unit. vandalism. coming to a sudden halt. modal masses and modal damping of the footbridge. Running.Stage 2: Crowd tests. Marching in step. but less accurate. in order to measure the damping of the footbridge. in normal service. using suitable computer tools.Stage 1: Characterisation of the natural vibration modes This stage shall be carried out using the dynamic exciter if high accuracy is required of the values measured. kneeling. as a second stage.3. tolerable in the case of a synchronised crowd.3. in order to test the footbridge under extreme stresses. these recommendations can be adapted. As far as equipment is concerned. The tests shall be prepared so as to define the apparent alarm levels. There must be a sufficient number of accelerometers. the decision will lie with the Owner. or continuously. Marching in step. or even from one end to the other. Several dozen pedestrians will be necessary for a representative test. it is possible to measure the vibrations of the footbridge. to infer from them the natural frequencies and also the natural vibratory shapes. the extent of the phenomena encountered and also on what is required to be measured. natural frequencies. In this case. The tests will have been passed if the vibrations felt during the passage of a desynchronised crowd result in vibrations acceptable to the Owner. unless a dynamic exciter is used. together with a data acquisition unit and a recording system. In addition. it is possible to measure the damping using the logarithmic decrement method. but not the generalised masses.1 . Depending on the size of the construction. its theoretical characteristics. locked into position so as to determine their actual effectiveness.2006 . running. In the absence of a dynamic exciter. who must judge the comfort level of his footbridge according to his requirements.

Footbridges .2006 .Assessment of vibrational behaviour of footbridges under pedestrian loading – Practical Guidelines 56 / 127 © Sétra .

2006 .Dynamic behaviour of footbridges – Annexes 57 / 127 © Sétra .

a single parameter being enough to describe it.1) is subjected to the return force of the spring k x: if the stiffness k is negative.1. connected to a fixed point via a linear spring of stiffness k. In addition.2 .1). Its study enables the characteristic phenomena of dynamic analysis to be understood.A simple oscillator 1. We will still assume that the oscillator carries out small movements around its balance configuration: x(t) designates its position around this balance configuration. The resonance phenomenon can also be shown: the natural pulsation of a system will be defined.Dynamic behaviour of footbridges – Annexes 58 / 127 1. A simple oscillator comprises a mass m. 1. Finally.1 . and a linear viscosity damper c (figure 1. This oscillator is fully determined when its position x(t) is known. An exhaustive study is carried out on this oscillator: free damped / non-damped oscillation. The mass may be subjected to a dynamic excitation F(t). The "non-damped" case is that with a damping value of zero.1: Simple oscillator The mass m (figure 1.Introduction A simple oscillator is the basic element of the mechanics of vibrations.1. © Sétra . the spring will travel in the direction of movement: this will lead to buckling.Formation of the equations Figure 1. The "free" case is that of zero excitation. It is then said that the system is conservative. Appendix 1: Reminders of the dynamics of constructions 1.2006 . forced damped / non-damped oscillation.1 . it can be seen that the dynamic analysis of constructions reduces them to simple oscillators. it is said to have 1 degree of freedom (1 DOF): it is also called an oscillator at 1 DOF. a brief description of damping will be given.

3 . or greater than 1 (figure 1. the construction is unable to dissipate energy: this will lead to dynamic instability.2006 .1.Dynamic behaviour of footbridges – Annexes 59 / 127 the viscous dissipation force c x : if the viscosity c is negative.2) becomes: x 0 2 x x 2 0 F m (Eq. this force may be sinusoidal. It is typical and practical to divide the equation (1.2). therefore. the external force F(t).2) In order to solve such an equation. the study of this oscillator can be separated into two rates: the natural or free rate: the system is excited only by non-zero initial conditions: the free vibrations of the system are then obtained. random or commonplace. a differential equation with constant coefficients and one entry. be available: There is. We will study here only the case that is typical in practice: is strictly smaller than 1. 2 km The equation (1. 1. We thus get: k 2 f 0 : natural pulsation of the oscillator (rad/s). equal to. periodic. c : critical damping ratio (without dimension).1) which is rewritten in the form: m x c x x ( 0) x ( 0) x0 x0 k x F (Eq. The fundamental relationship of the dynamics is therefore written: m x F cx k x (Eq. different rates are obtained depending on whether is less than. the forced rate: only the force F(t) is non-zero.Free oscillation The system is set to vibrate only by non-zero initial conditions: 0 x 2 0x 0 x x(0) x0 x(0) x0 The solution of this differential equation with constant coefficients is of the type: x(t) = A e(rt) with r checking the characteristic equation: 2 r2 + 2 0 r+ 0 = 0 in which the reduced discriminant equals: = 02 ( 2 1) It is assumed that is positive (dissipative system) or zero. the initial conditions must. 1. f0 being the natural 0 m frequency of the oscillator (in Hz).3) 1. the system being linear. From a practical point of view. Thus. the force F(t).2) by m. obviously. 1. © Sétra .

there would be a solution that would become divergent: the oscillations would have an amplitude that would increase exponentially.Dynamic behaviour of footbridges – Annexes 60 / 127 Figure 1.2006 . if the damping was negative. Comment: It should be noted that.2: Free oscillations The solution is therefore: x(t) = A sin( 0 t) + B cos( 0 t) A and B are two constants to be determined with the help of the initial conditions: x ( 0) x0 B x ( 0) x0 A 0 It should be noted that the solution can also be written in the form: x(t) = A cos( 0 t + ) The integration constants are now A and the phase . it is then possible to obtain © Sétra . The solutions of the characteristic equation are: 2 i 0 1 r1. This is the most interesting case in practice.2 = 0 The general solution of the movement equation is written according to one of the three following forms: x(t ) B1 e 0 A1 e t ( 0 0 i 1 2 )t A2 e 1 1 2 2 ( 0 i 0 1 2 )t t 0 cos( 0 0 t ) B2 e t ) sin( 0 1 2 t) C1 e 0 t cos( The pulsation a is known as the natural pulsation of the damped system or pseudo-pulsation: 2 a = 0 1 The constants are obviously determined thanks to the initial conditions. which are also obtained with the help of the initial conditions. This problem is encountered in wind-excited constructions: the wind can induce a force proportional to its speed and in the same direction.

Hx.Dynamic behaviour of footbridges – Annexes 61 / 127 negative damping. with X( ) ~t . 1. 1. random or commonplace. From a mathematical point of view.5) is injected into the equation (1. also known as the frequency response. between the excitation and the response (here in movement) by the ratio: H x. 1. If the dynamic amplification is defined by: A( ) H x.3) is the sum of the general solution of the equation without a second member (transitory movement or rate) and of a particular solution of the full equation (permanent movement or rate) xSP(t).Forced vibration The objective is to resolve the equation (1. It can therefore be written in the following form: xt a Ae 0t cos t xSP t (Eq. With the help of the harmonic excitation.3). 1.5) x R+. by the end of a few normal periods it has already been cancelled out). periodic. This notion is fundamental to the mechanics of the vibrations. Various cases can be examined for F(t): F may be harmonic.4 . The solution sought is therefore the actual part of ~ t : x t x x (1. it is not taken into account. we get: F 2 e i ( 2 X ( ) 2i 0 X ( ) 0 X ( )) 0 m The transfer function is then defined.3).6) = 2 0 This function characterises the dynamics of the system under study. the general solution of (1. 1. After simplification by the factor ei t. F ( ) X ( )e F0 / m i 2 1 2i 0 2 0 (Eq. F ( ) X( )e i F0 / k X( )e xstatic i / 1 1 2 0 is placed.F( ). A harmonic excitation is a sinusoidal function: F(t) = F0 cos( t) = (F0 ei t) A particularly interesting solution is in the form: x(t) = A cos( 0 t + ) In order to carry out the calculations more simply and more quickly. (Eq. as soon as this proportionality factor becomes greater than the damping of the construction.7) 2i © Sétra .2006 .4) The constants are determined with the initial conditions. the notion of transfer function will be introduced.1. The transitory rate is associated with the free movement: its influence quickly becomes negligible (in fact. complex numbers are used. A solution is therefore sought in the form: ~ (t ) X ( ) ei ( t ) (Eq. During the study with harmonic or periodic excitation.

which tends towards be seen from figure 1. this response is greater (or even much greater) than the static response. the maximum response of the oscillator will depend effectively on the frequency (figure 1.5 %.e. As far as the phase is concerned. Figure 1. at a given F0.3): A( ) allows a maximum for 1 1 2 1 R ) shows that. This is why it is so important to determine the resonance frequencies of a construction as soon as there is dynamic excitation. the resonance phenomenon is = 1 2 2 which equals: A( 2 R ) It can be seen that.3): for certain frequencies. with a rigidity of k: it is important to note that this value is taken by X( ) at zero frequency. resonance is reached when it equals /2. and thus a high level of stresses in the construction. if is between 0 and 1/ 2 . © Sétra .3 an amplification in the order of 100. it can 2 Comment 1: Using this oscillator at 1 DOF. Thus the intercept ordinate of the function Hx. the displacement resonance pulsation is combined with the natural pulsation of the system. the higher the 1 : thus. for = 0. only the amplitude of the excitation affects the amplitude of the response.2006 .3: resonance phenomenon The complex solution is therefore: F0 ~ x (t ) m ei 2 0 2 t 2i 0 A study of A( obtained (figure 1. However. for systems without much damping.Dynamic behaviour of footbridges – Annexes 62 / 127 xstatic is the value of the displacement that would be obtained if a static force F0 was exerted on the oscillator. Comment 2: It can be seen that resonance is reached at a pulsation value that is less than the natural pulsation of the system 0. tends towards 0). the lower the damping (i. This function A( ) shows that. illustrates immediately the difference between "static" and "dynamic". In static.F is determined with a static calculation: this value is often called static flexibility. in dynamic. frequency also has to be taken into account: exciting a construction to its resonance may produce (especially if the construction does not have much damping) major displacements. the more amplification to the level of resonance.

3).3). response transitory response transitory permanent Figure 1.2006 . spectral power density. only be understood by the use of values characterising this process (average. no allowance has been made for initial or transitory conditions. which is expressed using Duhamel's integral: t 1 F ( ) e ( t ) sin( 0 1 2 (t ) ) d (Eq. the study of random vibrations exceeds the objective of this annexe. However. We assume that the excitation F is modelised by a stochastic process.F ( ) X( ) 2 2 m 2i 0 0 in which X( ) and F( ) are the Fourier transforms of x(t) and of F(t). Taking the Fourier transform of the equation (1. by determining the solution xp(t) for each harmonic p (previous paragraph). standard deviation. as has been commented previously. as can be seen in figure 1. this component of the solution quickly becomes negligible. the higher the damping. therefore. etc. using a computer program. The response x of the system at 1 DOF is also a stochastic process: it can. The load function can be developed as a Fourier series: 2 F (t ) Cn exp(i t) T p 0 Thus. autocorrelation function.Dynamic behaviour of footbridges – Annexes 63 / 127 Comment 3: in the above. we then infer x(t). as.4. However.). we get the frequential equation: F ( )/m F( ) H x. © Sétra . we get the solution by superposing (summing) the xp(t). and the reader is referred to the reference works.4: transitory and damping We return to the previous case. it is clear that. we carry out a digital integration of the differential equation of the movement (1. the quicker the transitory component becomes negligible. 1. By inverse Fourier transform.8) x(t ) m 0 1 2 0 In practice.

Dynamic behaviour of footbridges – Annexes 64 / 127 A movement u(t) is now imposed on the support: F(t) and the initial conditions are therefore assumed to be zero. 1.u ( ) U 1 2 X( )e 1 2i 2i i (Eq. By injecting these two expressions into the equation (1. NB: x(t) and u(t) designate displacements within an absolute reference. Sandow shock cords). We are only interested in the harmonic movement of the support: u(t) = U ei t We are looking once again for a solution in the form: X( ) ei ( t ). If a system is excited. compared with a base excitation It is interesting to observe the existence of a fixed point: . after this value low values of are preferred.3): 2 2 x 2 0x 2 0u 0 x 0 u We can infer the transfer function between the base displacement and the mass displacement m: H x .9) The modulus of this function is plotted in figure 1. However. it transmits a signal (displacement) to its support: this then produces a wave that is transmitted (ground. an identical study is carried out to that for the forced excitation. all systems are excited by their support. © Sétra . If the base excitation is periodic or commonplace. etc. Figure 1.5: FRF in movement. wall.5. it is seismic excitation (therefore rare) that first comes to mind as a priority to illustrate a base excitation. air.2006 . Hx.u( 2 ) = 1 Because of this. This is the reason why we try and "isolate" them (elastic bases. during a base excitation a compromise must be found in order to optimise the damping: if high limit displacements up to 2 0. The base excitation is a much more common excitation than could be expected at first glance: indeed.) and causes the displacement of the support of another system.

6: Divergence induced by "negative damping" $ When an actual system is studied ( is apparently less than 1). its value is linked to temperature and the frequency of excitation damping by friction (Coulomb): it is linked to the joints between elements of the construction. In addition. which can lead to the collapse of the construction. which leads to oscillations of the construction that become greater and greater.Dynamic behaviour of footbridges – Annexes 65 / 127 1. there exist exciting forces that "agree" with the speed (certain modellings of pedestrian behaviour): this generates "negative damping". Experimentally. it can be seen that the critical damping ratio does not generally depend on frequency: the damping is then called structural damping: n= Traditionally. Instability then exists (see figure 1.1. thus.7): the period Ta between two successive maxima located at t1 and t2 (Ta = t2 t1): we then get: © Sétra .5 . 0 and . the maximum amplification is directly linked to the damping. The following are then measured (see figure 1. The most common sources of damping are: the internal damping linked with the material itself. it is induced by the relative displacement of two parts in contact. it becomes essential to identify its characteristic mechanical parameters. This can be done by means of a free vibration test: initial conditions are imposed on the system and its temporal response measured.6).2006 . response Figure 1. It is interesting to know that.Damping We have seen it: the resonance phenomenon causes displacements in constructions and. It is therefore essential for this parameter to be estimated properly in order to achieve correct dynamic sizing. stresses (very) much higher than would be obtained by a static calculation. sometimes. modelised by a simple oscillator. this damping is modelised by viscous damping: this is interpreted as a force that opposes the speed of the construction.

by carrying out a swept sine test.8). is the difference between 1 A( ) max A( 1. at -3 decibels. therefore.Dynamic behaviour of footbridges – Annexes 66 / 127 2 = (2 )/(Ta). this graph enables the bandwidth to be obtained and.e. 2 ) 2 It can be shown that the bandwidth is linked to the damping by the relationship: 2 = 2 2 (if the damping is low) 1 2 1 The more this bandwidth tends towards 0. As a result. to determine R (and therefore 0) and .: = Log (x(t1) / (x(t2)) (2 )/ 1 Therefore. the sharper the resonance and. The pulsation corresponding to the maximum of the graph gives R.2006 . © Sétra . 0 1 the value x(t1) and x(t2) of two successive maxima: their logarithmic decrement then be inferred: can exp ( ) 0 x( t1 ) exp( x( t 2 ) Ta ) exp( 1 2 2 ) I. therefore. In addition. . enable the mechanical 2 Figure 1. the lower the damping (figure 1. A( ) is plotted experimentally.7: response of the 1 DOF oscillator $ Definition: What is known as the bandwidth and 2 such that: . / (2 ) Thus the measurements of Ta and of the logarithmic decrement characteristics of a simple oscillator to be determined. if << 1 (which is typical).

a equivalent to can be defined in such a way as to conserve c x0 2 the energy dissipated per cycle: Wd = k x02 = Which leads to: 0 1. If the dynamic balance of the mass m is then written: m x + ks x = F(t) Under harmonic excitation. The general method consists of: writing the dynamic equations in a matrix form: the matrices of mass. However. This model consists of using a spring unit. whatever the DOF value n. © Sétra . the discrete systems at n DOF can be studied directly. from 2 DOF onwards. This model is very widely used to represent damping. During harmonic stresses.Dynamic behaviour of footbridges – Annexes 67 / 127 Figure 1.2 The damping used in our model is viscous damping: the damping force is proportional to the speed. in actual fact.2006 . the methods of understanding vibratory phenomena are the same. However. there are very few constructions that are subject to this type of damping: its use is due to its simplicity.Linear systems at n DOF Having described the simple oscillator. without loss of generality when the calculations are made.8: Bandwidth for = 0. studying systems at 2 DOF enables the calculations to be carried out up to the end: that is why. as. at a given amplitude of displacement. stiffness and damping are then defined. depends on the exciting frequency: experimentally. studying the non-damped system: the natural vibration modes and the modal base of the system are then defined. it can be seen that this is not the case.2 . When the stress is harmonic. the rigidity ks of which is complex: ks = k (1 + i ) in which is known as the structural damping factor. this damping model indicates that the energy dissipated in each cycle. Other models must be determined to overcome this problem. one simple advantageous model is the hysteretic damping model. they are carried out on systems at 2 DOF.

1. [F]: vector of external forces. Thus. in this chapter we will deal with the notions of mass. Thus.2. the notions of natural vibration modes of a construction. [K]: stiffness matrix.11) Which is therefore written: M X (t ) C X (t ) K X (t ) F (t ) (Eq.2006 .9: System at 2 DOF Obviously.9. 1.Dynamic behaviour of footbridges – Annexes 68 / 127 if necessary. The mass and stiffness matrices may be found using energy considerations: © Sétra . and of linking between the DOF. n simple oscillators have to be used. 1.. the displacement of a mass is dependant on the displacement of another mass: it is said that the n masses are coupled by means of the springs and the dampers.. For the system represented in figure 1.10) This system may easily be written in matrix form: C C C M x x1 (t ) 0 1 2 2 1 1 (t ) C C C x (t ) x (t ) 0 M K1 2 2 2 2 3 2 M C K2 K 2 K K2 K 2 K 3 x1 (t ) x (t ) 2 F1 (t ) F2 (t ) F (t ) (Eq.1 .12) This matrix equation recalls the equation governing the oscillator at 1 DOF.9): if this is not the case.n of the n masses. the dynamic equations are written: M 1 x1 (t ) (C1 C 2 ) x1 (t ) C 2 x 2 (t ) M 2 x 2 (t ) (C 2 C 3 ) x 2 (t ) C 2 x1 (t ) (K 2 ( K 1 K 2 ) x1 (t ) K 2 x 2 (t ) F1 (t ) K 3 ) x 2 (t ) K 2 x1 (t ) F2 (t ) (Eq. Figure 1.Formation of the equations Systems at n DOF are formed of n masses connected together by springs and dampers (figure 1.. the n DOFs are the positions xi(t)i = 1. The following names are used: [X]: vector of the DOFs. [C]: viscous damping matrix. [M]: mass matrix. determining the temporal solution (by going into the modal base. stiffness and damping matrices. for example).. 1.

For a system at n DOFs this typical equation has 2 n solutions: i k. The most frequent choices are the following: © Sétra . This particular case ([C] and [F] zero) is fundamental to the study of dynamic systems: it demonstrates the notion of natural vibration modes.Non-damped system It is assumed here that the damping matrix is zero.2. which is the sum of the kinetic energies of each mass: T = ½ M1 x 12 + ½ M2 x 22 = ½ t[ X ] [M] [ X ] The stiffness matrix is a matrix associated with the strain energy J of the system. n in which the k and the k are determined by the initial conditions. 1. Solutions are sought in the form: [X] = [ 0] er t By injecting into the above equation.. A modal vector is defined to within one multiplicative constant: a standard therefore has to be chosen for these vectors. The matrix equation (1...e.e. The components of the vector k are obviously linked.11) thus becomes: (Eq.2006 .9: K1 K 2 r 2 M 1 K2 01 0 0 K2 02 K2 K3 r 2M 2 This system permits the zero solution that cannot be satisfactory as soon as there are non-zero initial conditions. we get: ((K1+K2)+r 2 M1) ((K2+K3)+r 2 M2) K22 = 0 k = 1. the determinant of the system must be zero. We can therefore infer from it the following equation in r: (Eq.n. 1. 1. according to: [K + r2 M] [ 0] = [0] I. The solution is then written in the form: i kt i kt X (t ) k k e k e k 1.13) [M][ X ] + [K][X] = [0] The initial conditions are therefore not zero. solution of the system: 2 [K k M] [ k] = [0] I... To obtain an advantageous solution. for the system at 2 DOF above. the homogenous system at 0 i is obtained. The natural vectors [ k] are the modal vectors of the system.e. i. for the system at 2 DOF in figure 1.e. for the system at 2 DOFs: K1 K 2 2 k With each index M1 K2 K3 K2 2 k k1 K2 M2 k2 0 0 The pulsations k are known as the natural pulsations of the system.14) Det ([K + r2 M]) = 0 I.Dynamic behaviour of footbridges – Annexes 69 / 127 The mass matrix is a matrix associated with the kinetic energy T of the system.. therefore. there is an associated natural vector k. otherwise the system would obviously remain at rest.2 ... with the elastic potential energy of each spring: J = ½ K1 x12 + ½ K2 (x2 x1)2 + ½ K3 x22 = ½ t[X] [K] [X] These matrices are. symmetrical matrices associated with levels of energy.

non-damped. In summary.2006 . under free excitation: x1 (t ) 1 1 2 2 1 e 2 3 4 i t ei e t i t ei t i 1t i 2t ) 22 ( 3 e i 2 t ) x2 (t ) 21 ( 1 e i 1 t 2e 4e These expressions are real: x1 and x2 can be expressed in trigonometric form: x1 (t ) a1 cos( 1 t ) a2 sin( 1 t ) a3 cos( 2 t ) a4 sin( 2 t ) x2 (t ) 21 ( a1 cos( 1 t ) a2 sin( 1 t ) ) 22 ( a3 cos( 2 t ) a4 sin( 2 t ) ) The constants are determined using the initial conditions. j t[ i] [K] [ j] = ki ij in which ij is the Kronecker symbol and mi and ki are the mass and the generalised stiffness associated with the mode i: these values will depend on the standard of the natural vectors selected.Dynamic behaviour of footbridges – Annexes 70 / 127 If we select the first normalisation: [ i 2i one of the components is made equal to 1. The couple ( i . Once again. a standard is set according to the mass matrix: t[ i] [M][ i] = 1 1 ] It is now possible to express the general solution of the system at 2 DOF. before passing to a general forced excitation. j t[ i] [M] [ j] = mi ij i. i) is called a natural vibration mode. The natural forms of a discrete system are the natural vectors i associated with it. It is considered. we are interested in the harmonic excitation: it is thus that the notion of transfer function is generalised. therefore. in such a way that all the components of the vector of the DOFs are synchronous (they reach their maxima and their minima at the same time): X mod e 0 ei 0 t 0 The natural pulsations i are the square roots of the natural positive values of the matrix [M] 1 [K]. that the construction is excited by a force vector of which each component is harmonic and in phase with the other components: [F(t)] = ei t [F0] It is assumed therefore that the solution is in the form: [X(t)] = ei t [X0] The matrix equation of the dynamic therefore becomes: 2 ([K] [M]) [X0] ei t = ei t [F0] © Sétra . The modes are sought directly by writing: natural pulsation i: this is all of the solutions to the equation: 2 det([K i M]) = 0 natural vector [ [K 2 i i]: M] [ i] = [0] The natural vibration modes are orthogonal in relation to the mass matrix and in relation to the stiffness matrix: i. non-damped problem. each mode is unitary: i = 1. a mode of a linear system is one particular solution of the free.

This leads to a differential system of linked equations. the response becomes infinite. interprets the influence of the component j of the force on the component i of the displacement.15) where: [ ]= [ 1 2 n]: is a square matrix in which the columns are the natural vectors. In theory. we have: 0 0 0 mi 0 0 0 q(t ) 0 0 0 ki 0 0 0 q (t ) t F (t ) Thus. if all the components of [F0] are zero except the x component j. or even its frequency response function (FRF). as the two matrices t[ ][M][ ] and t[ ][K][ ] are diagonal. [q(t)] is the vector of new variables. we get: [ ][ M ][ ] [q (t )] t [ ] [ K ] [ ][q (t )] t [ ] [ F (t )] t In view of the orthogonality of the modes in respect of the mass and stiffness matrices. that form a base. ij. in the matrix equation of the dynamic. this peak is infinite. That is why it is standard practice to act in a different way: a modal method is used that unlinks the equations. we have: ij ( ) 0i F0 j This is not forgetting the transfer function of a system at 1 DOF between the excitation F0 j and the response x0 i. The principle is to be positioned in modal base: the solution [X(t)] is therefore expressed using a combination of the natural vectors. The change of reference is therefore carried out using the transformation: n [ X (t )] [ ][q (t )] i 1 [ i ] qi (t ) (Eq. We repeat that ij is the transfer function between F0 j and x0 i. known as modal variables. Remember the equation to be solved: [ M ][ X (t )] [ K ][ X (t )] [ F (t )] In order to determine the components of [X(t)].Dynamic behaviour of footbridges – Annexes 71 / 127 From which: [X0] = ([K] 2 [M]) 1 [F0] = [ ( ) ] [F0] the matrix [ ] is called admittance matrix or dynamic influence factor matrix: this is a response matrix at a frequency between the forced excitation [F] and the response [X(t)]. that would simplify their solution: there would be n differential equations of second order to solve (see previous chapter). If [X(t)] is replaced by its expression (Eq. 1. If the equations to be solved were not linked. we get a system of unlinked differential equations of the type: © Sétra . This also requires the inverse of the mass matrix to be determined (high digital cost if the number of DOFs is high).15). it is always possible to carry out a temporal (digital) integration of the equations. We can demonstrate the resonance phenomenon: if it is excited at a natural frequency. The general term of this matrix. This arises from the (unrealistic) assumption that the system is not damped. 1. Comment: There is a resonance peak for each natural frequency of the system. and if this equation is pre-multiplied by t[ ].2006 .

We will even see that this might be essential to make allowance for dissipation effects. In the following. the modal method would become more attractive. Particular case: For a harmonic excitation. We thus have to refer to the chapter on the oscillator at 1 DOF: Duhamel's integral allows each qi(t) to be obtained. obviously. we will apply the modal method to our system: we will then have to discuss its validity. the modal vectors would appear not to be orthogonal in respect of the damping matrix: t constant × ij ij = [ i][C][ j] Thus.16) The vector [p(t)] is the vector of the modal contribution factors of the force [F(t)]: it is the component per unit of generalised mass of [F(t)] in the modal base. However. following the same route as in the previous paragraph: [ ] [ M ] [ ] [q (t )] t [ ] [C ] [ ] [q (t )] t [ ] [ K ] [ ] [q(t )] t [ ] [ F (t )] t Although the matrices t[ ][M][ ] and t[ ][K][ ] are.15). still diagonal. The equation of the problem is thus the equation (1. the study turns from a system with n DOFs to a study of n simple oscillators.2006 . Its vector [X] can then be inferred by modal superposition.12) on the basis of natural vectors. 1.3 . 1. the couple ( i. it is possible to integrate the equations directly. If the linking due to damping was zero. the projection on the basis of natural vibration modes does not unlink the equations from the damped system: the modal method seems to lose a lot of its attraction for determining a digital solution of [X]. Using the principle of modal superposition. only the viscous damping is considered. in accordance with equation (1. [ i]) designates the mode i of the equivalent non-damped system.Damped systems In order to simplify the study.2. We will break down the equation (1. the same does not necessarily apply to the matrix t[ ][C][ ].12) already described: [M][ X ] + [C][ X ] + [K][X] = [F] Here again. This is only realistic if the distribution of the damping is similar to that of the mass and of the rigidity: such an assumption does not appear to be justifiable.Dynamic behaviour of footbridges – Annexes 72 / 127 t q (t ) i i 2 i q (t ) mi p i (t ) i F (t ) (Eq. we have: qi(t) = qi0 cos( t) with: qi 0 2 i 2 pi 0 From which n [X ] i i 1 pi 0 2 i 2 cos( t ) This expression clearly demonstrates the resonance phenomenon for each natural pulsation. © Sétra .

Thus. the solution to the problem. each critical damping ratio i must be recalibrated experimentally. traditionally. the principle of modal superposition is applied. very little is known about the distribution of the damping. 1. it can be shown that the diagonal damping assumption is reasonable: this is known as the modal damping assumption. Once again. However. in the form: [q] [ p] t [ q0 ] e i [ p0 ] e i t (Eq. if the construction is not very dissipative and the natural frequencies are clearly separated. It is normal to consider that the value of the critical damping ratio does not depend on the mode under consideration: i = This constant is then fixed by experiment or by the regulations. In this case. In practice.17) Each represents the equation of the movement of a simple damped oscillator under a forced excitation. if we take: i 2 i mi we find once more a system of uncoupled equations: i qi (t ) 2 i i qi (t ) mi 2 i qi (t ) pi (t ) t [ i ][ F ] (Eq. A resolution is carried out by modal superposition: we are looking for [X]. Caughey has shown that the natural vibration modes are orthogonal at any damping matrix that is expressed in the general form: N [C ] k 1 ak [ M ] [ M ] 1 [ K ] k 1 In the particular case where N equals 2. using modal variables: [X] = [ ] [q] Each modal variable qi is then subjected to the equation (1.18) [p] always being the vector of the factors of modal contribution to the force [F]: t t [ i ][ F ] [ i ][ F ] p0i e i t pi t [ i ][ M ] [ i ] mi The solution is therefore: n n [X 0 ] i 1 q0i [ i ] i 1 2 i 2 p0i [ i ] i2 i n i 1 i 1 mi [ 2 i i 2 ] t [ i2 i i ] i [F ] © Sétra . it can also be shown that the natural vibration modes of the non-damped construction are fairly close to the modes of the construction.Dynamic behaviour of footbridges – Annexes 73 / 127 In practice.17). it is then clear that. therefore. the natural vibration modes of the system are orthogonal in respect of the damping matrix. using the modes of the non-damped construction. in this particular case. The natural modes are therefore assumed to be orthogonal in respect of the damping matrix.2006 . In fact it is very rare to find a construction with damping that needs simple modelling. 1. The solution sought and the excitation are. Damping is often low if it is well spread out. proportional damping can be found: the damping matrix is expressed according to a linear combination of [K] and [M].

or even much less. the modal equations (1.17) become [ i ][ F ] 2 i qi mi and then the modal response is: © Sétra . than n. where N is less. for more thoroughness and accuracy. The response of an oscillator at n DOF is therefore expressed according to the response of n oscillators at 1 DOF % Thanks to the modal superposition method. 1. it can be shown that is possible to express the neglected modes according to the modes retained [X ] i 1 N N i N 1 & N n qi (t ) [ i ] mi mi i 1 [ i ] t[ ] i 2 i i 2 i [F ] [F ] qi (t ) [ i ] i 1 [K ] 1 [ i ] t[ ] Matrices with residual stiffness are known as Kres and residual flexibility Sres such that: t N [ i ] [ i ] [ S res ] [ K res ] 1 [ K ] 1 mi i2 i 1 It is very important to note that this matrix depends only on static characteristics ([K]) and on the first N modes: the neglected modes are thus taken into account without having to calculate them! Comment: The above expression comes from the modal decomposition of the inverse of the stiffness matrix: n [K ] 1 i 1 [ i ] t[ ] mi i 2 i n i 1 [ i ] t[ i ] ki t (Eq. by modal superposition: n [X ] [ i i 1 ] qi (t ) 1. we are led to the study of a system at a DOF under a commonplace excitation. If it is thought that these modes have an almost-static response.4 .19) If the system is loaded statically. where n is very large. it is normal to represent the physical DOFs by a linear combination of the first N modes (they are assumed to be ranged in increasing order of frequency). However.2006 . Such an operation must be carried out with certain precautions: a reliable criterion must be defined to justify this modal truncation.Modal truncation A complex construction may be modelised by n DOF. Also. it may be justified. The solution is then given by Duhamel's integral: t 1 q pi (t ) e i i ( t ) sin( i 1 i2 (t )) d i (t ) 1 i2 0 i which leads to the determination of the solution. it is essential to make allowance for these neglected modes. This approach will depend on the spectrum of the excitation: if it is restricted to a range of pulsation less than N.2.Dynamic behaviour of footbridges – Annexes 74 / 127 It can then be recognised that each term of this sum is the response of an oscillator at 1 DOF under harmonic excitation.

a displacement ~ vector [ X 0 ] : ~ ~ [ X (t )] [ X 0 ] d (t ) This vector must be kinematically permissible. we find once more equation (1. then the solution obtained is the first natural pulsation.2006 .Dynamic behaviour of footbridges – Annexes 75 / 127 t n n [X ] i i 1 i [ ]q [ ] i i 1 [ ][ F ] i 2 i i 1 n [ i ] t[ i ] 2 i mi mi [F ] [K ] 1 [F ] By identification. The strain and kinematic energies associated with this shape are then calculated. the displacement vector obtained under a static load proportional to the value of the masses is a vector close to the first natural vector. ~ The whole problem is to select a "good" vector [ X 0 ] in other words to have an idea of the natural vector. ' This is. i. This method is based on a principle of approximation of the field of displacement of the system being studied: we suggest the shape of the construction. to a certain extent.e. The first natural pulsation is then estimated by the ratio: ~ t ~ [ X 0 ][ K ][ X 0 ] 2 (Eq. here.2.5 .. 1.20) ~ i t ~ [ X 0 ][ M ][ X 0 ] ~ If the vector [ X 0 ] is the first natural vector. The principle is the same: we propose N vectors ~ { [ X 0 i ] }i = 1 N that are independent.Rayleigh-Ritz method We propose an energy approach in order to determine an estimate of the first natural frequency of a construction: this is the Rayleigh method. kinematically permissible: the displacement vector [X(t)] is then approached by: ~ ~ [ X (t )] [ X (t )] [ X 01 ~ X ] 0N d1 (t ) d N (t ) N i 1 ~ [ X ] d (t ) 0i i ~ [ X 0 ] [d (t )] n N We then get an estimate of the first N natural frequencies by solving the following problem: ~ ~ ~ ~ [ X ][ K ][ X ] [ X ][ M ][ X 0 ] 0 t 0 0 2 t i 0 N N N N If we suggest: ~ [K ] ~ [M ] t ~ t ~ [ X 0 ][ K ][ X 0 ] ~ ~ [ X 0 ][ M ][ X 0 ] © Sétra . 1. The closer one gets to the natural vector. In general. but also a close estimate of the natural vectors. the better the estimate.19). a generalisation of the Rayleigh method: it enables not only an estimate of the first N natural frequencies to be obtained (often N is much lower than n).

Thus.10. to the extent that N is (much) less than n.Conclusion To represent a system at n DOF by n simple oscillators: this is the fundamental idea of this chapter and of the mechanics of linear vibrations of discrete systems in general.3 .Continuous elastic systems 1. Figure 1.2.1 .2006 .3. There is a natural vector that matches each of these natural values: ~ i1 ~ i ~ iN We then get an approximation of the natural vector i by: ~ ~ ~ ~ ] [ X 01 X 0 N ] [ ~i ] [ X 0 ] [ ~i ] N [ j 1 i ~ [ X 0 j ] ~i j 1.Generally © Sétra .10: Modal approach to a system at 2 DOF 1.Dynamic behaviour of footbridges – Annexes 76 / 127 we therefore come to determine the values and natural vectors associated with the problem: ~ 2 ~ [K ] [M ] 0 This problem is more advantageous to solve than the original problem.6 . the ~i obtained are approximations of the first N natural pulsations i of the system. This is interpreted by the diagram in figure 1.

Dynamic behaviour of footbridges – Annexes 77 / 127

Real systems are rarely discrete. Thus, a beam is deformable at any point, has a certain inertia at any point and is capable of dissipating energy at any point. We are therefore in the presence of a problem of the dynamics of continuous environments: to the statics equations we must therefore add the term that matches the inertia forces. However, we must not believe that the study of discrete systems was useless: we will see that, from a practical point of view, the study of continuous systems leads finally to the study of discrete systems. Such an approach will necessarily be an approximation, as a discrete system has a finite number of DOFs, whereas a continuous system has an infinite number of DOFs. The dynamic study of a continuous elastic system requires: setting the vibrational problem (formulating the equations): determining the equation for the partial derivatives governing the field of displacement X (x,y,z,t): this is the equation of the movement, giving the equations expressing the boundary conditions: these equations may also be equations of partial derivatives, giving the initial conditions. solving the problem: calculating the natural modes of the continuous elastic system, calculating the response of the continuous elastic system. To do this, we assume that the systems studied check the following hypotheses: the continuous elastic system is only subject to small strains around its natural state, the continuous elastic system has (by definition) a law of linear elastic behaviour; its mechanical characteristics are E, its Young's modulus, and , its density, the continuous elastic system vibrates in a vacuum.

**1.3.2 - Formation of the equation
**

In this chapter, we study a beam (to be understood here in its widest sense): this is a rectilinear uni-dimensional continuous environment capable of deforming: by tension-compression (bar), by pure bending. We will study these two cases successively. We will only be studying bi-dimensional constructions with no real loss of generality: the tridimensional aspect only adds additional variables. In a general way, the field of displacement X(x,y,z,t) of a point M(x,y,z,t) on a beam is u ( x, y , z , t ) characterised by its DOFs: X ( x, y, z , t )

w( x, y, z , t ) ( x, y , z , t )

in which: u is the longitudinal displacement, i.e. along the direction of the beam, w is the transverse displacement (deflection), i.e. along a direction perpendicular to the beam, is the rotation around an orthogonal axis in the plane of the construction: this is the slope adopted by the beam, due to the bending moment, In addition, we are working within a reference linked to the beam: the axis of the beam is combined with (x ,x) and the system is in the plane (x,z)

(

(

(

(

(

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 78 / 127

We are only interested therefore in the component u(x,t) of X(x,t). The balance of a length of beam of thickness (assumed infinitesimal) dx and of sectional area S, subjected to: the normal force at x, the opposite of the normal force at x+dx, an external distributed force (pre-stressing) p(x) dx. leads to the equation of the movement of a bar of constant sectional area:

2

S t2 ES x2 p ( x, t )

u u

2

(Eq. 1.21)

We study a beam in pure bending. We use the Navier hypothesis: straight sections remain straight. We assume, in addition, that transverse shear is negligible (Euler-Bernoulli beam). We are only interested therefore in the component w(x,t) of X(x,t). The balance of a length of beam of thickness dx, of sectional area S and of inertia I, subjected to: the shear force and the bending moment at x, the opposite of the shear force and of the bending moment at x+dx, an external distributed force F(x,t). leads to the equation of the movement of a beam in bending of constant sectional area:

S t2 EI x4 F ( x, t )

2 w 4 w

(Eq. 1.22)

To solve movement equations, there have to be initial conditions (temporal) and boundary conditions (spatial). The boundary conditions are very important data, as they set the resonance frequencies of the construction. They are associated with: imposed displacements (geometric condition): housings, supports, hinges, etc. imposed forces: ends free (natural condition): normal force, zero shear, zero bending moment, etc. localised masses, localised stiffness. Example 1: Fixed beam in bending-tension-compression with x = 0 and free at x = L: x=0 x=L

**u=0 N=0 w=0 T=0 =0 M=0 Example 2: Supported beam in bending-tension-compression with x = 0 and supported at x = L : x=0 x=L N=0 N=0 w=0 w=0 M=0 M=0
**

Example 3:

(

(

(

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 79 / 127

u(L,t) m L

Figure 1.11: Fixed bar (E, S, L) in tension-compression with x = 0 with a localised mass m at x = L.

In order to obtain the boundary condition at x = L, the balance of the mass m is written: N ( L , t ) m u ( L, t ) I.e.:

E S u ( L, t ) m u ( L, t )

Example 4: L m w(L,t)

k

Figure 1.12: Fixed beam (E, I, L) in pure bending with x = 0 with a localised mass m at x = L and a transverse localised stiffness k at x = L.

In order to obtain the boundary condition at x = L, the balance of the mass M is written: T ( L, t ) k w( L, t ) m w ( L, t ) 0 M ( L, t ) I.e.: E I w( 3) ( L, t ) k w( L, t ) m w ( L, t ) w ( 2 ) ( L, t ) 0 It can be seen that, in the two examples above, there are six boundary conditions: Two associated with the tension-compression (second order equation) and four associated with the bending (fourth order equation).

1.3.3 - Natural modes of a continuous elastic system

In order to determine the natural modes of a continuous elastic system, we can use: the balance equation, the boundary conditions. One natural mode of a continuous elastic system is a solution u(x,t) to the problem that is written in the form:

© Sétra - 2006

The following solution is injected: u(x. therefore. A mode is thus a stationary wave: all points of the system vibrate in phase. As its solution cannot be a zero solution (all the constants zero).t) = (x) a(t) (Eq. 1. we get the general form of the natural shapes of a bar: ( x) A sin Thus. t ) a0 cos( t ) A sin E x B cos E x There therefore remains to be determined the integration and pulsation constants : a0 and are determined with the initial conditions.23) is injected into the balance equation without a second member and without damping. in order to have a physically acceptable solution. we have: E (x) a(t) constant 2 (x) a(t) The constant is necessarily negative. The general solution obtained will then permit the natural shapes (x) of the system to be determined. L. the solution of the characteristic equation then provides the natural pulsations of the system. physically unacceptable. the modal solution is: E x B cos E x u ( x. In order to determine the natural modes of the system. the integration constants are determined using boundary conditions: a homogenous system of equations is obtained.Dynamic behaviour of footbridges – Annexes 80 / 127 u(x. it can be shown that the temporal function is harmonic: a(t) = a0cos( t + ) A mode consists of the data of a natural shape (x) (shape of the construction) and a natural pulsation . ): 2 2 u u S 2 ES 0 t x2 After simplifications and separation of the variables. S. associated with the spatial value: integration constants appear. the equation (probably differential) is then solved. therefore. that a(t) is sinusoidal: a(t) = a0cos( t + ) In the same way. in order to determine the constants it will be necessary to impose the nullity of the determinant of the system: this condition gives the characteristic equation of the system. the sign of which is clear. Example: Mixed supported bar: the boundary conditions are written: © Sétra . B and are obtained with the boundary conditions. otherwise a solution is obtained that grows exponentially over time and is.2006 .23) In addition. the time and space variables are separated: there then appears a constant. We apply these principles to the two traditional cases: longitudinal vibration of a bar and bending vibration of a beam. It is found. A. the method is as follows: the solution (1.t) = (x) a(t) in the balance equation of a bar (E.

n 2L The natural shapes (defined to the nearest multiplicative constant) are therefore: 2n 1 A sin x n ( x) 2L In the same way as the discrete case.t) = (x) a(t) into the balance equation: 2 4 w w 0 S t2 EI x4 After simplifications and separation of the variables. a generalised stiffness kn and a generalised mass mn can be defined: mn kn 0 n L 0 L S ES 2 2 n ( x) dx ( x) dx We find the relationship: 2 n kn mn The following solution is injected: w(x. 2. the general solution of the above equation (therefore the shape) is: (x) = A sin(kx) + B cos(kx) + C sinh(kx) + D cosh(kx) We use the boundary conditions to determine the integration and pulsation constants. we have: ( 4) ( x) EI a (t ) 2 constante S ( x) a (t ) Here again. In addition. the sign of the constant arises from a physically acceptable solution.Dynamic behaviour of footbridges – Annexes 81 / 127 u (0. Example: Beam supported at each end: the boundary conditions are written: © Sétra . t ) 0 ES ( L. t ) 0 a (t ) (0) u N ( L. we find that a(t) is sinusoidal. in order to obtain a non-zero solution. the differential equation governing is of the fourth order: d4 S 2 ( x) 0 EI d x4 ( x) If we take k 4 S EI 2 .2006 . it must be: cos E L 0 We can therefore infer the natural pulsations from this: E (2n 1) n 1. t ) E S a(t ) ( L) x The following system can be inferred from this: 0 B E A cos E L 0 As B is zero.

to which a Rayleigh-Ritz method is applied.2006 . t ) M ( L. t ) B D B D 0 0 A sin(kL) B cos(kL) C sinh(kL) D cosh(kL) 0 A k 2 sin(kL) B k 2 cos(kL) C k 2 sinh(kL) D k 2 cosh(kL) As this system is homogenous. © Sétra .Introduction As has been mentioned. t ) E I w ( L.2 . finally. These methods are discretisation methods: we pass from an infinite number of DOFs for continuous systems to a finite number of DOFs. few modal analysis problems of continuous elastic systems have an analytical solution. That is why it is necessary to use approximation methods: direct mass discretisation method: the construction is approached by masses and springs. very important. t ) 0 0 E I w (0.4 . This method consists of: dividing the construction up geometrically. there can only be a non-zero solution if the determinant of the system is zero. t ) 0 0 0 Whence the system: 0 L S EI 2 2 n ( x) dx ( x) dx 1. 1. It also permits the mass and stiffness matrices to be obtained very simply. to the mode shapes: n x A sin n L A generalised mass and stiffness can also be defined: mn kn 0 n L w( L.4. t ) M (0.Direct discretisation of masses This method may depend heavily on the aptitude of the "modeliser" to discretise a construction: it is.4.Dynamic behaviour of footbridges – Annexes 82 / 127 w(0. Rayleigh-Ritz method: the spatial variation of functions is apparently postulated. however. finite element method: the construction is broken down into sub-sections. we can infer from that that characteristic equation which reduces to: sin (kL) = 0 The resolution of this equation leads to the natural pulsations: n2 2 E I n L2 S and. as it is widely used to carry out civil engineering modelisations (particularly for earthquake-protection design).1 .Discretisation of the continuous systems 1.

The choice of ki and mi is important and determines the accuracy that there will be in the natural frequencies. t) by the determination of a finite number of parameters {ai (t)}i = 1. Such a method leads to a diagonal mass matrix. the stiffness is determined by connecting 2 DOFs together: the rigidity matrix is not necessarily diagonal. the second at 2/3 of the length and the last one at the end. It is proposed therefore to find out about these continuous functions u(M. it is determined as soon as the continuous functions describing the DOFs of each point are known: translations. These are our unknowns to be determined: they are grouped together in the vector of generalised coordinates [a(t)].13: Direct discretisation of the mass of a bar 1. As far as rigidity is concerned. that each section "strictly follows" the DOF of the node.Dynamic behaviour of footbridges – Annexes 83 / 127 concentrating the mass of each divided part on a node (in general its centre of gravity) in the form of: mass acting in translation. therefore. the dimension n of the vector [a(t)] gives the number of DOFs of the discretised system.Rayleigh-Ritz method As the system to be calculated is continuous.N (comprising the vector [a(t)]) thanks to the approximation functions i(M): N t ~ u(M . t ) u (M . As a consequence. the DOFs of the discrete system are the "physical" DOFs associated with the point on which the mass is concentrated: it is supposed. A choice can be made for a division into three sections of the same length: ( ( Figure 1.13): the first node is located at 1/3 of the length. . It must be kinematically permissible and the whole construction. ) 0 3 2 1 0 ) © Sétra .3 . Each node is allocated half of the mass of each element to which it is connected. whence the importance of the choice of dividing up. over i(M): given spatial function.2006 . inertia of the mass acting in rotation.4. the stiffness of each spring is calculated from the rigidity of a bar element. rotations. t ) [ ][a (t )] i ( M ) ai (t ) i 1 with: ai(t): generalised coordinate. which occupies the domain V. Example: Discretisation of a fixed/free bar (figure 1.

and also the work of external ~ forces according to the field approached u . We do recall. we resolve: 2 [ K ][ ( x)] [ M ][ ( x)] Let us follow the procedure described in the introduction to this chapter: Calculation of the kinetic energy: T 1/ 2 V [M ] V u 2 ( M .2006 . however. by demonstrating the vectors of the generalised coordinates: T J Wext t 1 / 2 t [a(t )][ M ] [a(t )] 1 / 2 t [a(t )][ K ] [a(t )] [a(t )][ F ] t [ F ] [a(t )] The matrices [M] and [K] which then appear are the matrices of the mass and the stiffness of the discretised system: the governing equation [a(t)] is then: [ M ] [a(t )] [ K ] [a(t )] [ F (t )] In order to determine the natural modes ( . using the approximation functions i(M). [ (x)]). we find u at any point in the system. the system strain energy. a bar. t ) dV 1 / 2 t [a (t )] [a (t )] [ ] t [ ] dV We infer from this the expression of the mass matrix of the discretised system: [M ] V [ ] t [ ] dV L 0 In the simple case of a bar or a beam. Re-writing these energies in matrix form. we can then infer from them the balance equation: [ M ][a (t )] [ K ][a (t )] [ F (t )] The whole arsenal developed for systems at n DOF can then be applied: © Sétra . t )] dV t [a(t )][ F (t )] t [ F (t )][a (t )] Knowing all these matrices. we use the procedure described in the chapter on discrete systems with n DOF. the most typical energies: Tension-compression 1 / 2 E S L u ( x) 2 dx 0 Bending 1/ 2 E I L 0 L 0 w ( x ) dx 2 We can then infer from this the stiffness matrices: Tension-compression [ K ] E S Bending [K ] EI [ ( x)] t [ ( x)] dx L 0 [ ( x)] t [ ( x)] dx Calculation of the work: Wext t V [u ( M . the mass matrix therefore becomes: [M ] S [ ( x)] t [ ( x)] dx Calculation of the strain energy: The general calculation of the strain energy is not made here: depending on the RDM hypotheses.Dynamic behaviour of footbridges – Annexes 84 / 127 Thus. depending on whether the system is modelised by a beam. etc. we get different "simplified" expressions. when the generalised coordinates are known. The general method consists of: Writing the kinetic energy. t )][ f ( M .

it is normal to use polynomes for the functions i(x).4 . we choose: i(x) = x i with i 2. It is easy to check that these functions are all kinematically permissible. how do we choose the interpolation functions? The simple solution is to "break" the system: we turn therefore to the finite element method. if the natural vectors [ i] of the discrete system associated with the natural pulsations i have been determined. in general. response to an external force.Dynamic behaviour of footbridges – Annexes 85 / 127 calculation of the natural modes of the discrete system. the first n modes (close) of the continuous system are determined by the same natural pulsation i and by the mode shape ui(x): ui ( x) t [ ( x)][ i ] Comment 1: As many modes are determined as there are DOFs. a good approximation of the mode shape of the first mode. We can also determine a load vector by writing. We then apply the Rayleigh-Ritz method to each of the elements: we calculate the kinetic and strain energies for each element. for example. Once the discrete problem has been solved. we can return immediately to the continuous system. called here interpolation functions. Comment 3: When the system being studied is formed from beams. in matrix form.2006 . the work of the external forces applied to the system. We can then infer the kinetic and strain energies for the whole construction: we use the additivity of the energy. are selected in such a way that the generalised coordinates are the values of the displacements of the nodes: the generalised coordinates thus have a physical meaning that enables a continuous displacement field to be obtained over the construction. © Sétra . In practice. this method is known as Rayleigh's method. We can therefore study the response of a continuous system subjected to any load. it can be estimated that only the first n/2 modes are determined with sufficient accuracy. The construction is sub-divided into fields of a simple geometric shape (of which we know the analytical expression). two beams rigidly fixed at one end but non-collinear: in this case. the contour of which is supported at points (the nodes): these are the elements. 1. Comment 4: When n = 1. Comment 5: The static shape of the construction subjected to a load proportional to its selfweight gives. etc. such as. Thus.4. the same does not necessarily apply to systems that are hardly more complex. These elements have a sufficiently simple shape for it to be possible to determine functions that are kinematically permissible on each of these elements: these are generally polynomes. For example for a fixed/free beam.Finite element method Although it is easy to determine the functions i(x) for a straight fixed/free beam. NB: The approximation functions. The estimate of the natural pulsations by the Rayleigh-Ritz method is always made by excess. We can then infer the mass and stiffness matrices of the construction: we turn once again to the study of a system with n DOFs. Comment 2: Accuracy is improved when the number of DOFs is increased.

etc. which can be described sufficiently as F(t). are set out below for the various components of F in the case of walking and of running. that the values one can find in the literature are approximate. for the vertical component of walking. apart from the individual action of a pedestrian. we will indicate the recommendations specified in the Eurocodes covering dynamic models of pedestrian loading. swaying. experimental measurements show that it is preferable to make allowance for at least the first three harmonics to represent the stresses correctly.). deliberate excitation. in particular as far as the phase © Sétra . it being possible to break down the horizontal component into a longitudinal component (along the axis of the footbridge) and a transverse component (perpendicular to the axis of the footbridge). In addition. but the same does not apply when a person starts to run (figure 2. while. the force created comprises a vertical component and a horizontal component. These two types of loading must be treated separately. For activities without displacement (jumping. Appendix 2: Modelling of the pedestrian load Pedestrian footbridges are subjected mainly to the loads of the pedestrians walking or running on them. we will consider making allowance for a group of people (crowd) and also the stresses connected with exceptional loadings (inaugurations. [17]) Whatever the type of load (walking or running). Finally.1). when they exist. It should be noted. there is always one foot in contact with the ground. as there is a difference between them: when walking. arising from experimental measurements.1: changes of the force over time for different types of step (Ref. Figure 2. The values of the amplitudes and of the phase shifts.Dynamic behaviour of footbridges – Annexes 86 / 127 2. etc. however.).2006 . we can restrict ourselves to the first harmonic.

2. as the front foot touches the ground before the back foot leaves it. we have G1 = 0. orders of magnitude corresponding to an average displacement. in other words n = 3. etc. for slow walking (1 Hz). 2. therefore. In the same way. [3]) For normal (non-obstructed) walking. to differences between one person and another. to uncertainties when taking measurements and. this being due.1 G0): G 1 = 0. Figure 2. As can be seen in figure 2. the other values being unchanged. the first three terms are taken into account. © Sétra . Comment: for a frequency of walking of 2.Dynamic behaviour of footbridges – Annexes 87 / 127 shifts ( i). the vertical component.22 depending on the authors). the shape of the stress changes between walking and running. for one foot. the first maximum corresponding to the impact of the heel on the ground and the second to the thrust of the sole of the foot (see figure 2. G2 = G3 0.2). the values provided are.1.1G0. the factors of the terms of greater order being less than 0. /2.). At the average frequency of 2 Hz (fm = 2 Hz). the values of the Fourier transform coefficients of F(t) are as follows (as remarked previously. fast.4 G0 . the frequency may be described as a Gaussian distribution of average 2 Hz and standard deviation 0.1).1 G0 .Vertical component In normal walking.Walking Walking (as opposed to running) is characterised by continuous contact with the ground surface. the values of Gi and of i for the terms beyond the first are hardly precise. has a saddle shape. and also according to the type of walking or running (slow. corresponding to "normal" walking or running. until it is reduced to a semi-sinusoid when running (see figure 2.1. The amplitudes Gi are therefore linked to the frequency fm and the values indicated subsequently are therefore average values. on the other hand.2: force resulting from walking (fm = 2 Hz) (Ref.1 . the recommended value of G1 is 0.175 to 0. on the one hand.20 Hz approximately (from 0.2006 . 2= 3 As can be seen above. This shape tends to disappear when the frequency of the walking increases.5 G0.4 Hz.1 .

a static part (no constant term in the expression of F(t)). the change in F(t).Dynamic behaviour of footbridges – Annexes 88 / 127 We have plotted on figure 2. The transverse component. a fundamental aspect when calculating response. etc. the transverse and longitudinal components do not have. 2. We can see that only when taking the first three harmonics into account can the saddle shape be found. of course. but cannot. the solution generally used is to modify the presentation in the following form: n F (t ) i 1/ 2 Gi sin 2 if m t i therefore being able to have (non-whole) values of 1/2. of less intensity. © Sétra . for a person of 700 N walking at a frequency fm of 2 Hz and over 1 second.3. occurs. This was. Apart from the shape of the signal. Figure 2.4). two or three harmonics in the previous development. be neglected as it can be a source of problems. etc. 3/2. This model has since been deleted from the Eurocode.4 G0 with G0 = 700 N. the weight of the pedestrian. corresponding to changing from one foot to the other when walking. the difference lies in the frequential content of the excitation. as opposed to the vertical force. leading. fm. for example. the first term (dynamic) of Fourier's transform. As announced previously. therefore. 3fm/2) according to the basic frequency of walking fm. we will investigate successively being moved transversely and longitudinally (although longitudinal movement is more rarely a problem on footbridges). however. in the annex to Eurocode 1 covering dynamic models of pedestrian loads. In addition. 2.1. to 0.2006 . the longitudinal component is mainly linked to the frequency of walking (see figure 2. Finally. In order to present Fourier's transform of the transverse component (at the frequencies fm/2. It is apparent that people are very sensitive to being moved sideways and walking is rapidly disturbed. 1. at a frequency of half that of the frequency of walking (1 Hz for fm = 2 Hz). taking into account one.Horizontal component The horizontal component of the load is. as the phase shifts are close to 0. admittedly. to a feeling of insecurity or even the closing of the footbridge. they do not therefore appear in the previous expression.3: walking – vertical component (fm = 2 Hz) It should be noted that.2 . in fact. to keeping close to the handrail. the recommended load was (DLM1): Q pv 280 sin(2 f v t ) which corresponds. but its physical reality remains no less relevant. therefore. On the other hand.

As for the vertical component (G0 = 700 N.5 the transverse component. we have marked on figure 2.05 G0 . taking into account from one to four harmonics.Dynamic behaviour of footbridges – Annexes 89 / 127 Figure 2.01 G0. a frequency corresponding to the lateral oscillations of the centre of gravity of the body when walking. it can be seen that. G1 = G2 0. fm = 2 Hz). the frequency is half that of the vertical component (in other words the period is double) and. [22] As indicated above. the transverse (b) and longitudinal (c) components of the displacement (and of the acceleration) measured – walking at 2 Hz Ref. As expected. tests have shown that the main amplitudes of this component are located at a frequency of about half that of the vertical component. on the one hand. The corresponding values of the Fourier coefficients are thus as follows: G1/2 = G3/2 0.4 (b).2006 . on the other hand.4: changes in the vertical force (a). that taking into account the first four terms gives an appearance close to those measured in figure 2. © Sétra .

2 G0 .04 G0 .6. in practice.6: walking – longitudinal component (fm = 2 Hz) It should be noted that. depending on the number of harmonics taken into account. Its oscillations correspond.1 G0 .5: walking – transverse component (fm = 2 Hz) The main frequency associated with this component is approximately the same as for the vertical component (fm = 2 Hz). flexible piers.2006 . The influence is greater when the footbridge bears on overhanging. G1 0. the values of the Fourier coefficients are: G1/2 0. the previous comments remaining valid. G3/2 0.03 G0 . Figure 2. little influence on most footbridges. © Sétra . for which the longitudinal component of the pedestrian leads to the bending of the pier. then to the thrust exerted subsequently. in general. initially to the contact of the foot with the ground. G2 0. For this component. under the same conditions as for the other components (G0 = 700 N.Dynamic behaviour of footbridges – Annexes 90 / 127 Figure 2. the change in this component. the longitudinal component of the walking force of a pedestrian has. for each step. fm = 2 Hz). We have shown in figure 2.

: the period of the contact.3. fm being the frequency of Figure 2. the vertical component of the load can be approached by a simple sequence of semi-sinusoids.2 .2006 . represented using the following expression: F(t) k pG0 sin( t t p) for (j 1)Tm t (j 1)Tm t p F(t) 0 for j 1 Tm t p t jTm with: kp j Fmax G0 tp Tm running. : the maximum load. : the period (Tm = 1/fm. which enables the expression of F(t) to be written also in the following form: F(t) k pG0 sin(2 f mt) for (j 1)Tm t F(t) 0 for j 1 Tm t jTm 2 j 1 Tm 2 © Sétra . impact factor according to the relative period of contact (b) Ref [3] The period of contact tp in this model is simply the half-period Tm.Running As indicated in the introduction.2.4(b)). 2.7 (a)). : the weight of the pedestrian.7: running – change in the force according to time (a). running is characterised by a discontinuous contact with the ground (see figure 2.5 Hz. etc. the frequency fm generally being between 2 and 3.Vertical component In an initial approximation.Dynamic behaviour of footbridges – Annexes 91 / 127 2. 2. : the step number (j = 1.).1 . noted Tp on fig. : the impact factor (kp = Fmax / G0).

As for the values of the impact factor kp. the Fourier series transform.7 G0 . these are inferred from figure 2. © Sétra .8: Change in the period of contact according to frequency Figure 2.Dynamic behaviour of footbridges – Annexes 92 / 127 This approximation of the period of contact tp is an over-estimation of the values measured experimentally. which has the advantage of not revealing explicitly the impact factor kp. a semi-sinus approximation leads to a weaker amplitude signal (with the value of kp used) than taking only one harmonic into account in the Fourier series.33) gives tp = 0. only the positive part of the transform is retained. We can see that the various graphs are close. on the one hand.6 G0 . according to the relative period of contact (tp/Tm). which are represented according to the frequency of running on the graph in figure 2. G2 = 0.9: Amplitude of the various harmonics On figure 2. but at a frequency fm = 3 Hz). with the previous notations: n F(t) G0 i 1 Gi sin 2 if mt for (j 1)Tm t (j 1 )Tm 2 F(t) 0 for (j 1 )Tm t jTm 2 In this case. a function of the frequency of running (figure 2.4 when tp/Tp = 0.8. which may be written. It is also possible to use. these coefficients being.2006 .2 G0. running at a frequency of 3 Hz (thus a period Tm = 0. More precisely. As an example. the phase shifts are assumed to be negligible and the amplitudes of the first three harmonics are average values. we can therefore see from this example that the semisinusoidal approximation is conservative. G3 0. for running. in strict logic.61).17 s and kp = 3 when tp/Tp = 0. a Fourier transform. on the other hand.5 (instead of kp = 2. taking into account 1 to 3 harmonics.9) : G1 = 1.10 are represented (still over 1 second with G0 = 700 N. In order to make allowance for the natural discontinuous contact when running. As announced.7 (b). the determination of which is delicate. Figure 2. as kp. a semi-sinus approximation and.

Dynamic behaviour of footbridges – Annexes 93 / 127 Figure 2.Horizontal component To our knowledge. It is this value that could be used if a specific analysis of runners had to be made. that. either of its longitudinal projection. it is logical to estimate that the frequency of the transverse component will be half the vertical component. on the one hand. 2. as for walking. no measurements have been taken of the horizontal component during a running race. it is reasonable to think.2006 . or of its transverse projection. In addition. This guide does not cover specific load cases of runners on a footbridge. the transverse component (that to which the public is most sensitive) has a low relative amplitude compared with the vertical component (the race appearing to be a more "directed" progress). © Sétra . However. during a race. as it is considered that the effects of a crowd of pedestrians are clearly less favourable.10: running – vertical component (fm = 3 Hz) This load according to time can be broken down into a Fourier series.2. revealing a constant section at about 1250 N and a varying section of the same frequency as the running for the first harmonic and of amplitude 1250 N. whereas that of the longitudinal component will be of the same order of magnitude. whereas the longitudinal component will be larger (larger motor force).2 .

1.2.1. the relative displacement of the outer plates in respect of the central plate produces shear stresses in the visco-elastic layer. cos sin( t ) sin cos( t ) (Eq. out of phase: (t ) (t ) 0 0 sin( t ) sin( t ) (Eq. 0 and however. 3.1 . be re-written: (t ) 0 0 0 0 depend. on the frequency 0 0 .1 . the strain (t ) and the shear stress (t ) are also at the frequency . but. When this type of device is installed on a construction. 3.2006 . © Sétra .Visco-elastic dampers The use of visco-elastic materials for controlling vibrations dates back to the 1950s. the area of which is the energy dissipated by the material per unit of volume and per cycle of oscillation. The shear stress may. in general.Principle The visco-elastic materials used are typically polymers that dissipate energy by working in shear. Appendix 3: Damping systems 3. when they were used to limit fatigue damage caused by vibrations on aircraft frames.) The parameters 0 .2). in general. the following stress(Eq.) G1 ( ) sin( t ) G2 ( ) cos( t ) 0 By replacing the terms sin( t ) and cos( t ) by (t ) / strain ratio is obtained: (t ) G1 ( ) (t ) G2 ( ) (t ) and (t ) /( 0) . Figure 3. The application to civil engineering constructions dates from the 1960s. Figure 3. which dissipates the energy.1 – Visco-elastic damper Under a harmonic load of frequency . 3.Dynamic behaviour of footbridges – Annexes 94 / 127 3.) which defines an ellipse (Figure 3. 3.1 shows a visco-elastic damper formed from layers of visco-elastic materials between metal plates.3.

Dynamic behaviour of footbridges – Annexes 95 / 127

**Figure 3.2 – Stress-strain diagram This energy is given by:
**

2 /

E

0 0 0

(t ) (t ) dt

sin

(Eq. 3.4.)

The equation (3.3) reveals an initial term in phase with the displacement, representing the elastic modulus; the second term describes the dissipation of energy. The damping G2 ( ) / leads to expressing the damping factor by: G2 ( )

2G1 ( ) tan 2

(Eq. 3.5.)

The parameters G2 ( ) and tan determine completely the behaviour of the visco-elastic damper under harmonic excitation. The term tan is called the dissipation factor. These factors vary not only according to the frequency, but are also function of the ambient temperature. As the dissipation of energy is in the form of heat, the level of performance of the visco-elastic dampers must also be evaluated in relation to the variation in internal temperature that appears when they are working.

**3.1.2 - Layout and design of dampers
**

The layout of the dampers is an essential parameter in their effectiveness in relation to the dissipation of energy. However, technical constraints may prevent the most appropriate places from being selected. The effectiveness of a visco-elastic damper is measured by its capacity to work in shear; it is, therefore, recommended that the dampers be set out in such a way that this condition can be checked. This type of damper has been little used for footbridges.

3.2 - Viscous dampers

3.2.1 - Principle

Dry or visco-elastic friction dampers use the action of solids to dissipate the vibratory energy of a construction (Figure 3.3). It is, however, also possible to use a fluid. The first device to come to mind is one derived from a dashpot. In such a device, the dissipation takes place by the conversion of the mechanical energy into heat, using a piston that deforms a very viscous substance, such as silicone, and displaces it. Another family of dampers is based on the flow of a fluid in a closed container. The piston is not limited to deforming the viscous substance, but forces the fluid to pass through calibrated orifices, fitted or not with simple regulation devices. As in the previous case, the dissipation of the energy

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 96 / 127

leads to heat being given off. Very high levels of energy dissipation can be reached, but require adequate technological devices. The main difference between these two technologies is as follows. In the case of a pot or wall damper, the dissipative force will depend on the viscosity of the fluid, whereas, in the case of an orifice damper, this force is due mainly to the density of the fluid. Orifice dampers will therefore be more stable in respect of variations in temperature than pot dampers or wall dampers. In the case of pedestrian footbridges, the movements are very weak and dampers must be selected that will be effective even for small displacements, in the order of one millimetre. Because of the compressibility of the fluid, friction in the joints, and tolerance in the fixings, this is not easy to achieve. Viscous dampers must be located between two points on the construction having a differential displacement in respect of each other. The larger these differential displacements, the more effective the damper will be. In practice, these dampers will be located either on an element connecting a pier and the deck a few metres away from the pier (for horizontal or vertical vibrations), or on the horizontal bracing elements of the deck (for horizontal vibrations).

Figure 3.3 – Example of viscous damper

**3.2.2 - Laws of behaviour of viscous dampers
**

As part of the study of the behaviour of a construction, it is necessary to have a macroscopic behaviour model of the damper. For this, it is traditional to use a force-displacement law as described by a differential equation of order (Ref. [55]) :

f (t ) dt d f

(t )

C0

dx (t ) dt

(Eq. 3.6.)

in which f (t ) is the force applied to the piston, and x(t ) the resultant displacement of the piston. The parameters C 0 , , represent respectively the damping factor at zero frequency, the relaxation time and the order of the damper. These parameters are generally determined experimentally, although approximations do exist to estimate them analytically on the basis of the characteristics of the materials (Ref. [56]). Notes: When 0 , the case for a linear viscous damper is made. It should be pointed out that it is often the most-used, as, obviously, it simplifies an analysis of the behaviour of the construction. Another formulation may be used: dx f (t ) C (t ) dt is a coefficient generally between 0.1 and 0.4. (Eq. 3.7.)

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 97 / 127

Note that, for certain fluids, the kinematic viscosity also varies according to the rate of shear of the fluid, which is itself proportional to the speed gradient. Viscous dampers were installed on the Millennium footbridge in London.

**3.3 - Tuned mass dampers (TMD)
**

A tuned mass damper (abbreviated to TMD) comprises a mass connected to the construction by a spring and by a damper positioned in parallel. This device enables the vibrations in a construction in a given vibration mode, under the action of a periodic excitation of a frequency close to the natural frequency of this mode of vibration of the construction, to be reduced substantially. The first developments of TMDs applied essentially to mechanical systems for which a frequency of excitation is in resonance with the basic frequency of the machine.

3.3.1 - Principle

Consider an oscillator with 1 degree of freedom, subjected to a harmonic force f (t ) . The response of this oscillator may be reduced in amplitude by the addition of a secondary mass (or TMD) that has a relative movement in respect of the primary oscillator. The equations describing the relative displacement of the primary oscillator in respect of the TMD:

M m y2 (t ) c y2 (t ) k y2 (t ) m y1 (t ) m y2 (t ) f (t ) y1 (t ) C y1 (t ) K y1 (t ) c y2 (t ) k y2 (t ) f (t )

(Eq. 3.8.)

**which is reduced to the single equation:
**

M m y1 (t ) C y1 (t ) K y1 (t )

(Eq. 3.9.)

**3.3.2 - Den Hartog's solution
**

In the case where the primary oscillator has zero damping ( C 0 ), it is possible to determine the optimum characteristics of the TMD for a harmonic excitation f (t ) of amplitude f 0 and of pulsation . This optimum solution is the solution deduced by Den Hartog in his first studies. For a harmonic excitation, the impact factor A, defined as the ratio between the maximum dynamic amplitude y1 max and the static displacement y1stat , is expressed by:

A

2 2

y1max y1stat 1

2 2 2 2

2

2 2

2

2 ada

2

2 ada

1

k m

2

2

2

(Eq. 3.10.)

where:

;

osc 2 osc ada osc

;

2 ada

(Eq. 3.11.)

K ; M

c

ada

2m

;

ada

m M

(Eq. 3.12)

The dynamic amplification factor A therefore depends on four parameters: , , , ada . Figure 3.4 gives the changes of A according to the frequency ratio , for various damping factors and 0,05 . This figure shows that, for the case where 1 (agreement of frequencies) and for the damping ada is zero, the response amplitude has two resonance peaks. Contrary to when the damping tends towards infinity, the two masses are virtually fused, so as to form a single

© Sétra - 2006

16) 3.15) opt 3 8 (1 )3 (Eq.3.14) opt The optimum frequency f opt and the optimum damping factor can be determined for the tuned mass damper ( 1 ). therefore. Figure 3.3 . 3. Between these extreme cases. The minimum amplitude of the resonance is thus obtained (P and Q) on the graphs A f such that these two points are of equal amplitudes. The analysis proposed by Warburton also enables the results to be (1) The optimum frequency is very slightly less than that of the construction. 3.05 M with an infinite amplitude at the resonance frequency. in the case where the construction does not have its own damping: this is Den-Hartog's solution (Ref.Dynamic behaviour of footbridges – Annexes 98 / 127 oscillator of mass 1.13) and the amplitudes at points P and Q are: R 1 2 (Eq. [51]): f opt 1 1 fs (Eq.2006 .05 et 1. of bringing the resonance peak down as low as possible. [54]) in order to determine the optimum characteristics of the TMD for several types of excitations (in the presence of weak structural damping). © Sétra . there is a damping value for which the resonance peak is minimal.General solution A more detailed analysis has been carried out by Warburton (Ref. This optimum by selecting the ratio ratio is thus given by the expression: opt 1 1 m/M (Eq. Figure 3. the added mass always being in the order of a few hundredths of the mass of the construction. 3. 3.4 shows the independence in relation to ada of two points .0 ) ( The objective of a tuned mass damper consists.4 – Impact factor depending on 0.

17) 3 1 4 opt (Eq3. Figure 3. minimising the impact factor A. to cases with several degrees of freedom.7 2. [53]) : 2 2 ~ opt opt opt 0. The optimum characteristics of Den Hartog's tuned mass damper for a harmonic excitation have been established.15 . such as the minimisation of the displacement of the construction.19) in which osc is the damping factor of the primary mass..241 1. which is much easier to adjust than the stiffness of the spring. the theoretical optimum frequency is very slightly weaker than that given by the above formula (valid for a harmonic excitation). © Sétra . the optimum parameters opt and opt are given by the expressions: 1 2 opt 1 (Eq.5 gives an overview of such a device. 3. Optimality is not therefore sought on a given mode (modal component).0 osc 0. than to the value of the added damping. of the acceleration of the construction or of forces in the construction. These TMDs may be linked together to damp torsion vibrations. This is why it must be possible to adjust the natural frequency of the mass when fixing the device. 3. In particular. this section will be limited to dealing only with the case of oscillators with 1 degree of freedom. etc. then seeking the solutions minimising this criterion. all for various types of excitations. In addition.03 The effectiveness of a tuned mass damper is much more sensitive to the natural frequency of the added mass.40 and 0.6 osc 2 1.5 . which is nevertheless sufficient in usual cases (Ref. in our case. 3.13 0. This is the very principle of the determination of an optimum TMD: first defining a criterion of optimality A.01 0. but on a generalised coordinate (displacement or derivatives).12 osc 0. These expressions show less than 1% 0. numerous other criteria can be selected.0 1.3.3. This is generally done by varying the value of the mass itself.9 3 2 2 osc (Eq. when the excitation is random (the primary system is subjected to a random force – in other words. The reader is therefore referred to reference 54 for further details on these analogies. Apart from the impact factor.4 0.9 2 osc ~ opt 0.Case of a damped primary oscillator In the case where the construction has natural damping.Examples There are various types of tuned mass dampers. the crowd). of the displacement of the damper.18) 41 2 1 3.Dynamic behaviour of footbridges – Annexes 99 / 127 used for primary oscillators with one degree of freedom.4 . error for 0. The most typical type consists of a mass connected to the construction by means of vertical coil springs and one or more hydraulic or pneumatic dampers.2006 . These analogies are based on a breakdown on the basis of the standardised natural modes of the construction. so as fine-tune it to the actual natural frequency of the construction.

The first TLD prototype was proposed in the 1900s by Frahm to control rolling in ships. c.6b) thus formed has characteristic frequencies that can be tuned to optimise one performance criterion. the liquid acts as the secondary mass and the damping is provided by friction with the walls of the container. Another category of dampers consists of replacing the mass. a second mass is attached to the construction using struts and dampers (figure 3. The response of the "container/construction" system is thus dependant on the amplitude of the movements. However. As for a traditional TMD. 1 m F k M F M K 2 K 2 C K 2 C K 2 (b) TLD (a) TMD with mass Figure 3. the secondary system (figure.6a). c m. Since the 1970s. But it has only been since the 1980s that building applications have been considered.2006 . although the parameters of a TMD can be optimised and analytical formulae provided. associated with a horizontal hydraulic damper.3. the non-linear response of the fluid in movement in a container makes such optimisation very difficult.6 – Comparison between TMD with mass and TLD with fluid The principles used for the sizing of TMDs also apply to TLDs.Dynamic behaviour of footbridges – Annexes 100 / 127 Figure 3. these dampers have been installed on satellites to reduce long period vibrations. may be considered. © Sétra . As the action of gravity comprises a recall mechanism. strut and damper by a container filled with a liquid.5 – Tuned mass damper for the Solferino footbridge When the vibration to be damped is horizontal. 3.4 . a device consisting of a mass attached to the bottom of a pendulum.Tuned liquid dampers In the technology of tuned mass dampers.

3. the same way as equations (3. to our knowledge. The equations of the coupled system are written in exactly 2 2m adaf adaf . no footbridge has had tuned fluid dampers installed. 3. The total fluid mass was 4770 kg and the tuning frequency was 0.7 – Description of a TLD Let us consider a rectangular container of length 2a filled with a liquid of viscosity and of average depth h.21) h .5 .Dynamic behaviour of footbridges – Annexes 101 / 127 z x h Boundary layer hb 2a xs Figure 3.255 Hz.2006 . b in which b is the width of the tank. in order to dampen the horizontal vibrations of the pylons. t ) is constant over this free surface (figure 3.8. k m adaf .) and (3. 3. It is assumed that the free surface remains continuous (no waves breaking) and that the pressure p( x. the resolution of the coupled problem is more complex due to the fluid nature of the secondary oscillator.9. the natural vibration frequency of the fluid is: adaf g h .7).22) This equation shows the analogy that can exist between a tuned mass damper and a tuned liquid damper. [57]) : adaf adaf 1 2h 1 (Eq. They were.Comparative table Type of damper Visco-elastic Field of use Very little used Advantages Damps several modes Disadvantages Needs to be fixed to work in shear © Sétra . considered for the Ikuchi bridge (Japan).20) The damping factor may be approached by the expression (Ref. The container is subjected to a horizontal displacement x(t). In reality. (Eq. z. This analogy only makes sense subject to the validity of the expressions (3.) with m adaf 2 abh . tanh 2a 2a (Eq.20) and (3. c in other words: 1 osc osc osc 2 adaf y1 y2 0 0 adaf adaf 2 2 0 y1 y2 0 2 y1 y2 f t m 0 . As opposed to numerous high-rise towers.21) selected as equivalent pulsation and damping factor. 3. The fluid is assumed to be incompressible and irrotational. however. According to the linear theory of the boundary layer.

damps a given mode.Dynamic behaviour of footbridges – Annexes 102 / 127 Viscous pot or wall dampers Viscous orifice dampers Tuned mass dampers Widely used Easy to size Additional mass to be considered. damps a given mode. damps several modes Non-linear calculation Little used Damps several modes Temperature-sensitive. additional mass to be considered. nonlinear calculation Tuned liquid dampers Very little used © Sétra . needs adjustment in frequency Little used Independent of temperature. needs adjustment in frequency "Innovative".2006 .

Photograph 4.95 Hz This frequency was calculated without taking the mass of the pedestrians into account. from the main types of constructions. the mass of the framework is 18. with a central bay of 64 m span. The construction was erected with a crane. The modal analysis carried out using finite element design software gave the following results: Mode 1 Vertical bending central bay 1. The functional width is 3.20 m high. a steel box-girder construction with orthotropic deck. Appendix 4: Examples of footbridges We will be reviewing several different designs of footbridges recently constructed. The deck consists of a reinforced concrete slab bearing on floor beams 2. a suspended construction with one steel mast. The works were carried out in 2000 .1998. A modal analysis made using an analytical calculation gave the following results: Mode 1 Vertical bending 1.8 t per bay.Dynamic behaviour of footbridges – Annexes 103 / 127 4. the total mass of the framework is 595. The functional width is 11.97 Hz © Sétra . We will indicate the main features of each of the bridges and its method of construction. of a total length of 180 m.2 . a construction with a ribbed slab.2001.0 t.26 m apart. The structure of the deck is formed from a steel box girder with orthotropic topping. 4. a cable-stayed construction. and end bays of 54 and 50 m.Steel box-girder: Stade de France footbridge The bridge is a box-girder bridge with double intermediate supports.00 m. a bow-string arch with orthotropic deck. the results of tests. The works were carried out in 1997 . it comprises two independent bays of 49.00 m. and we will give the results of the dynamic studies and.2006 .1 .7 m span. The steel construction was erected with a crane.1 – Cavaillon footbridge 4.Warren-type lateral beams: Cavaillon footbridge The bridge is a Warren-type construction with lateral beams 3. We will be presenting a construction with lateral beams. as applicable. a steel lattice arch.

7 mm acceleration 0.2 – Stade de France footbridge 4. The functional width is 5 m. and end bays of 22 m. The two beams comprising the structure were constructed perpendicularly to their final position and positioned over the A4 motorway by rotation.48 Hz 4.06 Hz Photograph 4. The modal analysis carried out using finite element design software gave the following results: Mode 1 Vertical bending 1.5 mm acceleration 1 m/sec2 acceleration 10 m/sec2 Vertical bending of end bay 1 2.Dynamic behaviour of footbridges – Annexes 104 / 127 Mode 2 Mode 3 Vertical bending of end bay 2 2. with a central bay of 44 m span.1994.3 .4 t mass: 640 pedestrians 25 in phase deflection 6.86 Hz © Sétra .Ribbed slab: Noisy-le-Grand footbridge The bridge is a ribbed slab in pre-stressed concrete of varying depth of a total length of 88 m.2006 . The works were carried out in 1993 . the total mass of the deck is 860.20 Hz The dynamic study gave the following results: 1 pedestrian deflection 2.4 m/sec2 640 pedestrians 25 in phase deflection 70 mm The dynamic study with damper of 2. The depth of the slab varies from 1 m at the summit to 3.05 m at the piers.65 Hz Mode 2 Mode 3 Lateral bending Axi-symmetrical bending 3.0 t.

50 m. Photograph 4.0 t.Dynamic behaviour of footbridges – Annexes 105 / 127 Photograph 4.62 Hz Anti-symmetrical bending 2. and the mass of the bridge being increased by the mass of the pedestrians corresponding to a(l): Mode 1 Vertical bending 2.52 Hz Comment: without the pedestrians. The construction was erected with a crane. the frequencies increase by 0. The functional width is 3. The works were carried out in 1998 .51 Hz Mode 2 Mode 3 Lateral bending of deck 2. A modal analysis carried out using finite element calculation software gave the following results.Bow-string arch: Montigny-lès-Cormeilles footbridge The bridge is a bow-string with polygonal arch of 55 m span. The structure of the deck comprises two lateral steel tubes connected by an orthotropic topping. the total mass of the framework is 85.1999.3 – Noisy-le-Grand footbridge 4.4 .4 – Montigny-lès-Cormeilles footbridge © Sétra .2006 . with the tubes of the tie filled with cement.5 Hz.

The structure was erected with a crane on temporary legs.37 Hz 1. linked by cross-pieces supporting a lower deck.80 Hz Photograph 4.6 . The works were carried out in 1998 .1999. The works were carried out in 2000.03 Hz Torsion-bending Vertical bending 1. The double upper deck is supported on A-frames and braces bearing on the two arches. The functional width varies from 12 m to 14. the total mass of the steel framework is 105. The modal analysis carried out using finite element design software gave the following results: Mode 1 Lateral swing 0. The structure was erected with a crane on temporary legs before the mast was erected. The functional width is 3.Steel arch: Solferino footbridge The bridge is a steel arch of 106 m span.00 m. The structure of the deck is formed from a steel box-girder suspended from a mast consisting of four inclined tubes joining at mid-span.71 Hz Mode 2 Mode 3 Mode 4 Axi-symmetrical vertical bending 1.5 – Soissons footbridge 4. The modal analysis carried out using finite element design software gave the following results: Mode 1 Lateral bending of deck 1.0 t.Dynamic behaviour of footbridges – Annexes 106 / 127 4.5 .66 Hz © Sétra .80. The structure of the arch is formed from two double parabolic arches as a ladder beam.2006 .10 Hz Lateral bending of the arches 2.0 t.Suspended construction: Footbridge over the Aisne at Soissons The bridge is a suspension bridge of 60 m span.10 Hz Mode 2 Mode 3 Vertical bending of the deck 3. the total mass of the framework is 900.

59 Hz Axi-symmetrical vertical bending 1. The dynamic tests and measurements gave the following accelerations: 16 pedestrians balance mode 1 acceleration 0.9% in respect of mode 1 (lateral swing) to be achieved.5 m/sec2 16 pedestrians walking mode 6 16 pedestrians running mode 7 acceleration 2.3% to 0.7 m/sec2 The damping with more than 100 pedestrians on the footbridge varied from 0. acceleration 2.81 Hz Mode 2 Mode 3 Mode 4 Mode 5 Mode 6 Mode 7 Bending-torsion 3.5%.94 Hz Vertical bending 1.22 Hz © Sétra .5 m/sec2 106 pedestrians running mode 7 acceleration 5.0 m/sec2 Central torsion-swing 2.9 t allow damping of 3.5 m/sec2 The damping with 16 pedestrians on the footbridge varied from 0.4% to 0. Dampers 6 pendular systems supporting masses of 2.8%.Dynamic behaviour of footbridges – Annexes 107 / 127 Mode 5 Torsion-swing 1.6%.69 Hz Torsion-bending 1.5 t and 1.7% to 1.5 t allow damping of 2.2006 .75% in respect of modes 5 and 6 (central torsion-swing) to be achieved. 106 pedestrians balance mode 1 acceleration 1. 8 mass/spring systems supporting masses of 2.66 Hz The dynamic tests and measurements gave the following frequencies empty (without damper): Mode 1 Lateral swing 0.22 Hz Central torsion-swing 1.09 Hz The damping for the empty structure varied from 0.

The construction was positioned with a crane.Cable-stayed construction: Pas-du-lac footbridge at St Quentin The bridge is a dissymmetric cable-stayed bridge of a total length of 188 m.92 Hz Displacement-bending of deck 1.2006 . two bays. suspended from a single pylon.50 m.7 .95 Hz Bending of pylon 1. each of 42 m span. with. and.1992.85 Hz Photograph 4. The modal analysis carried out using finite element design software gave the following results: Mode 1 Lateral bending of deck 1. two bays with spans of 68 m and 36 m.38 Hz Mode 2 Mode 3 Mode 4 Vertical bending of the deck 1. on one side. The functional width is 2. The structure of the deck is formed from two steel beams. on the other.Dynamic behaviour of footbridges – Annexes 108 / 127 Photograph 4. The works were carried out in 1991 .6 – Solferino footbridge 4.7 – Pas-du-Lac footbridge © Sétra . linked by floor beams.

the mass of the framework is 22.2006 .Mixed construction beam: Mont-Saint-Martin footbridge The bridge has a mixed steel-concrete deck with a span of 23 m. The structure of the deck is formed from two slightly cambered steel beams linked by floor beams.8 . A modal analysis made using an analytical calculation gave the following results: Mode 1 Vertical bending 2.15 Hz Mode 2 Mode 3 Lateral swing 4. The functional width is 2.99 Hz © Sétra .50 m. The works were carried out in 1996. The structure was erected with a crane.0 t.Dynamic behaviour of footbridges – Annexes 109 / 127 4.50 Hz Vertical bending 3.

2: Cross-section of the lateral beam footbridge The distance between the centre lines of the beams is 2.85 m.Dynamic behaviour of footbridges – Annexes 110 / 127 5. If the results led to unacceptable accelerations.50 m 1. taking into account the various categories of traffic. Appendix 5: Examples of calculations of footbridges. This section sets out a full study of two standard footbridges based on actual examples.90 m Figure 5.90 m. together with a sensitivity study of the natural frequencies of typical footbridges. the characteristics of these constructions were modified to improve their dynamic behaviour and to try and satisfy comfort conditions. of a constant depth of 1. bears on these floor beams. 10 cm thick. 38. A precast reinforced concrete slab.Examples of complete calculations of footbridges 5. The longitudinal profile is curved to a radius of 450 metres.50 m.1 . Characteristics of the deck © Sétra . giving a width of passage for pedestrians of 2.1.Warren-type lateral beam footbridge The first footbridge studied is a construction with a mixed steel-concrete framework forming one independent bay with a span of 38.85 m Figure 5.1 . The two complete examples of dynamic calculation have been carried out using the methodology of the guide. These beams.215 m.215 m 2. 5.1: Warren-type lateral beam footbridge The framework is formed from two triangulated lateral beams. members 400 200 12 Useful width 2. are linked by floor beams located at the level of the bottom member.2006 .

1.85 2. we will consider class III. depending on the class of the footbridge.14 Hz 2 210 10 9 0. Moment of inertia of the deck: I 0. For class III.85 2 2 1 2.1.6 kg / m The linear density is: S 1456 87. which is calculated for each crowd density.Dynamic behaviour of footbridges – Annexes 111 / 127 The moment of inertia is calculated taking into account the reinforced concrete slab with a homogenisation coefficient of 6. 5.030 1456 1 210 10 9 0. in other words a normally-used footbridge that can sometimes be crossed by large groups.6% (mixed deck).6 The total mass of the pedestrians is: 70 48. 5.2 Calculation of the dynamic load of the pedestrians We will calculate the load for the first mode only. the high-bond and low bond frequencies are equal to: f1 2 38. The surface load to be taken into account for the vertical modes is: © Sétra .85 87. where the density d of the crowd is equal to 0.030 m 4 Natural linear density of the deck: m 1456 kg / m Young's modulus of the steel: E 210 10 9 N / m 2 First of all.030 f1 2.54 Hz 8.5 38.6 kg / m For the first mode.8 2 38. with a critical damping ratio of 0. the high-bond and low bond frequencies are equal to: f2 2 2 38.85 For the second mode.08 Hz 2 38.6 1534. The number of pedestrians on the footbridge is: n p 0.2006 .85 f2 2 2 210 10 9 0.030 1456 2 2 2 8.5 pedestrians/m2.030 1630.1 Calculation of natural modes n2 EI The natural frequencies are equal to: f n S 2 L2 S is the linear density of the deck increased by the linear density of the pedestrians.1. we are interested in a sparse crowd.1.8 Only the first mode is likely to cause uncomfortable vibrations.5 48.1. and the mass of the deck is calculated taking into account the floor beams and the reinforced concrete slab.1.85 210 10 9 0.6 3402 kg The linear density of the pedestrians is: m p 3402/38. but never loaded over its whole area.02 Hz 2 1630.

the high-bond and low bond frequencies are equal to: f1 2 38. in other words an urban footbridge linking populated zones. in other words at an unacceptable comfort level (acceleration > 2.85 2 2 1 2.5 77.8 8.2 Calculation of the dynamic load of the pedestrians © Sétra .2.1.85 Only the first mode is likely to cause uncomfortable vibrations.1.6/100 1534.02 Hz 2 1630.08 t 0. is within range 1 (1. as the frequency of the first mode. 5.8 2.1.8 n is equal to 1.6/100 0.120 1 16.8 48.54 Hz 210 10 9 0. 10.Dynamic behaviour of footbridges – Annexes 112 / 127 Fs d 280N cos2 fvt 10.6 n The surface load is equal to: Fs 0.1.030 1456 1 210 10 9 0.2006 .5 kg / m The linear density is: S 1456 140.1 Hz) with a maximum risk of causing resonance. which is 2.14 Hz 2 210 10 9 0.85 f2 2 2 2 2 2 2 210 10 9 0.8 2 38. the high-bond and low bond frequencies are equal to: f2 2 38. to be loaded over its whole area. subjected to a high level of traffic and likely.08 t N / m This load is applied to the whole of the footbridge.8 cos 2 2.030 1630.85 140. The number of pedestrians on the footbridge is: n p 0.7 to 2.7 78 The total mass of the pedestrians is: 70 78 5460 kg The linear density of the pedestrians is: m p 5460/38. where the density d of the crowd is equal to 0. 5.3 Calculation of dynamic responses The calculation of the acceleration to which the construction is subjected gives: 4 42.2. on occasions.120 We have 10.5 1596.08 t N / m2 The linear load is equal to: F Fs l p 16.8 0.0 cos 2 2. 5.5 m/sec2).1.5 280 cos 2 2.6 The maximum acceleration calculated is located within range 4 of the accelerations.08 Hz 2 38.8 38.030 f1 2.08 t 42.91 m/ s 2 1 Acc max 2 0.0 2.08 Hz.5 cos 2 2.85 For the second mode.1 Calculation of natural modes For class II.8 pedestrians/m2. we are interested in a dense crowd. We will next consider class II.85 2.1.5 kg / m For the first mode.1.030 1456 8.

85 f2 2 38.030 1630. the high-bond and low bond frequencies are equal to: f2 2 38.1.1.030 1456 210 10 9 0. as in the previous case.1.1.85 For the second mode.04 t N / m2 The linear load is equal to: F Fs l p 21.28 2.8 10. etc.0 pedestrian/m2. in other words an urban footbridge linking zones with high concentrations of pedestrians (presence of a station.85 2 2 2 2 2 2 210 10 9 0.85 2 2 1 2.5 cos 2 2. The surface load to be taken into account for the vertical modes is: © Sétra . the high-bond and low bond frequencies are equal to: f1 2 38. with 0. where the density d of the crowd is equal to 1.030 f1 2.8 2 38.04 t 53. we have: 10.3.85 174. tourists. we are interested in a very dense crowd.85 2.2 3.53 m/ s 2 1 Acc max 2 0.8 0.6% of critical damping.8 kg / m The linear density is: S 1456 174. We will finally consider class I.8 280 cos 2 2.6 The surface load is equal to: Fs 0.1.). 5.5 The maximum acceleration calculated is located within range 4 of the accelerations.28 cos 2 2.6/100 1596.08 Hz Only the first mode is likely to cause uncomfortable vibrations.6% of critical damping (mixed deck).2.120 n 48. 5.8 kg / m For the first mode.8 1630.02 Hz 2 1630.2 Calculation of the dynamic load of the pedestrians We will calculate the load for the first mode only.2 cos 2 2.14 Hz 2 210 10 9 0.3.5 97 The total mass of the pedestrians is: 70 97 6790 kg The linear density of the pedestrians is: m p 6790/38. or frequently used by dense crowds (demonstrations.6/100 0.095 1 21. for example).8 8. in other words at an unacceptable comfort level (acceleration > 2.1 Calculation of natural modes For class I.2006 .04 t 0.3 Calculation of dynamic responses 4 53.5 m/sec2). subjected to a very high level of traffic.030 1456 1 210 10 9 0.04 t N / m 5. The number of pedestrians on the footbridge is: n p 1 38.1.54 Hz 8.Dynamic behaviour of footbridges – Annexes 113 / 127 With 0.

90 m Figure 5. It can be seen that the accelerations are always higher than 2.415 m Characteristics of the deck The moment of inertia of the deck is modified and equals: I 0.02 t 131.55 m/ s 2 1 We have: Acc max 2 0.1.60 cos 2 2.Dynamic behaviour of footbridges – Annexes 114 / 127 Fs d 280N cos2 fvt 1. as the first natural frequency is in the middle of the range of maximum risk. members 400 200 14 1. We choose both to increase the thickness of the metal in the members and to increase the depth of the beams.415 m.5 cos 2 2. which is 2.02 t 1.1 Hz) with a maximum risk of causing resonance. The sheet metal of the members is increased from 12 to 14 mm.5 m/sec2).5 cos 2 2. In order to reduce the accelerations obtained.045 m4 © Sétra .3. The surface load equals: Fs 1. as the frequency of the first mode.02 Hz.85 1 n is equal to 1.215 m to 1.3 Calculation of dynamic responses 4 131.5 8.6/100 1630.3: Cross-section of the footbridge with members in 14 mm sheet metal and depth of 1.02 t N / m² 97 The linear load is equal to: F Fs l p 52. To do this. and then we recalculate the frequencies. the depth of the triangulated lateral beams can be increased (for example by 20 cm) and the thickness of the sheet metal of the members also increased (for example 14 mm). in other words at an unacceptable comfort level (acceleration > 2.60 2. It should be pointed out that this example is a particularly unfavourable case. whatever class is selected.2006 . the dynamic loads and the corresponding responses. is within range 1 (1. the stiffness of the construction must be increased.415 m 2. The depth between centre lines of the members is increased from 1.5 m/sec².1.8 The maximum acceleration calculated is located within range 4 of the accelerations.0 280 cos 2 2.02 t N / m 5.7 to 2.85 1 1 52.

which is compatible with a minimum comfort level (1 – 2. but if it stays around 2. it should not cause any problems. 40.21 for class I and 0. Useful width 3. The frequencies of the first modes are modified in the following way: Class III: 2. and almost maximum (0.7 Hz – 2.50 m/sec²) Acc = 1.6 Hz. This leads to the following accelerations: Acc = 0.Dynamic behaviour of footbridges – Annexes 115 / 127 The natural mass of the deck is slightly increased.00 m 0. with concrete deck topping.00 m Figure 5. but that is not significant. for example.10 m 1. Characteristics of the deck © Sétra .2006 .1. it would be possible. each of 40 m.6 is then zero. 5.52 Hz Class I: 2. A 10 cm thick pre-cast reinforced concrete slab bears on this girder.1 Hz.5 m/sec²) To make this footbridge even more comfortable. the second harmonic of the pedestrians must Hz. For class I and II. the width for pedestrian passage is 3.55 m/sec² for class II.00 m Figure 5.4: Steel box girder footbridge 40. which is compatible with a medium comfort level. calculations are necessary.5: Cross-section of a mixed box girder footbridge The width of the slab is 4.78 m/sec² for class I.50 Hz The frequency of 2. but with a coefficient 0.16 for class II.50 m.57 Hz for class III does not lead to any calculation. as this frequency is outside the range 1.50 m Concrete slab 4. to increase the thickness of the sheet metal to 16 mm so that the natural frequencies are greater than 2. In this case.2 .00 m The framework is formed from a steel box girder of a constant depth of 1 metre. The coefficient be taken into account.Box-girder footbridge The second footbridge studied is a steel box girder with two bays.00 m.00 m sheet 30 mm 2.57 Hz Class II: 2.

Dynamic behaviour of footbridges – Annexes 116 / 127

The mass of the deck is calculated taking into account the reinforced concrete slab and the mass of the balustrades. The moment of inertia is calculated taking into account the concrete slab with a homogenisation coefficient of 6. Moment of inertia of the deck: I 0,057 m 4 Natural linear density of the deck: m 3055 kg / m Young's modulus of the steel: E 210 10 9 N / m 2

First of all, we will consider class III, in other words a normally-used footbridge that can sometimes be crossed by large groups, but never loaded over its whole area. 5.1.2.1.1 Calculation of natural modes The natural modes of the deck were calculated using the Systus program. S is the total linear density, in other words the linear density of the deck (including the slab), increased by the linear density of the pedestrians, which is calculated for each crowd density, depending on the class of the footbridge. For class III, we are interested in a sparse crowd, where the density d of the crowd is equal to 0.5 pedestrians/m2. The number of pedestrians on the two bays is: n p 0.5 2 40 3.5 140 The total mass of the pedestrians is: 70 140 9800 kg The linear density of the pedestrians is: m p 9800/80 122.5 kg / m The total linear density is: S 3055 122.5 3177.5 kg / m For the first mode, the high-bond and low bond frequencies are equal to: f1 1.94 Hz f1 1.86 Hz For the second mode, the high-bond and low bond frequencies are equal to: f 2 3.04 Hz f 2 2.92 Hz 5.1.2.1.2 Calculation of the dynamic load of the pedestrians First of all, we calculate the load for the first mode, and a critical damping ratio of 0.6% (mixed deck). The surface load to be taken into account for the vertical modes is:

Fs d 280N cos2 fvt 10.8 n is equal to 1, as the frequency of the first mode, which is 1.90 Hz, is within range 1 of the frequencies (1.7 to 2.1 Hz). 10.8 0.6/100 0.071 We have: 10.8 n 140 The surface load is equal to: Fs 0.5 280 cos 2 1.90 t 0.071 1 9.94 cos 2 1.90 t N / m2 The linear load is equal to: F Fs l p 9.94 3.5 cos 2 1.90 t 34.79 cos 2 1.90 t N / m

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 117 / 127

5.1.2.1.3 Calculation of dynamic responses The acceleration under the vertical load is equal to: 4 34.79 1.16 m/ s 2 1 Acc max 1 4 F 2 S 2 0.6/100 3177.5 The maximum acceleration calculated is located within range 3 of the accelerations, in other words at the minimum comfort level (acceleration between 1 and 2.5 m/sec2). However, the medium comfort level is almost reached. For the second mode, the frequency of 2.96 Hz does not impose particular checks.

We will then consider class II. 5.1.2.2.1 Calculation of natural modes For class II, we are interested in a dense crowd, where the density d of the crowd is equal to 0.8 pedestrians/m2. The number of pedestrians on the footbridge is: n p 0.8 2 40 3.5 224 The total mass of the pedestrians is: 70 224 15680 kg The linear density of the pedestrians is: m p 15680 / 80 196 kg / m The total linear density is: S 3055 196 3251 kg / m For the first mode, the high-bond and low bond frequencies are equal to: f1 1.94 Hz f1 1.86 Hz For the second mode, the high-bond and low bond frequencies are equal to: f 2 3.04 Hz f 2 2.92 Hz 5.1.2.2.2 Calculation of the dynamic load of the pedestrians First of all, we calculate the load for the first mode, and a critical damping ratio of 0.6% (mixed deck). The surface load to be taken into account for the vertical modes is:

Fs d 280 N cos 2 f v t 10,8 n is equal to 1, as the frequency of the first mode, which is 1.879 Hz, is within range 1 of the frequencies (1.7 to 2.1 Hz).

We have 10,8

n

10,8 0.6/100 0.056 224

The surface load is equal to: Fs 0.8 280 cos 2 1.879 t 0.056 1.0 12.54 cos2 1.879 N / m2 The linear load is equal to: F Fs l p 12.54 3.5 cos 2 1.879 t 43.89 cos 2 1.879 t N / m 5.1.2.2.3 Calculation of dynamic responses The acceleration under the vertical load is equal to: 4 43.89 1.43 m/ s 2 1 4 F 1 2 S 2 0.6/100 3251 Acc max

© Sétra - 2006

Dynamic behaviour of footbridges – Annexes 118 / 127

The maximum acceleration calculated is located within range 3 of the accelerations, in other words at a medium comfort level (between 1 and 2.5 m/sec2). For the second mode, the frequency of 2.92 Hz requires us to take the second harmonic into account. But, with a force of one-quarter of that of the first harmonic, and taking into account the above acceleration result, the comfort level obtained is maximum.

We will finally consider class I. 5.1.2.3.1 Calculation of natural modes The natural modes of the deck were calculated using the Systus program. For class I, we are interested in a very dense crowd, where the density d of the crowd is equal to 1.0 pedestrian/m2. The number of pedestrians on the footbridge is: n p 1 2 40 3.5 280 The total mass of the pedestrians is: 70 280 19600 kg The linear density of the pedestrians is: m p 19600 / 80 245 kg / m The total linear density is: S 3055 245 3300 kg / m For the first mode, the high-bond and low bond frequencies are equal to: f1 1.94 Hz f1 1.86 Hz For the second mode, the high-bond and low bond frequencies are equal to: f 2 3.04 Hz f 2 2.92 Hz 5.1.2.3.2 Calculation of the dynamic load of the pedestrians First of all, we calculate the load for the first mode, and a critical damping ratio of 0.6% (mixed deck). The surface load to be taken into account for the vertical modes is: Fs d 280N cos2 fvt 1.86 1 n is equal to 1, as the frequency of the first mode, which is 1.86 Hz, is within range 1 of the frequencies (1.7 to 2.1 Hz). The surface load equals:

Fs 1.0 280 cos 2 1.86 t 1.85 1 1 30.956cos 2 280 1.86 t N / m² 1.86 t N / m

The linear load is equal to: F Fs l p 30.956 3.5 cos 2

1.86 t 108.35 cos 2

5.1.2.3.3 Calculation of dynamic responses The acceleration under the vertical load is equal to: 4 108.35 3.48 m/ s 2 1 4 F 1 2 S 2 0.6/100 3300 Acc max The maximum acceleration calculated is located within range 4 of the accelerations, in other words at an unacceptable comfort level (acceleration > 2.5 m/sec2). For the second mode, the frequency of 2.91 Hz requires us to take the second harmonic into account. But, with a force of one-quarter of that of the first harmonic, and taking into account

© Sétra - 2006

A 10 cm thick pre-cast reinforced concrete slab bears on this girder. If the medium comfort level is selected. the depth of the box girder can be increased (for example by 40 cm). 5.Dynamic behaviour of footbridges – Annexes 119 / 127 the above acceleration result. the two natural frequencies are outside the range of frequencies that have to be checked.85 m/ s 2 1 In class I: Acc max 1 4 F 2 S 2 0.Sensitivity study of typical footbridges. the comfort level obtained is medium (approximately 0. the project will have to be modified.47 Hz and f 2 4.48 1. The moment of inertia of the deck is modified and equals: I 0. Class III II I Table 5.106 m4 Natural linear density of the deck: m 3241 kg / m Young's modulus of the steel: E 210 10 9 N / m 2 The high bond and low bond frequencies are modified as follows: f1 2. in which the span of these typical footbridges is varied around the actual span.2 .6 1. In class III. To do this.2006 . It is medium for class I.6/100 3437 In class II: Acc max 4 28. The accelerations associated with the first mode are.90 0. are unchanged.16 Frequency in Hz Damping ratio Acceleration in m/sec2 The framework is formed from a steel box girder of a constant depth of 1.57 Hz f1 2.83 Hz For class III. © Sétra . the comfort level is maximum.40 metres. but remaining within their field of use. the comfort level is therefore automatically maximum.43 1.9 m/sec²). taking the second harmonic of pedestrians walking into account. for class I and II: 4 10.88 0. This section sets out a study of four types of standard footbridges based on actual examples. the stiffness of the construction must be increased.17 0. In order to reduce the accelerations obtained.86 0. but acceptable even so.6 1. The conclusions on comfort levels.1 It can be seen that the accelerations are higher than 1 m/sec² for all categories.02 Hz f 2 3.32 m/ s 2 1 4 F 1 2 S 2 0. 1.6/100 3437 For class II. for the second natural modes.36 0.6 3.

For the reinforced slab. two edge brackets and two balustrades must be added.03 x 24 = 2.80 m 0.1 .damp-proofing 3. are: . Only the depth h varies when the span changes.2. in pre-stressed concrete. the fixed dimensions are as follows: 3 m 50 3 m 50 h 0. 5. for which we will use an H-shaped reinforced footbridge.5 kN / m in total.e. For reinforced concrete. either in the 3. a distinction is made between two distinct fields: where L is between 20 m and m.Presentation of the footbridges studied and static pre-sizing Four main types of footbridges were distinguished: in reinforced concrete. Only the depth h varies when the span changes. a pre-stressed box girder is preferably to be used. for which a rectangular reinforced concrete slab is sufficient. or in the two 50 cm beams.04 x 24 = 4 kN /m .sundries 1 kN / m i. In the same way. the superstructures are comparable.50 m 20 m < L < 25 m 30 m < L < 45 m Figure 5.Dynamic behaviour of footbridges – Annexes 120 / 127 The first paragraph sets out the footbridges studied and the second states their natural frequencies. For each type. even though they have been adapted to each case. The superstructures of this footbridge.5 x 0. 16 kN / m in total. On the other hand. rough pre-sizing was carried out. for the slab. In order to allow a comparison between the various footbridges. the fixed dimensions are as follows: © Sétra . a distinction is made between two fields of span: small spans (20 to 25 m). i. the same net width of 3.20 h The depth of the footbridge to be used is that which enables the compression limit of the concrete (here 15 MPa) to be guaranteed.e. and to take the steels.6: Reinforced concrete footbridges 0. and longer spans (25 to 45 m).50 m was taken for all the footbridges. for the H-shaped footbridge. a pre-stressed rectangular slab is used. In the same way as for reinforced concrete. 7. as soon as L becomes longer than 35 m (up to 50 m). the assumptions and the procedure used in each case are given in the following section. The role of the balustrade is played by the H-beams themselves.5 kN /m . together with their weights.finish 3.2006 . for footbridges of a total length of between 20 and 80 m.50 m width. mixed steel-concrete box girder and steel lattice.5 x 0.

34) = 4.5 kN / m in total.025 m 0.04 x 24 = 4 kN /m .2006 .5 kN / m in total. If it does not.5 x 0. i. but the thickness of the concrete is still there and adds to the overall weight. the mixed section takes the superstructure (with an equivalence factor n = 18) and the static pedestrian loads (factor n = 6). it becomes a mixed section.damp-proofing 4 x 0. Only the depth of the footbridge varies when the span changes.2 h 2m 20 m < L < 30 m Figure 5.7: Pre-stressed concrete footbridges 4m 0.04 x 24 = 4 kN /m . are: .2 0.10 m 0.2 h 35 m < L < 50 m The superstructures of this footbridge. the only load-bearing section is the steel girder.Dynamic behaviour of footbridges – Annexes 121 / 127 4m 0. the sizing is done on the assumption that the steel section bears its own weight and also that of the wet concrete (assumed placed in a single pour).2 edge brackets 2 x 25 x (0.8: Mixed footbridge h We will study this type of footbridge for total spans ranging from 40 to 80 m. the fixed dimensions are as follows: 4m 0.e. This type of footbridge is a steel box girder (welded reconstituted beams) in grade S355 steel.sundries 1 kN / m 16. For the mixed box girder.2 balustrades 2 x 2 = 4 kN / m . The superstructures of this footbridge. A concrete slab bears on this box girder. together with their weights. are: . if it is connected to it. © Sétra .damp-proofing 4 x 0. For the steel box girder.finish 3.5 kN / m . i.25 x 0.025 m 2m Figure 5.03 x 24 = 3 kN /m .34) = 4.03 x 24 = 3 kN /m .5 kN / m .25 x 0.5 x 0.025 m 0. together with their weights.e.2 edge brackets 2 x 25 x (0.finish 3. the load-bearing section is the steel alone. On the other hand.sundries 1 kN / m 16.2 balustrades 2 x 2 = 4 kN / m .

S is the linear density of the bridge (self-mass and mass of the superstructures) in kg / m.Natural frequencies The four previous types of footbridges are considered to be iso-static bays. We have taken 1 pedestrian / m². freq. freq. freq.damp-proofing 3. are: . vertical or horizontal. freq.2. to which is added the mass of the pedestrians on the footbridge. I is the inertia in m4.2 .5 kN / m in total. hor. 1 2 1 2 vert.5 m 0. freq.e. i. together with their weights. hor. C 7 8 C 5 B 7 H 5 E E G E D F C 7 8 C 5 B A 6 @ 9 F 7 7 C 7 8 C 5 B A 6 @ 9 8 7 6 5 4 L vert. vert.Dynamic behaviour of footbridges – Annexes 122 / 127 The main field of span taken into account for this type of footbridge is 50-80 m. The depth was determined according to static sizing. 5. freq. depending on what is being sought. We therefore varied the length of the iso-static bay for each type of footbridge. S = permanent loads + 3.04 x 24 = 4 kN /m . with 70 kg per pedestrian. hor.finish 3. It is assumed that the footbridge is properly triangulated. freq.05 m h 0. freq. vert.03 x 24 = 2.23 m Figure 5. hor. the formula giving the natural frequency of the bridge is: 1 n 2 2 EI fn 2 S L2 fn is the natural frequency of the mode n. The following tables set out the four natural frequencies: the first two vertical and the first two horizontal.5 x 0.2 balustrades 2 x 2 = 4 kN / m . hor. vert. L is the length of the bay in m. 1 2 1 2 © Sétra . 1 2 1 2 vert. For an iso-static bay.5 kN /m . The values that remain fixed when the span changes are: 3 m 50 0.9: Steel lattice footbridge 0. freq. freq.sundries 1 kN / m 11.5 x 0. i. E is the Young's modulus of the material comprising the structure in N / m². hor.5 x 70 x 1 in kg / m.18 m The superstructures of this footbridge. freq.e.2006 . freq.

0 1.0 2. for lateral vibrations.4 12 11 10 10 11 8 6 5 44 33 26 21 box girder 30 35 40 45 50 2.2 1.1 0. For vertical vibrations. freq.4 2.7 25 3.3 The following comments may be inferred from these tables: Only the first mode need be used for these simple types of footbridges. vert.0 1.9 5. 2 11 10 2. freq. All others would appear to be classified as "at risk".7 5.4 1.4 1.2 6.4 13 15 61 30 1.9 12 8. Finally. Footbridges that are unable to vibrate under pedestrian excitation are rare.2 4.0 4.5 5. 1 15 10 hor.0 1.9 3.7 7.4 1.1 2.5 8.3 6.9 21 15 12 9.1 5. 2. for concrete footbridges. the first natural frequency of metal footbridges is around 1 – 1.0 7.1 1. these are Hshaped reinforced footbridges of 25 m.9 3.4 1.4 1.1 2.5 1.7 5.1 3. 1 19 13 hor.3 5. © Sétra .5 Hz.0 2.5 2.1 Table 5.1 hor.7 2.4 4. 1 freq. it is nearer 2 .9 4.9 7. freq.8 9.2006 Q R I 5 Q P I 5 vert.6 4.0 9.5 5.5 L slab RC Hsect.1 1.6 1.5 3.3 Hz for pre-stressed concrete). whereas.8 1.3 Hz (more precisely.2 1.3 2.3 Hz for reinforced concrete and 2 .5 20 25 2.6 1.0 2. 2 8.3 1.1 8. freq.7 2.8 hor.6 2.5 Hz).5 7.2 1.0 0.1 3. freq.6 4. it would be sufficient to study large-span steel footbridges (for all others.8 2.7 5.0 28 20 25 vert.4 6.5 1.5 . both for vertical and horizontal vibrations. 2 77 slab 50 PC Table 5.6 5. 2 58 39 . L freq.2 7.9 3.Dynamic behaviour of footbridges – Annexes 123 / 127 40 50 60 70 80 1.8 8.4 1.0 5.5 4.2 3. 30 35 40 45 3.0 14 9.4 1. vert. the first horizontal frequency is beyond 2.3 3.0 1. 1 freq.

Matsumoto.1 . 8.Design of concrete structures – Part 2: Concrete bridges . 1966. Leonard. The Structural Engineer. 6.4 . Dynamique des constructions. Armand et al. H. Dynamic design of footbridges. October 1999. Y. Recommendations for the calculation of the effects of wind on constructions.Behaviour analysis 15. Stade de France footbridge – Design of tuned mass dampers. 59B No. 1978. McGraw-Hill Book Company. 16. 3. Bachmann et al. PrENV 1995-2. 1997. B. D.Practical guidelines. P-17/78.3 .M. 5. R. R. Hanson.Jarret. CECM.Generally 1. Vol.2 . 10. Eye. 2. K. pp. Bachmann.Dampers 9.Part2 : Bridges. 14 January 1997. 1. Two case studies in the use of tuned vibration absorbers on footbridges. Annexe 6: Bibliography 6. IABSE Proceedings. R.J. No. Montens. Manufacturers' documentation: Taylor . Permissible vibrations for footbridges and cycle track bridges.Appendix C. Tuned mass dampers to control floor vibration from humans. 2. March 1992. © Sétra .Dynamic behaviour of footbridges – Annexes 124 / 127 6. Weber. ENSMP courses. Tuned vibration absorbers for lively structures. Matsuzaki. 209. BS 5400 . 14. H. Vibration problems in structures . R. Hanson. A. 7. Pretlove. CEB. 3. 1996. March 1992. Vol. M. August 1991. H. Shiojiri. Practical guidelines. Outteryck. ASCE Journal of Structural Engineering. No. T. ENV 1992-2. French Civil Engineering Review. 118.Design of timber structures . Vol. 2nd edition. C. Setareh.2006 . No. 1989. 6. Human tolerance levels for bridge vibrations. June 1964. 118.T. 11. R. Tuned mass dampers for balcony vibration control. June 1981. ASCE Journal of Structural Engineering. 13. J. Structural Engineering International. 6. Biggs. Jones. 1983. 1995. 12.. I. M. 1-15. Birkhäuser. S. Setareh. J. 3. Introduction to structural dynamics. Eurocode 5 . Nishioka.Regulations 4. 6. B. Eurocode 2 .

IABSE. Amman. 1980. Y. 12.Dynamic behaviour of footbridges – Annexes 125 / 127 17. 68. Willford. 2001. Development of a simplified design criterion for walking vibrations.2006 . A. R. Bachmann. Kurzberichte aus der Bauforschung. M. 120. 37. Earthquake engineering and structural dynamics. No. ASCE Structures Congress. 19. W. H. Mouring. M. Case studies of structures with man-induced vibrations. 21. pp. Ellingwood. T.C. A. Fitzpatrick. Hanagan. R. 18.E. Fujino. Tallin. part 3. Iwankiewicz.L. © Sétra .Evaluation of human exposure to whole-body vibration – Part 1: General requirements. Bachmann.1121-1136. 31. B. Bases for design of structures . G. J. ASCE Journal of structural engineering. 1997. Schneider. Vol. February 1984. 10. ASCE Journal of Structural Engineering. Vibration of a beam under a random stream of moving forces. ASCE. October 1987. 108.M. ASCE Structures Congress. R. Guidelines to minimise floor vibrations from building occupants. 29. J. G. Le Bourva. Vibration Upgrading of Gymnasia . B. S.M. Vibrations in structures induced by man and machines. The Structural Engineer. Murray. On minimum weight design of pedestrian bridges taking vibration serviceability into consideration. February 1992. USA. Design live loads for coherent crowd harmonic movements. 25. Price. H. No. ASCE Journal of Structural engineering. 34. Sack. 30. 1991. Mechanical vibration and shock . Vol. Synchronisation of human walking observed during lateral vibration of a congested pedestrian bridge. Durch Menschen verursachte dynamische Lasten und deren Auswirkungen auf Balkentragwerke. Y. R. 20 November 2001. Sugimoto. S-26/84.E. 3. Vol. Structural Engineering International. No. IABSE periodical. 27. Design criterion for vibrations due to walking. Vol. 24. H. 2.Serviceability of buildings against vibration. P. ST9. D.N. D. Vol. Vol. T. Proceedings ARRB. Eyre. An investigation into crowd-induced vertical dynamic loads using available measurements. Report No. 36. ASCE Journal of Structural Engineering. The Structural Engineer.Dance Halls and Footbridges. 79.W. Dynamic behaviour of footbridges. 118.E. 1984. Measuring and modelling dynamic loads imposed by moved crowds. Vanderplaats. February 1984. 15 April 1992. 32. Structural serviceability . R. T. February 1994. Dallard. 21-23 May 2001. December 1996. Kim. 1987. Pacheco. Willford. A. B. ASCE Journal of Structural Engineering. M. Nakamura. A. Vol. Washington DC. Smith.I. Prediction and control of pedestrian-induced vibration in footbridges. H. 12. 28. 35. No. Serviceability vibration evaluation of long floor slabs. ASCE. 20. 38. ISO 10137. J. 1993.Floor vibrations. 1993. 1993. March 1992. 22. Structural Engineering Documents. Flint. H. Allen. S. 1990. Engineering Journal/ American Institute of Still Construction. Structural Mechanics. 118-4. Bachmann. Grundmann. Wheeler. ASCE Journal of Structural Engineering. Schwingungsuntersuchungen für Fu gängerbrücken. Vol. 21-23 May 2001. Wheeler. Pawel Sniady. H. Pedestrian induced vibrations in footbridges. ASCE Journal of Structural Engineering. Warnitchai. 26.P. M. 2. first edition. Bauingenieur. Tilly. No. 23. International Standard ISO 2631-1. September 1982.M. Ellingwood. page 13 et seq. 33. Bachmann. Ebrahimpout. Various Authors. 110. November 1990. A. The London Millennium Footbridge. 22. S. Cullington. No. Low. Vol. Washington DC. P. L. June 2001. No. USA. Kajikawa. 1. Ridsdill Smith. H. 4th Quarter. 79.J. Kreuzinger. No. 125.

Using component mode synthesis and static shapes for tuning TMDs. Optimal absorber parameters for various combinations of response and excitation parameters. 1998. Youssouf – course leader F. Willford.8. K. Bulletin of the Japanese Society of Mechanical Engineering. Switzerland. R. Flint. 4/91. A briefing on pedestrian-induced lateral vibration of footbridges.Articles on either vibrating or instrumented footbridges 43. Paultre. M. G. 1994.C. M. Earthquake Engineering and Structural Dynamics. 151. FIP notes 98-1. F. T.P. Structural Engineering International. ASCE Journal of Structural Engineering. M. 4th quarter. No. Rebirth of the ribbon. Dallard. 1940. Mc Graw-Hill. Ioi. 6. 16th IABSE Congress.5 . Revue Ouvrages d'Art. Vol. 45. Fitzpatrick. 3. October 2000.10. Redfield. Vol. 40. 6. D. P. Aerodynamic behaviour of open-section steel deck cable-stayed bridges. No.A review. Various Authors. F. French Civil Engineering Review. 5 December 2000.Calculation methods 39. November 1993. Various Authors. Ikeda. 119. Study of vibratory behaviour of footbridges as pedestrians pass.2006 . U. N. 381-401.7 .1 . Cable dynamics . 54. 4th edition. 48. Ridsdill Smith. 23-24. Foucriat. 1978. Dynamic testing of the Sherbrooke pedestrian bridge. Legeron. A. M. 46. Roy.B. G. 2000. Straski. J. Warburton. J. Structural Engineer. 1992. Dynamic study of pedestrian footbridges. Various Authors. 52. Gardnermorse.8 . Vol. Pedestrian-induced vibration of footbridges. July 1999. A. Le Moine. Bridge. 49. Lucerne. 11. Legeron. No. Huston.Dynamic behaviour of footbridges – Annexes 126 / 127 6. Proulx. 78.Specific footbridges 47. 42. J. C. Low. 21. © Sétra . Den Hartog.Additional bibliography 6. Construction of the world’s longest pedestrian stress-ribbon bridge. Journées techniques AFPC. No. On the dynamic vibration damped absorber of the vibration system.J. 53. P. Légeron. Structural Engineering International 3/94. J.6 . 6. No. Mechanical vibrations. ASCE Journal of Structural Engineering. Modal identification of cable-stayed pedestrian bridges.Damping 51. Starossek. Sacramento River pedestrian bridge USA. 41. part I. 6. 1998. 1991. MSOA thesis of ENPC promotion 2000. Lemoine. SETRA. 4. A. 1978. 1982. pp. 44. 32.

Yokoyama. No. 61. Sun. ASCE Journal of Structural Engineering. Constantinou.2006 . Vibration control by multiple tuned liquid dampers. 8.Additional literature on pedestrian behaviour studies 60. 1993. 1993. No. Paris France. Constantinou. Model for lateral excitation of footbridges by synchronous walking. January 2004 62.Dynamic behaviour of footbridges – Annexes 127 / 127 55. 117. 12.8. 27082724. Fractional-derivative Maxwell model for viscous dampers. Makris. N. S. S. M. 57. Nakamura. 56. Dargush. Y. 59. September 2004. 6.8.M. Steel bridges bulletins. Journal of Engineering Mechanics.C. Nakamura. Vol. ASCE Journal of Structural Engineering. Vol.2 . November 2002. Vol.C. pp. H. Experimental study on lateral forces induced by pedestrians.OTUA Technical Bulletins 58. Fujino. © Sétra .3 . 1991. Katsuura. IABSE Symposium. 6. ASCE Journal of Structural Engineering. China. Proceedings and technical texts of the 2002 Footbridge International Conference. 119. N. Dynamic analysis of generalised visco-elastic fluids.F. K. M. G. Makris. 119. L. Shanghai. Steel structures bulletins.

gouv.setra.Design: Eric Rillardon (Sétra) The Sétra authorization is required for reproduction of this document (all or even part) © 2006 Sétra – Reference: 0644A .equipement. designers and engineers. The Sétra belongs to the scientific and technical network of the French Public Work Ministry (RST) This document is awailable and can be downloaded on Sétra website: http://www.setra. Supplementary theoretical (reminder of structural dynamics. examples of recent footbridges. The methodology is based on the footbridge classiﬁcation concept (as a function of trafﬁc level) and on the required comfort level and relies on interpretation of results obtained from tests performed on the Solferino footbridge and on an experimental platform. typical calculations) data are also provided in the guideline appendices. as well as structural response to loading.ISRN: EQ-SETRA--06-ED17--FR+ENG Le Sétra appartient au Réseau Scientifique et Technique de l'Équipement .fr These guidelines round up the state of current knowledge on dynamic behaviour of footbridges under pedestrian loading. • a presentation of the practical methods for calculation of natural frequencies and modes.fr Cover . These guidelines are aimed at Owners.service d'Études techniques des routes et autoroutes Sétra 46 avenue Aristide Briand BP 100 92225 Bagneux Cedex France téléphone : 33 (0)1 46 11 31 31 télécopie : 33 (0)1 46 11 31 69 internet : www. equipement. An analytical methodology and recommendations are also proposed to guide the designer of a new footbridge when considering the resulting dynamic effects.Photographers: Michel Leher (Matière TP). pedestrian load modelling) and practical (damping systems. Gérard Forquet (Sétra) . This document covers the following topics: • a description of the dynamic phenomena speciﬁc to footbridges and identiﬁcation of the parameters which have an impact on the dimensioning of such structures.gouv. • recommendations for the drafting of design and construction documents. Marc Mimram. • a methodology for the dynamic analysis of footbridges based on a classiﬁcation according to the trafﬁc level.

- 0071450173_ar003
- 0071450173_ar002.pdf
- 0071450173_ar001
- Td-r-ms-500 Miscellaneous - Clearance Diagrams for Bridges
- Bridge Design Eurocodes Worked Examples Annex Only
- H Rod Datasheet
- Bbr Hiam Cona en Rev2 0411 (1)
- Bbr Hiam Cona Damping en Rev1 0614
- Jacobsz - Modulus of Subgrade Reaction
- Concrete Pavements TU Delft
- SANS10005(Amdt1)
- 9554662
- AS4050196.pdf
- Chapter 3
- 425-889-1-PB
- 425-889-1-PB
- Semir Rigid Joints M Sadat HAYALIOGLU
- Subgrade Reaction Ks Pile and SPT Correction
- Metric Design Guide
- Acdmac 2013 Autolisp Developers Guide
- Cement and Concrete Presentation Lafarge
- CODE - Appendix 3 Design for Fatigue {American Institute of Steel Construction, Steel Construction Maunual 13th Ed.}
- TF2821_CH02
- Post Tensioning Tendon Installation and Grouting Manual.pdf
- Transit NZ Bridge Manual 2nd Edition

Sign up to vote on this title

UsefulNot useful- forced vibration
- Intro
- Damped
- Oscillation
- Dynamics-2 Marks Qb
- Unit3LectbyVRKDynofMachinery
- LAB 2 Free n Force Vibration
- Elements of Soil Dynamics and Machine Foundations
- m470hw
- Vibration - Wikipedia, the free encyclopedia.pdf
- Report 5 Forced Oscillation
- Vibration - One Degree of Freedom
- MVSyllabus
- Vibration
- 2 Marks
- Handouts on DOM
- Footbridge Guidelines EN03
- Vibration
- wikipedia Vibration
- ME3001 A4 Introductory Problems in Vibration
- Exp 3 Force Vibration (Finalized)
- L15
- Theory of Vibration
- Results & Discussion
- A1_labsheet
- Lecture by S.P. Singh
- ME010 706 L03 Theory of Vibration
- Vibration - Wikipedia, The Free Encyclopedia
- vibrasyon
- Damping - Wikipedia, the free encyclopedia.pdf
- US 0644A Footbridges