Sie sind auf Seite 1von 11

15th Symposium on Earthquake Engineering

Indian Institute of Technology, Roorkee


December 11-13, 2014

Paper No. A064

DYNAMIC RESPONSE OF A FOOTBRIDGE


SUBJECTED TO PEDESTRIAN INDUCED
VIBRATIONS

Ravi Chandra I.V.P1, Pandey A.D2, Bhattacharya K.1, Lahiri S. K.1, Vishnu B. K.1, Sinha A.1
1

Post graduate student, D.E.Q, Indian Institute of Technology Roorkee, India,ivpravichandraiitr@gmail.com


2

Assistant professor, D.E.Q, Indian Institute of Technology Roorkee, India,adpanfeq@gmail.com

ABSTRACT

Structures like footbridges, cantilever grand stands, office floors, etc. are often susceptible to excessive
vibrations under human movements. However aesthetics, availability of efficient materials, low live loads, and
design based on static analysis, make footbridges lighter and flexible. A consequence of the foregoing, the
resulting natural frequencies of the footbridge often fall in the dominant range of the pedestrian excitations,
triggering undesirable vibrations. In most of the cases, a vibration phenomenon is a serviceability problem,
rather than structural safety and stability. A representative suspension footbridge is analyzed, focusing on
acceleration and displacement values at critical points on the footbridge under different pedestrian load models.
The walking load model which is periodic is resolved into a Fourier series, and applied at the relevant points on
the footbridge. The loads due to random pedestrian crowds need to be applied as surface loads varying with time
on the deck, duly accounting for the synchronization effect. A parametric study on the crowd density is carried
out to simulate the response during rare events. Results indicate that the acceptance criteria for footbridges
should encompass both structural and pedestrian perspectives.

INTRODUCTION
Structures which are subjected to dominant human induced loads like stadium stands, gymnasiums,
footbridges, concert halls, office-floors, etc. are highly susceptible to vibrations imparted by human
movements. Reasons being, the live loads are relatively lower, enhancement in the construction
materials, dynamic design is overlooked, therefore resulting in light-weight and flexible forms. As a
consequence, the dynamic components of the pedestrian load are capable enough in triggering these
undesirable vibrations. Actions exerted by the humans on structures include walking, jumping,
bouncing, deliberate excitations. These synchronized crowd- excitations can give rise to resonant or
near resonant structural vibrations, more often treated as a serviceability issue. Very rare, the
structural failures due to fatigue, overstressing of materials were seen. Of all the pedestrian dynamic
loads, walking load model is appropriate for the footbridges, hence attempted in the present work.
The important parameters in the footbridge vibration are source, path and the receiver; Pedestrians are
identified as the source of vibration, due to their dynamic component of load and synchronized human
movements. The path of the vibration is observed to be the dynamic properties of the footbridge i.e.
stiffness, bridge mass, damping. The receiver of the footbridge vibrations are the pedestrians. The
human perception to these vibrations is highly sensitive, and feels discomfort. Moreover, in order to
balance ones own body weight, pedestrians have a tendency to adopt their walking gait to match with
deck vibration leading to the amplification of vibrations.

HUMAN INDUCED DYNAMIC LOADS

Human loads primarily comprises of both static and harmonic components. In most cases, pedestrian
live load is considered to be static and applied as load per surface area. Pedestrian exerted dynamic
load has three components i.e. vertical, lateral and longitudinal. The vertical component is obvious
due to continuous walking. The lateral component of force is by the oscillation of center of gravity of
ones own body motion. The longitudinal component arises from the push-off by the both heel and
toe. The exciting frequencies in the lateral direction are half the vertical footfall rate ivanovi 2005.
Therefore, lateral vibrations can be more sensitive due to resonance or near resonance phenomena.
The quantification of the pedestrian dynamic loads was obtained using force-plate measurements
Racic 2009. The continuous time histories of forces in all the three components were observed to be
periodic in nature as shown in Fig.1

Figure 1 Periodic walking time histories in all the three directions


The vertical force time history comprises of two peaks referring to heal strike and toe-off. The time
period for the vertical load time history is the time elapsed between two successive heel strikes or toeoffs. The vertical load is always summed up with the static self-weight and can be resolved using
Fourier-series. At the end of each toe-off, there exists a heel strike at different location. Therefore,
there are times at which both feet are in contact with the ground which should be considered in
loading modeling.
The lateral loading is primarily because of the oscillation of center of gravity of an individual to
maintain body-balance against the lateral vibrations. The time period of loading is twice that of the
vertical loading; hence, lateral vibrations exert a frequency half the vertical foot fall rate Dallard 2001.
The longitudinal component is developed because of continuous push-offs of heel and toe. However,
its effect on the response of structures is not considered to be critical.

MODELING OF THE FOOTBRIDGE

A representative suspension footbridge was considered for modeling and subsequent dynamic
analysis. The footbridge was designed as a suspended structure with unstiffened bridge deck. The
clear span of the footbridge was 142.5m, however for modeling purpose the span is considered to be
150m. The clear walk way width for pedestrians is considered as 2.0m. The superstructure of the
bridge consists of two towers on either ends of the bridge; carrying eight main steel suspension
cables-four on each tower leg, pre-tensioned with ends anchored into main anchor blocks. Four main
suspension cables pass over a pulley at the top of each tower leg spaced at 6.0m c/c. Towers with the
height of 14.85m were anchored through bolts in reinforced concrete foot blocks. The materials were
assumed to be linear isotropic and homogeneous in nature as per Table 1.
Modeling idealizations
3

1. The four main suspension cables resting on each tower leg were modeled as a single cable
element which is fixed at their anchorage points.
2. The horizontal movement of the main suspension cables at the top of tower legs is considered
to be negligible. Therefore modeled as rigid node.
3. The wind guy cables are fixed at both ends of the bridge at their anchorage points.
4. The details concerning to the joints and its connections are avoided in modeling.
5. The portion of the bridge deck which is resting on the soil mass near two approach ends was
considered to be fixed.
Table 1 Material Property of Different Components of the Bridge
Material property

Cables

Rolled sections

Bridge deck

Modulus of elasticity (kPa)

2.1E8

2.1E8

2.73E7

Shear modulus (kPa)

8.078E7

8.078E7

1.19E7

Density (Ton/m3)

3.736

7.85

2.4

Thermal expansion
coefficient

12E-6

12E-6

12E-6

Under gravity the maximum vertical upward and downward displacement of the deck was 1.31m and
1.05m respectively. The contour plots of the vertical displacement are shown as in Fig. 2. In the freevibration analysis, first 75 modes of the bridge were extracted. The cumulative mass participation for
the first 70 modes contributes 95% of the bridge mass out of which first 46modes are observed to
insignificant. Therefore modes from 47 to 70 are considered to be critical and the corresponding
frequencies along with time periods are listed below in Table 2
Table 2 Critical frequencies and time periods
Mode
47
49
54
56
57
58
60
61
62
65
70
Frequency (Hz) 0.28 0.31 0.38 0.42 0.44 0.61 0.67 0.68 0.68 0.74 0.83
Time Period (Sec) 3.52 3.19 2.63 2.40 2.25 1.64 1.48 1.47 1.47 1.34 1.20

Figure 2 Contour plot of vertical displacement of the footbridge and bridge deck under gravity (m)

LOAD MODELING- SINGLE PEDESTRIAN AND RANDOM PEDESTRIAN CROWD


An average walking speed of pedestrian was selected as 1.5 m/s for the present work with exciting
frequency of 2.0 Hz, step length as 0.75m and step width of 0.5 m. The load time histories induced by
the pedestrians during walking as shown in figure 1 are periodic in nature and therefore can be
expressed as Eq. (1, 2 and 3)
( )

Where ( )
respectively.

( )

(1)

( )

(2)

( )

(3)

( ) represents the dynamic loads in vertical, lateral, and longitudinal direction

= Static weight of the pedestrian = 0.7 kN.


The Fourier coefficients

s were adapted according to Bachmann 1987.

The total no .of harmonics considered were n=3, with


, respectively for ( ) .
Total no .of harmonics considered were n=2, with
( ).
Total no .of harmonics considered were n=2, with

=0.4,

1/2=0.05,

1/2=0.04,

= 0.1,

1=0.01,

1=0.2,

3=0.1

and

3/2=0.05

3/2=0.03

and

1,

2,

and
2=0.1

3 taken

2=0.01

for

as

for

( ).

Setra guidelines 2006 the loads due to random pedestrian crowds are applied as the surface area on the
deck and the lateral and longitudinal components are applied as tangential forces respectively. The
factor

is multiplied with the periodic wave-form generated by the single-pedestrian and the

load is distributed over the surface area varying with time to simulate the random pedestrian crowd.
If is the pedestrian arrival rate expressed in persons per sec and td is the duration time taken to
travel the bridge, the total no of pedestrians that can be present at any point of time on the bridge is
given as N which can be calculated as the product of the arrival rate and the duration of the pedestrian
on the bridge as Eq. (4)
N=

(4)

The static component of the pedestrian relating to the pedestrian density is applied as surface pressure
to the deck.
FIB guidelines 2005 Pedestrian density has an influence on the speed of an individual and the
relationship between pedestrian density q and the arrival rate is given as Eq. (5)
(5)

is expressed in persons per m2


is the velocity of the pedestrian in walking expressed in m/s.
is the effective deck width used for pedestrian walking.

SOLUTION TECHNIQUE

Newmarks-Beta time integration scheme Biggs 1964, has been used for the dynamic analysis of the
footbridge under pedestrian induced loads. A time step equal to 1/20th fraction of the time period
observed in pedestrian induced dynamic load is adopted to include the response of the footbridge at
high frequency components. Material Damping was idealized as Rayleigh damping and the
coefficients ( and ) were calculated assuming the damping of the structural system to be 2% and
2.5% for the first two modes.
The dynamic response of the footbridge in the form of displacements and accelerations at critical
sections (mid-span, quarter-span and three quarter span of the footbridge) under single pedestrian load
model are shown as in Figs. 3, 4, 5 and 6. The response under random pedestrian crowd model at midspan of the footbridge is shown as in Figs. 7, 8, 9, and 10 for a representative crowd condition
equivalent to 40% bridge occupancy as per Table 2 according to Bruno 2008. A parametric study on
the response of the footbridge under varying percentage of the bridge occupancy is evaluated and
maximum vertical and lateral accelerations are listed in Table 3 and 4 and are shown as in Figs. 11
and12 respectively. The maximum pedestrian density most likely to be possible is considered as 6
persons per m2 according to Abuarafah 2012.
Table 2 Details of the parameters with varying percentage bridge occupancy
% Bridge
Occupancy
10
20
30
40
50
60
70
80
90

Pedestrian density
(q)
(Persons per m2)
0.6
1.2
1.8
2.4
3.0
3.6
4.2
4.8
5.4

Walking
speed vs
(m/s)
1.31
1.10
0.85
0.62
0.46
0.34
0.25
0.18
0.12

Pacing
rate fp
(Hz)
1.90
1.76
1.56
1.29
1.05
0.83
0.65
0.48
0.34

No. of
pedestrians on
footbridge (N)
180
360
540
720
900
1080
1260
1440
1620

No. of
Synchronized
pedestrians
24.82 25
35.10 36
42.99 43
49.64 50
55.5 56
60.79 61
65.66 66
70.20 71
74.46 75

Figure 3 Variation of vertical displacements


under single pedestrian load model- linear
analysis

Figure 4 Variation of vertical accelerations under


single pedestrian load model- linear analysis

Figure 5 Variation of vertical displacements


under single pedestrian load model- nonlinear
analysis

Figure 6 Variation of vertical accelerations under


single pedestrian load model- nonlinear analysis

Figure 7 Variation of vertical (UY) and


lateral(UZ) displacements under random
pedestrian crowd model -40% bridge occupancy linear analysis

Figure 8 Variation of vertical (AY) and


lateral(AZ) accelerations under random
pedestrian crowd model -40% bridge occupancy linear analysis

Figure 9 Variation of vertical (UY) and


lateral(UZ) displacements under random
pedestrian crowd model -40% bridge occupancy nonlinear analysis

Figure 10 Variation of vertical (AY) and


lateral(AZ) accelerations under random
pedestrian crowd model -40% bridge occupancy
nonlinear analysis

Table 3 Resulting accelerations-linear analysis.


% Bridge
occupancy

Pedestrian
density
Persons/m2

10
0.6
20
1.2
30
1.8
40
2.4
50
3.0
60
3.6
70
4.2
80
4.8
90
5.4
Single Pedestrian(2 Hz)

Vertical acceleration
(av max-m/s2)
Quarter
span
0.4
0.7
0.63
0.5
0.4
0.4
0.3
0.3
0.1
0.12

MidSpan
0.25
0.4
0.65
0.6
0.5
0.4
0.3
0.3
0.1
0.12

Lateral acceleration
(ah max-m/s2)

3-Quarter
span
0.39
0.7
0.63
0.5
0.4
0.4
0.3
0.25
0.1
0.14

Quarter
span
0.15
0.14
0.2
0.2
0.22
0.23
0.13
0.2
0.32
0.016

MidSpan
0.19
0.25
0.2
0.5
0.25
0.48
0.35
0.5
0.41
0.01

3-Quarter
span
0.16
0.14
0.2
0.2
0.24
0.23
0.13
0.2
0.32
0.018

Table 4 Resulting accelerations-nonlinear analysis.


% Bridge
occupancy

Pedestrian
density
Persons/m2

Vertical acceleration
(av max-m/s2)
Mid- Span

Lateral acceleration
(ah max-m/s2)

10

0.6

Quarter
span
1.8

Quarter
span
0.15

Mid- Span

1.0

3-Quarter
span
1.2

0.2

3-Quarter
span
0.15

20

1.2

1.6

1.8

1.0

0.15

0.25

0.12

30

1.8

1.0

1.25

1.0

0.2

0.2

0.2

40

2.4

1.6

1.0

1.0

0.2

0.5

0.25

50

3.0

1.0

0.8

0.8

0.25

0.25

0.24

60

3.6

0.8

0.8

0.5

0.24

0.5

0.25

70

4.2

0.6

0.6

0.5

0.2

0.4

0.22

80

4.8

0.3

0.4

0.4

0.18

0.5

0.23

90

5.4

0.25

0.3

0.2

0.3

0.45

0.3

Single Pedestrian(2 Hz)

0.15

0.16

0.1

0.016

0.0125

0.016

Vertical accelerations
2
1.8

av max m/s2

1.6
1.4

L/2 NON-LINEAR

1.2
3L/4 -NON-LINEAR

L/4 LINEAR

0.8
0.6

L/2 LINEAR

0.4
3L/4 LINEAR

0.2
0
0

10 20 30 40 50 60 70 80 90 100
% Bridge occupancy

Figure 11 Variation of vertical acceleration (m/s2) with % of bridge occupancy

Lateral accelerations
0.6

ah max m/s2

0.5
0.4

L/4 -NON-LINEAR
L/2 -NON-LINEAR

0.3

3L/4 NON-LINEAR
L/4 -LINEAR

0.2

L/2 -LINEAR
0.1

3L/4 LINEAR

0
0

10 20 30 40 50 60 70 80 90 100
% Bridge occupancy

Figure 12 Variation of lateral acceleration (m/s2) with % of bridge occupancy.

Conclusion
1. The maximum vertical displacement and acceleration under single pedestrian walking load
model are observed to be 0.3 m and 0.16 m/s2 respectively.
2. The maximum lateral displacement and acceleration of the bridge deck under single
pedestrian walking load model are observed to be around 0.4mm and 0.016m/s2 respectively
which are smaller in comparison with vertical vibrations.
3. Under random pedestrian crowd model, with the % increase in the bridge occupancy beyond
50%, the vertical and lateral displacements gradually decrease as the stiffness and damping in
the structure increases.
4. The maximum vertical and lateral accelerations under random pedestrian crowd model are
observed to be 1.8 m/s2 and 0.45 m/s2 respectively.
5. The lateral vibrations are observed to be larger under random pedestrian crowd model in
comparison with the single pedestrian walking load model.
6. A complete dynamic analysis during the design stage needs to be performed to limit the
vibrations within specified limits.
References

[1] Abuarafah, Adnan Ghazi, Mohamed Osama Khozium, and Essam AbdRabou (2012). Realtime Crowd Monitoring using Infrared Thermal Video Sequences, Journal of American
Science., 8.3, 133-140.
[2] Bachmann Hugo., and Walter Ammann., (1987), Vibrations in structures: induced by man
and machines, 3, IABSE.
[3] Biggs John M., Introduction to Structural Dynamics, McGrew-Hill, Inc., New York, 1964.
[4] Bruno Luca., and Fiammetta Venuti (2008), The pedestrian speeddensity relation: modeling
and application, Proceedings of Footbridge.
[5] Dallard, P., (2001). London Millennium Bridge: pedestrian-induced lateral
vibration, Journal of Bridge Engineering, 6.6, 412-417.
[6] Einar Thr Inglfsson, 2011. Pedestrian induced lateral vibrations of footbridges,
experimental studies and probabilistic modeling, Ph.D. Thesis, Department of Civil
Engineering, Technical University of Denmark.
[7] Fdration internationale du bton (2005). Guidelines for the Design of Footbridges: Guide
to Good Practice, 32.
[8] Racic, V., Pavic, A., and Brownjohn., J.M.W., (2009), Experimental identification and
analytical modelling of human walking forces: literature review," Journal of Sound and
Vibration, 326.1, 1-49. [9] Stra, Footbridges, (2006), Assessment of vibrational behavior of
footbridges under pedestrian loading, Technical guide SETRA, Paris, France.
[9] ivanovi Stana., Pavic, A., and Paul Reynolds, (2005), Vibration serviceability of
footbridges under human-induced excitation: a literature review, Journal of sound and
vibration, 279.1, 1-74.

10

11

Das könnte Ihnen auch gefallen