Influence of Vibration Modes On Flutter Analysis For Long-Span Suspension Bridges

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

INFLUENCE OF VIBRATION MODES ON FLUTTER ANALYSIS FOR LONG-SPAN SUSPENSION BRIDGES

Shigeki Kusuhara1, Ikuo Yamada2 and Naoki Toyama3


P P P P P P

ABSTRACT
U

Since the aerodynamic stability is one of the most important issues in the wind resistance design for the long-span bridges, various studies have been carried out. The studies on the Akashi-Kaikyo Bridge, the worlds longest suspension bridge with the center span of 1,991m, revealed that not only the three-dimensional effect of the structure and the wind characteristics but also the influence of multi-vibration modes have to be considered in checking the aerodynamic stability1). If the characteristics of dominant vibration mode changes, the critical flutter velocity also may change. This paper describes the results of flutter analysis for an assumed long-span bridge, which has the center span is approximately 1500m or 2300m, and the influence of vibration mode for each bridge.
P P

1. INTRODUCTION
U

There are several plans or ideas of strait crossing road projects in Japan (Fig.1)2). In these projects, super long-span bridges, which would be longer than the Akashi-Kaikyo Bridge, are included. In order to make these super long-span bridges to come true, the aerodynamic stability is one of the most important issues. In the advanced studies on the Akashi-Kaikyo Bridge, special attentions were paid to the following considerations in order to ensure the aerodynamic stability for a long-span bridge. 1) The three-dimensional effect of the structure Honshu-Shikoku Bridges and the wind 2) The influence of multi-vibration modes According to the later consideration, it can be Kanmon Strait Kitan Strait possible to improve the aerodynamic stability of long span bridges by controlling the dominant vibration modes. Therefore, the influence of the vibration mode on flutter characteristics of assumed suspension bridges was examined. This Tokyo Bay Mouth study is base on the Wind resistant design code Ise Bay Mouth for Honshu-Shikoku Bridges (2001)3). Hoyo Strait
P P P P

Shimabara Amakusa, Amakusa - Nagashima Shimabara, Nagashima

Honshu-Shikoku Bridges

Fig.1

Strait crossing road projects in Japan

1 Sub-Leader, Wind and Structural Engineering Group, Long-span Bridge Engineering Center, Honshu-Shikoku Bridge Expressway Co., Ltd. (former Honshu-Shikoku Bridge Authority), Japan 2 Leader, ditto 3 Engineer, ditto

2. OUTLINE AND ASSUMPTION


U

Start The procedure of the flutter analysis in this paper is shown in Fig.2 Fig.3 shows two suspension bridges assumed in the flutter analysis. These bridges are designed for the above-mentioned strait crossing road projects. For simplicity, a two-span suspension bridge with the center span of 1,480m is called Bridge-A, and a three-span suspension bridge with the center span of 2,250m is called Bridge-B. The cross sections of girders are shown in Fig.4. One-box girder is applied to Bridge-A and slotted-box girder with superior aerodynamic stability is applied to Bridge-B. Assumption of Suspension Bridges Choose the Cross section of girder (Spring-Supported test) Coefficients of Three Components Forces and Unsteady Aerodynamic Forces Flutter Analysis Evaluation of Aerodynamic Stability End Fig.2
360 1480

Procedure of flutter analysis


730

1A
860

2P (a) BridgeA
2250

3P

4A
860

1A

2P (b) BridgeB Fig.3 Analysis objects of assumed suspension bridge

3P

4A

(Unit: m)

4.0

3.0

26.1

26.7

(a) One-box girder (Bridge-A) Fig.4

(b) Slotted-box girder (Bridge-B) (Unit: m)

Cross sections of girders for four lanes

3. RESULTS OF SPRING-SUPPORTED TEST


U

The results of the two-dimensional spring supported test at the angles of attack of -3, 0 and +3 degrees are shown in Table 1. Both of the two cross sections showed good aerodynamic stabilities except at +3 degree (one-box girder) and -3 degree (slotted-box girder). Fig.5 shows coefficients of three components forces. The coefficients of unsteady aerodynamic forces, which coordinate system is defined by Fig.6, is shown in Fig.7. Coefficients of the unsteady aerodynamic forces were defined as follows: L = B 2 LZR 2 z + LZI z + B 3 LR 2 + LI (1) (2) where, L:lift, M:aerodynamic moment, z:vertical displacement, :torsional displacement, :circular frequency, ( )':d( )/dt, Lxx or Mxx: coefficients of unsteady aerodynamic forces (Z: caused by vertical vibration, :caused by torsional vibration, R:in phase with displacement, I:in phase with velocity)
ZR
B B B B B B B B

{ ( M = {B (M
3

z + M ZI z ) + B (M R + M I )}
2 4 2

)}

Table 1 Results of the two-dimensional spring-supported test Angle of Attack (deg.) -3 0 +3 Bridge-A (One-box) 82 m/s+ 80 m/s 76 m/s Bridge-B (Slotted box) 81 m/s 100 m/s+ 100 m/s+
Note: + indicates that the critical flutter velocity is larger than the tabulated value

CD , CL
3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 -15 -10 -5 0 5 CD CL CM 10

CM
0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 15

CD, CL
3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 -15 -10 -5 0 5 CD CL CM 10

CM
0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 15

Angle of attack(deg.)

Angle of attack(deg.)

(a)

Bridge-A Fig.5

(b) Bridge-B Coefficients of Drag, Lift and Moment

D L

Fig.6

Coordinate system for unsteady aerodynamic forces

2.0 1.5 1.0 One Box Girder Slotted Box Girder Theodorsen

0.0 -2.0 -4.0


LZI

LZR

0.5 0.0 -0.5 0.00 0.05 0.10 0.15 0.20 0.25 fB/V

-6.0 -8.0 -10.0 0.00 0.05 0.10 0.15 0.20 0.25 fB/V One Box Girder Slotted Box Girder Theodorsen

(a) LZR
B B

(b) LZI
B B

0.1 0.0 -0.1

2.0 0.0 -2.0


M ZI

M ZR

-0.2 -0.3 -0.4 0.00 0.05 0.10 0.15 0.20 0.25 fB/V One Box Girder Slotted Box Girder Theodorsen

-4.0 -6.0 -8.0 0.00 0.05 0.10 0.15 0.20 0.25 fB/V One Box Girder Slotted Box Girder Theodorsen

(c) MZR
B B

(d) MZI
B B

50.0 40.0 30.0


LR

4.0

One Box Girder Slotted Box Girder Theodorsen


LI

3.0 2.0 1.0 0.0 -1.0

One Box Girder Slotted Box Girder Theodorsen

20.0 10.0 0.0 0.00 0.05 0.10 0.15 0.20 0.25 fB/V

0.00

0.05

0.10

0.15

0.20

0.25 fB/V

(e) LR
B B

(f) LI
B B

4.0 3.0 2.0


R

0.5
One Box Girder Slotted Box Girder Theodorsen

0.0 -0.5
I

1.0 0.0 -1.0 0.00 0.05 0.10 0.15 0.20 0.25 fB/V

-1.0 -1.5 -2.0 0.00 0.05 0.10 0.15 0.20 0.25 fB/V One Box Girder Slotted Box Girder Theodorsen

(g) MR
B B

(h) MI
B B

Fig.7

Coefficient of unsteady aerodynamic forces

4. FLUTTER CHARACTERISTICS OF BRIDGE-A


U

1) Influence of high mode Before selecting the important modes in flutter analysis, the influence of high natural modes was examined in order to evaluate the number of modes in the analysis. The critical flutter velocities were calculated for Bridge-A with various combinations of natural vibration modes. These calculations applied a multi-mode flutter analysis, using the mode combination method. Static displacements by wind load, which were calculated from the measured three component forces (Fig.5), were also considered. Table 2 shows the assumption in the flutter analysis. The preliminary analyses by using the lowest 20, 30, 40, and 50 modes were carried out. The analysis concluded that the right solution was equivalent to the approximation by considering at least the lowest 30 modes. In addition, a mode, which had an influence on the flutter characteristics, was expected to exist between the 20th mode and the 30th mode. Therefore, additional analyses were carried out to identify such mode. The conclusion was that the 21st mode affected largely, and the analysis including modes up to the 22nd mode or more could obtain a good approximation of flutter characteristics of Bridge-A as shown in Fig.8. Table 2 Assumptions in flutter analysis
U

Item Analytical method Air density Structural damping Static deformation in wind condition

Coefficient of aerodynamic forces

Analysis condition Mode combination method.( Using lower 50 modes.) 1.23 kg/m3 =0.02 for all modes Considered Main girder Direction Vertical Torsional Horizontal Lift Moment Drag ;Unsteady aerodynamic forces ;Quasi-steady aerodynamic forces Cable: Quasi-steady drag force and lift force (CD=0.7) Tower: Not considered
P P

Forces

150
Critical flutter velocity (m/s)

100

85 59 57 57 57

50

0 10 20 30 Numbers of the lowest modes used in the flutter analysis 40

Fig.8

Influence of the lowest natural modes (Bridge-A: angle of attack=+3 deg.)

However, the critical flutter velocities derived from these flutter analyses (Fig.8) were significantly smaller than the results based on the two-dimensional spring-supported test (Table 1). Consequently, it might be difficult for the long-span suspension bridge with center span of 1480m to evaluate the aerodynamic stability by the results of the ordinary two-dimensional wind test. 2) Selection of dominant modes Since it had been predicted that dominant modes existed between the 1st to the 21st of natural vibration modes, a series of analyses were carried out with various combinations of modes. The analysis resulted that the 2nd (1st symmetric vertical mode), the 8th (2nd symmetric vertical mode), and the 21st (1st symmetric torsional mode) dominate the flutter characteristics of Bridge-A. The combination of these three modes is identified with the combination from the 1st to the 21st mode in the flutter characteristics (shown in Fig.9). Results of natural vibration analysis without wind load are shown in Table 3. According to the table, both the 20th and the 21st modes are symmetric torsional mode having close frequencies to each other. Based on previous experiences, the 20th mode had been applied to spring-supported test condition because of its lower frequency and smaller equivalent mass. Though, the result indicated that the 21st mode affected more than the 20th mode. Dominant vibration modes are shown in Fig.10. The 21st mode, which is basically a torsional mode, was combined with not only horizontal mode but also 2nd symmetric vertical mode. Consequently, the 21st mode is considered to have a relatively large contribution to the flutter characteristics. In addition, the proportion of center span length and side span length is exactly 2 to 1 in Bridge-A. This was expected to excite the vertical vibration in high modes.
U

0.15

Logarithmic decrement

0.1

1st - 21st mode 2nd + 8th + 21st

0.05

2nd + 21st 8th + 21th 21st

-0.05 0 20 40 60 80 100

Wind velocity(m/s)

Fig.9

Selection of dominant modes

(Bridge-A: angle of attack=+3 deg.)

Mode 2

(f=0.078Hz)

Mode 8

(f=0.175Hz)

Mode 20

(f=0.294Hz)

Fig.10 Table 3
Mode 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Frequency (Hz) 0.046 0.078 0.083 0.092 0.115 0.116 0.132 0.175 0.182 0.186 0.193 0.220 0.228 0.243 0.245 0.252 0.263 0.265 0.272 0.294 0.295 0.295 0.305 0.324 0.334 Period (sec) 21.589 12.806 12.025 10.893 8.684 8.592 7.569 5.711 5.485 5.370 5.177 4.552 4.395 4.113 4.081 3.974 3.800 3.769 3.674 3.401 3.391 3.389 3.277 3.082 2.993

Mode 21 (f=0.295Hz) Dominant vibration modes (Bridge-A)

Natural vibration modes without wind load (Bridge-A)


Longitudinal Equivalent mass (kN/m) Vertical Horizontal 166.7 196.1 304.0 147.1 156.9 382.5 186.3 176.5 156.9 176.5 186.3 2177.1 176.5 1147.3 Torsional Mode shape SH-1 SV-1 AV-1 (1) AH-1 AH-2 AV-1 (2) SV-2 (1) SV-2 (2) SH-2 (1) SV-2 (1) AV-2 (2)

353.0

264.8 4020.7

SV-3 SH-2 (1)

892.4 1304.3 490.3 294.2 176.5 205.9 1088.5 509.9

SH-2 (2) 18485.5 38079.2 ST-1 (1) ST-1 (2) AV-3 (1) AV-3 (2)

S: symmetric, A: asymmetric, H: horizontal, V: vertical ,T: torsional

5. FLUTTER CHARACTERISTICS OF BRIDGE-B


U

1) Influence of higher mode The influence of high natural modes was examined in order to evaluate how many number of modes in the analysis for Bridge-B in the same way as Bridge-A. The analysis including modes up to the 20th mode could obtain a good approximation of flutter characteristics of Bridge-B as shown in Fig.11. Therefore, the influence of the lowest 20 modes of Bridge-B was investigated.
U

150 130

Critical flutter velocity (m/s)

115 118 100 114 114 114

114

50

0 10 20 30 Numbers of the lowest modes used in the flutter analysis 40

Fig.11
U

Influence of the lowest natural modes (Bridge-B: angle of attack=0 deg.)

2) Selection of dominant modes Flutter analyses with various combinations of vibration modes of Bridge-B were carried out by the same way as Bridge-A. Fig.12 shows the result of analysis. The combination of the 4th mode (1st symmetric vertical mode) and the 13th mode (1st symmetric torsional mode), which usually applied to spring supported test, did not excite the flutter. Furthermore, the result of flutter analysis with the lowest 20 modes can be represented by the analysis with four dominant modes, 4th, 11th (2nd symmetric vertical mode-a), 13th and 14th (2nd symmetric vertical mode-b). The critical flutter velocity, calculated by removing the 11th mode from these four modes, was lower than that calculated with these four modes. As mentioned above, it was found that the 11th mode is the important mode which suppresses the excitation of flutter characteristics.
0.30

Logarithmic decrement

0.20

1st - 20th mode 4th + 11th +13th + 14th

0.10

4th + 11th + 13th 4th + 13th

0.00

4th + 13th + 14th

-0.10 70 80 90 100 110 120

Wind velocity(m/s)

Fig.12

Selection of dominant modes (Bridge-B: angle of attack=0 deg.)

Mode 4

(f=0.073Hz)

Mode 11

(f=0.117Hz)

Mode 13

(f=0.145Hz)

Mode 14 Fig.13 Table 4


Mode 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Frequency (Hz) 0.037 0.066 0.072 0.073 0.084 0.086 0.086 0.086 0.098 0.103 0.117 0.118 0.145 0.147 0.156 0.157 0.165 0.171 0.176 0.178 0.181 0.194 0.194 0.198 0.198 Period (sec) 26.851 15.084 13.924 13.661 11.875 11.692 11.611 11.610 10.168 9.717 8.544 8.502 6.904 6.794 6.411 6.353 6.078 5.850 5.687 5.615 5.530 5.162 5.152 5.059 5.056

(f=0.147Hz)

Dominant vibration modes (Bridge B)

Natural vibration modes without wind load (Bridge-B)


Longitudinal Equivalent mass (kN/m) Vertical Horizontal 19.6 19.6 29.4 19.6 323.6 470.7 19.6 19.6 39.2 19.6 19.6 39.2 3834.4 19.6 3 19.6 Torsional Mode shape SH-1 AH-1 AV-1 (1) SV-1 L L SH-1 (1s) SH-1 (2s) AV-1 (1) SH-2 (1) SV-2 (1) AV-1 (2) ST-1 SV-2 (2) AH-2 (1) AV-2

49.0 509.9 19.6 19.6

49.0 4520.9 39.2 5403.5 50720.0

2226.1

107.9 78.5 68.6 205.9 137.3 43257.1 19.6

229642.3 4118.8 5364.2

SH-2 (2) AT-1 (1) AT-1 (2) SH-2 (3) AH-2 (2) SV-3

S: symmetric, A: asymmetric, H:horizontal, V: vertical T: torsional, L: longitudinal, (s):side-span mode

6. EVALUATION ON ENERGY
U

Besides three-dimentional analysis, the investigation was carried out in order to identify the part of span exciting the flutter. The energy of aerodynamic forces which work on the girder was examined. The energy excited by aerodynamic forces are defined in the following equation.
WL = LR dy R WM
B B

= M
B B B

R d R

(3)

where WL and WM are the energy in vertical and torsional directions,


B B B B B B B

is the path integral during

vibration, LR and MR are the real part of the lift and pitching moment of unsteady aerodynamic forces, yR and and R are the real part of the vertical and torsional displacements, respectively. The aerodynamic stability of the bridge depends on plus or minus of Eq.(4), the integral of energy on each nodal points.

span

Wdl =

(W
span 0

+ WM )dl

(4)

Fig.14 shows the spanwise distributions of energy on girders of Bridge-A and Bridge-B at the critical flutter velocity. It is assumed that the spanwise distribution of energy is strongly associated with the flutter mode shape (shown in Fig.15). When we focus the energy distribution in the center span, Bridge-A indicates a high symmetric mode (the 3rd symmetric mode). On the other hand, Bridge-B is the 1st symmetric mode shape. Therefore, each bridge is evaluated to have a different characteristic. Furthermore, the distribution of energy in the side span, as shown in Fig.14, seems to be minus, and it means that the side span has a damping effect regularly. Bridge-A has only single side span, and this fact seems to reduce the aerodynamic stability of the Bridge-A.
()
Lift Moment Drag Total Energy

Energy

2P

-750

-250

250

750

3P

1250

()

Length (Bridge-A: angle of attack=+3 deg.)


Lift Moment Drag Total Energy

()

Energy

2P

-1985

3P

1985

()

Length (Bridge-B: angle of attack=0 deg.) Fig.14 Spanwise distribution of energy

Akashi-Kaikyo
T/8

Bridge-A

Bridge-B

Horizontal Vertical Torsional

3T/8 5T/8 7T/8

T/8 3T/8 5T/8 7T/8

T/8 3T/8 5T/8 7T/8

T: period Fig.15 Flutter mode shape

7. CONCLUSIONS
U

The flutter analyses were carried out for the assumed long-span suspension bridges with the center spans of 1,500m and 2,300m. The results of the flutter analyses are summarized as follows, 1) The combination of the lowest flexural mode and torsional mode, which usually used in the two-dimensional spring supported test, might have the possibility of underestimating the flutter instability for super long-span bridges. 2) The characteristics of multi-mode flutter with dozens of vibration modes can be represented by the analysis with only three or four dominant modes in flutter analysis. 3) Some of the dominant modes can control the excitation of flutter, which implies that the adjustment of vibration modes can enhance the critical flutter velocity. It is necessary to obtain the influence of high natural modes in more detail by using the same analytical method for other suspension bridges.

REFERENCES 1) Honshu-Shikoku Bridge Authority: The Akashi-Kaikyo Bridge: Design and Construction of the Worlds Longest Bridge, pp.97-100, 1999.8 2) H. Sato, N. Hirahara, K. Fumoto, S. Hirano and S. Kusuhara: Full Aeroelastic Model Test of a Super Long-span Bridge with Slotted Box Girder, Journal of Wind Engineering and Industrial Aerodynamics, No.90, pp.2023-2032, 2002 3) Honshu-Shikoku Bridge Authority: Wind resistant design code for Honshu-Shikoku Bridges (2001), 2001.8 4) N. Toyama, K. Hata and S. Kusuhara: The Influence of the Vibration Mode on Flutter Characteristics of Long-span Bridges, The 6th Asia-Pacific Conference on Wind Engineering, 2005,9
U U

You might also like