You are on page 1of 8

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27,

2010

Study on aerodynamic forces of stay cables under rainwind induced vibration by a hybrid method of test combining with computation
Chen Wen-Lia, Li Huia , Xu Fenga, Li Feng-Chenb, Ou Jin-Pinga,c
a b

School of Civil Engineering, Harbin Institute of Technology, Harbin, 150090, China School of Energy Science and Engineering, Harbin Institute of Technology, Harbin, 150001, China c School of Civil and Hydraulic Engineering, Dalian University of Technology, 116024,China

ABSTRACT: This paper presents a hybrid approach of test combining with computational fluid dynamics (CFD) numerical simulation for investigation on the aerodynamic forces of rain-wind induced vibration (RWIV) of a stay cable, based on the structural response to be easily measured and the flow field around the structures to be difficultly obtained, in particular for RWIV of a stay cable. The stay cable and the flow field are regarded as two substructures of a system. The motion of the stay cable measured through wind tunnel tests and the rivulets forming on the cable are quantitatively measured by an ultrasonic transmission thickness measurement system (UTTMS) are regarded as known moving boundary conditions and applied to the flow field. The flow around the moving boundary is then numerically simulated. The result indicates that the hybrid approach can accurately simulate the transient aerodynamic coefficients of the stay cable. The equivalent damping ratios induced by aerodynamic forces are obtained at various wind speeds. Furthermore, the nonlinear aerodynamic force coefficient is identified by a nonlinear least square technique. 1 INTRODUCTION Rain-wind induced vibration (RWIV) is a fluid-structure interaction phenomenon that occurs when rain and wind act simultaneously on the inclined cables, which was firstly reported by Hikami & Shiraishi (1988) for the Meikonishi Bridge, and then was observed worldwide in other bridges. Many wind tunnel experiments (Hikami & Shiraishi, 1988; Matsumoto, 1990, 1992, 1995, 1998, 2003;, Flamand, 1995; Bosdogianni, 1996; Verwiebe, 1998; Cosentino et al, 2003; Gu & Du, 2005; Li et al, 2009) are carried out for this vibration. The experiments can reproduce the RWIV of the cable segment models under the condition of artificial rainfall and obtain the vibration response, but the flow fields of the models are not easy to be measured. So, the synchronous aerodynamic force can not be obtained when measuring the vibration response of the cable models. For the RWIV of a stay cable, the cable response is easily measured and the flow field around the cable is difficultly obtained. Based on this background, the hybrid approach of tests combining with CFD is carried out. The stay cable and the flow field are regarded as two substructures of a system. The motion of the stay cable measured in the test is regarded as a known moving boundary condition in the flow field. The flow around the moving boundary is then numerically simulated. The transient aerodynamic coefficients

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

are obtained and applied to a single degree of freedom model (SDOF) of the stay cable. The comparison is carried out between the SDOF model and the experimental results. The equivalent damping ratios to the stay cable induced by aerodynamic forces are obtained at various wind speeds. Furthermore, the nonlinear aerodynamic force coefficient is identified by a nonlinear least square technique. 2 THE HYBRID APPROACH FOR INVESTIGATION ON RWIV OF A CABLE 2.1 Description of RWIV measurement results in wind tunnel tests The RWIV of a stay cable is reproduced in the wind tunnel tests and the rivulet oscillations are synchronously measured by an ultrasonic transmission thickness measurement system (UTTMS) (Li et al, 2010). But the aerodynamic forces are difficultly measured. In this study, the hybrid approach is employed to investigate the aerodynamic forces. The stay cable used in this study has the same geometry parameter with the experimental condition, i.e. the length L and diameter D of the cable are 2.0 m and 0.1 m, respectively. The yaw and inclined angles are 22.5and 30, respectively, which are also the same with the experimental condition, as shown in Fig. 1.
y

Fy Cy

U sin U sin cos U sin sin


U cos

Fx Cx

U
U cos U sin

U0

& y
U sin sin

(b) Photo of testing set-up (c) Relative velocity to model (a) Spatial location of cable model Fig. 1. Schematic diagram of the inclination angle and wind yaw angle of cable model

As the cable vibrates, the wind direction 0 is changed in the experiment. The x positive direction is set the wind direction for defining the aerodynamic coefficients. The lift coefficient C y and drag coefficient C x can be written as:
Fy Fx , Cx = 1 (1) 1 2 (U cos ) D (U cos )2 D 2 2 where Fx and F y are the aerodynamic forces of the cable in x and y directions, respectively; is the air density and U is the outlet mean velocity of wind tunnel. Cy =

The definition of the rivulet position and schematic of rivulet measurement by UTTMS is shown in Fig. 2. The vibration amplitude of the stay cable changing with the wind speed is denoted in Fig. 3.
0.10

First test Second test


0.08

Ampitude (m)

0.06

0.04

0.02

0.00 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0

Wind speed (m/s)

Fig. 2. Schematic of rivulet measurement by UTTMS

Fig. 3. Vibration amplitude versus wind speed

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

When the velocity is over 8.04 m/s or below 6.76 m/s, the RWIV of the stay cable does not occur; and only the wind speed is in the range of 6.76-8.04 m/s, such as 7.4 m/s and 7.72 m/s, the RWIV occurs. Meanwhile, the water rivulets present three different states in these three wind speed ranges, as shown in Fig. 4a, b, c, respectively. Below 6.76 m/s, the continuous upper rivulets do not flow along axial direction of the cable and slide to the windward side. Fig. 4a is the result of water rivulet distribution on the cable surface measured by UTTMS, which is similar to a snapshot in Fig. 5a. When the wind speed exceeds the critical wind speed of 6.76 m/s, the RWIV of the cable model occurs and the continuous upper rivulet forms and comes into a steady circumferential oscillation as shown in Fig. 4b at 7.72 m/s. The continuous lower rivulet also forms but nearly un-moves in the circumferential direction. The snapshot given in Fig. 5b also shows a through upper rivulet forming on the cable surface. As the wind speed reaches 8.04 m/s, the upper rivulet still remains, but is disordered and unsystematic as shown in Fig. 4c which is different with the approximate simple harmonic vibration in Fig. 4b.
360

Lower rivulet
Position (deg)

315

Lower rivulet

270

225

Upper rivulet

180

135

Upper rivulet

90

Upper rivulet

45

10

Time (s)

(a) 6.12 m/s (b) 6.76 m/s (c) 8.04 m/s Fig. 4.Water rivulet contour along circumferential direction measured by UTTMS

Upper rivulet Upper rivulet

Upper rivulet

(a) 6.12 m/s (b) 6.76 m/s Fig. 5.A snapshot of water rivulet distribution

(c) 8.04 m/s

2.2 Models The lower rivulet is neglected in the hybrid approach according to the analysis of the rivulet states above. The statistical mean shape of upper rivulet is about an isosceles trapezoid with a hemline of 8 mm, a height of 0.509 mm and the bottom angle of 45 degree (Li et al, 2010). The wind speed range can be classified into three conditions: 6.12 m/s and 6.76 m/s are defined as condition 1 with no continuous upper rivulet, 7.40 m/s and 7.72 m/s are defined as condition 2 with a regular rivulet oscillation, and 8.04 m/s, 8.69m/s and 9.94 m/s are defined as condition 3 with a disordered rivulet oscillation. For conditions 2 and 3, the system is classified into: the flow field, the stay cable and the upper rivulet substructures, respectively, and the model of cable and rivulets are shown in Fig. 6. For condition 1, the system is only classified into the fluid substructure and the stay cable substructure with many small sliding rivulets, respectively. The model of cable and rivulets are shown in Fig. 7. The sliding rivulets are fixed on the surface of the stay cable and no relative motion exists between the stay cable and sliding rivulets.

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

Fig. 6. Model of cable and grid partition for conditions 2 and 3

Fig. 7. Model of cable and grid partition for condition 1

2.3 Moving boundary of the flow field There are seven cases: 6.12 m/s, 6.76 m/s, 7.40 m/s, 7.72 m/s, 8.04 m/s, 8.69 m/s and 9.94 m/s. Through the reproduction the RWIV, the cable vibration and the upper rivulet oscillation are synchronously measured. Figs 8, 9 indicate that the frequencies of the cable vibration and the rivulet oscillation are identical to the natural frequency 0.952 Hz of the stay cable when the RWIV occurs at the wind speed of 7.72 m/s. Figs 8, 9 are added to the flow field as the moving boundaries, which are interfaces between flow field with the stay cable and the upper rivulet, respectively. The condition of wind speed of 7.40 m/s is close to that of 7.72 m/s, so the moving boundaries of 7.40 m/s are not shown here.
0.06

0.05

0.9521 Hz
Displacement (m)

0.15

0.9528 Hz
165

Position of upper rivulets (deg)

0.10

0.04

0.05

160

155

Amplitude

0.00

Amplitude

150

0.03

-0.05

145

140 135

-0.10

0.02
-0.15 0 10 20 30 40 50

130

Time (s)

10

20

30

40

50

0.01

1.721 Hz

Time (s)

0.00 0 1 2 3 4 5

0 0 1 2 3 4 5

Frequency (Hz)

Frequency (Hz)

Fig. 8 Transverse displacement response of cable model at a wind speed of 7.72 m/s

Fig. 9 Upper rivulet oscillation response along circumferential direction at a wind speed of 7.72 m/s

Figs 10, 11 indicate that the cable vibration frequency is just the natural frequency of the stay cable with small vibration amplitude, but with no dominant frequency for the rivulet oscillation at 8.69 m/s. It is the same characteristic at 8.04 m/s and 9.94 m/s.
0.5
0.005
Position of rivulets (deg)
175 170 165 160 155 150 145 140 135

0.952 Hz
0.020

0.4

0.004
Displacement (m)

0.015 0.010 0.005 0.000 -0.005 -0.010 -0.015 -0.020

Amplitude

Amplitude

0.003

0.3

0.002

0.2

10

15

20

25

30

35

40

45

50

Time (s)

0.001

10

15

20

25

30

35

40

45

50

0.1

Time (s)

0.000 0 1 2 3 4 5

0.0 0 1 2 3 4 5

Frequency (Hz)

Frequency (Hz)

Fig. 10 Transverse displacement response of cable model at a wind speed of 8.69 m/s

Fig. 11 Upper rivulet oscillation response along circumferential direction at a wind speed of 8.69 m/s

Fig. 12 indicates that the RWIV of the stay cable at 6.12 m/s. The vibration amplitude is very small, but with a dominant frequency of 0.952 Hz. The cable and rivulet have the same boundaries and the similar set can be used for case of the wind speed of 6.76m/s.

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

0.012 0.009

0.0020

Displacement (m)

0.006 0.003 0.000 -0.003 -0.006 -0.009 -0.012 0 5 10 15 20 25 30 35 40 45 50

0.0015

0.952 Hz

Amplitude

0.0010

0.0005

0.0000 0 1 2 3 4 5

Time (s)

Frequency (Hz)

(a) Time histories (b) Frequency spectrum analysis Fig. 12 Transverse displacement response of cable model at a wind speed of 6.12 m/s

3 RESULTS AND ANALYSIS 3.1 Aerodynamic coefficients The aerodynamic behavior of the flow field with moving boundary is calculated step by step and then the aerodynamic coefficients of the cable are obtained by Fluent 6.3. Fig. 13 indicates the time histories and their frequency responses of the lift coefficients (the respective means have been subtracted). The initial large variant in the lift coefficient histories may be induced by the numerical simulation being not into a steady state.
0.09
0.0003

0.06

Lift coefficient

0.00

Amplitude

0.03

0.0002

0.952 Hz

-0.03

0.0001

-0.06
0.0000

-0.09 0 5 10 15 20 25 30 35 40 45 50

Time (s)

Frequency (Hz)

(a) 6.12 m/s


0.6
0.05 0.04

0.3

Lift coefficient

0.952 Hz

Amplitude
0 5 10 15 20 25 30 35 40 45 50

0.03

0.0

0.02

-0.3

0.01

-0.6

0.00 0 1 2 3 4 5

Time (s)

Frequency (Hz)

(b) 7.72 m/s


0.4
0.010 0.008

Lift coefficient

0.2

Amplitude
0 5 10 15 20 25 30 35 40 45 50

0.006

0.0

0.004

-0.2

0.002

0.000

-0.4
0 1 2 3 4 5

Time (s)

Frequency (Hz)

(c) 8.69 m/s Fig. 13. Time histories lift coefficients of the stay cable at various wind speeds

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

3.2 Cable vibration response The lift force is then applied to the SDOF model of the stay cable in the cross-flow direction, and the cable vibration response is obtained and compared with the result of wind tunnel tests. The SDOF model of the stay cable can be expressed as:
& + Cy & + Ky = Fy , Fy = M& y

1 U 2 DC y 2

(2)

where M , C and K are the mass, damping coefficient and stiffness coefficient of the cable, respectively; C y is the lift coefficient history, as shown in Fig. 13, Fy is the lift force history of the cable and U is the inlet velocity in the numerical simulation. The cable vibration response is obtained by Eq. (2) and very close to experimental results at 7.72 m/s as shown in Fig. 14b. However, some difference between them at other wind speeds can be observed. The phenomenon may be attributed to that the upper rivulets around the circumferential direction are not entirely synchronically oscillating or sliding and may behave randomly. It induces irregularly aerodynamic forces.
0.05

0.2

0.05

0.04

Numerical result Experimental result

0.15

Numerical result Experimental result

0.04

Numerical result Experimental result

0.03

0.03

0.1

Displacement(m)

Displacement(m)

0.02

Displacement(m)

0.02

0.01

0.05

0.01

-0.01

-0.05

-0.01

-0.02

-0.02

-0.1
-0.03
-0.03

-0.04

-0.15

-0.04

-0.05

10

15

20

25

30

35

40

45

50

-0.2

-0.05

10

15

20

25

30

35

40

45

50

10

15

20

25

30

35

40

45

50

Time (s)

Time (s)

Time (s)

(a) 6.12 m/s

(b) 7.72 m/s

(c) 8.69 m/s

Fig. 14. Comparison of cable resopnses by numerical and experimental investigations

3.3 Aerodynamic damping The aerodynamic force will add an equivalent damping ratio to the stay cable. The occurrence of RWIV may be caused by the aerodynamic damping which may result in the total damping of the stay cable into a negative value. The most common method for defining the equivalent damping ratio is to let the energy dissipated in a vibration cycle of the actual structure be identical to that of an equivalent damping system. The equivalent damping ratio can be expressed by the following equation (Chopra, 2005): E eq = D (3) 4Es 2 where E D is the dissipated energy by the aerodynamic force in one cycle; Es = Ku0 2 is the elastic strain energy in one cycle of the cable, u 0 is the amplitude of the cable transverse oscillation y in Eq. (2). The system damping ratio of the stay cable under a free vibration with zero wind speed is 0.0017.The equivalent damping ratio added by the aerodynamic forces at different speeds are shown in Fig. 15. The horizontal axial is the periodic numbers; the vertical axial is the equivalent damping ratio and system damping ratio at each period of cable oscillation. The equivalent damping ratio by the aerodynamic force is negative and their absolute values are larger than system damping ratio of 0.0017 in most oscillating periods at 7.72 m/s where the RWIV of the stay cable occurs. At other wind speeds, the equivalent damping ratio by the aerodynamic force is smaller than system damping ratio in most oscillating periods. It is indicated that the occurrence of RWIV has a close relationship with the equivalent damping ratio by the aerodynamic force.

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

0.01

0.01

0.01

System damping ratio Equivalent damping ratiio

System damping ratio Equivalent damping ratiio

System damping ratio Equivalent damping ratiio

Equivalent damping ratio

Equivalent damping ratio

0.005

0.005

Equivalent damping ratio


0 5 10 15 20 25 30 35 40 45 50

0.005

0.0017 0

0.0017 0

0.0017 0

-0.005

-0.005

-0.005

-0.01

-0.01

-0.01

-0.015

10

15

20

25

30

35

40

45

50

-0.015

-0.015

10

15

20

25

30

35

40

45

50

Periodic numbers

Periodic numbers

Periodic numbers

(a) 6.12 m/s (b) 7.72 m/s (c) 8.69 m/s Fig. 15. Comparison of equivalent damping ratio applied by aerodynamic forces and system damping ratio at different speeds

3.4 Aerodynamic force coefficient


& , t ) in Eq. (2) is a nonlinear function of cable vibration, &,, The lift coefficient C y ( y, y rivulet oscillation and time, and is not easy to be obtained in the wind tunnel tests or field measurement. The lift coefficient histories in the seven velocity cases are obtained through the hybrid approach, and the seven cases include the whole process of RWIV of & , t ) can be expressed as following: &,, the cable. The nonlinear lift coefficient C y ( y, y 3 3 & & + c2 y + c3y & + c4 y & + c5 y + c6 + c7 C y = c1 y (4) & are the displacement and velocity of the cable vibration, respectively, where y and y & is the angular velocity of the rivulet oscillation, and c1 , c2 , c3 , c4 , c5 , c6 and c7 & , t ) is identified through a & , , are coefficients. The nonlinear lift coefficient C y ( y, y nonlinear least square method and can be expressed as following: & & + 0 .0823 y + 0 .755 y & 0 .246 y & 3 + 13 .933 y 3 + 0 .294 + 0 .00253 C y = 0 .169 y (5) The comparison between the results by the hybrid approach and parameter identification is shown in Fig. 16. The identification results of 6.12 m/s, 6.76 m/s, 7.40m/s and 7.72 m/s are better than other cases.
0.4

Lift coefficient

0.2

CFD numerical results Identification results

-0.2

6.12m/s
-0.4 0

6.76m/s
50

7.40m/s
100

7.72m/s
150

8.04m/s
200

8.69m/s
250

9.94m/s
300 350

Time (s)

Fig. 16. Lift coefficient histories comparison of seven cases between CFD and identification

4 CONCLUSION This paper presents a hybrid approach of tests combining with CFD to obtain the aerodynamic force of the stay cable when the RWIV occurs. The following conclusions are obtained from this study: (1) Over the wind speed range of the RWIV occurrence, more accurate aerodynamic force can be obtained and it has a dominant frequency (the same as that of cable) and larger amplitude; below above wind speed range, the aerodynamic force has a main frequency but a very small amplitude; and larger than the wind speed of the RWIV occurrence, the aerodynamic force has a larger amplitude but no dominant frequency.

The Fifth International Symposium on Computational Wind Engineering (CWE2010) Chapel Hill, North Carolina, USA May 23-27, 2010

(2) When RWIV occurs, the equivalent damping ratio added by the aerodynamic force is negative and its absolute is larger than system damping ratio in most oscillating periods. Over other wind speeds in which RWIV cannot occur, the equivalent damping ratio achieved by the aerodynamic force is also negative, but its absolute value is smaller than system damping ratio in most oscillating periods. (3) The nonlinear aerodynamic force coefficient is identified through a nonlinear least square technique based on the results obtained by the hybrid approach and it agrees well with the calculated results for most of test cases. 5 ACKNOWLEDGEMENTS This research is funded by National Natural Sciences Foundation of China (NSFC) (90815022), (50910290), (90715015) and Natural Scientific Research Innovation Foundation in Harbin Institute of Technology (HIT.NSRIF.2009099). The authors gratefully appreciate to Prof. Yaojun Ge, Dr. Lin Zhao and staff of the Wind Tunnel Laboratory of TongJi University in China for their great help in the wind tunnel test. 6 REFERENCES
Bosdogianni A., Olivari D., 1996. Wind- and rain-induced oscillations of cables of stayed bridges. Journal of Wind Engineering and Industrial Aerodynamics 64, 171-185. Chopra A. K., 2000. Dynamics of structures: Theory and applications to earthquake engineering (Second edition), Prentice-Hall, Englewood Cliffs, NJ, USA. Coesentino N., Flamand O., Ceccoli C., 2003. Rain-wind-induced vibration of inclined stay cables Part I: experimental investigation and physical explanation. Wind and Structures 6(6), 471-484. Flamand O., 1995. Rain-wind-induced vibration of cables. Journal of Wind Engineering and Industrial Aerodynamics 57, 353-362. Gu M., Du X.Q., 2005. Experimental investigation of rainwind-induced vibration of cables in cablestayed bridges and its mitigation. Journal of Wind Engineering and Industrial Aerodynamics 93, 79-95. Hikami Y., Shiraishi N., 1988. Rain-wind-induced vibrations of cables in cable stayed bridges. Journal of Wind Engineering and Industrial Aerodynamics 29, 409-418. Li F. C., Chen W. L., Li H., Zhang R., 2010. An ultrasonic transmission thickness measurement system for study of water rivulets characteristics of stay cables suffering from wind-rain-induced vibration. Sensors & Actuator: A. Physical 159:12-23. Matsumoto M., Shiraishi N., Kitazawa M., Knisely C., Shirato H., Kim Y., Tsujii M., 1990. Aerodynamic behavior of inclined circular cylinders-cable aerodynamics. Journal of Wind Engineering and Industrial Aerodynamics 33, 63-72. Matsumoto M., Shirashi N., Shirato H., 1992. Rain-wind-induced vibration of cables of cable-stayed bridges. J. Wind Eng. Ind. Aerodyn. 41-42, 2011-2022. Matsumoto M., Saitoh T., Kitazawa M., Shirato H., Nishizaki T., 1995. Response characteristics of rainwind-induced vibration of stay-cables of cable-stayed bridges. Journal of Wind Engineering and Industrial Aerodynamics 57, 323-333. Matsumoto M., Daito Y., Kanamura T., Shigemura Y., Sakuma S., Ishizaki H., 1998. Wind-induced vibration of cables of cable-stayed bridges. Journal of Wind Engineering and Industrial Aerodynamics 7476, 1015-1027. Matsumoto M., Yagi T., Goto M., Sakai S., 2003. Rain-wind-induced vibration of inclined cables at limited high reduced wind velocity region. Journal of Wind Engineering and Industrial Aerodynamics 91, 1-12. Verwiebe C., Rucheweyh H., 1998. Recent research results concerning the exciting mechanisms of rainwind-induced vibrations. Journal of Wind Engineering and Industrial Aerodynamics 74-76, 1005-1013.

You might also like