Suppression of Vortex Shedding from a Circular Cylinder by using a Suction Flow Control Method

Size: px

Start display at page:

Download "Suppression of Vortex Shedding from a Circular Cylinder by using a Suction Flow Control Method"

Jasmine Cain
6 years ago
Views:

1 51st AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition January 013, Grapevine (allas/ft. Worth Region), Texas AIAA Suppression of Vortex Shedding from a Circular Cylinder by using a Suction Flow Control Method Wenli Chen 1 epartment of Aerospace Engineering, Iowa State University, Ames, Iowa, 50011, USA; School of Civil Engineering, Harbin Institute of Technology, Harbin, Heilongjiang, China, , China Hui Hu ( ) epartment of Aerospace Engineering, Iowa State University, Ames, Iowa, 50011, USA Hui Li 3 School of Civil Engineering, Harbin Institute of Technology, Harbin, Heilongjiang, , China An experimental study was conduct to suppress the vortex shedding from a circular cylinder by using a suction flow control method. The suction flow control was accomplished using two suction holes located on the test cylinder model at an angle of 90.0 degrees in relation to the oncoming flow direction. In addition to measuring the pressure distributions around the test model, a high-resolution digital particle image Velocimetry (PIV) system was used to conduct detailed flow field measurements to quantify the transient shedding behavior of the unsteady vortex and turbulent flow structures in the wake of the test cylinder model under different suction flow control conditions. The effectiveness of the vortex shedding suppression using the suction flow control method was evaluated based on the quantitative measurement results, which include the mean and fluctuating pressure coefficient distributions around the test cylinder model, the lift and drag coefficients of the test cylinder model with and without suction flow control, the instantaneous velocity and vorticity fields, the mean velocity and turbulence kinetic energy distributions around the cylinder model, and the mean velocity profiles in the wake of the test model under different suction flow control conditions. The measurement results of the present study revealed that the suction flow control method exhibited excellent performance in suppressing the alternative vortex shedding from the circular cylinder model (i.e., change the alternative vortex shedding mode into symmetric mode). As a result, the fluctuations of the lift coefficients and drag coefficients of the cylinder model were found to be reduced dramatically by using the suction flow control method. Nomenclature C pi = wind pressure coefficient of the point i on the surface of the cylinder model, i = 1 6 C x C y F x F y L = drag coefficient of the test model = lift coefficient of the test model = diameter of the cylinder model = aerodynamic drag force of the test model = aerodynamic lift force of the test model = length of the test model p i = wall static pressure of the point i on the surface on the cylinder model 1 Visiting scholar, epartment of Aerospace Engineering; Assistant professor, School of Civil Engineering. Associate Professor, epartment of Aerospace Engineering, AIAA Associate Fellow, huhui@iastate.edu 3 Professor, School of Civil Engineering. 1 Copyright 013 by Wenli Chen, Hui Hu and Hui Li. Published by the, Inc., with permission.

2 p Re U 0 x y = static pressure in front of the cylinder model = Reynolds number = velocity at the inlet = coordinate along the in-flow direction = coordinate along the cross-flow direction ρ = air density, assumed to be 1.05kg/m 3 θ i = pressure tap angle relative to the x coordinate = angle difference between two neighbouring pressure taps, = θ i θ i I. Introduction n recent years, long-span suspension bridges and cable-stayed bridges have been widely constructed worldwide I due to their superior structural performance and elegant appearance. As key components of cable-stayed bridges, inclined cables often vibrate under wind, rain and traffic loads, such as the vortex-induced vibration (VIV) (Main and Jones, 1999; Matsumoto et al., 003; Zuo et al., 008; Zuo and Jones, 010) and rain-wind induced vibration (RWIV) (Hikami and Shiraishi, 1988). Although the VIV of the cable is a self-limiting vibration, it frequently occurs at lower wind speeds. Zuo et al. (008) observed the VIVs of cables on the Fred Hartman Bridge in Houston, Texas, USA through a long full-scale measurement investigation. Frequent wind-induced vibration may induce fatigue damages to the cables, and mitigation is necessary to reduce the vibration. Viscous damper devices have been widely used to control the VIVs and RWIVs of the stay cables of cablestayed bridges (Main and Jones, 001; Persoon and Noorlander, 1999; Zhou and Xu, 007). Main and Jones (001) obtained an additional optimal damping ratio to the stay cable provided by the viscous damper. Wang and Xu (007) employed an active stiffness control method to suppress the RWIVs of prototype stay cables. Li et al. (007) and Liu et al. (007) investigated the theory, algorithm and model test of the semi-active control of stay cables windinduced vibrations using magnetorheological fluid dampers, which can effectively reduce the cable vibrations and increase the stay cables system damping ratios. The control mechanisms of each type of damper device provided additional damping ratios to the stay cables and increased their ability to resist wind-induced vibrations, but the methods did not change or reduce the wind loads on the stay cables. Aerodynamic methods, including passive and active approaches, can be used to reduce the wind loads on the stay cables. They can simultaneously increase the natural damping ratios of the stay cables at certain times. Matsumoto et al. (199) and Flamand (1995) proposed the use of protuberances to suppress the RWIVs of stay cables. Gu and u (005) conducted wind tunnel tests using double-spiral wires to mitigate vibration. Owen et al. (001) attached hemispherical bumps to the circular cylinder surface to control the vortex-induced vibration amplitudes of the cable models. Many active aerodynamic approaches were also used to control the flow separation and flow-induced vibration. Wu et al. (007) developed a moving-wall, e.g., appropriate travelling transverse wave, control strategy to manage the unsteady separated flow over a circular cylinder. This method allows the global flow to remain attached when there is a strong adverse pressure gradient and eliminates vortex shedding. Modi (1997), Munshi et al. (1997) and Patnaik et al. (00) investigated the use of momentum injection to control the flow field around the airfoils, flat plates, rectangular prisms, -section prisms and circular cylinders; the authors found that this method could effectively resist the wind-induced vortex and galloping instabilities. Grager et al. (011) used a dynamic burst control plate to suppress stall on an airfoil by preventing the low-reynolds-number leading edge separation bubble from bursting. Suction flow controls were also widely used to suppress the flow separation of airfoils (Qin et al., 1998; Greenblatt et al., 006; Gbadebo et al., 008; Chng et al., 009) and circular cylinders (Seal and Smith 1999; Fransson et al., 004; Patil and Ng, 010) by tests or CF numerical simulations. In this paper, a steady suction flow control method is adopted to mitigate the vortex shedding of a circular cylinder. Further details and comprehensive study on this method to control the vortex shedding of a circular cylinder will be submitted as a journal paper in the near future.

II. Experimental Setup and PIV System The experimental study was conducted in a closed-circuit low-speed wind tunnel located in the Aerospace Engineering epartment of Iowa State University.

3 II. Experimental Setup and PIV System The experimental study was conducted in a closed-circuit low-speed wind tunnel located in the Aerospace Engineering epartment of Iowa State University. The tunnel has a test section with a ft ( cm) cross section and the walls of the test section are optically transparent. The test model used in the present study was a circular cylinder. The circular cylinder had a diameter d= m and a width equal to the test section width of the wind tunnel. Its roots were mounted on two sides of the wind tunnel wall. uring the experiments, the wind 4 tunnel speed was adjusted to 8.0 m/s, which corresponds to a Reynolds number Re = Figure 1 shows the schematic of the experimental setup used in the present study for the two-dimensional PIV measurements. The flow was seeded with 1~5 µm oil droplets. Illumination was provided by a double-pulsed Nd:YAG laser (New Wave Gemini PIV 10-15) adjusted on the second harmonic and emitting two pulses of 10 mj at the wavelength of 53 nm with a repetition rate of 8 Hz. The laser beam was shaped to a sheet by a set of mirrors, spherical and cylindrical lenses. The thickness of the laser sheet in the measurement region is about 1.0 mm. The high-resolution 1-bit ( pixels) CC camera (SensiCam, CookeCorp) was used to perform twodimensional PIV image recording. The CC camera and double-pulsed Nd:YAG lasers were connected to a workstation (host computer) via a igital elay Generator (Berkeley Nucleonics, Model 565), which controlled the timing of the laser illumination and image acquisition. Figure 1. Schematic of the experimental setup Figure shows the assemblage of the test model used in present study. The test model was employed in present test as shown in figure (a). For the test model, four different suction sections were employed with the spaces of L/8@L/4@L/8. The suction holes have the same diameter of 3 mm. Two suction holes were symmetrically distributed about the x axis at each suction section, i.e., the angle of suction holes: 90.0 (70.0) degree, as shown in figure (b) which illustrates the suction hole distribution. Omega FMA-613A was used in the testing to control the suction flow rates of the main suction tube which was separated into 8 branches: 8 suction tubes which were connected with 8 suction holes. The flow rate was corrected to a set of standard conditions which are 5ºC and one standard atmosphere pressure of kpa. Five suction flow rates of main suction tubes: 0, 40, 60, 80 and 95 L/min were adopted and the corresponding mean velocities of each suction hole were 5.89, 11.79, 17.68, 3.58, 8.00 m/s, respectively. To measure the pressure distribution on the cylinder model, a digital pressure measurement system based on the model SA317 pressure scanner (Scanivalve Corporation) was used in the wind tunnel tests. As shown in figure (b) and (c), total 6 pressure taps were arranged on the model in two sections, 3 for pressure section 1 and only 30 for pressure section because two pressure taps are replaced by two suction holes at each suction angle. Pressure section 1 was located on the middle of the test model and pressure section was shared with one suction section as shown in figure (b). The sampling rate was 300 Hz and the sampling time was 100 s for the pressure measurement of each tap. The blockage ratios for all section models were % and the measured pressures on the cylinder had 3

4 been corrected combining the solid blockage and the wake blockage (Barlow et al. 1999). All pressure taps are connected by polyvinyl chloride (PVC) tubes of 0.5 m long to the pressure measurement system. The effect of the phase distortion and magnitude variation alone on measured peak pressures was found to be small for short PVC tubes of length less than ft. (0.61 m) (Irwin et al. 1979). Wind speed direction Suction holes Suction holes 1 and 3 and 4 Suction holes Suction holes 5 and 6 7 and Pressure section 1 Pressure section Pressure tap Suction hole Pressure Pressure section 1 section (a) Test models Wind U 0 y θ i Cylinder section (c) Pressure section 1 Figure. Setup of the test model (b) Pressure taps and suction holes III. Experimental results and discussions Figure 3 shows the distributions of the mean pressure coefficients of the test model at two suction pressure sections, respectively. It can be seen that the separation point should be at least in front of 90.0 degrees and the suction flow control induces great change to the distributions of the mean pressure coefficients, especially in the flow separation region. As the flow rate increases, the absolutes of the mean pressure coefficients in the flow separation region decrease. It is indicated that the drag forces of the test model decrease under the suction flow control, even at a small flow rate of 0 L/min. But, the control effect will not be improved and tends to be steady when the suction flow rate exceeds 40 L/min. For pressure section 1, the position with the minimum mean pressure coefficient has no change under suction flow control with different flow rates. But, for pressure section, the point with the minimum mean pressure coefficient moves to next point when the flow rate is larger than 40 L/min. It can be deduced that the separation point may move backward or the flow separation is delayed at this suction section. It should be further investigated combined with the results of PIV measurement. C y θ i 1 θ i x C x 4

5 Figure 3. Mean wind pressure coefficient distribution of the test model Figure 4 shows the distributions of the fluctuating pressure coefficients of the test model at two suction pressure sections, respectively. Similar with the analysis of the mean pressure coefficients, the small flow rate of 0 L/min can obtain better control effect, but the control effect tends to be steady when the suction flow rate exceeds 40 L/min. The fluctuating pressure coefficients of most taps decrease to 1/3 compared with the case without the suction flow control. Figure 4. Fluctuating wind pressure coefficient distribution of the test model The coordinate system of the model and the directions of the aerodynamic coefficients are schematically illustrated in Fig. (c). The aerodynamic coefficients can be obtained by integrating the wind pressures over all the taps; the lift and drag coefficients, C and C, can be calculated with the following expression: C C C y x 1 1 ρu C θ cosθ 0 pi i i Fx i 1 x = = = pi i cos i 1 1 C θ θ i ρu 0 ρu ρu C θ sinθ 0 pi i i Fy i 1 y = = = pi i sin i 1 1 C θ θ i ρu 0 ρu 0 pi pi p = 1 ρu 0 Figures 5 and 6 show the time histories of lift and drag coefficients of the cylinder model, respectively. The fluctuations of lift coefficient histories are dramatically suppressed. If the cylinder model can freely vibrate, the oscillation amplitude will also be suppressed under a suction flow control. So the suction flow control should have the ability to suppress the flow-induced vibration according to the analysis of the lift coefficients. Figure 6 shows that not only the amplitudes of drag coefficients are greatly suppressed, but also the means of drag coefficients are obviously decreased. It is shown that the suction flow control has the function of the flow drag reduction. (1) 5

For further statistical analysis, the control effects of the root mean squares (RMSs) of lift coefficients and the means of the drag coefficients are obtained as shown in figure 7.

reduction. But, for the drag coefficient reduction, the best control effect is about 30% on pressure section, but only 5% on pressure section 1.

direction. It will be further studied in present paper by analyzing flow structures of different planes along the axial direction through PIV measurements.

6 Figure 5. Time histories of lift coefficients of the test model (left: section 1 and right: section); the gray lines denote the cases without suction control and the red lines denote those with suction control Figure 6. Time histories of drag coefficients of the test model (left: section 1 and right: section); the gray lines denote the cases without suction control and the red lines denote those with suction control For further statistical analysis, the control effects of the root mean squares (RMSs) of lift coefficients and the means of the drag coefficients are obtained as shown in figure 7. The control effect of the lift coefficients of pressure section is just slightly better than that of pressure section 1; and the lift coefficients at two sections can reach about 80% the amplitude reduction. But, for the drag coefficient reduction, the best control effect is about 30% on pressure section, but only 5% on pressure section 1. It is indicated that the control effect of the drag coefficients along the axial direction is different; the reason may be that the flow should have a three-dimensional feature along the axial direction. It will be further studied in present paper by analyzing flow structures of different planes along the axial direction through PIV measurements. (a) RMS of lift coefficients 6 (b) Mean of drag coefficients Figure 7. Comparison of control effects for aerodynamic coefficients at two pressure sections The PIV measurements were simultaneously performed when the wind pressures on the cylinder were measured for the test model. Two measured PIV planes are located on pressure section 1 and. Figure 8 shows the instantaneous vorticity field on section for the test model. Under the suction flow control, the alternating vortex shedding was changed into a symmetric mode. Figure 9 shows the mean velocity fields on section for the test model. The results indicate the velocities in the wake have the same direction with the main flow. The maximums of T. K. E. under the flow rates of 40 and 95/min are smaller than that under the flow rate of 0 L/min.

7 (a) 0 L/min (b) 0 L/min (c) 40 L/min (d) 95 L/min Figure 8. Instantaneous vorticity field of the test model for section (a) 0 L/min (a) 0 L/min (b) 40 L/min 7 (c) 95 L/min Figure 9. Mean velocity field of the test model for section

Figure 10. Velocity profiles in the wake of the test model For further investigating the suction flow control for the wake velocities, the velocity profiles along the y axis behind 1.

When the cylinder model is without the suction flow control, some velocities are negative values which indicate these velocity directions are opposite to the main flow direction.

8 Figure 10. Velocity profiles in the wake of the test model For further investigating the suction flow control for the wake velocities, the velocity profiles along the y axis behind 1.5 from the center of the cylinder models have been extracted as shown in figure 10. When the cylinder model is without the suction flow control, some velocities are negative values which indicate these velocity directions are opposite to the main flow direction. Under a small flow rate of 0 L/min, the negative velocities are changed into the positive ones; but as the flow rates continuously increase, the control effects have no distinct improvement. (a) Section 1 (b) Section 11 (c) Section 1 (d) Section Figure 11. Mean velocity field of the test model under the flow rate of 95 L/min To investigate the effects of the suction flow control for the wake along the axial direction, two measured PIV planes are not enough, so another two planes are added between section 1 and section, named section11 and section1 along the direction of section 1 to section. The PIV measurements are performed on these four planes at the flow rates of 0, 0, 40 and 95 L/min for the test model. 8

9 Figure 11 shows the mean velocity fields on four sections for the test model. The results indicate that the values of the T. K. E. are greatly reduced and the turbulence intensities of the wake are decreased under the suction flow control on which section. But only for the section, the velocities in the wake have the same direction with the main flow and the velocities close to the centreline in the near wake of other three sections have the opposite direction with the main flow. Furthermore, the separation point of section has been moved backward compared with those of other three sections. It can explain that the point with the minimum mean pressure coefficient moves to next point as shown in figure 3(a). So, combining the results of the distribution of the mean pressure coefficient, it can be deduced that the separation point may also move backward along the flow direction, i.e. the flow separation is delayed on the section. The velocity profiles behind 1.5 from the center of four sections of the test model have been extracted as shown in figure 1. Same with the analysis of figure 11, only for section, there are no negative velocities in the wake. Under the suction flow control, the absolutes of negative velocities increase with increasing flow rate for other three sections; besides the peak of negative velocities is less than that without suction flow control for section 1. Comparing the velocity profiles, the control effects are gradually improved along the direction of section 1 to section. (a) Section 1 (b) Section 11 (c) Section 1 (d) Section Figure 1. Velocity profiles in the wake of the test model IV. Conclusions In this study, the suction flow control for vortex shedding of a circular cylinder based on a multi-point suction type was investigated combining a PIV measurement system and a pressure measurement through a wind tunnel test. The following conclusions were obtained from this study: 1. The angle of suction hole of 90.0 degrees which is close to the flow separation point can achieve good control effects: reducing the average drag and the lift fluctuation of the test model, changing the alternating vortex shedding into a symmetric mode and delaying the flow separation on the suction section. 9

10 . As a multi-point suction type, the control effects of the average drag and the lift fluctuation reduction are improved from the section without suction to the suction section. The flow separation delaying is only achieved on the suction section. Acknowledgments This research was funded by the National Natural Sciences Foundation of China (NSFC) ( and ). The authors also want to thank Mr. Bill Richard and Mr. Kai Zhang of Iowa State University for his help in manufacturing the test models. References Barlow. J. B., Rae, W. H., and Pope. A., Low-speed wind tunnel testing. 3rd edition. New York: John Wiley & Sons, Inc., 1999, Chaps. 9. Chng, T. L., Rachman, A., Tsai, H. M., and Zha, G. C., Flow control of an airfoil via injection and suction,. Journal of Aircraft, Vol. 46, No. 1, 009, pp Fransson, J.H.M., and Alfredsson, P.H., On the disturbance growth in an asymptotic suction boundary layer, Journal of Fluid Mechanics, Vol. 48, 13 May, 003, pp Fransson, J. H. M., Konieczny, P., and Alfredsson, P. H., Flow around a porous cylinder subject to continuous suction or blowing, Journal of Fluids and Structures, Vol. 19, No. 8, 004, pp Gbadebo, S. A., Cumpsty, N. A., and Hynes, T. P., Control of three-dimensional separations in axial compressors by tailored boundary layer suction, Journal of Turbomachinery, Vol. 130, No. 1, , 008, pp Greenblatt,., Paschal, K. B., Yao, C. S., Harris, J., Schaeffler, N. W., and Washburn, A. E., Experimental investigation of separation control Part 1: Baseline and steady suction, AIAA Journal, Vol. 44, No. 1, 006, pp Hikami, Y., and Shiraishi, N., Rain-wind induced vibrations of cables in cable stayed bridges, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 9, No. 1-3, pp Irwin, H. P. A. H., Cooper, K. R., and Girard, R., Correction of distortion effects caused by tubing systems in measurements of fluctuating pressures, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 5, No. 1-, 1979, pp Main, J. A. and Jones, N. P., Full-scale measurements of stay cable vibration, 10th International Conference on Wind Engineering (10ICWE). Wind Engineering into the 1st century, Vol. 1 3, 1999, pp Matsumoto, M. Shirato, H. Yagi, T. Goto, M. Sakai, S. and Ohya, J., Field observation of the full-scale wind-induced cable vibration, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 91, No. 1-, 003, pp Patil, S. K. R., and Ng, T. T., Control of separation using spanwise periodic porosity, AIAA Journal, Vol. 48, No. 1, 010, pp Qin, N., Zhu, Y., and Poll,. I. A., Surface suction on aerofoil aerodynamic characteristics at transonic speeds, Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, Vol. 1, No. 5, 1998, pp Seal, C. V., and Smith, C. R., The control of turbulent end-wall boundary layers using surface suction, Experiments in Fluids, Vol. 7, No. 6, 1999, pp Zuo,., Jones, N. P., and Main, J. A., Field observation of vortex- and rain-wind-induced stay-cable vibrations in a three-dimensional environment, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 96, No. 6-7, 008, pp Zuo,., Jones, and N. P., Interpretation of field observations of wind- and rain-wind-induced stay cable vibrations, Journal of Wind Engineering and Industrial Aerodynamics, Vol. 98, No., 010, pp

Proceedings of the 4th Joint US-European Fluids Engineering Division Summer Meeting ASME-FEDSM2014 August 3-7, 2014, Chicago, Illinois, USA

Proceedings of the 4th Joint US-European Fluids Engineering Division Summer Meeting ASME-FEDSM4 August 3-7, 4, Chicago, Illinois, USA FEDSM4-38 SUPPRESSION OF UNSTEADY VORTEX SHEDDING FROM A CIRCULAR CYLINDER