-
1 Copyright © 2014 by ASME
Proceedings of the 4th Joint US-European Fluids Engineering
Division Summer Meeting ASME-FEDSM2014
August 3-7, 2014, Chicago, Illinois, USA
FEDSM2014-21318
SUPPRESSION OF UNSTEADY VORTEX SHEDDING FROM A CIRCULAR CYLINDER
BY USING A PASSIVE JET FLOW CONTROL METHOD
Wenli Chen and Hui Li
School of Civil Engr., Harbin Institute of Technology Harbin,
Heilongjiang, 150090 P.R. China
Email: [email protected]
Hui Hu Department of Aerospace Engineering
Iowa State University, Ames, Iowa 50011-2217, USA Email:
[email protected]
ABSTRACT A passive jet flow control method was employed to
suppress the unsteady vortex shedding from a circular cylinder
at the Reynolds number level of Re= (0.18~1.1)×105. The passive jet
flow control was achieved by blowing jets from the holes near the
rear stagnation point of the cylinder, which are connected to the
in-take holes located near the front stagnation point through
channels embedded inside the cylinder. Since a part of the oncoming
flow would inhale into the in-take holes, flow through the embedded
channels, and blow out from the holes near the rear stagnation
point to suppress/manipulate the alternating vortex shedding in the
wake flow behind the circular cylinder, the present passive jet
flow control method does not require any additional energy inputs
for the flow control. In the present study, the aerodynamic force
(i.e., both lift and drag) acting the circular cylinder model with
and without the passive jet flow control were compared
quantitatively at different Reynolds numbers (i.e., different inlet
mean speed). It was found that, in addition to almost eliminating
the fluctuations of the lift forces acting on the cylinder, the
passive jet flow control method was also found to reduce the mean
drag acting on the cylinder model greatly. The instantaneous
vorticity distributions and corresponding streamline patterns were
used to reveal the underlying physics about why and how the passive
jet flow control method can be used to suppress the alternating
vortex shedding and induce a symmetrical wake pattern behind the
cylinder model. Keywords: Passive jet flow control; suppression of
alternating vortex shedding; vortex-induced-vibration; wake flows
behind circular cylinders.
INTRODUCTION
Vortex-induced-vibrations (VIVs) of circular cables are closely
related to the vortex shedding in the wakes behind the bluff bodies
(i.e., circular cylinders); a number of studies have been performed
in recent years to manipulate the vortex
shedding process in order to reduce the VIVs of the bluff
bodies. Several passive flow control methods, which depend on the
modifications of the geometry shapes of the bluff bodies, have been
suggested to manipulate the vortex shedding process behind the
bluff bodies. For examples, Roshko (1955, 1961) suggested to mount
a splitter plate behind a circular cylinder to change the
alternative shedding mode of the vortex structures in the wake flow
into a symmetric mode. Owen et al. (2001) attached hemispherical
bumps to circular cylinders to control the fluctuating amplitudes
of the vortex-induced vibrations of cable models.
Glezer and Amitay (2002) investigated the interactions of
synthetic jets with the cross flows and reported that synthetic
jets could delay flow separation and reduce the drag acting on
bluff bodies. While synthetic jets were usually placed near the
separation points for flow control, Feng et al. (2010, 2011) and
Feng and Wang (2010, 2012) found that synthetic jets located at the
rear stagnation points of circular cylinders would also be used to
suppress flow separations and reduce the resultant drags acting the
circular cylinders.
Active jet flow control need input external energy to induce the
jet flow; however, the present passive jet flow control needn’t
consume the external energy and realize as follows: the partial
oncoming flow blew into the inlet holes, went through the ring
channels and blew from the jet holes into the near wake as a result
of breaking/suppressing the alternating vortex shedding in the
wake.
MESH GENERATION FOR CFD SUMLATION The diameter D of the circular
cylinder is 100 mm. The
pipe height of the ring channel around the cylinder is 5 mm; as
a result, the diameter of the ring channel is 110 mm. Five inlet
holes are employed near the front stagnation point and five outlet
holes are employed near the back stagnation. The inlet and outlet
holes’ location is shown in Fig. 1 and the interval angle between
two neighboring holes equals 7.5 deg.
-
2 Copyright © 2014 by ASME
The size of each inlet or outlet hole is the same rectangular to
be 8.0mm×6.9mm which are the dimensions along the axial and
circumference direction of the cylinder, respectively.
Fig. 1. Mesh generation for the CFD simulation of the
flow around a circular cylinder Figure 1 shows the computation
domain for the CFD
numerical simulation of the flow field around a circular
cylinder. The grid partition was achieved in the pre-processor code
ICEM-CFD 13.0, and the structural mesh was chosen in the
discretization process. The number of grid cells of the cuboid
computational region was roughly 250,000. The first grid normal to
the cylinder surface was set as 0.6 mm and 200 grids were
equidistantly arranged on the cylinder surface along the
circumferential direction. The boundary conditions are defined as
follows: left side is the “velocity inlet”; right side is “outflow”
which is fully developed outlet boundary; front, back, upper and
lower boundaries are “symmetry”; the circular cylinder surface is
“wall”; and the surface of the ring channel is “wall”. For the
baseline cases (i.e., no control cases), the inlet and outlet holes
are set as “wall”; for the control cases, the inlet and outlet
holes are set as “opening” which means the flow can enter into or
come out from the ring channel.
The shear stress transport (SST) k turbulent model (Menter,
1993) based on the RANS method was employed to simulate the
behavior of turbulent flow in the CFD numerical simulation. The
time interval was set as 0.0025 s, the mean wind speed of the inlet
was uniform and changed at 2.5, 5.0, 10.0 and 15.0 m/s with a low
turbulence intensity of 1%, the corresponding Reynolds number is
about (0.18-1.1) ×105. The
simulation Reynolds number range frequently occurs in the actual
engineering.
RESULTS AND DISCUSSIONS To verify the above-mentioned numerical
method and
meantime provide baseline cases for the passive jet flow
control, the flow past an uncontrolled circular cylinder was first
simulated and the aerodynamic coefficients (i.e., the lift and drag
coefficients) as shown in Fig. 2. Meantime, the corresponding
frequency spectra of the time histories of the lift and drag
coefficients were calculated as shown in Fig. 3. The mean drag
coefficient and the Strouhal number were compared with well-known
benchmark results by previous researchers as shown in Fig. 4. The
mean drag coefficient is consistent with the previous results. The
Strouhal number is little larger than the previous experimental
results, however, is close to the simulation results (Franke,
1990). Therefore, the feasibility and reliability of proposed
numerical method and grid can be accepted by the baseline
simulation examples.
The numerical simulations of the passive jet flow control were
then performed and the controlled aerodynamic coefficients were
shown in Fig. 2. Compared with the baseline cases, the fluctuations
of the lift and drag coefficients are dramatically suppressed and
the mean drag coefficients are simultaneously decreased. As shown
in Fig. 3, Comparing with the baseline cases, there is no dominant
frequency component in the frequency spectra of the lift and drag
coefficients for the controlled cases. The present aerodynamic
forces of the controlled cases (i.e., lift or drag) will not bring
the flow-induced vibration whatever in the cross-flow direction or
the in-line direction.
Based on the time sequences of the unsteady aerodynamic force
measurements as shown in Fig. 2, the standard deviations of the
instantaneous drag and lift forces acting on the cylinder model can
be calculated, which can be used to assess the effectiveness of the
passive jet flow control in reducing the unsteadiness of the wind
loads acting on the cylinder model for potential VIV suppression
more clearly and quantitatively. In the present study, a
non-dimensional parameter, called force fluctuation reduction
factor, jet-controlf , which is defined as the ratio of the
standard deviation values of the dynamic aerodynamic forces (either
drag or lift) acting on the cylinder for the cases with passive jet
flow control, with-control, to those of the baseline cases without
control, without-control, i.e.,
jet-controlf =with-control/without-control, is used as a
quantitative indicator to assess the effectiveness of the passive
jet flow control under different simulation conditions (i.e.,
different Reynolds number).
-
3 Copyright © 2014 by ASME
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6-1
-0.5
0
0.5
1
1.5
2
Time (s)
Aer
odyn
amic
coe
ffic
ient
s
Cl-Baseline Cd-Baseline Cl-Control Cd-Control
0 0.5 1 1.5 2 2.5 3 3.5 4-1
-0.5
0
0.5
1
1.5
2
Time (s)
Aer
odyn
amic
coe
ffic
ient
s
Cl-Baseline Cd-Baseline Cl-Control Cd-Control
(a) Re=0.18×105 (b) Re=0.36×105
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2-1
-0.5
0
0.5
1
1.5
2
Time (s)
Aer
odyn
amic
coe
ffic
ient
s
Cl-Baseline Cd-Baseline Cl-Control Cd-Control
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2-1
-0.5
0
0.5
1
1.5
2
Time (s)
Aer
odyn
amic
coe
ffic
ient
s
Cl-Baseline Cd-Baseline Cl-Control Cd-Control
(c) Re=0.71×105 (d) Re=1.1×105
Fig. 2. Comparison of aerodynamic coefficients between the
baseline and passive control cases
0 50 1000
0.05
0.1
0.15
Frequency (Hz)
PSD
Cl-BaselineCl-Control
0 50 1000
0.5
1
1.5
2x 10-3
Frequency (Hz)
PSD
Cd-BaselineCd-Control
11.13 Hz5.56 Hz
0 50 1000
0.02
0.04
0.06
0.08
0.1
Frequency (Hz)
PSD
Cl-BaselineCl-Control
0 50 1000
0.2
0.4
0.6
0.8
1x 10-3
Frequency (Hz)
PSD
Cd-BaselineCd-Control
21.29 Hz10.74 Hz
(a) Re=0.18×105 (b) Re=0.36×105
0 100 2000
0.005
0.01
0.015
0.02
0.025
0.03
Frequency (Hz)
PSD
Cl-BaselineCl-Control
0 100 2000
1
2x 10-4
Frequency (Hz)
PSD
Cd-BaselineCd-Control
42.58 Hz
21.48 Hz
0 100 2000
0.005
0.01
0.015
0.02
0.025
0.03
Frequency (Hz)
PSD
Cl-BaselineCl-Control
0 100 2000
1
2x 10-4
Frequency (Hz)
PSD
Cd-BaselineCd-Control
31.25 Hz
62.50 Hz
(c) Re=0.71×105 (d) Re=1.1×105
Fig. 3. Comparison of PSD of aerodynamic coefficients between
the baseline and passive control cases
-
4 Copyright © 2014 by ASME
102 103 104 105
0.14
0.16
0.18
0.20
0.22
0.24
0.26
Experiment, Kondo(1993)
Simulation, Cao(2001) Simualtion, Franke(1990) Present
simulation
St
Renolds number
Experiment, Franke(1990)
Fig. 4. Comparison of the present mean drag coefficient and St
with the previous results versus Reynolds number
It should be noted that the force fluctuation reduction
factor, jet-controlf , would have a value of 1.0 for the
baseline cases (i.e., the cases without control). When the value
of
jet-controlf is smaller than 1.0, it indicates that the
fluctuating amplitudes of the dynamic wind loads acting on the
cylinder can be reduced by using the passive jet flow control
method, thereby, it would offer a great potential for VIV
suppression. Smaller the jet-controlf value is; better performance
the passive jet flow control method will have for VIV suppression.
Fig. 5 gives the jet-controlf values based on the simulated lift
coefficients of the cylinder with and without the passive jet flow
control. The control effectiveness of the
jet-controlf values are 3.28%, 4.24%, 0.04% and 0.01%
corresponding to the Reynolds number of Re=0.18×105, Re=0.36×105,
Re=0.71×105 and Re=1.1×105.
0 2 4 6 8 10 12x 104
0
0.2
0.4
0.6
0.8
1
Renolds number
Lift
fluct
uatio
n re
duct
ion
fact
or
Fig. 5. The force fluctuation reduction factor vs. the
Reynolds number
0 2 4 6 8 10 12x 104
0
0.2
0.4
0.6
0.8
1
Renolds number
Mea
n dr
ag re
duct
ion
fact
or
Fig. 6. The mean drag reduction factor vs. Reynolds number
Figure 6 shows the measured mean drag reduction factor, which is
defined as the ratio of the mean drag forces acting on the cylinder
for the cases with the passive jet flow control to that of the
baseline case without control (i.e., CD_with_passive jet_control/
CD_without_control), as a function of the Reynolds number. The mean
drag coefficients of the cylinder were found to decrease rapidly
under the passive jet flow control. The drag coefficients of the
cylinder were found to become 68.74%, 72.05%, 73.60%, and 75.62% of
those of the corresponding baseline cases.
From the simulation results of the force fluctuation reduction
factor and the mean drag reduction factor, the passive jet flow
control exhibits excellent control effectiveness, meantime, the
control effectiveness is stable and no further significant
improvements were observed when the Reynolds number changed.
The simulated aerodynamic coefficient results given above reveal
many interesting features and global characteristics about the
passive jet flow control method in reducing the unsteadiness of the
dynamic wind loads acting on the simulation model. Moreover, the
quantitative information about the corresponding flow fields around
the cylinder with and without passive jet flow control is highly
desirable to elucidate the underlying physics to gain further
insight about the fundamental mechanism of the passive jet flow
control method with isolated jet holes. The vorticity and
streamline results were utilized in the present study to quantify
the characteristics of the wake flow behind the cylinder.
Figure 7 shows the comparison of the instantaneous vorticity
contour with and without the passive jet flow control. The left
column is the baseline results and the right one is the controlled
results. As the Reynolds number was located in the subcritical
region, the wake patterns are “2S” mode for four simulated cases.
Under the passive jet flow control, the wake becomes the completely
symmetrical pattern for the cases of Re=0.71×105 and Re=1.1×105.
Comparing with the results of the force fluctuation reduction
factor shown in Fig. 5, the lift fluctuation was suppressed to be
nearly zero. For the cases of Re=0.18×105 and Re=0.36×105, the wake
doesn’t completely becomes the symmetrical mode, i.e., an
approximately symmetrical pattern; however, the lift fluctuation
was suppressed to be about 5% of those of the corresponding
-
5 Copyright © 2014 by ASME
baseline cases. Whatever the case, the instantaneous vorticity
contours indicate that the vorticity near the outlet holes are the
opposite directions from the main vorticity distribution. It means
that the outlet holes can generate small scale of reverse vortices
again the main wake vortices. The symmetrically generated small
scale of reverse vortices can destroy/suppress the alternating
vortex shedding and induce a symmetrical wake pattern as shown in
right column of Fig. 7.
In the wall turbulence or shear flow, the strong shear layers
existing will cover the virtual turbulence eddies and the shearing
motions will make the reliable calculation of the virtual vortices
difficultly as shown in right column of Fig. 7. Figure 7 shows that
whole shear layer in the far wake are computed as the vorticity
under a passive flow control. To overcome this shortcoming, we
prefer to use the imaginary part of the complex eigenvalue of the
velocity gradient tensor to distinctly visualize vortices (Zhou et
al., 1999). Swirling strength is closely related to vorticity but
it can discriminate between vorticity due to shear motion and
vorticity resulting
from rotation as shown in Fig. 8. Figure 8 indicates that the
wake vortices are almost eliminated in the far wake for the cases
of Re=0.71×105 and Re=1.1×105 and the vortex shedding is greatly
weakened in the wake for the cases of Re=0.18×105 and
Re=0.36×105.
The streamline distributions in the flow field were then studied
as shown in Fig. 9 and Fig. 10; the left column is the baseline
results and the right one is the controlled results. The controlled
results indicate that the partial fluid (i.e., the air) flows into
the inlet holes, goes through the ring channel and then flows out
from the outlet holes. The flow from the outlet holes stops the
streamline alternate in the vertical direction and separates the
wake into the upper and lower parts. Under the passive jet flow
control, the recirculation region was suppressed and most of the
fluid moves along the main flow direction. Similarly with the
results of Fig. 7, the streamline distributions are completely
symmetrical for the cases of Re=0.71×105 and Re=1.1×105 and
approximately symmetrical for the cases of Re=0.71×105 and
Re=1.1×105.
(a) Re=0.18×105
(b) Re=0.36×105
(c) Re=0.71×105
(d) Re=1.1×105
Fig. 7. Comparison of the instantaneous vorticity contour with
and without the passive jet flow control, left: the baseline
results, right: the controlled results
-
6 Copyright © 2014 by ASME
(a) Re=0.18×105
(b) Re=0.36×105
(c) Re=0.71×105
(d) Re=1.1×105
Fig.8. Comparison of swirling strength contour with and without
control (left: the baseline ; right: controlled results)
(a) Re=0.18×105
(b) Re=0.36×105
(c) Re=0.71×105
(d) Re=1.1×105
Fig. 9. Comparison of streamlines around the cylinder with and
without control (left: baseline; right: controlled results)
-
7 Copyright © 2014 by ASME
(a) Re=0.18×105
(b) Re=0.36×105
(c) Re=0.71×105
(d) Re=1.1×105
Fig. 10. Comparison of the instantaneous streamline distribution
in the near wake with and without the passive jet flow control,
left: the baseline results, right: the controlled results
Inlet holes
Outlet holes
Inlet holes
Outlet holes
Inlet holes
Outlet holes
Inlet holes
Outlet holes
-
8 Copyright © 2014 by ASME
For further investigation of the control effectiveness of the
passive jet flow control method with the Reynolds number, following
the work of Chen et al. (2013), the suction/jet momentum
coefficient, Cµ, which is defined as the ratio of the suction/jet
momentum flux to the oncoming free-stream
momentum flux, i.e., 2
suc suc
0
U SC
U DL
, where sucS is
the area of the suction/jet holes (i.e., the total area of the
five suction/jet holes); 0L is the thickness of the circular
cylinder and is chosen as 0.2D in the simulation. The suction/jet
momentum coefficients C for each simulated case are shown in Fig.11
where _InC indicates the suction momentum coefficient of the inlet
holes and _OutC denotes the jet momentum coefficient of the outlet
holes. The result show when the Reynolds number increases, the
suction momentum coefficient increases and the jet moment
coefficient decrease. However, the change of suction/jet momentum
coefficient is very small with the Reynolds number.
2x104 4x104 6x104 8x104 1051.2x1050.005
0.010
0.015
0.020
0.025
0.030
Suct
ion/
jet m
omen
t coe
ffici
ent
Reynolds number
C_In C_Out
Fig. 11. The suction/jet momentum coefficients of the
passive
jet flow control vs. Reynolds number
Figure 12 indicates the velocity of inlet/outlet holes of the
passive jet flow control changing with the Reynolds number. The
result indicates that the improvement of the control effectiveness
is more related to the absolute velocities of the outlet holes.
2x104 4x104 6x104 8x104 1051.2x10501234567
Vel
ocity
of i
n/ou
tlet h
oles
(m/s
)
Reynolds number
C_In C_Out
Fig. 12. The velocity of inlet/outlet holes of the passive jet
flow
control vs. Reynolds number
CONCLUSIONS
This paper investigated the passive jet flow control method for
suppressing the unsteady vortex shedding from a circular cylinder
through numerical simulations. The following conclusions were
obtained from this study:
(1) The passive jet flow control method can dramatically
decrease the lift fluctuation and achieve the control effectiveness
exceeding 95%, which means this method can potentially suppress the
VIV of the cylinder; simultaneously, this flow control method can
decrease the mean drag coefficient to be less than 75% of those
corresponding baseline cases. The dominant frequencies in the
aerodynamic forces are eliminated under the passive jet flow
control.
(2) The passive jet flow control can generate small scale of
reverse vortices which are symmetrical. These symmetrically reverse
vortices can destroy/suppress the alternating vortex shedding and
induce a symmetrical wake pattern in the near-wall region. In the
far region, the wake vortices are greatly suppressed for smaller
Reynolds number cases or entirely eliminated for larger Reynolds
number cases.
(3) In the present Reynolds number region which locates in the
subcritical region, the wake pattern (either vorticity or
streamline distribution) will change from the approximate
symmetrical mode into the completely symmetrical mode with the
Reynolds number increasing. As a result, the lift fluctuations will
decrease to nearly zero.
(4) The control effectiveness of the passive jet flow control
method is improved with increasing the absolute velocities from the
outlet holes when the Reynolds number increases.
ACKNOWLEDGMENTS The research work was funded by National
Natural
Sciences Foundation of China (NSFC) (51378153, 51008093,
51161120359 and 91215302).
REFERENCES Cao, F. C., Xiang, H. F., 2001. Calculation of
unsteady flow
around circular cylinder and vortex-induced vibration. Journal
of Hydrodynamics 16(1), 111-118(in Chinese).
Chen, W. L., Xin, D. B., Xu, F., Li, H, Ou, J. P., Hu, H., 2013.
Suppression of vortex-induced vibration of a circular cylinder
using suction-based flow control, Journal of Fluids and Structures,
42(10), 25-39.
Feng, L. H., Wang, J. J., Pan, C., 2010. Effect of novel
synthetic jet on wake vortex shedding modes of a circular cylinder.
Journal of Fluids and Structures, 26, 900–917
Feng, L. H., Wang, J. J., 2010. Circular cylinder wake vortex
synchronization control with synthetic jet positioned at
-
9 Copyright © 2014 by ASME
back stagnation point. Journal of Fluid Mechanics, 662,
232–259
Feng, L. H., Wang, J. J., Pan, C., 2011. Proper orthogonal
decomposition analysis of vortex dynamics of a circular cylinder
under synthetic jet control. Physics of Fluid, 23, 014106-1–13
Feng L. H., and Wang J. J., 2012. Synthetic jet control of
separation in the flow over a circular cylinder. Experiments in
Fluids, 53, 467–480
Franke, R., Rodi, W., Schönung, B., 1990. Numerical calculation
of laminar vortex-shedding flow past cylinders. Journal of Wind
Engineering and Industrial Aerodynamics 35, 237-257.
Glezer, A., Amitay, M., 2002. Synthetic jets. Annual Review of
Fluid Mechanics, 34, 503–529.
Menter, F. R., 1993. Zonal two-equation k turbulence models for
aerodynamic flows. AIAA 24th Fluid Dynamics Conf., American
Institute of Aeronautics and stronautics,
Washington, DC. Munson, B. R., Young, D. F., Okiishi, T. H.,
Huebsch, W. W.,
2009. Fundamentals of fluid mechanics, 2nd edition. John Wiley
& Sons, New York.
Kondo, N., 1993. Direct third-order upwind finite element
simulation of high Reynolds number flows around a circular
cylinder. Journal of Wind Engineering and Industrial Aerodynamics
46, 349-356.
Roshko, A., 1955. On the wake and drag of bluff bodies. Journal
of the Aeronautical Sciences, 22, 124–132.
Roshko, A., 1961. Experiments on the flow past a circular
cylinder at very high Reynolds number. Journal of Fluid Mechanics,
10, 345–356
Zhou, J., Adrian, R. J., Balachandar, S., Kendall, T. M., 1999.
Mechanisms for generating coherent packets of hairpin vortices in
channel flow, Journal of Fluid Mechanics 387, 353–396
