Home Search Collections Journals About Contact us My IOPscience

Flow analysis of vortex generators on wing sections by stereoscopic particle image

velocimetry measurements

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2008 Environ. Res. Lett. 3 015006

(http://iopscience.iop.org/1748-9326/3/1/015006)

View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 14.139.209.19
The article was downloaded on 01/02/2013 at 13:43

Please note that terms and conditions apply.

IOP PUBLISHING ENVIRONMENTAL RESEARCH LETTERS
Environ. Res. Lett. 3 (2008) 015006 (11pp) doi:10.1088/1748-9326/3/1/015006

Flow analysis of vortex generators on wing

sections by stereoscopic particle image
velocimetry measurements
C M Velte, M O L Hansen and D Cavar
Technical University of Denmark, Department of Mechanical Engineering,
Fluid Mechanics Section, 2800 Kongens Lyngby, Denmark

E-mail: cve@mek.dtu.dk

Received 5 July 2007

Accepted for publication 7 January 2008
Published 24 January 2008
Online at stacks.iop.org/ERL/3/015006

Abstract
Stereoscopic particle image velocimetry measurements have been executed in a low speed wind
tunnel in spanwise planes in the flow past a row of vortex generators, mounted on a bump in a
fashion producing counter-rotating vortices. The measurement technique is a powerful tool
which provides all three velocity components in the entire measurement plane. The objective of
this study is to investigate the effect of vortex generators in a turbulent, separating, low
Reynolds number ( Re = 20 000) boundary layer over a geometry which is generating an
adverse pressure gradient similar to the flow past a wind turbine blade. The low Reynolds
number is chosen on the basis that this is a fundamental investigation of the structures of the
flow induced by vortex generators and the fact that one obtains a thicker boundary layer and
larger structures evoked by the actuating devices, which are easier to measure and resolve. The
flow behaves as expected, in the sense that the vortices transport high momentum fluid into the
boundary layer, making it thinner and more resistant to the adverse pressure gradient with
respect to separation. The amount of reversed flow is significantly reduced when vortex
generators are applied. The idea behind the experiments is that the results will be offered for
validation of modeling of the effect of vortex generators using various numerical codes. Initial
large eddy simulation (LES) computations have been performed that show the same qualitative
behaviour as in the experiments.

Keywords: vortex generators, separation control, stereoscopic PIV, LES

1. Introduction tip creates a longitudinal vortex that convects high momentum

fluid down into the boundary layer. The positive effect of
Vortex generators (VGs), as described in [1], have been used vortex generators on an airplane wing is to delay separation
for more than 50 years in applied aerodynamics on airplane to higher angles of attack, thus increasing the maximum lift
wings. Vortex generators belong to the category boundary coefficient so that an airplane can land at a lower speed.
layer manipulators. Their function is to reenergize an adverse However, this increased maximum lift is paid for by a so-
pressure gradient boundary layer that is about to separate by called drag penalty, where the drag for small angles of attack
transporting high momentum fluid from the outer part of the is increased. Vortex generators are also used to increase the
boundary layer down to the low momentum zone closer to the maximum speed of commercial jets flying at approximately
wall. A vortex generator is commonly a small triangular or Mach 0.8. Locally on the wing, the flow is accelerated and
rectangular plate that is mounted on a surface at an angle to can exceed the speed of sound, Mach 1, which can cause so-
the incoming flow; see figure 1. A vortex generator acts like called shock induced separation even for very low angles of
a small wing with a low aspect ratio, and the flow past the attack. This is easily felt by the pilot and determines the speed

1748-9326/08/015006+11$30.00 1 2008 IOP Publishing Ltd Printed in the UK

Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Figure 1. Triangularly shaped vortex generators.

Figure 2. (a) Counter-rotating and (b) co-rotating vortices.

limit on many subsonic passenger airplanes. However, vortex
generators can also be used to delay this type of separation
to allow the airplane to fly slightly faster. Since commercial rotor plane. This twisting of the relative wind speed could in
jets have no problems in increasing the lift coefficient at for principle be compensated for by twisting the blades, but there
example landing configuration by using flaps and slats, the is, for structural reasons, a limit to the extent to which this can
main reason that they often are equipped with vortex generators be done. Since the relative wind speed increases towards the
is shock induced separation. tip, the highest aerodynamic loads are also situated here, giving
For small airplanes and wind turbines, the flow is far from large bending moments at the root of the blades. To be able to
reaching Mach 1, and there is thus no problem with shock carry these loads, the blade must be very thick close to the root,
induced separation. However, sometimes a high lift coefficient and eventually it often ends with a cylindrical cross section.
is needed to decrease the landing speed for STOL (short take The last real airfoil (i.e., with cross section with a sharp trailing
off and landing) airplanes. A classical solution has been to use edge) often has a thickness to chord ratio of more than 30%.
so-called slots in front of the leading edge. The slot is a small, This is very bad from an aerodynamic point of view in the
stationary list in front of the leading edge of the main wing that sense that the flow may separate even for very small angles
for large angles of attack allows high momentum flow from the of attack. This, in combination with the fact that the angles of
pressure side to be accelerated in the slot between the list and attack for these inboard sections, for the reasons given above,
the main wing, creating a high momentum flow to the suction are large, almost guaranties a separated flow for all wind speeds
side of the wing and thus making the flow more resistant to even for pitch regulated machines. Since vortex generators
the adverse pressure gradient. Vortex generators have the same are known to delay and to some extent suppress separation,
effect on the flow in the sense that they also delay separation it is natural to apply vortex generators on the inner part of
and increase the maximum lift coefficient. Because they can wind turbine blades. Because vortex generators give some drag
be mounted/glued directly to the pressure side of a bare wing, penalty also for low angles of attack, they are normally not used
a less complicated wing can be built, and therefore they are on the outer part of wind turbine blades. The fact that vortex
often used on modern sport planes instead of slots. Vortex generators can have quite a dramatic effect on the performance
generators are sometimes used to prevent separation of the of wind turbines is shown in [2], where a comparison between
flow past the rudder or ailerons to improve the controlling of the measured power curve on the 1 MW experimental Avedre
the airplane, which could otherwise be drastically worsened. wind turbine is shown with and without vortex generators on
Vortex generators are used on wind turbine blades for two the inner part of the blades. The maximum power occurring
reasons. One is to decrease separation and increase lift at at a wind speed of around 17 m s1 for a pitch of 1 for
high angles of attack and the other is to attempt to increase the wind turbine running in stall regulation is approximately
the maximum ratio between the lift and the drag coefficients, 850 kW for the clean configuration and 1050 kW when using
which directly influences the aerodynamic efficiency of the vortex generators.
turbine. The relative wind seen by a wind turbine blade To have an effect over a section of a blade, vortex
is composed by the incoming wind speed and the rotational generators are mounted in arrays and often in pairs that can
velocity of the blade. There exists an optimum value for the tip produce either co- or counter-rotating vortices; see figure 2.
speed ratio, opt = R/Vo , where the wind turbine operates Also, the size, geometry, aspect ratio, angle of incidence
most efficiently. To run the turbine at this optimum value, to the flow, chordwise position in relation to the separation
it is necessary to increase the angular velocity of the rotor, line, configuration of co- or counter-rotating vortices etc are
, proportionally to an increasing wind speed, Vo . However, parameters that must be optimized in order to use vortex
there is for noise reasons a limit on the speed of the tip of generators in a most efficient manner.
approximately 70 m s1 , that gives a limit on the rotational In order to find these optimal parameters, one needs to do
speed of max = 70/R , where R is the rotor radius. The experiments. In [3] a review of many experiments and their
rotational speed of approximately 70 m s1 at the tip decreases conclusions made by Lin et al is given. In a recent paper by
linearly further inboard with the decreasing radius. Thus, the Godard and Stanislas [4], a parametric study of the effect of
flow at the tip will mainly be in the rotor plane, and near the hub vortex generators was conducted in which the devices were
the flow seen by the blade will almost be perpendicular to the placed on a bump designed to keep the boundary layer flow

2
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

a circular sector, extended in the spanwise direction, creating

a cylindrical sector with radius 390 mm. The bump height
is 30 mm and the chord length and bump width are 300 mm
and 600 mm respectively. The presence of this model in the
flow induces an adverse pressure gradient strong enough to
generate separation. The experiments were conducted at a free
stream velocity of U = 1 m s1 , corresponding to Reynolds
number Re = 20 000 based on the bump chord length. The
turbulent inflow is assured by a turbulence generating inlet grid
Figure 3. Sketch of vortex generator geometry. with mesh length M = 39 mm situated at the beginning of
the test section. In [6], the turbulence intensity at the inlet
has been estimated to 12% and the boundary layer thickness
at the vortex generator position has been estimated from laser
on the verge of separation. An optimum vortex generator
Doppler anemometry (LDA) measurements to be 25 mm.
geometry was sought using hot film sensor skin friction
The boundary layer is here defined as the part of the flow
measurements and thereafter characterized using stereoscopic
contained by vorticity derived from the presence of the wall.
particle image velocimetry (PIV). The main conclusions were
The principle of determining the boundary layer thickness is
that triangular vanes in a counter-rotating configuration give
to determine where the vorticity is reduced to a value much
the best effect on skin friction increment. The optimal device
smaller than the maximum value (z = 0.01 z max ). This is
configuration showed good concurrence with results found in here defined as the boundary where the viscous effects cease to
the literature for all parameters except for the device height, influence the flow. When calculating the vorticity, it is tedious
which appeared to behave differently from results found in to do measurements that give you the data to calculate both
the literature. The characterization study done on the most terms. In the potential flow free stream, the first term and the
optimal case showed counter-rotating vortices, stable in their second term must cancel out since there is no vorticity. In the
downstream development, transferring momentum from the boundary layer, however, vorticity is present, which means that
outer flow and in towards the near-wall region. In a recent the terms must be unequal. An order of magnitude analysis
study by Angele and Grewe [5], some experimental results gives that if U and x are of order 1, then V and y are of order
on wind tunnel experiments using stereoscopic PIV on the . Therefore the second term is quadratically larger than the
flow past vortex generators are discussed, focusing on both first term.  
the mean and instantaneous behaviour of the flow induced by 1 V U
z = . (1)
the devices. The test section consisted of a concentric inner x
2  y

cylinder ending in a cone made of wire gauze, through which 1
suction was applied, creating an adverse pressure gradient. A 1

center body was in turn placed downstream of this cone to Therefore one can argue that the second term should
generate a favorable pressure gradient. The vortex generators be sufficient in determining the vorticity. The boundary
succeeded in obstructing the separation created by the adverse layer thickness was therefore extracted from a velocity profile
pressure gradient. It was also found that the vortices were measured by LDA at the position of the trailing edges of
nonstationary, with larger movements in the spanwise direction the vortex generators with no actuators present. U/y was
than in the wall normal one. The spanwise movements were of obtained from the velocity profile and thereby also the estimate
the same extent as the size of the vortices. of the vorticity.
In an ongoing research project, the flow field behind
different vortex generators mounted on top of a cylindrical 2.2. Actuators
bump was investigated experimentally. The purpose is to Triangular vanes of the same height as the boundary layer
create a database that can be used for validating various thickness, 1 , were applied. represents the boundary layer
numerical codes to correctly calculate the physics contained thickness at the position of the trailing edges of the devices.
in the production of the vortices and their dynamic behaviour The vortex generators were positioned with their trailing edges
embedded in the viscous boundary layer. Some of the results at 50% bump chord. The resulting parameters from the
obtained so far are given in this paper. optimal vortex generator configuration found as a result of
the optimization study performed by Godard and Stanislas [4]
2. Method have been used and can be found in figure 3, where h is the
device height, l is the device length, s is the distance between
2.1. Wind tunnel setup the trailing edges of two vortex generators within one pair, z
is the distance between two vortex generator pairs and is
The measurements were carried out in a closed-circuit wind the device angle of incidence. This optimum geometry will,
tunnel with an 8:1 contraction ratio and a test section of cross of course, depend on the application and is therefore most
sectional area 300 mm 600 mm with length 2 m. The suction likely not universal, but has nevertheless also been used in
side of a wind turbine wing is represented by a bump mounted this experiment. The distance between the devices and the
vertically on one of the test section walls with the leading edge separation line showed a weak dependence on the wall shear
positioned 600 mm downstream of the inlet grid. The bump is stress and was therefore not considered in this study.

3
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Figure 4. Sketch of the wind tunnel test section, the positioning of measurement planes and definition of coordinate system.

2.3. Measurement planes

The measurements were conducted in spanwise planes at
various positions downstream of the vortex generators. A
sketch of the wind tunnel test section and the positioning
of measurement planes is shown in figure 4, where also the
coordinate system is defined. The equipment was mounted
on a rigid traverse, traversing in the axial and normal
directions. This configuration enables one to calibrate only
once and perform measurements accurately at the different
streamwise positions using the same calibrated configuration.
All measurement planes are parallel to each other, and their
normal component is parallel to the test section walls.

2.4. Setup of stereoscopic PIV (SPIV) equipment Figure 5. Sketch of the SPIV setup at the test section.
The experimental setup is illustrated by a sketch in figure 5.
A laser was placed at the top of the test section, illuminating
the measurement volume without significantly disturbing the
the vertically mounted bump from the side. Two cameras
flow. This kind of global seeding of the complete tunnel is
were placed on the same side of the light sheet, resulting in
possible since the tunnel is a closed-circuit wind tunnel. The
one camera being placed in the forward scattering direction
size of the seed particles has been measured by an APS TSI
and one in the backward scattering one. The angle of
3320 time-of-flight spectrometer. The droplets are produced by
each respective camera to the laser sheet was 45 . The f -
blowing pressurized air over a thin-walled pipe with diameter
numbers were set to between 8 and 16 for the camera in
of 0.5 mm. The other end of the pipe is submerged in glycerol.
the forward scattering direction and 4 or 5.6 for the camera
A calibration target was aligned with the laser sheet. This
in the backward scattering direction, depending on the light
target has a well defined pattern, which can be registered by the
budget of reflections from the particles and the bump and
two cameras to obtain the geometrical information required for
devices at each individual plane position. The stereoscopic PIV
reconstructing the velocity vectors received from each camera
equipment included a double cavity NewWave Solo 120XT
to obtain a full description of all three velocity components
Nd-YAG laser (wavelength 532 nm), capable of delivering in the plane. Calibration images were recorded with both
light pulses of 120 mJ. The pulse width, i.e., the duration of cameras at five well defined streamwise positions throughout
each illumination pulse, was 10 ns. The light sheet thickness at the laser sheet. A linear transform was applied to these images
the measurement position was 2 mm and was created using a for each camera respectively to perform the reconstruction.
combination of a spherical convex and a cylindrical concave This procedure was executed both previous to and after the
lens. The equipment also included two Dantec Dynamics measurements.
HiSense MkII cameras (1344 1024 pixels) equipped with
60 mm lenses and filters designed to only pass light with
2.5. Data processing
wavelengths close to that of the laser light. Both cameras were
mounted on Scheimpflug angle adjustable mountings. In order The images were processed using Dantec Flowmanager
to obtain a smaller measurement area, the cameras were, in software version 4.7. Adaptive correlation was applied using
some of the tests, equipped with teleconverters. The seeding, refinement with an interrogation area size of 32 32 pixels.
consisting of glycerol droplets with a diameter of 23 m, Local median validation was used in the immediate vicinity
was added to the flow downstream of the test section. The of each interrogation area to remove spurious vectors between
seed particles were added at this position in order to obtain each refinement step. The overlap between interrogation areas
a more homogeneous distribution of the tracers throughout was 50%. For each measurement position, 500 realizations

4
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Figure 6. The effect of sample size on accuracy of U , V , W and based on a single measurement point (z/ h, y/ h) = (1, 1) in N = 500
vector maps (() indicating average). Plotted points represent averages of the number of independent samples given in the figure. Additionally,
10 statistically independent averages based on 5, 10, 20, 30, 40 and 50 ensembles and 5 averages based on 100 ensembles are included in the
figures.

were acquired. The recording of image maps was done with of the velocity derivative estimate and half the uncertainty
an acquisition rate of 1.0 Hz. of for example the forward or backward difference. The
scheme cancels out the effect of oversampling from the 50%
2.6. Mean fields, spatial resolution and accuracy overlap between interrogation areas, in the estimation of the
velocity gradient, since neighboring data are not correlated (see
The velocity vector maps contain 73 by 63 vectors. The e.g. Raffel et al [7]).
interrogation areas have linear dimensions 1.15 mm in the y -
The effect of sample size on accuracy has been
direction and 1.73 mm in the z -direction. In the PhD thesis of
investigated in one single point ((z/ h, y/ h) = (1, 1), x =
Schmidt [6], the Taylor microscale was estimated from LDA
225 mm) for the first moment for the x -, y - and z -component
measurements of time series to be in the order of f 9 mm at
of the velocity. The same has been done for the first moment of
a position 330 mm upstream of the bump leading edge at mid-
the vorticity. The results are shown in figure 6. The maximum
channel. The estimate of the Kolmogorov length scale was in
number of realizations was N = 500. The mean based on
the order of 0.5 mm. LDA velocity profile measurements
this sample size is shown by a filled straight line. The dashed
had also been used to obtain the wall shear stress at the same
lines indicate the theoretical standard error estimates for the
downstream position. From this an estimate of the viscous
lengthscale is obtained, = (/w )1/2 = 0.2 mm. The first moment.

spatial resolution of the velocity vector fields is defined by s(U ) = . (2)
N
the size of the interrogation area, which also limits the spatial
resolution of the estimation of velocity gradients and hence Additionally, 10 statistically independent averages based
the vorticity. In order to resolve the smallest scales in the on 5, 10, 20, 30, 40 and 50 ensembles and 5 averages based on
flow, one needs to fulfill the Nyquist criterion. Therefore, 100 ensembles are included in the figures. The result does not
scales smaller that half of the size of the interrogation area vary significantly for other measurement points.
can not be resolved. Streamwise vorticity was calculated Since the PIV velocity measurement points are local
using central differencing, which gives second-order accuracy averages over the area of interrogation, one can never obtain

5
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Table 1. Control parameters for evaluation of correlation noise.

Control parameters
NI  1020 Particle image density
NI mean number of particle images in an interrogation area
FO  0.75 Loss of particles due to out-of-plane motion
FO fraction of particles present inside the interrogation
area in both images
M|U |t/d  0.5 Velocity gradients
M (pix m1 )image magnification
U (m s1 )difference in velocity across the interrogation area
t (s)time between acquired PIV images
d mean particle image diameter

measurements closer to the wall than half of the height of is of importance for accuracy issues, originating from the
the interrogation area size. And even then, because of the difficulties of determining the position of the particle image.
high velocity gradients in this region, one needs to be careful A larger dynamic range will minimize the effect of this
in interpreting the measurements at these points. Therefore, uncertainty. Therefore, the measurements need to be a tradeoff
the position of the wall was determined by extrapolating the between the laser sheet thickness and the time between laser
mean velocity profiles towards the wall and using the no-slip pulses.
condition. Another source of error, which is not commonly
Keane and Adrian [8] used an analytical model and Monte considered, is the ability of the particles to follow the flow.
Carlo simulations to investigate the effects of experimental This might be an issue, especially in regions of accelerating
parameters to optimize PIV performance. They showed that flow such as in a large scale rotation. An example of this in
double-pulsed systems were optimal when the interrogation the longitudinal vortices behind vortex generators. Due to the
area particle image density exceeded 1020. They also showed excess density of the particles in relation to the surrounding
that velocity gradients reduce the valid data rate and introduce medium, there is a centrifugal force acting on them, forcing
velocity bias. The primary source of error in two-component them to move outwards from the vortex center. This is
PIV measurements is correlation noise. Westerweel [9] and however discarded from having any significant effect in the
Foucaut et al [10] have shown that if the values of the control flow under consideration because of two reasons: the velocities
parameters in table 1 are satisfied, then the root mean square are very low and the distribution of particles in the PIV images
(RMS) of the measured particle image displacements corr is homogeneous. At higher velocities, a reduced particle
0.1 pixel. concentration in the vortex core is commonly observed (see
The dominating factor from table 1 is the effect of e.g. Stanislas et al [11]).
velocity gradients, which was only altered by changing the
image magnification and the time between pulses. If there 2.7. Reflection reduction
are large velocity gradients in the flow, variations of particle
displacements across the interrogation areas will lead to Reflections from vortex generators and bump surface entering
a broadening of the correlation peak. This will in turn the CCD cameras constituted a problem for two reasons.
lead to deterioration in the precision of the estimate of The reflections appeared in the most interesting part of the
the position of the correlation peak. If one lets the time measurement region, corrupting the signal in this area. The
separation between acquired images be relatively large, this high power of the laser light also creates reflections harmful
precision error will be small in relation to the displacement. for the CCD chip. If these high power reflections reach
However, if velocity gradients are present, this strategy will the CCD chip, one is risking damage to the cameras in the
also lead to a broadening of the correlation peak, since the form of dead pixel elements. Light scattering from objects
variation in displacement will increase as the time separation other than the glycerine particles was removed by painting
increases. The choice of experimental design will therefore by the reflecting surfaces with a mixture of Rhodamine 6G and
necessity have to be a tradeoff between precision and relative varnish. Rhodamine 6G is a fluorescent dye, absorbing light
displacement error. If the velocity gradients vary across the with the wavelength of the laser and reflecting light which
measurement plane, one cannot design the experiment so that it has a wavelength slightly shifted from the absorbed one. The
is optimized over the entire measurement volume. The number cameras were equipped with green-pass filters, which only
of spurious vectors was always less than 5%, which shows that permit the wavelengths of the laser to pass, allowing the
the effect of large velocity gradients in the flow on accuracy scattering from the particles to pass through and preventing
was not unwieldy. strong reflections from surfaces to reach the CCD chip.
Since the main flow component is perpendicular to the Rhodamine 6G has its absorption peak at around 530 nm
measurement plane, the finite thickness of the laser sheet and its emission peak at about 552 nm. This method was
will be critical in limiting the dynamic velocity range, successful in reducing most of the unwanted reflections. An
i.e., the ratio of the maximum velocity to the minimum additional approach used in some of the measurement positions
resolvable velocity. Maximizing the dynamic velocity range was to take pictures with no particles present in the flow at

6
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

the same position as that at which the measurements were spanwise directions respectively (300 mm 600 mm in the
taken. These pictures were subtracted from the images used for experiment) and a total length in the streamwise direction of
measurements prior to processing, removing scattering from all 1.2 m, where 300 mm upstream of the bump leading edge and
objects other than the particles. 600 mm downstream of the bump trailing edge are included
Before applying Rhodamine 6G, acquiring images in the domain. In the spanwise direction, periodic boundary
including the wall and receiving a good quality measurement conditions (BCs) are applied, while in the wall normal
signal was not possible since, even at relatively low laser direction the solid wall (bottom bump surface) and symmetry
intensity, the reflections from the wall were too strong, risking boundary conditions (upper domain surface) are applied. Note
damage to the CCD chip. The effect of applying Rhodamine that the upper wall boundary layer is not included in the
6G to the bump surface made measurements with maximum computations. Additionally the inlet and (convective) outlet
laser power close to the surface possible for all measurement BCs are applied in the streamwise direction. Vortex generators
positions. The largest reflection reducing effect is, however, are modeled as baffles i.e., surfaces without thickness with
obtained by utilizing the reflection subtraction technique, wall BCs applied to them. The choices of boundary conditions
removing almost all of the reflections from the wall and in and domain dimensions are largely connected to the inlet
particular the reflections from the vortex generators as close inflow database used in the computations. The inlet database
up as 6 mm from the device trailing edges. Previous to representing the instantaneous snapshots of a fully developed
applying this technique, reflections from the vortex generators turbulent boundary layer flow was generated on a precursor
obstructing the signal were detected as far as up to 100 mm basis using the method of [13]. In the recent work of [14],
downstream of the devices. the LDA measurements of the fully developed boundary layer
flow past a 2D bump without vortex generators mounted on
2.8. Large eddy simulation (LES) computations it are reasonably well reproduced by LES. Using the same
inlet database and very similar geometrical and numerical
The eddyviscosity based LES calculations were performed
computational layout as in [14], the present LES computation
utilizing the flow solver FASTEST-3D (flow analysis solving
is conducted.
transport equations simulating turbulence), [12]. The
The subgrid-scale (SGS) stresses are modeled using
code is based on a finite-volume numerical method for
solving both three-dimensional filtered and Reynolds-averaged the eddyviscosity assumption employing the Smagorinsky
NavierStokes equations on block-structured, body-fitted, model [15]. The grid consists of 584 162 224 (21.3
non-orthogonal meshes. Block interfaces are treated in a million) cells in the streamwise, wall normal and spanwise
conservative manner, consistent with the treatment of inner cell directions respectively. Special consideration has been made
faces. A cell-centered (collocated) variable arrangement and to obtain the proper mesh resolution in the area of the vortex
Cartesian vector and tensor components are used. The well- generators. Approximately 3.8 million cells are located in
known SIMPLE algorithm is applied for coupling the velocity the immediate vicinity of the vortex generators, potentially
and pressure fields. The convective and diffusive transport facilitating the computations with the possibility to properly
of all variables is discretized by a second-order central resolve various turbulent and vortical structures in this area.
differencing scheme, whose stability is enhanced through the
so-called deferred correction approach. Time discretization is 3. Results and discussion
accomplished by applying the second-order implicit Crank
Nicolson method. FASTEST-3D is parallelized based on 3.1. Downstream development of vortices
domain decomposition in space using the MPI message passing
library. In order to be able to quantify the induced effect of the
The computational setup regarded in LES calculations is vortex generators on the flow, both the controlled and the
somewhat different from the experimental one. The same uncontrolled flow has been measured for all positions. In
geometrical bump is considered, with vortex generators placed figure 7, for four positions downstream of the streamwise
at the same positions as in the experiment; see figure 3. placement of the vortex generators, the arithmetic averages of
As the FASTEST code utilizes block-structured meshes, a the velocity vector maps obtained from SPIV measurements of
trapezoidally shaped vortex generator pair is used in the the uncontrolled (left column) and controlled (middle column)
numerical setup in order to avoid grid discontinuity at the flows are displayed. The estimator of the mean is based on 500
leading edge of the vortex generator vanes. Therefore, the recorded independent velocity field realizations. The rightmost
leading edge of the vortex generator pair has a height of column shows the projected longitudinal vorticity of each
1/5 , while the trailing edge height corresponds to the value corresponding mean velocity vector map for the controlled
applied in the experiments (1 ). It should also be noted case. In the left and middle columns, the secondary velocities
that the inlet conditions in the conducted LES computations are illustrated as arrows, whereas the primary velocities are
do not match the experimental counterpart, as precomputed plotted as colour contours. For the sake of clarity, only every
snapshots of the spatially developing fully turbulent boundary fourth vector has been displayed in these plots, but the data
layer flow are used as an inlet database for the computations, in the contour plots have not been reduced. The positions of
and only one device pair is considered in the computations. the planes are given in absolute coordinates, referring to the
The computational domain considered consists of a cross- coordinate system given in figure 4, as well as in coordinates
sectional area of 240 mm 288 mm in the wall normal and relative to the vortex generators. The y -coordinate is always

7
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Figure 7. Plots of averaged velocity fields. The left column displays the uncontrolled case, the middle column shows the corresponding
controlled case and the right column displays the corresponding projected longitudinal vorticity for the controlled case, represented by
contours. In the velocity vector plots, the in-plane components are represented by vectors and the out-of-plane component is represented by
contours. These quantities are displayed for four positions downstream of the vortex generators ((a) xVG / h = 1, (b) xVG / h = 2,
(c) xVG / h = 4 and (d) xVG / h = 8), h/ = 1, showing the streamwise development of the flow field. The largest secondary velocities are of
the order of 0.5 m s1 , which can be compared to the free stream velocity of approximately U = 1 m s1 .

zero at the wall surface. The four downstream positions of the generators on average give rise to primary vortices, which in
measurement planes are x VG / h = 1, 2, 4 and 8, where x VG has turn generate shear layers due to the presence of the wall.
the same direction as the x -component, with its origin at mid One can also see that, throughout the downstream development
chord of the bump (at the trailing edges of the devices), i.e., of the longitudinal vortices, the mean distance between the
x VG = 0 at x = 150 mm. vortices is almost constant. This distance is close to the
It can be seen in the velocity fields in figure 7 that the distance between the trailing edges of two vortex generators
uncontrolled flow displays a 2D boundary layer, experiencing within one pair. The mean of the longitudinal vortices therefore
separation in the area around the bump trailing edge. One does not move substantially in the spanwise direction. One can
can also see that the vortex generators have a quite substantial also see a progression of the vortices away from the wall as one
impact on the boundary layer. The presence of the vortex moves through the downstream direction, which is expected
generators causes the flow to rotate, creating counter-rotating from inviscid theory.
longitudinal vortices distorting the flow so that the high The induced effect of the vortex generators on the
momentum fluid is transported from the outer flow into the boundary layer can also be seen in figure 8, which shows
near-wall region in the downwash region (in the vicinity of the estimated mean of the measured axial and normal velocity
z = 0). The boundary layer becomes significantly thinner profiles at different streamwise positions in the controlled and
in this zone, whereas the low momentum fluid is transported uncontrolled configuration in the symmetry plane, situated in
upwards in the upwash region between two vortices as seen the downwash region (z = 0 mm). It is clearly seen that the
on the sides in the velocity plots. One can also see the transfer of high momentum into the boundary layer from the
downstream development of the flow field, where the vortices vortex generators decreases the separation behind the bump at
start off more or less isolated from effects due to the wall. this spanwise position. The increase in downwash due to the
The vortices are then gradually integrated with the remaining vortex generators can be seen in the normal velocity profiles in
part of the boundary layer flow, eventually submerging into figure 8(b) and the effect on the reduction of recirculating flow
the boundary layer. The vorticity plots reveal that the vortex can accordingly be traced in figure 8(a).

8
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

(a)

(b)

Figure 9. Reversed volumetric flow rate for every measurement

position normalized by the total flow rate in each respective
measurement plane.

Figure 8. Streamwise (a) and normal (b) velocities in the downwash

region (z = 0). The normal velocity has been scaled with a factor of triangular vanes of height h/ = 1.0. The planes correspond
four compared to the streamwise one. to positions within and in the vicinity of the separation region
in the clean configuration. One can see that the longitudinal
vortices have the same qualitative structure and downstream
3.2. Volumetric flow rate development in the measurements and the computations. It is
The volumetric flow rate has been calculated by integrating the seen that LES has the ability of capturing similar large scale
streamwise velocity component, U , for each vector over its structures in the flow due to the effect of the vortex generators
respective interrogation area and summing all of these values as observed from the PIV measurements. The vortices start out
up for each respective measurement plane. By performing this as isolated vortices and gradually submerge into and unite with
calculation for all axial vector components that are negative the boundary layer further downstream. However, one can see
in the streamwise direction, one obtains an estimate of the that the agreement is not perfect and it is not possible to make
reversed flow rate. Figure 9 shows the magnitude of the a direct comparison since the geometry and inlet conditions
reversed volumetric flow rate calculated from the estimator used in the experiments are not exactly identical to the ones
of the mean for the controlled and the uncontrolled case. used in the computations. Still, the concurrence between the
This flow rate has been non-dimensionalized with the total measurements and the computations shows that the sensitivity
flow rate through each respective measurement plane, i.e., the in modeling to geometry of this flow by LES is not high for
volumetric flow rate for both the negative and the positive this particular case. Beyond these matters, a perfect agreement
streamwise vectors have been included. The plot thus shows is still not expected. Inherent assumptions in the computations
the streamwise development of the separation for the two also give rise to non-correspondence between measurements
cases with and without devices. One can clearly see that the and modeling. LES assumes that the boundary layer flow
vortex generators reduce the amount of reversed volumetric obeys the log law and that there exists an inertial subrange,
flow rate substantially. However, they do not eliminate the independently of the Reynolds number. It has been shown
backflow. The largest backflow in the clean case is found at from experiments, [16], that the Reynolds-stress profiles in the
measurement position x = 300 mm, which can be explained overlap region depend on the Reynolds number. Therefore
from the discontinuity in the geometry in the model at the the turbulent processes can not be completely independent of
border between the trailing edge of the bump and the test Reynolds number, and the boundary layer flow predicted by
section wall. At x = 300 mm, the reduction in the normalized LES will never concur completely with experiments because
reversed flow rate is 89% of the case without vortex generators. of the low Reynolds numbers.
The unsteady point of separation, which on average is located
at a streamwise position of approximately x = 210 mm 4. Conclusions
(known from [6]), is not substantially affected by the vortex
generators; however, it is evident that the reattachment point at It has been shown that it is possible to measure and resolve
x = 350 mm (also known from [6]) is significantly affected by the flow created by a row of vortex generator pairs inducing
the vortex generators. counter-rotating vortices and their effect on the boundary layer
using SPIV. It is apparent from the results that the vortex
3.3. Vortex generator modeling generators have the expected effect on the flow in the sense that
they create large scale mixing near the wall. The measurements
Figure 10 shows the time averaged velocity fields of the SPIV clearly show a structured vortex behind each vortex generator,
measurements (left column) and the LES simulations (right whose development can be traced throughout the downstream
column) for four planes (x = 175, 200, 250 and 350 mm) for planes. The effect of the vortex generators on the amount

9
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

Figure 10. Plots of averaged velocity fields with in-plane components represented by vectors and out-of-plane component represented by
contours for four positions downstream of the vortex generators ((a) xVG / h = 1, (b) xVG / h = 2, (c) xVG / h = 4 and (d) xVG / h = 8),
h/ = 1. The left column shows results from SPIV measurements and the right column shows corresponding results from large eddy
simulation (LES). The largest secondary velocities are of the order of 0.5 m s1 , which can be compared to the free stream velocity of
approximately U = 1 m s1 .

of reversed flow is significant and shows a clear reduction in References

the amount of recirculation. The separation point seems not
to be substantially affected, whereas the reattachment point is [1] Taylor H D 1947 The elimination of diffuser separation by
influenced, curtailing the extent of the separated region. The vortex generators United Aircraft Corp. Report No.
R-4012-3
expected behaviour of the streamwise and normal velocities in
[2] ye S 1995 The effect of vortex generators on the performance
the downwash region was observed. of the ELKRAFT 1000 kW turbine 9th IEA Symp.
A qualitative comparison with LES simulations of a Aerodynamics of Wind Turbines (Stockholm) ISSN
slightly different case of the flow behind vortex generators 0590-8809
[3] Lin J C 2002 Review of research on low-profile vortex
shows that LES is capable of reproducing the overall flow generators to control boundary-layer separation Prog.
structure. LES has the ability of capturing the large scale Aerosp. Sci. 38 389420
structures in the flow induced by the vortex generators. [4] Godard G and Stanislas M 2006 Control of a decelerating
Previous tests with Reynolds-averaged NavierStokes (RANS) boundary layer. Part 1: optimization of passive vortex
generators Aerosp. Sci. Technol. 10 18191
models have shown not to possess this capability for the case [5] Angele K P and Grewe F 2007 Instantaneous behavior of
under consideration. In the future these measurements will streamwise vortices for turbulent boundary layer separation
be used to further validate various numerical methods for control J. Fluids Eng. 129 22635
calculating the flow behind vortex generators. [6] Schmidt J J 1997 Experimental and numerical investigation of
separated flows PhD Thesis Department of Mechanical
Engineering ET-PHD 97-01, MEK-DTU
[7] Raffel M, Willert C and Kompenhans J 1998 Particle Image
Acknowledgment VelocimetryA Practical Guide 3rd printing (Berlin:
Springer)
[8] Keane R D and Adrian R J 1990 Optimization of particle image
The Danish Research Council, DSF, is acknowledged for their velocimeters. Part I: double pulsed systems Meas. Sci.
financial support of the project under Grant 2104-04-0020. Technol. 1 120215

10
Environ. Res. Lett. 3 (2008) 015006 C M Velte et al

[9] Westerweel J 2000 Theoretical analysis of the measurement [12] Durst F and Schafer M 1996 A parallel block-structured
precision in particle image velocimetry Exp. Fluids multigrid method for the prediction of incompressible flows
Int. J. Numer. Methods Fluids 22 54965
29 S312
[13] Lund T S, Xiaohua W and Squires K D 1998 Generation of
[10] Foucaut J, Miliat B, Perenne N and Stanislas M 2004 turbulent inflow data for spatially-developing boundary layer
Characterization of different PIV algorithms using the simulations J. Comput. Phys. 140 23358
EUROPIV synthetic image generator and real [14] Cavar D 2006 Large eddy simulation of industrially relevant
images from a turbulent boundary layer Particle Image flows PhD Thesis Department of Mechanical Engineering
Velocimetry: Recent Improvement. Proc. EUROPIV 2 MEK-FM-PHD 2006-02, MEK-DTU
[15] Smagorinsky J 1963 General circulation experiments with the
Workshop on Particle Image Velocimetry (Zaragoza, primitive equations Mon. Weather Rev. 91 99165
MarchApril 2003) ed M Stanislas et al (Berlin: Springer) [16] Gad-el-Hak M and Bandyopadhyay P R 1994 Reynolds number
[11] Stanislas M, Okamoto K and Khler C 2003 Main results of the effects in wall-bounded turbulent flows Appl. Mech. Rev. 47
first Int. PIV challenge Meas. Sci. Technol. 14 R6389 30765

11