Content uploaded by M. Raffel
Author content
All content in this area was uploaded by M. Raffel on Jul 24, 2014
Content may be subject to copyright.
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 1 -
Airborne Application of the Background Oriented Schlieren Technique
to a Helicopter in Forward Flight
André Bauknecht1,*, Christoph B. Merz1, Markus Raffel1
1: Institute of Aerodynamics and Flow Technology, German Aerospace Center, Göttingen, Germany
* Correspondent author: andre.bauknecht@dlr.de
Abstract Blade-tip vortices and their interaction with the helicopter blades play an important role in the generation of
noise on rotorcrafts. Full-scale vortex visualization is essential for the understanding of these effects and the validation
of sub-scale experiments and numerical simulations. In the present work, the reference-free Background Oriented
Schlieren (BOS) method was used to visualize blade-tip vortices of a full-scale BO 105 helicopter. Two flight tests
were conducted with a dual-camera BOS system. In the first experiment, the ground-based camera system was tested on
the helicopter in hovering flight with an artificial background pattern. The main and tail rotor tip vortices and the
exhaust gases from the helicopter engines were visualized during take-off and hovering flight. Main rotor tip vortices up
to a vortex age of =450°, vortex instability effects and interactions with the tail boom were detected. In the second
flight test, a modified version of the dual-camera BOS system was used to visualize the main rotor blade-tip vortices of
the BO 105 during forward, curve and accelerating forward flight with velocities over the ground of 60 80 knots. The
camera system was deployed aboard a microlight airplane flying above and to the side of the helicopter with fields and
meadows serving as natural backgrounds. It was shown that, during curve flight, vortices up to a maximum vortex age
of =630° at distances of up to one rotor diameter behind the rotor plane could be visualized in some parts of the
rotor wake. For accelerating forward flight, blade-vortex interaction effects and deformations of the vortex system were
observed. In general, vortex visibility varied greatly between different parts of the rotor wake and different
measurement images. This mainly resulted from the background noise level, which can vary based on the suitability of
the natural background for the BOS method. A variety of natural backgrounds was analyzed and compared to the
artificial dot pattern used for the hover test. Image contrast, homogeneity and structure size were found to vary between
different natural backgrounds and to be significantly lower than for the artificial background pattern, but still within a
useable range. The visualization results obtained during the inflight measurements show that the reference-free BOS
method is highly suitable for the detection of helicopter blade-tip vortices during full-speed forward and maneuvering
flight, therefore removing the restrictions imposed by ground-based measurements.
1. Introduction
The blade-tip vortices in the wake of a helicopter rotor play an important role in the generation of noise and
structural vibrations on rotorcrafts. These adverse effects are produced by the interaction of blade-tip
vortices with rotor blades; a phenomenon which is referred to as blade-vortex interaction (BVI). The key
factors involved in these flow-structure interactions are the strength of the vortex filaments and their
orientation and distance to the rotor blades (Hardin and Lamkin, 1987). Several studies have been conducted
to determine the trajectories of model rotor blade-tip vortices (Landgrebe, 1972; Rauleder and Leishman,
2013), but to date, only a small number of studies are available that are based on full-scale experiments
(Raffel et al, 2000a; Klinge et al, 2006; Heineck et al, 2010; Kindler et al, 2011; Wadcock et al, 2011) or on
flight tests under realistic conditions (Raffel et al, 2014; Bauknecht et al, 2014a). These full-scale results are
essential for the validation of subscale experiments and numerical codes and to gain a better understanding
of the flow-structure interaction effects in rotorcrafts.
Most vortex visualization techniques that work well in the laboratory are not suited for the inflight detection
of helicopter blade tip vortices. Measurement techniques like laser light sheet visualization or Particle Image
Velocimetry (PIV) heavily rely on seeding particles and powerful laser illumination system and are therefore
restricted by complicated authorization processes for airborne testing. There have been attempts to apply PIV
for full-scale flight tests on a fixed-wing aircraft (Politz et al, 2010) and on a hovering helicopter close to the
ground (Raffel et al, 2001; Kindler et al, 2011). However, instantaneous measurements with PIV are still
restricted to a small part of the rotor wake.
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 2 -
Optical measurement techniques like Schlieren and shadowgraph techniques have successfully been applied
to model rotors in wind tunnels (Light et al, 1992) and model rotor hover tests (Norman and Light, 1987).
While highly suited for ground based vortex visualization, these measurement techniques require an
extensive optical setup with artificial backgrounds and illumination, which is impracticable for flight tests
that take place above the ground level.
On the other hand, the Background Oriented Schlieren (BOS) technique (Raffel et al, 2000a; Dalziel et al,
2000) has been shown to be a reliable and highly scalable method for visualizing full-scale rotor tip vortices
in wind tunnel tests (Heineck et al, 2010) and helicopters in hover flight (Raffel et al, 2000a; Klinge et al,
2006). Recently, the reconstruction of 3D vortex positions on a full-scale helicopter with a multi-camera
BOS setup has also been demonstrated (Bauknecht et al, 2014b). For dynamic flight tests, Raffel et al
(2000b) proposed a reference-free BOS setup suitable for full-scale inflight measurements. The reference-
free BOS technique has been applied to quantitative, onboard, small field-of-view vortex characterizations
on a helicopter in hover flight (Kindler et al, 2007) and the entire vortex system of a helicopter in
maneuvering flight out of ground effect (Bauknecht et al, 2014a). Recently, Raffel et al (2014) published an
overview of possible applications of the reference-free BOS method including an airborne application. In the
current paper, we use the reference-free BOS method to visualize the blade tip vortices of a BO 105 in two
different test setups. In the first experiment, a ground-based reference-free dual-camera BOS setup is used to
study the blade tip vortices of the BO 105 in hovering flight. The results of these tests show the quality of
visualization that can be achieved with the BOS method and are also used for the preparation of airborne
flight tests. In the second measurement campaign, a reference-free dual-camera BOS system is deployed
aboard a microlight observer airplane flying above the helicopter and natural backgrounds such as fields and
meadows. The visualization of the main rotor vortex system during forward and maneuvering flight
demonstrates that the reference-free BOS method is highly suitable for airborne operation, therefore
removing the restrictions imposed by ground-based measurements.
2. Materials and Methods
2.1. The reference-free BOS method
Schlieren-based methods visualize and quantify the spatial gradient of the fluid’s density (i.e. a phase
object) by the consequent optical distortions. These distortions can be attributed to the change of the fluid’s
refractive index , which is influenced by the density, as described in the Gladstone-Dale equation
1 = , (1)
where is the Gladstone-Dale constant. The optical setup for the standard BOS method (Richard and Raffel,
2001) consists of a single digital camera which is focused on a background pattern behind the phase or
density gradient object, see Fig. 1a. The angular deflection due to a variation in the refractive index can
be written as (Venkatakrishnan and Meier, 2004):
tan=1
(2)
The method yields an integral measure of the density gradients along the line-of-sight , from the observer
position to the background at =. For small deflection angles, the apparent displacement in the image
plane can be expressed as
=, (3)
with the background magnification factor =/, the focal length and the distance between the
density variation and the background . The sectional cross-correlation of the undistorted background
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 3 -
image (reference) with a measurement image containing the density variation returns a two-dimensional
displacement field. These displacements are proportional to the density gradients, integrated along the line-
of-sight. The best results are achieved for spatially delimited density variations (as in case of tip vortices)
within an undisturbed environment. Typical background structures comprise artificial dot patterns printed on
retro-reflective foil or naturally occurring structures with a sufficiently high contrast and spatial frequency,
as shown e.g. by Hargather and Settles (2010). From Eq. 3 follows that the sensitivity of the camera setup is
directly proportional to the distance between the background and the density object. For all practical
purposes, the choice of is restricted by the size of the measuring object and the depth of field of the
camera lens, which has to be focused on the background in order to maintain the high spatial resolution
necessary for the cross-correlation. As a consequence, the density variations are out of focus and the
corresponding displacement profiles in the BOS images are biased by smearing effects.
Raffel et al. (2000b) proposed a variation of the standard BOS setup, the reference-free BOS method shown
in Fig. 1b, which is more suitable for full-scale airborne measurements, where the background and the
camera perspective vary with time. This method can be applied when the density variation moves relative to
the background with time, as is the case for the blade-tip vortices of a helicopter. The density variation has to
cover only a small part of the field of view, thus leaving some parts of the photograph almost free of
distortions. Two measurement images of the same background are acquired with a short time delay
= in between. Therefore, the density variation is located in front of different parts of the
background in both images. The mainly undistorted parts of the first photograph are used as a reference for
the second measurement image and vice versa. Cross-correlation between the two measurement images
returns a displacement field, containing the signal of the density variation at both instants of time and ,
with opposite sign. For practical applications, the interframing time has to be large enough to achieve
spatial separation of the phase object in the two measurement images due to the convectional velocity. If the
camera system is moved or pivoted, the time delay between the two measurement images also has to be kept
small enough for the same background to be recorded.
For the realization of the reference-free BOS technique, either a single camera with a short interframing time
or a pair of adjacent, paraxial cameras that are focused on the same background is required. The dual-camera
setup is a workaround for an optimal single-camera setup which would have both a high spatial resolution
and a short interframing time. The two cameras have to be set up with a small distance of the optical axes
and a large distance to the measuring object in order for the alignment error to be negligible. While the
undisturbed reference image of the standard BOS technique is advantageous for the reconstruction of
quantitative density information, Bauknecht et al (2014a) demonstrated that the reference-free BOS method
yields comparable results for the vortex visualization and is more suitable for unsteady measurement
environments, where undisturbed reference images cannot be acquired, as is the case for airborne
applications.
a) Standard BOS setup b) Reference-free BOS setup
Fig.
1 Principles of a) the standard BOS setup with an artificial background and b) the paraxial reference-
free BOS
setup with a natural background (from
Bauknecht et al, 2014a)
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 4 -
2.2. Data processing
The application of a dual-camera system in a highly unsteady measurement environment requires the raw
photographs to be pre-processed before the cross-correlation. In the current study, the measurement images
were recorded in raw format and converted to the grayscale portable network graphics (png) format based on
luminance-preservation. A Color-BOS approach (Leopold, 2007) was discarded for the present camera setup
due to axial chromatic aberration effects that decreased the image sharpness in the red and blue color
channels of the camera chips. A dewarping algorithm was used to correct the misalignment between pairs of
measurement images due to the camera movements between their acquisitions. A transformation matrix was
computed based on the displacements between areas in the corners of the images, as calculated by cross-
correlation. The transformation matrix was applied to one of the measurement images to correct for the
residual image translations and rotations.
The displacement fields between the corrected image pairs were computed by means of sectional cross-
correlation using a commercial PIV software (DaVis 8.1, LaVision). A multi-grid evaluation scheme with a
final window size of 16 ×16 pixels and an overlap of 87% was selected to resolve the small vortex-induced
displacements and in the horizontal and vertical directions, respectively. In general, blade tip vortices
manifest in the BOS data as curved lines with perpendicular displacements on either side of these lines.
Vortices with peak-to-peak displacements of = 0.2 1.5 pixels were detected in the present study. For
the vortex visualization, the displacement gradients / and / and the divergence of the
displacement field / +/ were applied as robust indicators for the vortex positions. These
indicators emphasize the high spatial gradients of the vortex signals compared to the moderate changes in the
surrounding flow field.
3. Hover test
3.1. Hover test setup
In preparation of the airborne measurements, a ground-based experiment was conducted in order to optimize
the reference-free BOS dual-camera setup and to study the vortex system of a BO 105 helicopter in hovering
flight. An apron in front of a hangar at DLR Brunswick was chosen as a test site, see Fig. 2a. Retro-reflective
foil was printed with a random dot pattern and installed on the hangar wall. The artificial background had a
width of 12 m, a height of 6.3 m and contained black dots with a diameter of 8 mm, see Fig. 2b. On the
opposite side of the apron, a car on a rigid support was used as a basis for the dual-camera system, depicted
in Fig. 2c. The camera system consisted of two Nikon D3X digital single-lens reflex (DSLR) cameras
featuring 24.5 Mpx CMOS image sensors and Nikon Nikkor lenses with a focal length of 135 mm that
provided a resolution of 0.7 0.9 px/mm within the measurement region. The cameras were mounted on a
Fig. 2 Photographs of the hover flight tests. a) artificial background pattern, b) dual-camera setup, c) test overview
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 5 -
plate with a horizontal distance of 150 mm between the sensors. The optical axes of the cameras were
aligned almost collinearly with a small angle to each other to cover the same part of the background. The
measurement system was positioned at a height of 2.25 m above the ground and at a distance of
= 46 m from the background. The ground had a slope of 1.5°, resulting in an effective camera height of
2.8 m relative to the background. The background pattern was illuminated by commercially available high
power flashes, resulting in camera exposure times of 1 ms with f/16 apertures and ISO values of 100. A time
delay of 32 ms between the cameras (corresponding to a rotor rotation of 80°) was realized with a custom-
made triggering system.
3.2. Test helicopter and maneuvers
The MBB BO 105 helicopter of DLR Brunswick was employed for the hover flight tests (see Fig. 2c). The
BO 105 has a hingeless rotor with four blades (= 4), cut-off blade tips, a radius of = 4.92 m, a chord
length of = 0.27 m and a solidity of = /() = 0.07. The blades rotated with an angular velocity
of =44.4 rad/s, corresponding to a hover tip Mach number of = 0.64. During the test flights, the
helicopter was manned with two pilots and had an approximate take-off weight of 2200 kg. In addition to
hovering flight, landing and take-off maneuvers were flown, which required thrust in the range of
=21 26 kN, corresponding to a thrust coefficient of =/() = 0.005 0.006 and a blade
loading of /= 0.068 0.085. The fixed test setup required the helicopter to be operated in ground
effect at a rotor height of 0.7 above the ground. A distance of =15 20 m to the hangar wall was
chosen to ensure the safety of the helicopter maneuvers and sufficient capture of the main part of the rotor
wake. The tests took place at an elevation of 85 m above sea level under sunny weather conditions with an
ambient temperature of 25°C.
3.3. Ground test results
In Fig. 3, a photograph of the BO 105 in hovering flight is overlaid with the divergence of the displacement
field obtained by cross-correlation between the measurement images of the reference-free BOS camera
system. The rotor hub of the BO 105 is located 4 in front of the background and 0.89 above the ground,
Fig.
3 Photograph of the BO 105 in hovering flight with overlaid contour plot of the
divergence of the displacement
field. Vortices are visible up to an age of =450°
and vortex instability effects can be observed
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 6 -
which is well within the ground effect. The image covers the part of the wake in front of the helicopter. The
influence of the engine exhaust gases is visible as an area with an increased noise level on the left side of the
image, indicated by “1”. The main rotor tip vortices at both acquisition times are clearly distinguishable from
the low background noise level. As mentioned before, the time delay between acquisitions of the first image
(left camera, white vortex lines, denoted by “2”) and the second image (right camera, black vortex lines,
denoted by “3”) is 32 ms, corresponding to a rotor blade rotation of 80°. With an approximate vertical
convection rate of 12.5 m/s, the corresponding vertical displacement of the vortex filaments between the
image acquisitions is of the order of 1.5, leading to an apparent grouping of adjacent vortex pairs. The
perspective error due to the horizontal distance between the cameras causes an apparent horizontal shift of
about 6 cm between the two vortex systems, as can be seen at the location marked with “2”. For both vortex
systems, the main rotor tip vortices are visible for vortex ages of between =90° and =450°. On the
right-hand side of the helicopter, a reduced maximum vortex age of up to =270° is visible. Without
knowledge of the pilot controls, the cause for this difference cannot be determined with certainty. For both
instantaneous vortex systems, the contraction of the rotor wake and the entrained vortex filaments can be
observed for young vortex ages of up to =270° (denoted by “4”). Further downstream, the wake starts to
radially expand again, indicating the influence of the ground effect. Simultaneously, the onset of vortex
instability effects can be seen. These disturbances become most articulated for vortex ages of about
=450° (denoted by “5”). At this location, maximum amplitudes perpendicular to the main vortex axis of
±0.5 are found. For older wake ages, the vortices become too weak to be detected by the BOS method.
During hovering flight close to the ground, the BOS method suffers from obliterations caused by the slowly
escaping engine exhaust gases, as evident in Fig. 4. The figure shows a side view of the BO 105 during a
rearward take-off maneuver and the overlaid divergence field. The depicted section of the image covers the
region above the tail boom including part of the main and tail rotors and their wakes. It shows the helicopter
flying in ground effect with a momentary hub height of 0.74 above the ground level and 3 in front of the
background. The momentary velocity is 0.3 / in the vertical and 0.1 / in the rearward horizontal
direction, resulting in a take-off angle of 70° to the horizontal. The corresponding rotor wake is dominated
by the engine exhaust, which covers the entire region below the tail boom and large areas above it, including
the region around the rotor hub, denoted by the number “1”. The area affected by the blurring effect due to
the exhaust gases extends up to 0.17 above and 0.29 in front of the rotor hub (not depicted here).
Fig.
4 Photograph of the BO105 during rearward take-off with overlaid contour plot of the divergence
of the
displacement field. The tip vortex positions of the main and tail rotor as well as the engine exhaust are visualized
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 7 -
As before, the interframing time of 32 ms between the cameras causes a large separation of the two
instantaneous representations of a single vortex. This can be seen the vortex filament marked with the
numbers “2” and “3” at the instants of the first and second image acquisition, respectively. The main rotor tip
vortices are visible up to a vortex age of =300° (denoted by “4”) on the observer side of the helicopter
and obliterated for older vortex ages, either by the helicopter fuselage or the exhaust gases. On the left-hand
side of the helicopter, the oldest detectable vortices have an age of about =180°. The corresponding
displacement signals are considerably weaker than on the right-hand side, which cannot solely be explained
by the altered distance ratio of the BOS setup (see Eq. 3). For vortex ages around =180°, small-scale
disturbance effects occur and a break-up of the characteristic helical vortex shape is visible, as denoted by
“5”. The proximity to the tail rotor and the tail rotor tip vortices (denoted by “6”) suggests an interaction of
the two rotor wakes to be the reason for the change of the vortex shape. The main rotor tip vortices are also
affected by interactions with the close-by tail boom and the stabilizers. The amplitude of the vortex
instability effects increases with wake age up to the maximum visible vortex age of 300°. The large distance
between the adjacent vortices denoted by “4” and “7” is an indication for the large fluctuation of the vortex
trajectories with time. The tail rotor tip vortices at location “6” appear to be of much smaller width than the
main rotor vortices. The corresponding thin lines in the displacement field combined with the
inhomogeneous vortex visibility that varies between =90° and =450° over the tail rotor
circumference indicate that the measurement resolution was just high enough to capture these vortices.
The results of the ground-based experiments, i.e. the visualization of main and tail rotor tip vortices in
combination with the obliterations by the engine exhaust, demonstrate the restrictions and the potential of the
BOS method for the investigation of dominant wake features and interaction effects in rotorcrafts. The
findings of these tests were also used for the planning and optimization of the airborne experiments,
described in the next section.
4. Airborne test
4.1. Airborne test setup
One of the goals of the airborne application of the reference-free BOS technique is to circumvent the
operational restrictions imposed on the helicopter by a ground-based measurement system. In the second part
of the current study, an airborne test with a reference-free dual-camera BOS system was conducted. The
camera system was placed inside a Comco Ikarus C42 microlight aircraft, which was operated above and
with a lateral offset to the BO 105 helicopter in forward and maneuvering flight. Various fields and meadows
near Brunswick, Germany served as natural backgrounds for the BOS system. The suitability of these
backgrounds for BOS is discussed in detail in section 4.3. A schematic illustration of the measurement setup
is given in Fig. 5. It depicts the two aircrafts during a typical measurement situation, but at a reduced
altitude. For the actual test runs, the airplane was operated at altitudes of between 250 m and 400 m and the
helicopter at altitudes of between 220 m and 300 m above the ground level. The right-hand side door of the
microlight airplane was removed to allow for steeper observation angles of up to -30° to the horizontal. This
resulted in increased distances for the measurement system from the natural backgrounds of
=700 1500 m for the airplane and =500 1000 m for the helicopter.
For the airborne experiments, a modified version of the ground-based dual-camera BOS setup was used, as
shown in Fig. 6. It consists again of two Nikon D3X DSLR cameras with 500 mm lenses, mounted on a
common base plate. A custom-made timing unit provides sequential triggering of both cameras with
interframing times of 516 ms. During these time delays, the blades of the main rotor turn by 13° 41°
and the camera system is displaced by 0.2 0.6 m, corresponding to an average speed over the ground of
the airplane of 67 knots. The time delay is selected to minimize these interframing shifts, while providing
sufficient spatial separation between the vortex systems in both images. Both cameras exhibit a residual
temporal jitter of the acquisition times of about ± 1 ms. The two cameras are adjusted to focus on the same
background at the anticipated measurement distances. Beside the necessary optical components, the
measurement system also includes a reflex sight for easier tracking of the helicopter and a GPS receiver for
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 8 -
flight path recording and accurate time information in the raw images. During the flight tests, the camera
system was handheld and attached to the cameraman with a climbing harness.
In the course of an initial flight test, the lens apertures were varied (/8, /16, /22) in combination with
different exposure times (1/1000 s, 1/1600 s, 1/2000 s) to find an optimal setting for the available natural
illumination. During a second flight test, the cameras were set up using optimized settings with an aperture
of /8, exposure times of 1/1600 s 1/2000 s and an ISO value of 800.
Fig.
5
Schematic representation of the measurement setup featuring the observer airplane, the BO105 helicopter, and a
natural background. During actual measurement runs, both aircrafts were
flying at higher altitudes
Fig. 6 Photograph of the handheld dual-camera system
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 9 -
4.2. Test helicopter and maneuvers
The BO 105 helicopter that had been used for the hover tests was also chosen as a test object for the airborne
experiments. During the test flights, the microlight airplane and the helicopter both had an average velocity
of 67 knots and maximum velocities of 80 knots and 102 knots, respectively. For the helicopter rotor, this
resulted in tip Mach numbers in the range of = 0.49 0.54 on the retreating and = 0.73 0.80
on the advancing blade side. Various maneuvers were flown by the helicopter, including forward flight,
curve flight and accelerating forward flight. During these maneuvers, the speed and heading of the observer
airplane were adapted to optimize the field of view of the cameras.
4.3. Natural backgrounds
In the current study, natural backgrounds are used because artificial backgrounds are impractical for airborne
flight tests. To assess the suitability of different natural backgrounds, Bauknecht et al (2014a) introduced a
quality indicator for reference-free BOS measurements. It is based on the inverse of the variance of the
rotation between the two measurement images of the dual-camera BOS setup. The commercial software
PIVview 3.5 (PIVTEC) was used to calculate the displacement field (,) between undisturbed areas of the
two measurement images for various interrogation window sizes. From the displacement field, the rotation
=// was derived. The variance of the computed rotation is an indicator for the
measurement noise. The inverse of the variance can therefore be interpreted as a signal-to-noise ratio.
In Fig. 7, photographs of some backgrounds encountered during the flight tests (Fig. 7a-f) and plots of the
resulting quality indicators over interrogation window size (Fig. 7g,h) are depicted. A detail from the plot in
Fig. 7g for the relevant interrogation window sizes in this paper is given in Fig. 7h. For comparison, the
quality indicator calculated for the artificial background in Fig. 2b is plotted in the graphs as well. It is
immediately apparent that the artificial background outperforms the natural structures by far. Beside the
optimal structure size of the artificial background with a size of 4 pixels for the recorded dots, the signal-to-
noise ratio is improved by the lower ISO value and the higher grade lenses of the ground-based BOS setup.
There are also notable differences between the natural backgrounds, which can be observed in detail in
Fig. 7h. Small scale structures with sufficient contrast are best suited for BOS measurements and it is
therefore not surprising that, with the current optical setup, small crops with a high contrast between sunlit
parts and shaded areas (Fig. 7a) gave the best results. The structures of the fallow in Fig. 7f on the other hand
had a relatively low contrast and were too small to be properly resolved from the distance of the observer
airplane. In general, larger window sizes increase the signal-to-noise ratio. However, small interrogation
windows are necessary to resolve the vortices within the wake of a full-scale helicopter. Interrogation
Fig.
7 a) – f) example cutouts of the natural backgrounds, g) plot of the resulting quality indicator
over interrogation
window size and h) detail from g)
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 10 -
windows with a size of 16 ×16 pixels proved to be the optimum between the noise level and the spatial
filtering effect and were therefore used throughout this paper for both artificial and natural backgrounds.
4.4. Airborne test results
A photograph of the BO 105 in accelerating forward flight including the overlaid divergence of the
displacement field is shown in Fig. 8. The corresponding images were acquired with a time delay of 8 ms
and an exposure time of 1/1600 s. The instantaneous velocity of the helicopter was 62 knots at an altitude
of 205 m above the ground. The microlight airplane had a constant velocity of 61 knots and was operated
65 m above, 150 m behind and 110 m to the left of the helicopter. The distance between the aircrafts was
=200 m and the distances to the background were =740 m and =540 m for the observer
airplane and the helicopter, respectively. During the interframing time of 8 ms, the airplane traveled a
distance of 0.3 m. In combination with the movement of the hand-held camera system, this led to a global
displacement of 90 pixels between the two measurement images, which was corrected for by the image
mapping algorithm described in section 2.2. Again, the helicopter, its rotor blades, and the vortex system are
depicted twice in the processed divergence image, once for each recorded picture of the dual-camera system.
The measurement images that Fig. 8 is based on cover a brown field with cut straw as a background, as
shown in the lower left corner of Fig. 8. Compared to the artificial backgrounds of the ground-based test, this
natural background is less homogenous, which causes an increased background noise level, but still permits
the detection of vortices for large parts of the image. Aside from the background noise, the vortices are
obstructed by the helicopter fuselage and the engine exhaust. The hot exhaust gases cause a strong de-
correlation of the displacement signal, as visible at the location marked with the number “1” in Fig. 8. The
two tip vortex systems can be distinguished by their color in the plot: the white lines belong to the earlier
image (denoted by “2”) and the black lines to the later image (denoted by “3”). Vortex visibility varies
between individual rotor quadrants. For the location denoted by “4”, the vortex signal vanishes entirely. The
maximum visible vortex age of about =270° can be observed at the location marked with the number
“5”. Behind the helicopter, the vortex system takes on the form of a skewed helix, due to the flight velocity
and the downward flow of the rotor wake. At the lateral edges of the rotor disk, the vortex system is also
affected by the roll-up of the rotor wake. The tip vortices are affected by this roll-up as well and merge into a
single and accentuated vortex, marked with the number “6”. On the retreating blade side, a disturbance of the
Fig.
8 Photograph of the BO105 in
accelerating forward flight with overlaid divergence of the displacement field as
imaged from the observer aircraft. The blade tip vortices and the engine exhaust gases are visualized
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 11 -
vortex filaments can be observed at the location denoted by “7”. Due to the un-instrumented rotor blades, the
cause for this flow feature is not apparent. Its location however suggests that it originates from an interaction
of the adjacent rotor blade with the vortex shed by the blade 90° ahead.
Fig. 9 depicts the photograph and overlaid divergence of the displacement field of the helicopter in forward
flight with 66 knots at an altitude of 185 m above the ground and at the beginning of a rolling maneuver to
the left side. The observer airplane had a velocity of 69 knots and was operated 80 m above, 200 m behind
and 165 m to the left of the helicopter. The distance between the aircrafts was =270 m and the
distances to the background were =860 m and =590 m for the observer airplane and the helicopter,
respectively. The settings for the cameras were the same as for the measurement images of Fig. 8 and a shift
of 0.28 m due to the movement of the airplane between image acquisitions was determined, resulting in a
global displacement of 45 pixels between the two measurement images. The natural background of the
measurement images consisted of a field of green crops, as depicted in the lower left corner of Fig. 9. The
background had a higher contrast and was more homogenous compared to the one of Fig. 8, resulting in a
reduced background noise level. However, this effect is not evident in Fig. 9, since an adjusted scale was
used for the plot in order to amplify the vortex signals. As for Fig. 8, the displacement gradient field contains
the skewed helical vortex systems corresponding to the two image acquisition times. Again, vortex visibility
varies between rotor quadrants and is lowest in front and on the right-hand side of the helicopter, as denoted
by “1”. Behind the helicopter, more vortex filaments are visualized. However, the camera perspective leads
to the obstruction of the rear left-hand side vortices (marked by “2”) by the engine exhaust, as denoted by
“3”. The large field of view permits tracking of the slowly expanding cloud of exhaust gases up to 1.5 rotor
diameters downstream of the exhaust outlets. Data from other evaluated images suggests that the hot gas
stream remains coherent up to distances of 3 rotor diameters behind the helicopter. On the right-hand side of
the helicopter, the roll-up of vortex filaments into a single tip vortex can be observed again, denoted by “4”.
This strong vortex and the vortex filaments shed by the blades in the rearward half of the rotor plane remain
visible at a large distance of up to one rotor diameter behind the rotor. A maximum vortex age of =630°
can be detected at the location marked with “5”, which shows the potential of the vortex visualization with
the reference-free BOS method. A better visualization of the tip vortex system is hindered by the relatively
large distance of 270 m between the aircrafts, which results in the helicopter rotor only covering 50% of the
image width, compared to 70% for Fig. 8.
Fig.
9 Photograph of the BO105 at the beginning of a rolling maneuver to the left side with overlaid divergence of the
displacement field. The wake on the right side is visible up to a wake age of 630°
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 12 -
Another example for the BO 105 in accelerating forward flight is depicted in Fig. 10. The helicopter was
captured with an instantaneous velocity of 80 knots during an accelerating maneuver at an altitude of 225 m.
The microlight airplane was operated 45 m above, 105 m behind and 150 m to the left side of the helicopter
at a speed of 67 knots. For the camera system, this resulted in distances of =1080 m and =900 m to
the ground and a distance of 180 m between the aircrafts. The natural background in the corresponding
measurement images consisted of a brown field with cut straw and tire marks, as depicted in the lower left
corner of Fig. 10. Unlike the regular structure of the tire marks, the edges of the corresponding tracks show
up in the evaluated displacement field, as denoted by “1” in Fig. 10. The camera magnification, in
combination with the short distance between the aircrafts, resulted in the helicopter rotor covering 73% of
the image width. During the interframing time of 8 ms between the two image acquisitions, the airplane
moved by 0.3 m, resulting in a global displacement between the measurement images of 65 pixels. The high
flight speed during the image acquisitions causes the exhaust gases to remain coherent and convect almost
horizontally away from the helicopter, see number “2”. On the advancing blade side and in front of the rotor
mast, only young blade tip vortices are visualized, with a maximum vortex age of =180°, as denoted by
“3”. On the retreating blade side and downstream of the rotor, vortex visibility is increased and vortices can
be distinguished from the background noise up to a vortex age of =360°, see position number “4”. The
convection of the vortices in the front half of the rotor plane happens predominantly in the horizontal
direction, thus causing small miss-distances between these vortices and the rearward-facing rotor blades. The
consequent interactions, in combination with the previously mentioned roll-up of the rotor wake, lead to
large-scale deformations of the vortex system, visible e.g. at the position marked with “5”. From the time
history of contiguous image pairs, it is evident that these deformation effects of the vortex system increase in
amplitude with flight speed and are still present during the transition to steady forward flight.
Aside from the example images shown in Figs. 8-10, there are several other image pairs that returned good
vortex visualization results. However, only about 5% of the acquired 380 image pairs were useable, due to a
combination of unsuitable backgrounds, sub-optimal distances between the aircrafts and unaligned flight
paths. Other influences on the quality and accuracy of the measurements are discussed in the next section.
Fig.
10 Photograph of the BO
105 during accelerating forward flight with overlaid divergence of the displacement field.
A deformation of the
vortex system can be observed on the retreating blade side
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 13 -
4.5. Measurement accuracy and limitations
During the present inflight measurements, some short-comings of the lenses of the airborne dual-camera
system were discovered. Beside the low aperture that restricts exposure times below the 1/2000 s used here,
chromatic aberration effects were found for both lenses. At the edges of the photographs, lateral chromatic
aberrations were visible in the form of color bands around the background structures. For the entire
photograph, the focus was found to be different for the individual color planes of the image due to axial
chromatic aberrations. This prevented the application of the Color-BOS method and decreased the overall
sharpness of the measurement images. These short-comings clearly outweighed the light weight of the
lenses, which therefore have to be replaced by achromatic lenses in future measurements.
Due to the nature of a single-observer camera setup, only two-dimensional projections of the density
variations onto the camera image planes can be obtained by the standard and the reference-free BOS method.
Without exact measurements of the distance and an accurately determined camera perspective,
vortex positions and other geometric properties of the rotor wake can only be determined relative to known
helicopter dimensions and with an error of about ±5%. The same holds true for the estimation of the vortex
ages, which can only be determined to within ±10°. For the three-dimensional reconstruction of the vortex
system in object-space, at least two different camera perspectives and a complete camera calibration are
necessary. For details about the 3D reconstruction of blade-tip vortices, see Bauknecht et al (2014b).
The present study aimed at capturing the entirety of the helicopter vortex system, resulting in a low
resolution of individual vortices. With a chord length of = 0.27 m, the initial vortex core diameter can be
estimated to be about 14 mm. For the ground-based tests, this leads to resolutions of 10 12 pixels within
the vortex core or 56 vectors using an interrogation window size of 16 ×16 pixels with 87% overlap.
For the inflight tests, smaller resolutions of 48 pixels or 24 vectors within the vortex core can be
determined. As a result of these relatively small resolutions in combination with a spatially averaging
evaluation algorithm such as the cross-correlation technique, the BOS data is biased by smoothing, leading
to reduced maximum vortex displacement amplitudes. The finite camera exposure times of 0.5 1 ms cause
additional spatial averaging in the order of 510 mm for the azimuthal blade movement and < 3 mm for
the vortex movement.
For the dual-camera system, the vortex positions corresponding to a single instant of time are unbiased and
show the projection of the vortex system as seen by the corresponding camera. Between the two vortex
systems corresponding to the two different acquisition times, a horizontal shift is observed. This shift is
caused by the lateral distance between the cameras and also by the movements of the aircrafts for the case of
the inflight tests. The shift between the vortex systems is of the order of 6 cm within the measurement
volume for the ground-based tests, as can be observed in Fig. 3. For the inflight measurements, the vortex
shift of up to 20 cm due to the camera movement has a larger influence and partially counteracts the vortex
separation due to the flight speed of the helicopter (of the order of 30 cm), which is essential for the
reference-free method to work. To reduce the lateral vortex shifts, the velocity of the observer camera system
can be decreased, e.g. by placing the camera system aboard a second helicopter in slow forward flight to
avoid contamination of the measurement images by its own blade-tip vortices.
In summary, the airborne application of the reference-free BOS method suffers from a reduced signal-to-
noise ratio and thus a reduction of vortex visibility. These restrictions, however, are outweighed by the
removed constraints of a steady, ground-based measurement setup and the associated increased
maneuverability and possibility to measure under completely realistic flight conditions.
5. Conclusions
The reference-free BOS method was used to visualize the blade-tip vortices of a BO 105 helicopter. Two
measurement campaigns were conducted with a dual-camera BOS system. In the first campaign, the ground-
based camera system was tested on the helicopter in hovering flight with an artificial background pattern and
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 14 -
artificial illumination. The main and tail rotor tip vortices and the exhaust gases from the helicopter engines
were visualized during take-off and hovering flight. Main rotor tip vortices up to a vortex age of =450°,
vortex instability effects and interactions with the tail boom were detected.
In the second measurement campaign, a modified version of the dual-camera BOS system was used to
visualize the main rotor blade-tip vortices of the BO 105 during forward, curve and accelerating forward
flight with velocities over the ground of 60 80 knots. The camera system was deployed aboard a
microlight airplane flying above and to the side of the helicopter with fields and meadows as natural
backgrounds. It was shown that during curve flight, vortices up to a maximum vortex age of =630° at
distances of up to one rotor diameter behind the rotor plane could be visualized for parts of the rotor wake.
For accelerating forward flight, blade-vortex interaction effects and deformations of the vortex system were
observed. In general, vortex visibility varied greatly between different parts of the rotor wake and different
measurement images. This mainly resulted from the background noise level, which can vary based on the
suitability of the natural background for the BOS method.
The variety of available natural backgrounds in the present study was analyzed and compared to the artificial
dot pattern used for the hover test. Image contrast and homogeneity were found to vary between different
natural backgrounds and to be significantly lower than for the artificial background pattern, but still in a
useable range. The results obtained during the inflight measurements show that the reference-free BOS
method is highly suitable for the detection of helicopter blade-tip vortices during full-speed forward and
maneuvering flight, therefore removing the restrictions imposed by ground-based measurements.
Acknowledgments
The authors would like to thank the flight test team, foremost Gabriel Ertz and Markus Krebs for their
support during the preparation and execution of the flight tests and the pilots Sebastian Soffner and Uwe
Göhmann.
References
Bauknecht A, Merz CB, Landolt A, Meier AH, Raffel M (2014a) Blade-tip vortex detection in maneuvering
flight using the Background-Oriented Schlieren technique. Journal of Aircraft, accessed April 29, 2014.
doi: 10.2514/1.C032672
Bauknecht A, Ewers B, Wolf C, Leopold F, Raffel M (2014b) Three-dimensional reconstruction of blade tip
vortices of a BO 105 using a multi-camera BOS system. In: Proceedings of the American Helicopter
Society 70th Annual Forum, Montréal, Canada.
Dalziel SB, Hughes GO, Sutherland BR (2000) Whole-field density measurements by ‘synthetic Schlieren’.
Experiments in Fluids, 28:322-335. doi: 10.1007/s003480050391
Hardin JC, Lamkin SL (1987) Concepts for Reduction of Blade/Vortex Interaction Noise. Journal of
Aircraft, 24(2):120-125.
Hargather MJ, Settles GS (2010) Natural-background-oriented schlieren imaging. Experiments in Fluids,
48:59–68. doi: 10.1007/s00348-009-0709-3
Heineck JT, Kushner LK, Schairer ET, Walker LA (2010) Retroreflective Background Oriented Schlieren
(RBOS) as applied to Full-Scale UH-60 Blade Tip Vortices. In: Proceedings of the American Helicopter
Society Aeromechanics Specialists’ Conference, San Francisco, CA.
Kindler K, Goldhahn E, Leopold F, Raffel M (2007) Recent developments in background oriented Schlieren
methods for rotor blade tip vortex measurements. Experiments in Fluids, 43:233–240.
doi: 10.1007/s00348-007-0328-9
Kindler K, Mulleners K, Richard H, van der Wall BG, Raffel M (2011) Aperiodicity in the near field of full-
scale rotor blade tip vortices. Experiments in Fluids, 50(6):1601–1610. doi: 10.1007/s00348-010-1016-8
Klinge F, Hecklau M, Raffel M, Kompenhans J, Göhmann U (2006) Measurement of the position of rotor
blade vortices generated by a helicopter in free flight by means of stereoscopic Background Oriented
Schlieren Method (BOS). In: Proceedings of the 13th International Symposium on Applications of
Laser Techniques to Fluid Mechanics, Lisbon, Portugal, June 26-29.
17th International Symposium on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 07-10 July, 2014
- 15 -
Landgrebe AJ (1972) The wake geometry of a hovering helicopter rotor and its influence on rotor
performance. Journal of the American Helicopter Society, 17(4):3–15.
Leopold F (2007) The Application of the Colored Background Oriented Schlieren Technique (CBOS) to
Free-Flight and In-Flight Measurements. In: Proceedings of the 22nd International Congress on
Instrumentation in Aerospace Simulation Facilities, ICIASF, Pacific Grove, CA.
doi: 10.1109/ICIASF.2007.4380894
Light J, Swanson A, Norman T (1992) Application of the Wide-Field Shadowgraph Technique to
Helicopters in Forward Flight. Journal of the American Helicopter Society, 37(2):23-28.
Raffel M, Richard H, Meier GEA (2000a) On the applicability of background oriented optical tomography
for large scale aerodynamic investigations. Experiments in Fluids, 28:477–481.
doi: 10.1007/s003480050408
Raffel M, Tung C, Richard H, Yu Y, Meier GEA (2000b) Background oriented stereoscopic schlieren
(BOSS) for full scale helicopter vortex characterization. In: Proceedings of the 9th International
Symposium on Flow Visualization, Edinburgh, UK, August 22-25.
Raffel M, Richard H, Agocs J, Otter D, Mattner H, Göhmann U (2001) Experimental aspects of PIV applied
to a Bo 105 helicopter in hover-flight condition. 4th International Symposium on Particle Image
Velocimetry, Göttingen, Germany, September 17-18.
Raffel M, Heineck JT, Schairer E, Leopold F, Kindler K (2014) Background-Oriented Schlieren Imaging for
Full-Scale and In-Flight Testing. Journal of the American Helicopter Society, 59:1-9.
Rauleder J, Leishman JG (2013) Turbulence Modifications and Phase Couplings in Ground Effect under
Simulated Brownout Conditions. In: Proceedings of the American Helicopter Society 70th Annual
Forum, Phoenix, USA, May 21-23.
Richard H, Raffel M (2001) Principle and applications of the background oriented Schlieren (BOS) method.
Measurement Science and Technology, 12(9):1576-1585. doi: 10.1088/0957-0233/12/9/325
Venkatakrishnan L, Meier GEA (2004) Density measurements using the Background Oriented Schlieren
technique. Experiments in Fluids, 37:237–247. doi: 10.1007/s00348-004-0807-1
Wadcock AJ, Yamauchi GK, Solis E, Pete AE (2011) PIV measurements in the wake of a full-scale rotor in
forward flight. In: Proceedings of the 29th AIAA Applied Aerodynamic Conference, Honolulu, USA.
