Content uploaded by Robert Bleischwitz
Author content
All content in this area was uploaded by Robert Bleischwitz on Jul 28, 2016
Content may be subject to copyright.
Fluid-structure Interaction of Membrane Wings in Ground Effect
R. Bleischwitz
a
, R. de Kat, B. Ganapathisubramani
Engineering and the Environment, University of Southampton, Southampton, SO17 1BJ, UK
1 Introduction
Demand for expanding the flight capabilities of Micro-Air-Vehicles (MAVs) has lead into increasing
interest in biological inspired wings. These wings show strong flow modifications for improved
flight performance at low to medium Reynolds numbers of Re = 10.000 to 100.000. For example,
bats show with their thin and flexible membrane wings strong fluid-structure interactions, where
vibrations in the membrane are cable to modify the flow around the wing. The key requirement of
such wings is the capability to get energy into the flow, allowing energy entrainment into the
boundary layer for lower angles of attack and excited leading edge vortex shedding at higher angles.
As a result, wing stall can be suppressed into higher angles of attack and it becomes possible to
increase maximum lift [1].
However, the aerodynamic efficiency remains still very limited for wings at such low to moderate
Reynolds numbers, constricting fundamental problems such as flight distance and mission time of
MAVs. Recent attempts focused on active camber adjustment or even modulated membrane
vibrations [2]. However, this challenging concept comes with electro-mechanic complexity for an
implementation in MAVs and probably payload constrains. The usage of MAVs with flexible wings
in ground effect could be one operational option to combine benefits of flexible membrane wings
with additional efficiency enhancement by flying in the vicinity to the ground. Ground effect is
known to get important for heights over ground below one chord length, measured from the trailing
edge. At close ground proximity, the flow starts to interact with the ground, causing stagnation of the
flow below the wing surface and as a result pressure increase (ram-pressure). Additionally, tip
vortices are pushed out and reduce in size and strength. Earlier flow separation appears at the
leading edge due to suppressed downwash flow within ground effect [3].
Most studies focused on rigid and high-Reynolds-number wings in ground effect. The physics of
flexible membrane wings in ground effect are challenging due to coupling effects between
membrane vibrations, interacting flow structures and resulting force dynamics. A recent study
focused on the ground/membrane/force interaction and found that membrane wings change their
vibration behaviour towards lower mode shapes, the closer the wing is placed to the ground [4]. The
dynamic behaviour was found to behave similar to an increase in angle of attack for a given height,
where the membrane is excited into lower mode shapes, probably due to low frequency leading edge
vortex shedding. However, the coupling of specifically the flow structures to membrane oscillations
and ground interference remain still unknown.
a
e-mail : R.Bleischwitz@soton.ac.uk
Workshop on Non-Intrusive Measurements for unsteady flows and aerodynamics - 2015
Fig.1. Setup sketch
f)
e)
d)
b)
a)
c)
U∞
Therefore, the current study tries to gain knowledge about the flow dynamics above a rectangular,
perimeter reinforced membrane wing and its interaction with membrane motions and wing forces.
Wind tunnel experiments are conducted at Re = 56.000, applying time-resolved planar high-speed
flow visualisation with simultaneously recorded deformation and force measurements at an angle of
attack of 15° in and out of ground effect. Time series are used to show coupling effects between
lift/drag dynamics, membrane fluctuations and instantaneous flow structures above and below the
membrane wing.
2 Experimental Setup
The experimental study on fluid-
structure-ground interaction was
conducted in an open loop, low speed
blow-down wind tunnel at the
University of Southampton. The
experimental equipment (see Fig.1)
involves a rolling road system (a),
ensuring 98 % free stream velocity 1.5
mm above the ground. A 2-axis robotic
sting arm (b) is used to change altitude
and angle of attack. The full membrane
wing model (c) of aspect ratio 2 consist
of a 0.2 mm thick latex membrane
which is attach to a rectangular 3 mm
thick steel frame of 100 mm chord and
200 mm span. No pre-strain or slack is
applied. The membrane is capable to
rotate around the steel frame. This
attachment method was previously
found to enable higher stall angles and
enlarged maximum lift for membrane
wings [5].
The membrane is designed with a light sheet (d) accessible, translucent, 10 mm wide latex region at
¼ span, allowing flow visualisation with particle image velocimetry (PIV) both above and below the
wing. The remaining area of the membrane area is covert with a blue illuminated speckle pattern to
resolve membrane deformations with digital image correlation (DIC).
A dynamic force transducer (ATI-Nano 17, 25 N) is integrated in the nose of the sting arm (c),
allowing time resolved lift and drag measurements with a mounted wing model. The uncertainty in
the load cell was validated and met the supplying company given range of 0.006 N. The natural
frequency of the force transducer is found at 5 kHz with an underdamped behaviour, showing a
damping ratio of 0.47. At the same time, membrane deformations (e) and the flow (f) around the
membrane are captured at 800 Hz with high speed DIC and PIV (wing + wake). Blue/green bandpass
filters are used for the DIC/PIV optics to avoid laser light/background interference. The
synchronisation of DIC and PIV cameras and the (dual) laser pulses (dt = 40 μs) was ensured with a
LaVison high speed controller which was externally triggered from a data acquisition system. This
external trigger also started the recording of the load cell. The commercial software from LaVision
is used for DIC and PIV image acquisition as well as post-processing. The measurement cases
involved three heights over ground, h/c = [0.1 0.25 2] and three angles of attack α = [10° 15° 25°].
The highly energetic case at h/c = 0.1 and α = 15° will be further discussed for brevity.
Workshop on Non-Intrusive Measurements for unsteady flows and aerodynamics - 2015
Fig.2. Time series of force + membrane + flow dynamics
3 Results
3.1 Exemplary case: Membrane wing in ground effect, h/c=0.1, α=15°
t2
t1
t3
a.) Time-series:
-Lift
-Drag
-Wing motion
b.) Flow at t1:
c.) Flow at t2:
d.) Flow at t3:
Slowed down flow in GE
LE vortex forming, membrane forward cambered
LE-vortex roll down on wing, backward cambered membrane
TE vortex
LE-vortex roll up and interaction with TE-vortex
Workshop on Non-Intrusive Measurements for unsteady flows and aerodynamics - 2015
3.2 Discussion
The experimental results in Fig.2 show the interaction between forces, membrane and flow
dynamics. Fig.2a illustrates the time series of lift (blue), drag (red) and membrane oscillation (black)
within a time period of 100 ms. The (black) curve with membrane oscillation z’(t) bases on the point
of the membrane with the highest vibration intensity (rear part of mode 2 oscillation, close to trailing
edge). Additionally, instantaneous membrane fluctuations (mean subtracted) are shown for three
exemplary time steps t = [t1 t2 t3]. Green shading regions illustrate moments of instantaneous
efficiency benefits.
Maxima in lift can be correlated with peaks in drag, enforcing reduced time instant aerodynamic
efficiency. At this time period, the membrane moves from a rearward located position into a forward
location. Maximum efficiency is reached at the time period when the membrane moves from a
forward to a backward orientated shape. It remains unclear if the temporal shape or the pure
movement of the membrane is the source of time instant efficiency gains.
The flow above the wing shows additionally clear correlation with membrane mode shapes. At the
time step t1 (Fig.2b), a leading edge vortex is formed and the membrane reacts with a forward
located maximum camber (Fig.2a at t1). At that period, the membrane exhibits probably vortex core
induced suctioning, resulting in a vertical membrane deformation close to the leading edge (LE). At
a later period t2 = t1 + 7ms (Fig.2c) , the leading edge vortex has moved downstream and is located
at ¾ chord. The vortex size increases and the membrane oscillation experiences a backward
orientated maximum camber (Fig.2a at t3). This correlates again with the low pressure location of
the leading edge vortex, which is placed very effectively close to the membrane surface. Finally, the
leading edge vortex reaches the trailing edge (Fig.2d), where it starts to interact with trailing edge
vortices of counter rotating spin. The period of the membrane oscillation is found within 65-70 Hz
which matches with the shedding frequency of the leading edge vortex.
The vicinity of the ground is found to slow down the flow below the wing surface to 50% free
stream velocity. As a result, more flow is diverted over the wing upper surface, resulting in a steep
increase in velocity at the leading edge and an increased flow separation bubble with a size of 30%
chord. The separation bubble itself is found to oscillate strongly in time and reducing time instant its
size by half. The flow directly below the membrane wing surface is found to behave very steady
within time. At the trailing edge region (narrowest region to the ground), the flow below the wing is
found to gain up to 90% of the previous free stream velocity. The given flow in this region shows
strong velocity fluctuations due to leading and trailing edge vortex roll ups. Future study will focus
on modifications given with height and angle of attack changes.
References
1. P. Rojratsirikul, M.S. Genc, Z. Wang, I. Gursul, Flow-induced vibrations of low aspect ratio
rectangular membrane wings, Journal of Fluids and Structures, 27:1296-1309 (2011)
2. O.M. Curet, A. Carrere, R. Waldman, K.S. Breuer, Aerodynamic Characterization of a Wing
Membrane with Variable Compliance, AIAA Journal, 52(8) : 1749-1756 (2014)
3. K.V. Rozhdestvensky, Wing-in-Ground effect vehicles, Progress in Aerospace Sciences
42(3) :211-283 (2006)
4. R. Bleischwitz, R. de Kat, B. Ganapathisubramani, Aeromechanics of Membrane Wings in
Ground-Effect, 45th AIAA Fluid Dynamics Conference, Dallas (2015)
5. R. Bleischwitz, R. de Kat, B. Ganapathisubramani, Aspect-ratio Effect on Aeromechanics of
Membrane Wings at Moderate Reynolds Numbers Effects, AIAA Journal, 53(3):780-788 (2015)
