Content uploaded by Patrick Bot
Author content
All content in this area was uploaded by Patrick Bot on May 11, 2015
Content may be subject to copyright.
5th High Performance Yacht Design Conference
Auckland, 10-12 March, 2015
AN INVESTIGATION OF THE DYNAMIC BEHAVIOUR OF ASYMMETRIC
SPINNAKERS AT FULL-SCALE
Dario Motta1, dmot267@aucklanduni.ac.nz
Richard Flay1, r.flay@auckland.ac.nz
Peter Richards1, pj.richards@auckland.ac.nz
David Le Pelley1, d.lepelley@auckland.ac.nz
Patrick Bot2, patrick.bot@ecole-navale.fr
Julien Deparday2, julien.deparday@ecole-navale.fr
Abstract. This paper presents new results obtained from analysing on-the-water pressure, shape, force, speed and direction data that
were obtained from experiments in the Hauraki Gulf in Auckland, New Zealand in April 2014. A fully instrumented Stewart 34 Class
yacht sailing downwind was used for the tests. Details of the analysis of the results from simultaneous time-resolved measurements of
pressure, sail shape and loads are presented. The dynamic behaviour of the fluid-structure system made up of a light sail cloth and
highly curved flow is investigated. Aerodynamic forces on the asymmetric spinnaker are determined from the combination of point
pressure measurements across the sail with simultaneous shape measurements. Simultaneous time histories show a strong correlation
between the variations of pressure distributions, flapping sail shape and the forces at the corners. Periodic curling and filling of the
spinnaker luff influences suctions, in particular at the leading edge, and forces, which can dynamically change on the order of 40-
50%. The results are similar to, and extend, those that were presented by the authors at the 2013 Innov’sail Conference in Lorient,
France. It is expected that the results from this work will give reliable benchmark data which may be used to validate unsteady fluid-
structure interaction numerical simulations of downwind sails.
1 Yacht Research Unit, Department of Mechanical Engineering, University of Auckland, New Zealand
2 Naval Academy Research Institute, France
NOMENCLATURE
A Total sail area (m²)
Amain Main sail area (m²)
Aspi Spinnaker or gennaker area (m²)
AWA Apparent wind angle (°)
AWS Apparent Wind Speed (m/s)
CFTOT Total Aerodynamic Force Coefficient
CFX Total Driving Force Coefficient
CFXmain Driving Force Coefficient for the main
sail only
CFXspi Driving Force Coefficient for the
spinnaker only
CMh Total Heeling Moment Coefficient
CMhmain Heeling Moment Coefficient for the
main sail only
CMhspi Heeling Moment Coefficient for the
spinnaker only
ΔCP Differential pressure Coefficient
FEPV Force Evaluation via Pressures and
VSPARS
Fx Total Driving Force (N)
Mh Total Heeling Moment (N.m)
TWS True Wind Speed (m/s)
VSPARS Visual Sail Position and Rig Shape
Vs Boat Speed (m/s)
1. INTRODUCTION
Wind tunnel testing [1-2] and numerical simulations [3-
4] are widely used to improve the understanding of
downwind sail aerodynamics. Recently, numerical
models [5] in particular have provided fresh insights into
fluid-structure interactions. However, both methods have
various drawbacks [6]. Full-scale testing is usually
required to validate results from these methods, since this
allows the investigation of sail aerodynamics in real
sailing conditions.
On-water experiments can focus on different aspects of
sail aerodynamics and yacht design in general.
Determination of the pressure distributions on the sails
[7-9], measuring loads on the rigging [10], investigation
the influence of rigging on yacht performance [11],
determining the total aerodynamic forces and loads by
means of sailing dynamometers [12-13], are all possible
applications of full-scale testing.
At the Yacht Research Unit of the University of
Auckland significant efforts are being made to develop a
method for investigating sail aerodynamics at full scale.
This method, which has been given the name FEPV
(Force Evaluation via Pressures and VSPARS) combines
simultaneous on-water pressure and sail shape
measurements to obtain the aerodynamic forces and
moments produced by sails at full scale, where the
VSPARS system is used to ascertain the sail shape. Le
Pelley et al. [7] presented the results of the first full-scale
test carried out using the FEPV system and a validation
of the full system through wind tunnel testing for upwind
sailing. Bergsma et al. [11] describe an application of the
FEPV system to upwind sailing, where the effects of
shroud tension on upwind sailing performance were
investigated. Motta et al. [8] extended the application of
FEPV from upwind to downwind sailing, presenting the
results from downwind full-scale tests on a Stewart 34
Class yacht in very light winds in New Zealand, and on a
J80 Class yacht in stronger winds in France. In all the
mentioned works, sail aerodynamics were studied by
analysing time-averaged values of pressures, sail shape
76
and forces. Averaging the results over 20 seconds
allowed the calculation of characteristic values of the
parameters of interest for given wind direction and speed.
A step forward in the development of the FEPV method
and the full-scale technique in general has been the
investigation of the dynamic behaviour of asymmetric
spinnakers in real sailing conditions [14].
The present paper illustrates new results obtained from
experiments on the Hauraki Gulf in Auckland, New
Zealand in April 2014. A fully instrumented Stewart 34
Class yacht sailing downwind was used for the tests.
Details of the analysis of the results from simultaneous
time-resolved measurements of pressure, sail shape and
loads are presented.
In the current work, in order to better assess the
aerodynamic forces, a higher number of taps, compared
to Deparday et al. [14], were placed on the sail (72
instead of 42), as described in section 3. In addition,
pressure transducers were placed at the foot and very
close to the head of the sails.
2. COMPONENTS OF FEPV SYSTEM
2.1 VSPARS and sail shape measurements
VSPARS is a system that can be used to capture the sail
shape both in the wind tunnel and at full scale. It was
developed in the Yacht Research Unit (YRU) at the
University of Auckland by Le Pelley and Modral [15]. It
uses cameras mounted at deck level looking upwards at
the sails and rig. The system determines the global
locations in Cartesian coordinates of specific targets on
the sails and rig. For the rig, these targets comprise
coloured dots which are placed at different heights on the
mast, typically under the spreaders or at diagonal crosses.
On the sails, coloured horizontal stripes are applied to the
mainsail, jib and downwind sails. The system is able to
dynamically track the stripes, calculate the stripe
coordinates in 3D space, and link the stripe positions to
the rig deflection.
The main advantage of VSPARS over other systems is
that it is able to deal with large perspective effects. Even
systems that look up or down at the stripes from the
centre of the chord can still have significant perspective
effects at the luff and leech ends of the stripes. By
accounting for these effects, it is possible to place a
camera in the optimum position to see as much of the sail
as possible whilst still producing an accurate sail shape,
as is done in the VSPARS system. This also enables the
system to cope with large changes in sheeting angle. It
has been shown to work well even for the highly curved
stripes in off-wind sails [15]. The main steps of the
software can be seen in Figure 1. The program
essentially takes images using the required camera(s),
automatically finds the sail stripes and rig targets, and
then combines the results of all the data to give the global
X, Y and Z coordinates of the sail stripes and rig relative
to the boat fixed origin of the coordinate system.
Further details on VSPARS performance can be found in
[15].
Figure 1: VSPARS system flow chart
2.2 Pressure measurement system
The pressure measurement system was custom-built by
the Yacht Research Unit at the University of Auckland.
The generic layout of the system, as applied to each sail,
is shown in Figure 2a. Ultra-low range differential
pressure sensors (Honeywell XSCL04DC) are the core of
the system. The sensors’ resolution and range fit the
criteria for sailing applications. The pressure sensors are
mounted in custom plastic housings, approximately 40
mm in diameter and 10 mm thick. On one side, they are
stuck to the sail, with a small hole melted through the sail
to a pressure port on the bottom surface of the housing.
On the other side, a light sail cloth patch, approximately
150 mm x 150 mm in area, is applied with another hole
through to the opposite pressure port, as shown in Figure
2b. The differential pressure between the two sides is
measured. Using this setup, transducers are placed
directly at the measuring locations, thus avoiding the
issues associated with the use of long tubing and the
recording of a reliable static reference pressure [16].
An operational amplifier (op-amp) is connected directly
onto each transducer which amplifies its analogue output
from a few mV to a signal in the ±2.5 V range. Using
IDC (Insulation Displacement Contact) connectors the
transducers are connected to a ribbon cable running
along the chord of the sail. Each chord-wise cable
terminates on an analogue-to-digital (ADC) converter
chip which converts the analogue voltage signal into a
77
12-bit digital signal. A maximum of 8 taps can be
connected to each ADC. For upwind sails this seems to
be a sufficient number of taps to catch an accurate chord-
wise pressure distribution. For downwind sails it is
necessary to increase the number of taps per row because
of the more highly varying pressure distributions, and
thus two separate systems have been mounted “in
parallel” on the sail in the present measurements.
Figure 2: a) Generic layout of the YRU pressure system as
applied to a sail; b) Example of application of pressure
sensors on the sail showing the pockets containing the
sensors
The current system handles 9 ADCs and therefore 9 sets
of 8 transducers. The ADC chips are connected to a
continuous ribbon cable along the luff of the sail which
terminates at a USB-driven microcontroller box placed at
the tack. The microcontroller combines the data from all
of the taps on the sail and sends them in a single sentence
back to the data acquisition PC. The system can sample
over 150 sensors at a rate of 20 Hz, which is higher than
required for sailing applications where 4-5 Hz is
probably sufficient. If necessary, in order to reduce the
effect of electrical noise, the signal can be averaged over
a number of readings from each transducer, resulting in a
lower effective frequency.
2.3 FEPV Data Analysis
The FEPV analysis was coded in Matlab, and uses the
output files from VSPARS and the pressure system to
obtain the aerodynamic forces and moments. The whole
sail surface is created from the recorded stripe shapes and
the known tack and head positions from physical
measurements a priori. The head is assumed to be flat
(with no camber) and to have a small finite length. A
spline curve, joining the leech points of the recorded
stripes, is extrapolated upwards to the known head height
position and also downwards using the known leech
length of the sail, to give the head and foot twists
respectively, together with the first estimate of the clew
position. Unfortunately the foot shape cannot be captured
by the camera as it is out of the viewing area with the
present VSPARS setup. Therefore an initial foot shape is
estimated by fitting a spline curve through the known
tack and clew positions together with a 3rd point given by
an estimated foot depth and draft position, obtained by
extrapolating the depth and draft position of the known
stripes. This foot shape is then scaled in both the
longitudinal and transverse directions to match the
known foot length. Starting from the “low resolution”
sail shape defined by the VSPARS stripes and the foot
and head positions, a fine quadrilateral mesh is then
interpolated over the sail surface.
The sail pressure distributions are obtained from the
discrete pressure values recorded by the pressure system
by interpolation. The interpolation scheme is based on
the Radial Basis Function of order 1 (linear), which is a
real-valued function whose value depends only on the
distance from the reference points, called centres (the
pressure taps in this application). Pressure tap positions
are defined intrinsically to the sail shape in terms of
chord-wise and span-wise percentages. Moreover the use
of this interpolation scheme allows a scattered set of
pressure measurements to be extrapolated over the sail.
The pressures are interpolated to the centre of each
geometrical cell in order to obtain a pressure map
distribution over the entire surface of both sails. Forces
in specified directions are computed by integrating the
known pressures acting over the cell areas taking into
account their surface normal directions. Moment
contributions from each cell are calculated about the
specified yacht moment reference centre. In the present
case the moment reference centre was fixed at the base of
the mast.
2 a
2
b
pressure transducer in differential housing
Op-amp
PCB
IDC connector
ribbon cable
8 channel AD converter
Arduino microcontroller
USB output to PC
78
3. FULL SCALE TEST SETUP
A Stewart 34 class yacht was used for the full-scale
testing in New Zealand. Class rules impose the use of a
symmetric spinnaker, but for these tests the yacht has
been equipped with an asymmetric spinnaker designed
by North Sails New Zealand.
A GPS unit, sampling at a rate of 2.5 Hz, was used to
record the speed over ground, course over ground and
boat location. An Inertial Measurement Unit (IMU) was
placed in the yacht cabin and logged the boat motion at
10 Hz. A sonic anemometer placed 1.2 m above the
masthead recorded the wind speed and direction in 3
dimensions. In addition two GoPro cameras, placed at
the stern and looking forwards, recorded the crew
activity and trimming of the spinnaker.
The asymmetric spinnaker was equipped with the YRU
pressure measurement system (described in Section 2.2).
Pressures were recorded at approximately 10 Hz. Taps
were placed at six different chord-wise stripes along the
sail, which are named: foot, S1/5, S2/5, S3/5, S4/5 and
S15/16 respectively. The name S1/5 indicates that the
stripe starts on the luff at 20% (i.e. 1/5) of the luff length
from the foot and ends on the leech, at 20% of the leech
length from the clew; sensors are placed on the curve
joining these two points. Tap locations as a percentage of
the curve length from the luff are shown in Table 1.
Based on previous experience [8], the authors decided to
use the maximum number of pressure taps available,
namely 72. Some taps were placed as close as possible to
the sail foot and close to the sail head, since these have
never been measured at full-scale to best of the authors'
knowledge.
The VSPARS system recorded the sail shape for the
spinnaker by means of two GoPro cameras (one camera
for each tack), used in video mode. Images were then
extracted from the video at 10 frames per second.
The spinnaker was equipped with four VSPARS orange
stripes, placed at 1/4, 1/2, 3/4 and 7/8 of the luff length
from the tack respectively. The stripes were laid on the
sail in order to fly approximately horizontally (in the
yacht’s reference system) whilst sailing. A further orange
stripe was placed at height 1/8 but, as the authors
expected, it fell outside the camera’s field of view and so
was not used during the data post-processing. The
general specifications of the spinnaker and VSPARS
stripe lengths are shown in Table 2.
A custom-made data acquisition unit recorded all data
streams, each one at its own sampling rate, and so the
data were all time stamped to enable subsequent
synchronous processing. During the post-processing, all
measurements were resampled at 10 Hz and smoothed
with an exponential moving average in order to remove
some unexpected peaks.
Table 1: Taps locations expressed as percentage of the
curve length
Table 2: Stewart 34 asymmetric spinnaker characteristics
Spinnaker area 88.6 m2
Luff length 13000 mm
Leech length 12260 mm
Mast height 12800 mm
VSPARS stripe 7/8
length
2740 mm
VSPARS stripe 3/4
length
5142 mm
VSPARS stripe 1/2
length
8022 mm
VSPARS stripe 1/4
length
8628 mm
The measurements were performed in the Hauraki Gulf,
Auckland, NZ, with a breeze of 10-15 knots, in an area
with insignificant tidal flow with almost flat water.
4. RESULTS
Investigating the dynamic behaviour of an asymmetric
spinnaker at full scale is not an easy task. The wind is
constantly changing, the yacht is moving and the
spinnaker itself has very large displacements that
influence pressures and forces. In order to understand
pressure and force variation it is important to identify
time windows in which as many parameters as possible
are constant, so that the changes can be related to the
remaining fluctuating variables.
Foot
S
1/5
S
2/5
S
3/5
S
4/5
S
15/16
length
/
tap no
7780
mm
8684
mm
8614
mm
7140
mm
4040
mm
1730
mm
1 7 0.6 0.6 0.7 1.2 10
2 10 3 3 3.0 3 35
3 15 5 5 5.0 5 65
4 25 7.5 7.5 7.5 10 90
5 40 10 10 10 15
6 60 15 15 15 25
7 80 20 20 20 35
8 94 25 25 25 45
9 30 30 30 60
10 35 35 35 75
11 50 50 50 90
12 60 60 60 99
13 70 70 70
14 80 80 80
15 90 90 90
16 99.4 99.4 99.4
79
In section 4.1 pressure averages over 20s are shown in
order to give an overview of the spinnaker behaviour. In
section 4.2 and 4.3 two different cases are shown and
discussed in order to explain the dynamic behaviour of
the spinnaker. In section 4.2, the spinnaker was slightly
overtrimmed and the sail was quite stable (no curling
occurring near the luff), while in section 4.3 the
spinnaker was on the verge of luffing, as is common
practice among sailors.
4.1 Overview of time averaged pressures
Sailing downwind is unsteady by nature, due to the wind
conditions, waves, helming and the sails' movement. To
determine averaged forces and pressure distributions at
different AWAs, time periods of about 20 seconds were
chosen where the standard deviation of the AWA was
below 5° and of the AWS below 0.5 m/s. Sail trim
(optimal sail trim with gennaker on the verge of luffing)
was kept constant for each run and the boat heading was
kept as straight as possible to enable the results to be
averaged.
Figures 3 shows an example of averaged differential
pressure coefficient (ΔCP) distributions for all the
pressure stripes on the spinnaker for AWA equal to 98.
ΔCP is plotted against the sail curve length percentage.
The dynamic pressure was calculated from the apparent
wind speed (AWS), and the pressure differences are
leeward minus windward, thus giving negative values.
The pressure coefficient plots have the negative direction
upwards, as is common in presenting pressure
distributions on wings. The standard deviation for ΔCP at
stripe 3/5 is shown as well for reference.
Figure 3 shows that suctions generally increase from the
foot upwards to section 4/5 and then reduce to the head.
For the lower stripes (1/5 and 2/5) there is a suction peak
at about 5-7% from the leading edge followed by a
recovery down to the trailing edge. A plateau in the
pressure distribution or a second suction peak can
sometimes be seen at about 20-30% of the curve, such as
with section 1/5.
At stripes 3/5 and 4/5 the flow is probably separated, as
indicated by the weak pressure recovery after the leading
edge pressure peak. This difference in behaviour between
the higher and lower sections confirms what has been
already seen in earlier studies of asymmetric spinnakers
on both the Stewart 34 and J80 yachts [8].
Even in stable wind conditions (standard deviation of
AWA and AWS less than 5 deg and 0.5 m/s respectively)
pressure variations along the curve are big, particularly
near the leading edge and at higher stripes. Standard
deviation of ΔCP in the proximity of the leading edge is
equal to approximately 1.7 at S3/5 and S4/5 and 1.2 at
S1/5.
One novel aspect of the present study is information on
the pressures at the foot and very close to the head
(namely at 880 mm from the head). Time-averaged
differential pressure coefficients at the foot as high as
-3.5 are obtained. It appears that these differential
pressure increase from near zero at the tack, reach a
maximum about a quarter of the way along the foot and
then decrease again. Also the ΔCP variations along the
foot are not negligible, having a maximum standard
deviation of about 1.04 at 15% from the luff. The first
pressure tap at the foot is placed at some distance (7%)
from the luff and therefore a suction peak, if present, was
not measured, either because too small or too close to the
leading edge. This trend is repeated for all the range of
AWAs investigated.
Figure 3:
ΔCP
distributions at different heights along sail
for AWA = 98.
As mentioned in Section 3, a line of 4 pressure taps was
placed at 15/16th of the sail height, thus very close to the
sail head. Other than giving some new interesting
information on the pressures in such locations,
knowledge of these pressures was very beneficial in
improving the interpolation of pressure over the whole
sail in the FEPV code. Unfortunately, limitations on the
number of taps meant that only 4 taps were placed closed
to the head. As expected in this case, the suctions are not
close to zero, but have significant values at all AWAs. In
the example shown in Figure 3, maximum suction is
achieved at about 35% of the curve and has a ΔCP value
of -3, while standard deviations are approximately equal
to 0.4-0.6 over the whole stripe. This happens regardless
of the presence of a leading edge suction peak at 4/5 of
the sail height.
4.2 Dynamic variation of pressures and forces with a
slightly overtrimmed spinnaker.
When sailing downwind, it is common practice to trim
the spinnaker on the verge of luffing. In this section we
discuss pressure and shape variation for a run in which
the spinnaker was “overtrimmed”, i.e. the luff is not
curling. During the period when the data were recorded,
trimming, steering and crew positions were fixed.
The pressure and shape behaviour explained herein
occurs several times throughout the test; the run
80
presented has been chosen as representative of all the
runs.
In Figure 4 the ΔCP variation for each of the pressure
stripes along the sail for a 10 second run are shown. The
horizontal axis is time while the vertical axes represent
the percentage of sail curve and ΔCP’s are shown through
the colourmap. Figure 5 shows CFX, AWA and AWS
variations for the same time period. Four times are
highlighted in Figures 4 and 5 and they are named A, B,
C and D. Sail shapes for frame A is shown in Figure 6.
Shapes at the other frames are very similar, and
differences are hardly noticeable, so they are not
presented in the paper.
As shown in Figure 4, suctions are near maximum at
frames A and C for all stripes. A leading edge suction
peak is present at all stripes and CFX is a maximum. Also
the AWS has a local maximum, while AWA has little
relationship with the driving force. At frames B and D
CFX, AWS and suctions for all stripes are minimum; ΔCP
distributions have similar trends to the ones at frames A
and C, but the leading edge suction peaks disappear or
become less pronounced.
Figure 4: Space-time diagrams of
ΔCP
: distribution along
the curves (vertical axis), variation with time (horizontal
axis) for all sections.
Figure 5: CF
X
, AWA and AWS variation with time.
Figure 7 shows the ΔCP distributions at stripe 2/5 as an
example and highlights the change from low suction to
high suction areas and vice-versa. A leading edge suction
peak is often present, and it is more pronounced at
frames A and C where suctions are highest. Note that in
this situation (sail not luffing) suctions increase/decrease
simultaneously along the chord (even if changes are
biggest near the luff). This leads to big changes in CFX,
which varies as much as ±35% from its average value
and, for instance, it increases of almost 80% of its initial
value from B and C. The whole sail is “inflating” and
“deflating” periodically, possibly driven by the changes
in AWS and yacht motions, but the shape doesn’t change
significantly and is similar to that shown in Figure 6
throughout.
Figure 6: Sail shape at frame A.
Figure 7:
ΔCP
distributions at frames A, B, C and D for
stripe 2/5.
81
4.3 Dynamic variation of pressures and forces with
spinnaker luff curling
In this section we discuss pressure and shape variation
for a run in which the spinnaker was slightly luffing
while sailing, as is usual practice when sailing
downwind. During the recorded time period, trimming,
steering and crew position were fixed. The results shown
herewith are extracted from a time period of about 20
seconds where the standard deviation of the AWA was
below 5° and of AWS below 0.5 m/s. During this period
the behaviour of the spinnaker was quite different from
that analysed in section 4.2. The pressure and shape
behaviour explained herewith occurs several times
throughout the test; the run presented has been chosen as
representative.
Similarly to section 4.2, Figures 8 and 9 show the ΔCP,
CFX, AWA and AWS variation with time. Four instants
are highlighted and named E, F, G and H in the figures.
Figure 10 shows the sail shape at these four times.
Figure 10 shows that the spinnaker is not curling
simultaneously throughout the whole luff, but curling
occurs alternatively at two separate locations, one at the
top part and one at the lower part of the luff respectively,
which are named P1 and P2. Curling at P1 directly
affects pressure at stripes S15/16, S4/5 and sometimes
S3/5, while curling at P2 concerns pressure at S2/5 and
S1/5.
Figure 8:
ΔCP
variation with time.
Figure 9: CF
X
, AWA and AWS variation with time.
Consider a full sail shape (no curling) as the starting
point for this investigation, as it is at t=113.0s (before
frame E):
- At frame E, the luff is curling at P1. Suctions are low
at S15/16, S4/5 and S3/5. CFX is minimum at this
stage, probably due to the highly reduced contribution
of pressures at higher heights.
- From E to F, the curling “travels down” from P1 to P2.
Suctions increase at stripes 15/16 and 4/5, while they
decrease where curling occurs, at stripes 3/5 and 2/5.
CFX is recovering from its local minimum at this stage.
- At G the curling has “moved” to P2. Suctions are now
lowest at S2/5, S1/5 and at the foot, while they have a
maximum at S4/5 and S3/5. The overall aerodynamic
force (which is not displayed on this paper) on the sail
is still increasing compared to time F. In this particular
case CFX instead maintains a value similar to that in
time F.
- From G to H, the shape recovers its full shape, and this
can be considered as the completion of a cycle.
Suctions recover at S1/5 and S2/5 near the luff, and all
stripes settle at their new ΔCP chord-wise distribution.
Maxima of CFX and CFTOT occurs during the
recovering of the sail shape. In this example, CFX at
time H is 21% higher than at time E.
During this period curling at parts P1 and P2 appear to
alternate, as do the suctions. High suctions at part P1
correspond to low suctions at part P2 and vice-versa. The
changes in sail shape and pressures appear to be related
to AWA variations. AWA is maximised when the sail is
full, it reduces when the sail starts folding at P1, it is
minimised between frames F and G (during the curling)
and it increases again when the sail shape recovers its
shape. The AWS has little variation during this process.
When the top part of the sail is curling, a chord-wise
“evolution” of suctions with time occurs at S4/5. ΔCP
distributions for stripe 4/5 at times E, F, G, H and at
some intermediate instants are shown in Figure 11. These
show that:
82
- At time E the luff is curling at S4/5 and the ΔCP’s are
low, especially near the leading edge.
- At time F the suctions increase in the first 40% from
the leading edge due to the top part of the sail filling up
again after the curling. There is no leading edge
suction peak yet.
- From time F to t=113.9s there is a big increase in
suction in the first 5% of the chord.
- From t=113.9s to t=114.7s: the high suction region
appears to travel towards the trailing edge. At the same
time the region becomes broader in extent but
gradually weaker in strength.
- At time H the cycle is almost completed. There is no
leading edge suction peak, indicating that the top part
of the sail is going to fold again.
A similar phenomenon occurs at section 15/16, with a
suction increasing near the leading edge and travelling
towards the trailing edge. In this case only 4 taps are
available, so a less detailed analysis can be carried out.
As shown in Figure 8, suctions increase at the leading
edge (the first tap is placed at 10% of the chord) at time
F. From times F to G the highest suction shifts towards
the leech, reaching 25% of the curve by time G.
However, in contrast to what happens at stripe 4/5, the
suctions keep travelling on to reach 65% by time H. In
other similar situations the high suctions seem to travel
on across the whole section.
Therefore the resulting effect of the uncurling of the luff
is a “suction shedding” or “evolution” towards the leech
at stripes S4/5 and S15/16, which repeats quite
consistently throughout the test. It can sometimes be seen
at lower stripes, but it is at S4/5 and S15/16 that it occurs
periodically throughout the whole test. The source of this
suction development is unknown but it could be
associated with the formation and shedding of a leading
edge vortex [4]. Alternatively it may be created by the
movement of the tip vortex shed behind the spinnaker
head [4].
Figure 10: Sail shape at frames E, F, G and H.
Figure 11: Chord-wise “suction evolution” at pressure
stripe 4/5.
P1
P2
a
b
c
d
83
It is interesting to discuss the evolution of chord-wise
pressures following to curling of the sail. A first example
of this behaviour had been published in [14]. Figure 12
shows ΔCP’s for S3/5 at three different time instants
across a curling and recovering of the shape, as it is been
between instants E and H. The average ΔCP distribution
across the time period considered is shown with a dotted
line as reference. Minimum suctions along the curve at a
given stripe occur at the beginning of the curling. When
the curling is at its maximum, suctions are lowest near
the luff (the extent depending on how much the luff is
folding), but bigger everywhere else. The recovery of the
shape induces a sudden increase in suctions in the first
20% of the chord, and a pronounced leading edge suction
peak. This is probably responsible for the increase in
aerodynamic force as was discussed in regard to Figure
9. This phenomenon might be due to the quick
deceleration of the luff when recovering its shape, which
creates a “whiplash” giving a consequent added mass of
air. It can also be related to a local increase of the flow
speed amplifying locally the ΔCP. Similar behaviour is
found at lower stripes, but peaks in suctions are not so
pronounced, as can be seen in figure 12.
Figure 12:
ΔCP
distributions at stripe 3/5 at start of curling,
maximum curling, recovered shape and average
ΔCP
5. CONCLUSIONS
On water experiments have been carried out to
investigate the dynamic behaviour of an asymmetric
spinnaker in real sailing conditions. Yacht motions, wind
characteristics, pressures and flying shape on an
asymmetric spinnaker were simultaneously measured.
Aerodynamic forces developed by the spinnaker were
computed by the FEPV system.
Pressures were measured using 72 pressure sensors,
distributed along six stripes at approximately 1/5th, 2/5th
3/5th, 4/5th of the sail height and, for the first time at full-
scale, at the foot and close to the head (15/16th of the sail
height).
Time averaged chord-wise pressure distributions at
stripes S1/5 to S4/5 agree with previous full-scale
publications [2, 8, 14]. Pressures along the foot were
found to be non-zero at all AWAs, and mean differential
pressure coefficients as high as -3.5 occurred at 25% of
the chord. Suctions very close to the head were also non-
zero, with the minimum mean ΔCP being approximately
-3.5 at 35% from the leading edge. Suctions close to the
head have significant high values over the whole chord.
The dynamic behaviour of the asymmetric spinnaker has
been analysed for two different spinnaker trims. In the
first case the spinnaker was slightly overtrimmed, and
sail was full at all times, while in the second situation the
spinnaker was on the verge of luffing, as is common
practice in downwind sailing.
When the spinnaker is slightly overtrimmed, suctions
increase/decrease simultaneously along the chord and at
all heights. This affects the entity of CFX, which varies as
much as ±35% from its average value despite little
variation of AWA. Chord-wise ΔCP distributions present
a leading edge suction peak at all time, which is more
pronounced when the suctions are highest over the chord.
When the trim is optimal, the sail is luffing almost
periodically. The curling alternates at two separate
locations, at the top and at the lower part of the luff
respectively, as do the suctions. When the sail is luffing
at the top part, the suctions are highest in the bottom
(full) part and vice-versa. The CFX variations are driven
by the different contributions from the pressures at the
top and bottom part of the spinnaker. Generally CFX is a
minimum when the top part of the sail is curling and a
maximum after the bottom part is curling, when the sail
is recovering its full shape.
The uncurling of the luff leads to the development of a
leading edge suction peak in the first 5% from the luff for
stripe S4/5, and then to a “suction shedding” towards the
leech up to 25% of the curve length. Similar pattern is
found at S15/16, but suctions keep traveling on to reach
65% of the curve and, sometimes, across the whole
section.This phenomenon could be associated with the
formation and shedding of a leading edge vortex.
Alternatively it may be created by the movement of the
tip vortex shed behind the spinnaker head.
At stripes 3/5 and 4/5, the recovery of the shape after a
curling induces a sudden increase in suction in the first
20% of the chord, and a pronounced transient leading
edge suction peak. This phenomenon might be due to the
quick deceleration of the luff when recovering its shape,
which creates a “whiplash” giving a consequent added
mass of air. It can also be related to a local increase of
the flow speed amplifying locally the ΔCP.
Findings from this work show that full-scale
measurement of pressure and shape in dynamic
conditions is a feasible task, and important insight into
the aerodynamics of sails can be made.
84
References
1. Le Pelley, D., Richards, P., (2011), “Effective Wind
Tunnel Tesing of Yacht Sails Using a Real-Time
Velocity Prediction Program”,20th Cheasapeake
Sailing Yacht Symposium, March 18th-19th,
Annapolis, Maryland, USA.
2. Viola, I.M., Flay, R.G.J., (2009),“Force and pressure
investigation of modern asymmetric spinnakers”,
Internation Journal of Small Craft Technology,
Transaction RINA, 151(2):31-40.
3. Richards, P.J., Lasher, W., (2008), “Wind Tunnel
and CFD Modelling of Pressures on Downwind
Sails”, BBAA VI International Colloquium on Bluff
Bodies Aerodynamics & Applications, July 20th-
24th, Milano, Italy.
4. Viola, I.M., Bartesaghi, S., Van-Renterghem, T.,
Ponzini, R.,(2014), "Detached Eddy Simulation of a
sailing yacht", Ocean Engineering, 90: 93-103.
5. Durand, M., Lothode, C., Hauville, F., Leroyer, A.,
Visonneau, M., Floch, R., (2013), "FSI Investigation
on Stability of Downwind Sails with an Automatic
Dynamic Trimming", in 3rd International
Conference On Innovation in High Performance
Sailing Yachts, June 26th-29th, Lorient.
6. Wright, A.M., Claughton, A.R., Paton, J., Lewis, R.,
(2010). “Off-Wind Sail Performance Prediction and
Optimisation”, The Second International Conference
on Innovation in High Performance Sailing Yachts,
June 30th- July 1st, Lorient, France.
7. Le Pelley, D.J., Morris, D., Richards, P.J., (2012).
”Aerodynamic force deduction on yacht sails using
pressure and shape measurement in real time”, 4th
High Performance Yacht Design Conference, March
12th-14th, Auckland, New Zealand.
8. Motta, D., Flay, R.G.J., Richards, P.J., Le Pelley,
D.J., Deparday, J., Bot, P., (2014), “Experimental
Investigation of Asymmetric Spinnaker
Aerodynamics Using Pressure and Sail Shape
Measurements”, Ocean Engineering, 90: 104-118.
9. Graves, W., T. Barbera, J.B. Braun, L. Imas, (2008)
“Measurement and Simulation of Pressure
Distribution on Full Size Sails”, 3rd High
Performance Yacht Design Conference, December
1st-4th, Auckland.
10. Augier, B., Bot, P., Hauville, F., Durand, M., (2012),
“Experimental validation of unsteady models for
fluid structure interaction: Application to yacht sails
and rigs”. Journal of Wind Engineering and
Industrial Aerodynamics 101(0): 53-66.
11. Bergsma, F., Motta, D., Le Pelley, D.J., Richards,
P.J., Flay, R.G.J., (2012). “Investigation of shroud
tension on sailing yacht aerodynamics using full-
scale real-time pressure and sail shape
measurements”, 22nd International HISWA
Symposium on Yacht Design and Yacht
Construction, November 12th-13th, Amsterdam.
12. Hochkirch, K., (2000). “Design and Construction of
a Full-Scale Measurement System for the Analysis
of Sailing Performance”, Technical University of
Berlin, PhD Thesis.
13. Masuyama, Y., Fukasawa,T., (2009). “Database of
sail shapes versus sail performance and validation of
numerical calculations for the upwind condition”,
Journal of Marine technologies, Vol.14(2), pp.137-
160.
14. Deparday, J., Bot, P., Hauville, F., Motta, D.,Le
Pelley, D.J., Flay, R.G.J., (2014), “Dynamic
measurements of pressures, sail shape and forces on
a full-scale spinnaker”, 23rd International HISWA
Symposium on Yacht Design and Yacht
Construction, November 17th-18th, Amsterdam.
15. Le Pelley, D.J., Modral, O., (2008), “VSPARS: A
combined sail and rig shape recognition system
using imaging techniques”, 3rd High Performance
Yacht Design Conference, March 12th-14th,
Auckland, New Zealand.
16. Flay, R.G.J., Millar, S., (2006), “Experimental
Considerations Concerning Pressure Measurements
on Sails: Wind Tunnel and Full-Scale”, 2nd High
Performance Yacht Design Conference, December
1st-4th, Auckland, New Zealand.
85
