Está en la página 1de 12

An Investigation of Flow Fields

Over Multi-Element Aerofoils


This paper presents results obtained from a combined experimental and computational
study of the flow field over a multi-element aerofoil with and without an advanced slat.
Detailed measurements of the mean flow and turbulent quantities over a multi-element
aerofoil model in a wind tunnel have been carried out using stationary and flying hot-wire
S. R. Maddah (FHW) probes. The model configuration which spans the test section 600 mm
600 mm , is made of three parts: 1) an advanced (heel-less) slat, 2) a NACA 4412 main
H. H. Bruun aerofoil and 3) a NACA 4415 flap. The chord lengths of the elements were 38, 250 and 83
mm, respectively. The results were obtained at a chord Reynolds number of 3 105 and a
Department of Mechanical and Medical free Mach number of less than 0.1. The variations in the flow field are explained with
Engineering, reference to three distinct flow field regimes: attached flow, intermittent separated flow,
Fluid Flow Division, and separated flow. Initial comparative results are presented for the single main aerofoil
University of Bradford, and the main aerofoil with a nondeflected flap at angles of attacks of 5, 10, and 15 deg.
Bradford, BD7 1DP, United Kingdom This is followed by the results for the three-element aerofoil with emphasis on the slat
performance at angles of attack 10, 15, 20, and 25 deg. Results are discussed both for
a nondeflected flap f 0 deg and a deflected flap f 25 deg . The measurements
presented are combined with other related aerofoil measurements to explain the main
interaction of the slat/main aerofoil and main aerofoil/flap both for nondeflected and
deflected flap conditions. These results are linked to numerically calculated variations in
lift and drag coefficients with angle of attack and flap deflection angle.
DOI: 10.1115/1.1431267

Keywords: Turbulent Flow, Flying Hot-Wire, Advanced and Conventional Slats, Multi-
Element Aerofoil

1 Introduction appears to be the most detailed and extensive experimental data


set covering attached as well as separated flows both for single
High-lift systems are used on aircraft to provide adequate low
and two-element aerofoils. However, there remains a need for
speed performance in terms of approach speed and take-off and
similar detailed measurements of mean velocity and turbulent
landing field lengths. Leading-edge devices, such as slats and the
quantities of the flow field over three-element aerofoils with em-
Kruger flap, are commonly used to delay stall to higher angles of
phasis on the effect of the flap and the slat in particular a heel-
attack and consequently increase the maximum lift performance,
less slat at low to high angles of attack ranging from 5 to 25
C lmax . For aviation purposes, the significant features are the total deg. This paper presents results for a single, two-element, and
integrated lift and drag forces. Consequently, there are many three-element aerofoil and highlights the variation in the mean
reported direct measurements of the lift and drag coefficients for and turbulent quantities as the flow changes from attached flow to
two- and three-element aerofoils e.g., Innes et al. 1 and Moens intermittently separated flow and finally to separated flow. The
and Capbern 2. However, such integrated quantities cannot pro- interpretation from the observed flow phenomena in this investi-
vide direct insight into the occurring velocity and pressure fields. gation is combined with related flow field results and vortex con-
As the primary fluid mechanics feature is the flow field over the cepts to highlight the varying interaction between upstream/
aerofoil, the emphasis of this paper is on the mean and turbulent downstream elements with angles of attack and flap deflection
flow fields. There are a number of other related investigations. angle in a high lift system, and the resulting effect on the lift and
Two-element configurations have been studied by e.g. 35 and drag of multi-element aerofoils.
the performance of a conventional slat in high lift systems has
been investigated e.g., Nakayama et al. 6 and Braden et al. 7.
The slat and flap gap flows have been reported by Savory et al. 8 2 Experimental Arrangement and Test Conditions
and Alemdaruglu 9. These studies confirm the presence of slat
The experiments were conducted in a low speed wind tunnel
cove separation, which has some effects on the boundary layer with a cross section of 600 mm600 mm located at the Univer-
development over the wing and the mean velocity and turbulence sity of Bradford. The model configuration spans the test section
fields over the main aerofoil, Maddah et al. 10. The slat cove and it is made of three parts: 1 an advanced heel-less slat, 2 a
separation is minimized for the advanced slat due to its smooth NACA 4412 main aerofoil and 3 a NACA 4415 flap. The chord
and heel-less lower surface and it therefore produces less drag lengths of the elements were 38, 250 and 83 mm, respectively.
compared to a conventional slat with lower surface discontinuity, The slat gap and overlap and the flap gap and overlap were set to
Jones 11. 0.02%c, 0.025%c, 0.015%c, and 0.04%c, respectively. The
The majority of experiments on multi-element aerofoils have free-stream velocity was 18 m/s and the corresponding Reynolds
been conducted for attached flow conditions with only limited
number based on the main aerofoil chord length was 3105 . By
data at near stall conditions. Even if there was separation, it was
using extra screens, at the inlet to the wind tunnel, the free stream
on the flap at low angles of attack. The work of Braden et al. 7
turbulence intensity was reduced to 0.4%.
The main experimental technique of the present research was a
Contributed by the Fluids Engineering Division for publication in the JOURNAL
OF FLUIDS ENGINEERING. Manuscript received by the Fluids Engineering Division
flying hot-wire mounted on a precise computer-controlled mecha-
January 23, 2001; revised manuscript received August 27, 2001. Associate Editor: nism. The technique is based on moving the probe along a pre-
T. Gatski. scribed path with a large enough velocity to avoid hot wire signal

154 Vol. 124, MARCH 2002 Copyright 2002 by ASME Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


technique to be used for cases with or without separation. Details
of Bradford Universitys flying hot-wire system are given by
Maddah 13.
Two-Dimensionality. The flow over the multi-element aero-
foil in the wind tunnel may deviate from two-dimensional flow.
The deviation is associated with the growth of the secondary flows
in the corners formed by the model and the side walls of the wind
tunnel and it will increase directly with angle of attack and flap
deflection angles. Aerofoil models with aspect ratios of about 2 or
less have been used 6,9,14 with consequently significant three-
dimensional effects. For the present investigation, the aspect ratio
for the main aerofoil was 2.5 which resulted in much weaker
three-dimensional effects. Comparative measurement of the mean
velocity profiles at different spanwise locations showed negligible
variation with and without the slat. For the nondeflected flap case
( f 0 deg), the two-dimensionality of the flow was validated for
about 90% of the span. However, for the highly deflected flap
( f 25 deg) cases, in particular for 20 deg, splitter plates
were used to maintain two-dimensional flow for 80% of the span.

3 The Measurement Matrix and Flow Regime Classi-


fication
The test matrix is shown in Table 1. It includes the two-element
Fig. 1 Four-bar flying hot-wire mechanism, notation, and wind aerofoil main aerofoil and the flap cases and the three-element
tunnel measurement coordinate system. The geometry is that aerofoil cases incorporating the advanced slat. Four configurations
which is used at the University of Bradford: r 60 mm, a are considered for the two-element aerofoil: three cases with the
160 mm, c 146 mm, b 468 and 548 mm for old and new main aerofoil set at 5, 10, and 15 deg with a nondeflected flap and
flying arm, respectively. one case for 10 deg with a highly deflected flap. The flow field
at each angle of attack, i.e., 5, 10, and 15 deg, is compared with
the corresponding flow field measured by Mahmood et al. 15,16
for the same single NACA 4412 aerofoil.
rectification associated with stationary hot-wire probes in revers- The effect of the leading edge slat on the aerofoil flow field is
ing flow, Bruun 12. The FHW probe was calibrated in the sta- explained with reference to the three-element aerofoil configura-
tionary mode as explained by e.g., Bruun 12 and Maddah 13. tion at angles of attack of 10, 15, 20, and 25 deg, using both a
Temperature effect and calibration drift were minimized by cali- non-deflected flap ( f 0 deg) and two deflection angles f 10
bration checks before and after test runs. and 25 deg see test matrix, Table 1.
The installation of the multi-element aerofoil in the wind tun- For the description in this paper, the flow over the aerofoils will
nel, and the principle of the mechanical implementation for the be classified as attached flow, intermittent separated flow and
bean shaped curve path used at Bradford University is illustrated small and large scale separated flow. The notation, attached flow,
in Fig. 1. Measurements were obtained from a number of points is self explanatory. Separated flow is identified as a spatially per-
on the lower part of the probe curve path during a single sweep, manent flow region usually toward the rear part of the main aero-
and for the study of the slat and multi-element aerofoil perfor- foil with reversed negative mean velocities. Intermittent sepa-
mance, 20 evenly spaced points were selected for detailed analy- rated flow is explained below.
sis. Each sweep was repeated 200 times. Using the X,Y wind Intermittent Separation. If a small positive streamwise
tunnel aligned coordinate system shown in Fig. 1, ensemble aver- and a relatively high turbulence in-
and mean velocity component U
aging was used to calculate the mean velocity components, U tensity occur near the trailing edge of the main aerofoil, then a
V , and the turbulent quantities, u 2 , v 2 , and u v . Using a plated separated flow condition might exist for short periods of time in
DANTEC X hot-wire probe, the minimum safe distance from this region. For example, with reference to the test matrix in Table
the center of the probe to the aerofoil surface was 2.5 mm. The 1 for case B5 three-element aerofoil at 25 deg and f
minimum mean velocity U min discussed in Sections 4 to 6 refers 0 deg, the acquired data for the streamwise component of the
to this probe path position. The maximum mean velocity obtained velocity, U, at a point located 90% c from the leading edge of the
within the measurement region near the front of the main aerofoil main aerofoil demonstrated that no steady reverse flow situations
is denoted by U max . The attached flow over the nose of the aero- exist. From corresponding pdf results, it was observed that about
foil was not included in this investigation. The capability of the one sixth of the total sampling size correspond to negative values
FHW for acquiring high quality experimental data enabled this of U. This can indicate the existence of an intermittent small sepa-

Table 1 Test configurations

Two-element aerofoil Three-element aerofoil


Single main
aerofoil f 0 deg f 25 deg f 0 deg f 10 deg f 25 deg
5 deg S1 A1
10 deg S2 A2 AD2 B2 C2 D2
15 deg S3 A3 B3 C3 D3
20 deg B4 C4 D4
25 deg B5 C5 D5

Journal of Fluids Engineering MARCH 2002, Vol. 124 155

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 2 Velocity profiles for a position near the trailing edge of
the aerofoil containing an intermittent unstable separation
bubble: i curve a: velocity profile when the boundary layer is
attached, and ii curve b: velocity profile when the flow is
separated. A hot-wire probe placed at position y will detect ve-
locities U 1 and U 2 corresponding to the two flow states.

ration bubble which starts to grow and remain attached for a short
period of time before it is swept downstream with the flow in the
streamwise direction, temporarily creating an attached unstable
boundary layer flow in this region. That the separation region was
of a limited axial extent was demonstrated by considering a point
at 60% c from the leading edge of the main aerofoil. The related
acquired data for the same configuration showed no negative
value for U, indicating the limited spatial extent of the intermittent
separation bubble.
Consequently, it is justifiable to suggest that the flow field var- Fig. 4 Mean velocity values for: a Single aerofoil at
ies between two extreme conditions: 1 an attached unstable 5 deg; b two-element aerofoil at 5 deg and f 0 deg
boundary layer flow and 2 a separation bubble containing re-
versed flow negative U values. These two conditions can be
illustrated by the velocity profile curves a attached and b sepa- ditions of 1 an attached unstable boundary layer flow and 2 a
rated in Fig. 2. A single hot-wire probe placed in the outer part of temporal intermittent separation bubble. This flow situation will
the boundary layer at a distance yy R from the surface where no be referred to as intermittent separation.
negative velocities occur can be used to indicate the existence of
the two flow states corresponding to curves a and b. As a single
normal hot-wire probe cannot detect negative velocities it is im- 4 Three Distinct Flow Regimes
portant that it is placed at a distance y greater than the maximum As described above the flow field over the investigated aerofoil
vertical extent y R of the reverse flow region. For such a position combinations will be described in terms of three types of flow
the average level of the hot-wire output should change related to regimes. Examples are given below.
the velocities U 1 and U
2 as shown in Fig. 3. Measurements were
carried out for a number of probe positions corresponding to this
4.1 Attached Flow
flow condition, and the expected change in the output voltage
level from E 1 to E 2 was observed as shown in Fig. 3. These and Single and Two-Element Aerofoils at 5 Deg (Cases S1 and
similar results demonstrate the existence of the two extreme con- A1) Mean Velocity. Figures 4a and 4b show the mean veloc-
ity vectors for the single aerofoil and the two-element aerofoil at
an angle of attack of 5 deg respectively. In general, the mean
velocity vectors for the single aerofoil, presented in Fig. 4a, are
similar to the corresponding velocity field over the main aerofoil
for the two-element case A1, depicted in Fig. 4b. These velocity
vector plots for both aerofoils, indicate that the flows are attached
to the model surfaces and there are no signs of separation. For
both configurations, accelerated flows over the leading edge of the
main aerofoil are responsible for creating suction and lift. The
measured maximum velocities in this region reach peak values of
1.27U and 1.30U for the single and the two-element aero-
foil configurations respectively. The velocity deficits in a small
region close to the trailing edge of the main aerofoil are slightly
larger for the single aerofoil case S1 compared to the two-
element aerofoil case A1. Comparing the vector plots for these
Fig. 3 Output signal from a single normal hot-wire probe lo-
configurations, it can be concluded that introduction of the flap
cated just outside the edge of an intermittent unstable separa- creates higher velocities over the trailing edge of the main aero-
tion bubble, demonstrating two quasi-steady flow conditions foil, thus slightly enhancing the flow field. The total lift for the
corresponding to 1 E 1 : an attached unstable boundary layer two-element aerofoil is higher than for the single aerofoil due to a
flow and 2 E 2 : an intermittent separation bubble. slightly enhanced flow field and the extra area of the flap.

156 Vol. 124, MARCH 2002 Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 5 Mean velocity values for: a Single aerofoil at Fig. 6 Mean velocity values for: a Single aerofoil at
10 deg; b two-element aerofoil at 10 deg and f 0 deg 15 deg; b two-element aerofoil at 15 deg and f 0 deg

Reynolds Stresses. The values of the Reynolds stresses for reaches 35103 , corresponding to a local turbulence intensity
these configurations S1 and A1 are small. For example, the peak Tu20%, located in a region over the trailing edge of the main
values of the normalized streamwise Reynolds stresses, (u 2 /U 2 ) aerofoil and the flap. This value is about four times higher than the
corresponding values for 5 deg. The local turbulence intensity
103 occurring over the trailing edge of the main aerofoil are of
in the low mean velocity region measured closest to the aerofoil
the order of 10103 for both the single and the two-element surface exceeds 30% which is consistent with intermittent separa-
aerofoils. This corresponds to a local turbulence intensity Tu tion. Similar trends are observed for the cross normal Reynolds
10%. The development of ( v 2 /U 2 ) and (u v /U 2 ) exhibits simi- stress v 2 and turbulent shear stress u v with peak values of 15
lar trends with smaller peak values. The small magnitudes of these 103 and 16103 , respectively. For the single aerofoil
stresses are consistent with the mean velocity results indicating
the existence of an attached boundary layer for both cases S1 and case S2, the peak values of u 2 /U 2 are observed over the trailing
A1. edge of the main aerofoil with almost the same magnitudes as the
corresponding values for the two-element aerofoil.
4.2 Intermittent Separation
4.3 Separated Flow
Single and Two-Element Aerofoils at 10 Deg (Cases S2 and
A2) Mean Velocity. The mean velocity vectors for the single 4.3a Small Scale Separation
aerofoil and the two-element aerofoil at an angle of attack of 10
deg are shown in Figs. 5a and 5b, respectively. Similar flow Single and Two-Element Aerofoil at 15 Deg (Cases S3 and
fields are observed for both cases S2 and A2 with high velocities A3) Mean Velocity. The mean velocity vector fields for the
over the top front and low velocities over the trailing edge of the single and the two-element aerofoils at an angle of attack
main aerofoil. The measured streamwise components of mean ve- 15 deg are shown in Fig. 6a and 6b. The experimental data
locity reach peak values of 1.38U and 1.40U over the front for the flow over the two-element aerofoil clearly shows a larger
of the main aerofoil for the single and the two-element aerofoil region of low velocities over the main aerofoil including reversed
respectively. The extent of the low velocity region over the trail- flow close to the surface compared to the corresponding mean
ing edge of the main aerofoil is larger for the single aerofoil velocity field for the single aerofoil case S3. The commence-
compared to that of the two-element aerofoil, but the minimum ment of flow detachment is at 50%c and 35%c from the leading
values of the measured mean velocities are as low as 10% of the edge of the main aerofoil for cases S3 and A3, respectively. The
free-stream velocity for both cases. As for the attached flow case, large velocity deficits are seen to persist downstream of the aero-
it can be concluded that the total lift for the two-element aerofoil foil, beyond a distance of a half chord length from the trailing
with a nondeflected flap is higher than for the single aerofoil. edge of the main aerofoil for the single case and twice that for the
two-element aerofoil case. For the two-element case, the flap gap
Reynolds Stresses. The highest values of the normalized flow is observed to be responsible for a limited improvement in
streamwise Reynolds stress u 2 /U 2 for the two-element case the velocity field over and beyond the flap.

Journal of Fluids Engineering MARCH 2002, Vol. 124 157

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 7 Mean velocity vectors for three-element aerofoil at
25 deg and f 25 deg

Reynolds Stresses. For the two-element aerofoil at an angle of


attack of 15 deg, the spatial distribution of the turbulence field has
been changed substantially, when compared to the attached flow
case e.g., case A1 and to the flow with a small intermittent
separation e.g., case A2. The separation bubble causes the free
shear layer surrounding the separation bubble to move a sub-
stantial distance in the cross-stream direction away from the two-
element aerofoil surface. Consequently, the viscous-inviscid inter-
action region and the locations of the maximum stresses have also
changed. The normalized stress u 2 /U 2 reaches a peak value of
180103 corresponding to a local turbulence intensity Tu of
70%, while the normalized v 2 and u v have increased to 70
103 and 55103 , respectively. These results are consis- Fig. 8 Three-element aerofoil at 25 deg and f 25 deg: a
tent with a region of separated flow. Normalized streamwise normal Reynolds stress u 2 U
2
103 ;
Similar trends are observed for the peak values of the Reynolds b normalized cross-stream normal Reynolds stress v 2 U
2

stresses for the single aerofoil at the same angle of attack of 15 103
deg. However, the extent of the separated region is smaller and the
peak locations are closer to the aerofoil surface, which is consis-
tent with the corresponding mean velocity results. the free shear layer above the rear part of the flap and wake. These
high values represent vertical oscillations of the free shear layer.
4.3b Large Scale Separation Stall
Main Aerofoil at 25 Deg Slat at s15 Deg, Flap at 5 Three-Element Aerofoil Configurations
f25 Deg (Case D5) Mean Velocity. The velocity vectors for
This section presents the experimental results for the mean ve-
the three-element aerofoil at 25 deg with a flap deflection
locity U and u 2 Reynolds stress for the three-element aerofoil
angle set to 25 deg are shown in Fig. 7. Almost 80% of the main
aerofoil and the whole flap are stalled. The large region of recir- placed with a slat at s 15 deg and the main aerofoil at angles of
culating flow over the main aerofoil and the flap is an indication attack of 10, 15, 20, 25 deg. Section 5.1 contains results for the
of a drastic loss of lift. As shown in Section 5.1b for an intermit- nondeflected flap ( f 0 deg) while results for the deflected flap
tent separated flow case B5 over the three-element aerofoil, the at f 25 deg are described in Section 5.2. For the related at-
mean velocity U reaches values as high as 1.65U . However, tached and intermittent separated flow cases, the peak values of
for the flow with large scale separation case D5, the accelerated the other two measured Reynolds stresses v 2 and u v occurred at
flow over the leading edge of the main aerofoil has been reduced the same locations as u 2 max , thus only the results for u 2 are pre-
and the maximum measured mean velocity has dropped to sented. For the separated flow case, e.g., the three-element aero-
1.40U , which indicates a loss of lift over the leading edge of foil at an angle of attack of 25 deg with a highly deflected flap
the main aerofoil. ( f 25 deg), as discussed in section 4.3b, the peak locations of
the Reynolds stresses are different, hence results were presented
Reynolds Stresses. The normalized u 2 and v 2 contour plots for u 2 and v 2 in Figs. 8a and 8b.
for the three-element aerofoil at 25 deg with a highly deflected
flap are shown in Figs. 8a and 8b. It was found that except for 5.1 Nondeflected Flap, f 0 Deg
a spatially much larger separated flow stall, and slightly larger
Reynolds stresses the turbulence structure is similar to the small 5.1a Attached Flow Cases
scale separated flow observed on the two-element aerofoil at
15 deg with a nondeflected flap, case A3. The normalized Main Aerofoil at 10, 15, and 20 Deg With a Slat and a Flap
(Cases B2, B3, and B4) Mean Velocity. The vector plot of the
streamwise Reynolds stress u 2 /U 2 reaches a peak value of mean velocity for the three-element aerofoil at angles of attack of
220103 and it occurs towards the front end of the aerofoil. 10 deg case B2, 15 deg case B3 and 20 deg case B4 are
This large value represents violent axial oscillations of the front shown in Fig. 9a11a. For all three configurations, the flow is
part of the separation bubble. The development of v 2 , depicted in attached. When the angle of attack is increased the maximum
Fig. 8b has a maximum value, 80103 and it is located in measured velocity over the front of the main aerofoil reaches val-

158 Vol. 124, MARCH 2002 Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 9 Three-element aerofoil at 10 deg and f 0 deg: a Fig. 10 Three-element aerofoil at 15 deg and f 0 deg: a
Mean velocity vectors; b normalized streamwise normal Rey- Mean velocity vectors; b normalized streamwise normal Rey-
nolds stress u 2 U
2
103 nolds stress u 2 U
2
103

ues of 1.41U , 1.47U and 1.55U for configurations B2, nondeflected flap, case B3. At this angle of attack the single and
B3, and B4, respectively. The following additional features for two-element aerofoils are stalled, while an enhanced lift coeffi-
each of these three cases are discussed below: cient is observed for the three-element aerofoil.
10 deg. The introduction of a deflected advanced slat for
the three-element aerofoil at 10 deg, is observed, in general, Reynolds Stresses. The contour plot for u 2 for the three-
to have only a small effect on the flow field compared to the element aerofoil at angles of attack of 10 deg case B2, 15
corresponding two-element flow case A2, except over the trailing deg case B3 and 20 deg case B4 are shown in Figs. 9b
edge of the main aerofoil Fig. 9a. In this region, small velocity 11b. The normalized streamwise normal Reynolds stress reaches
vectors are found to occur for case A2, as depicted in Fig. 5b, peak values of 9103 , 12103 and 20103 over the
while the corresponding velocity vectors for the three-element trailing edge of the main aerofoil and the flap for cases B2, B3,
aerofoil case B2 have improved to as high as 50% of the free- and B4, respectively. Compared to the results without the slat, the
stream velocity. However, overall the lift coefficients for the two- peak value for u 2 has decreased significantly from 35103 at
and three-element aerofoils are very similar as discussed in Sec- 10 deg and 180103 at 15 deg. Similar trends were also ob-
tion 6. served from contour plots of v 2 and u v . These slatted results
15 deg. The introduction of a deflected slat at 15 deg, is indicate a thinner attached boundary layer which is consistent
seen to enhance the whole flow field compared to the two-element with narrower wakes as observed from the velocity vectors plots,
case A3 and to re-establish large velocity vectors over the multi- in Figs. 9a11a.
element aerofoil and in the wake, Fig. 10a. The difference in the
streamwise component of the mean velocity for this case and for 5.1b Intermittent Separation. As the angle of attack is
with-slat U
the one without the slat, U without-slat , is shown in Fig. increased further, the extent of the low velocity region increases
component of the leading to an intermittent separated flow condition.
12. It is observed that the slat causes the U
mean velocity at the rear of the main aerofoil to recover up to Main Aerofoil at 25 Deg With a Slat and a Flap (Case B5)
75% of the free stream velocity with a resulting thin shear layer Mean Velocity. The mean velocity vector plot for the three-
and a narrow wake. element aerofoil at 25 deg is shown in Fig. 13a. Increasing
20 deg. The three-element aerofoil at 20 deg is more the angle of attack to 25 deg is seen to have deteriorated the flow
heavily loaded and consequently, the velocity deficit over the trail- field over the trailing edge of the main aerofoil resulting in a
ing edge of the main aerofoil and over the flap see Fig. 11a is moderate intermittent separated flow region. However, due to the
larger than that of the three-element aerofoil at 15 deg with a general enhancement of the flow field a higher lift is achieved.

Journal of Fluids Engineering MARCH 2002, Vol. 124 159

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 11 Three-element aerofoil at 20 deg and f 0 deg: a Fig. 13 Three-element aerofoil at 25 deg and f 0 deg: a
Mean velocity vectors; b normalized streamwise normal Rey- Mean velocity vectors; b normalized streamwise normal Rey-
nolds stress u 2 U
2
103 nolds stress u 2 U
2
103

The measured mean velocity over the front of the main aerofoil ized stress u 2 /U 2 reaches a peak value of 130103 , corre-
reaches a maximum of 1.65U which is 7% higher than the sponding to a local turbulence intensity Tu55%, which is ap-
corresponding value for 20 deg, case B4. proximately three times higher than the corresponding value for
Reynolds Stresses. Figure 13b shows the normalized Rey- the three-element aerofoil at 20 deg case B4. Similar trends
nolds stresses u 2 for the three-element aerofoil at 25 deg with were also observed for v 2 and u v . The observed trends in the
a nondeflected flap. The presence of intermittent separation and mean velocity and Reynolds stresses for the three-element aerofoil
small mean velocities over the rear part of the main aerofoil are associated with the change of the flow regime from attached flow
consistent with high values for the Reynolds stresses. The normal- ( 20 deg) to intermittent separation ( 25 deg) are similar to
the changes that occurred for the two-element aerofoil at
5 deg attached flow, case A1 and 10 deg intermittent
separation, case A2.

5.2 Deflected Flap, f 25 Deg


Main Aerofoil at 10, 15, and 20 Deg With a Slat and a Flap
(Cases D2, D3, and D4) Mean Velocity. For the three-element
aerofoil at angles of attack of 10 deg case D2, 15 deg case
D3 and 20 deg case D4 with a highly deflected flap ( f
25 deg), the flow is attached for cases D2 shown in Fig. 14a
and D3, and is intermittently separated for case D4. Compared to
the nondeflected flap cases B2, B3, and B4, setting the flap
deflection to 25 deg creates larger velocities over the main aero-
foil and the flap, resulting in lift enhancement. Consequently, the
measured mean velocity over the front of the main aerofoil
reaches larger peak values of 1.5U , 1.55U , and 1.63U
for configurations D2, D3, and D4, respectively, compared with
the corresponding results for the nondeflected flap described in
Section 5.1a. The measured minimum velocity over the trailing
Fig. 12 The difference between streamwise mean velocity edge of the main aerofoil for the deflected flap cases also exhib-
components, U with - slat U
without - slat , for multi-element aerofoil ited higher values than those observed for the corresponding non-
at 15 deg and f 0 deg deflected flap cases.

160 Vol. 124, MARCH 2002 Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


u 2 has increased significantly for the three-element aerofoil at 20
deg due to the change of the flow regime from attached case B4
to intermittently separated flow case D4. Similar trends were
also observed from contour plots of v 2 and u v . The results for
10 deg are compared with the following investigation:
10 deg. A comparison of the u 2 results shown in Fig. 14b
with the comparative data reported by Nakayama et al. 6 dem-
onstrates similar trends regarding the spatial distribution of the
turbulence intensity and the width of the shear layer. However, the
peak values of the Reynolds stresses in their experiment are ap-
proximately 60% higher than the corresponding values in the
present research. This is most likely caused by the heel effect of
the conventional slat used.

6 Summary of Aerofoil Measurements


As has been shown, the measured mean and fluctuating velocity
fields have clearly identified three types of flow fields over the
multi-element aerofoil, i.e., attached flow, intermittent separation
over parts of the aerofoil/flap and separated flow including stall.
The following features are particularly helpful in describing the
observed flow fields:
Mean Velocities

i The value of the maximum mean velocity U


max obtained
within the measurement region near the front of the aero-
foil.
ii The measured minimum velocity U
min close to the surface
of the trailing edge of the main aerofoil, as defined in
section 2.
An increase in both values will correspond to enhanced lift
conditions. In contrast, low values at the rear of the aerofoil would
Fig. 14 Three-element aerofoil at 10 deg and f 25 deg: indicate adverse flow conditions such as intermittent separation.
a Mean velocity vectors; b normalized streamwise normal The corresponding measured values for U max and U
min are plotted
Reynolds stress u 2 U
2
103 in Figs. 15a and 15b, respectively, as functions of angle of
attack and flap deflection angle f .
For the front of the aerofoil, the results for U max corresponding
The results for 10 and 15 deg are compared with the fol- to f 0 deg in Fig. 15a, show that the data for the two-element
lowing two investigations: A1 and A2 and the three-element aerofoil B2 to B5 lay on the
10 deg. The mean velocity field over a three-element aero- same straight line, demonstrating lift enhancement as a function of
foil incorporating a conventional slat has been reported by Na- due to the increase in U on the front part of the main aerofoil.
kayama et al. 6 for 10 deg using a smaller flap deflection For the same flap setting, at the rear of the aerofoil, there is a
angle, f 15 deg. Their results demonstrated a similar trend as major difference between the two-element and the three-element
for case D2 shown in Fig. 14a in both the spatial velocity results see Fig. 15b. The two-element results show a rapid
distribution and the wake width. deterioration in the flow field with , resulting in a separated flow
16 deg. The mean velocity field over a three-element aero- region when 15 deg. In contrast, for the three-element aerofoil
foil reported by Braden et al. 7 at an angle of attack min 6 m/s are maintained for
substantial positive values for U
16 deg incorporating a conventional slat with a higher flap de-
flection angle, f 30 deg, demonstrated a similar overall trend 10 20 deg, and even at 25 deg, U min is as high as 2 m/s.
as for case D3 in the flow field. However, the higher flap deflec- Consequently, for f 0 deg the lift will increase with over the
tion angle and the conventional slat caused a small separated re- complete angle range listed. Compared with a single aerofoil, the
gion commencing at the trailing edge of the main aerofoil and flap has the advantage of providing an additional surface area for
extending downstream into the wake and this resulted in a wider creating lift and the utilization of a slat ensures that both the main
wake. aerofoil and the flap contribute positively to the lift by maintain-
20 deg. Although, for the three-element aerofoil at ing an attached boundary layer over both.
20 deg, with a highly deflected flap, higher velocities were ob- For a flap deflection angle f 25 deg, the U max data in Fig.
served over the trailing edge of the main aerofoil compared to the 15a demonstrate an enhancement in the velocity field in the
corresponding nondeflected flap case, the direction of the flow range 1020 deg. However for 25 deg, we notice a rapid de-
indicated that the flow did not follow the surface in this region. crease in U max and hence in the lift performance. This is consis-
This boundary layer behavior is indicative of an intermittent flow tent with the minimum velocity U min at the rear of the aerofoil,
separation. It was also observed that the flow does not follow the Fig. 15b, having values as high as 9 m/s for 10 17.5 deg
surface of the flap. and poor separated flow conditions for 25 deg. Consequently,
the aerofoil will experience an enhanced lift in the range 10 to
Reynolds Stresses. For the three-element aerofoil at angles of 20 deg and a reduction in lift at 25 deg due to virtually stall
attack of 10 deg case D2, 15 deg case D3 and 20 deg case conditions.
D4 with a highly deflected flap ( f 25 deg), the normalized
streamwise normal Reynolds stress u 2 /U 2 reaches peak values of Turbulent Quantities. The maximum value for u 2 plotted as
13103 , 16103 , and 60103 , respectively, over the (u 2 /U 2 )103 can be used to describe the occurring turbulent
trailing edge of the main aerofoil and the flap. The peak value for flow fields. From this value the corresponding local turbulence

Journal of Fluids Engineering MARCH 2002, Vol. 124 161

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 16 Representation of the flap and slat by vortices: a
Flap representation; b slat representation

The change in the flow conditions and related turbulence field


2
can be summarized using a plot of (u max /U2 )103 as a function of
the angle of attack for four aerofoil conditions: two-element
aerofoil and three multi-element aerofoil conditions with f 0,
10, and 25 deg. These results are shown in Fig. 15c.
From the Turbulence Results in Fig. 15(c), it was found in gen-
eral:

i When u max 2
/U2 40103 , the corresponding local turbu-
lence intensity Tu30% and the flow over the aerofoil would be
attached. It is observed that these conditions apply to the configu-
rations A1, B2-B4, C2-C4, and D2-D3.
ii When u max2
/U2 (40 60)103 , then small scale inter-
mittent separation occurred flow cases S2, A2 and D4. This was
observed to correspond to a relatively small intermittent separa-
tion region towards the trailing edge of the main aerofoil. As the
2
value of u max increases, the extent of the intermittent separation
region grows, the longitudinal oscillations become more violent
and u 2 /U 2 reaches values of 130140 cases B5 and C5.
iii When the value of normalized u max2
reaches values of 180
103 case A4, then the flow is separated with a permanent
2
region of reversed flow. For u max values larger than this case D5,
the main aerofoil is, for all practical purposes, stalled.

7 General Discussion and Conclusion


The observed flow fields described in this paper can be related
to the interaction between the slat/main aerofoil and the main
aerofoil/flap and the consequent variations in the lift and drag. In
aerofoil theory it has been common practice to describe the circu-
lation and related lift of an aerofoil according to vortex theory. It
is therefore useful to replace the flap and slat with vortices as
shown in Figs. 16a and 16b and to consider the change in the
circulation and hence in the velocity fields for the interacting slat/
main aerofoil or main aerofoil/flap. Using a fixed aerofoil configu-
ration and a fixed angle of attack , Smith 17 has used this
Fig. 15 a Maximum mean velocity over the top front of the vortex concept to explain the upstream/downstream element inter-
main aerofoil for four aerofoil configurations: Three-element action. However no explanation, based on vortices, has been pre-
aerofoil, 10, 15, 20, 25 deg: series 1: , f 0 deg; series 2: sented for the variation of this interaction with and f .
, f 10 deg, series 3: , f 25 deg. Series 4: , Two-element
aerofoil, 5, 10, 15, and f 0 deg. b Minimum mean veloc-
First, the effect of a flap on the flow field over the main aerofoil
ity over the trailing edge of the main aerofoil for four aerofoil will be considered. Then the effect of the flap deflection angle is
configurations: legend as in Fig. 15a. c u 2 max U
2
as a
ascertained. Finally, an explanation for the slat effect is provided.
function of angle of attack for four aerofoil configurations: leg- Flap. The simulation of the flap by a clockwise vortex as
end as in Fig. 15a shown in Fig. 16a, will create a stronger circulation over the
main aerofoil which will attempt to create higher velocities over
its top surface. The following observations can be made from this
intensity Tuu /U
where u (u 2 ) 1/2 and U
is the local mean
figure:
velocity could be calculated. If the turbulent flow field has a
Gaussian distribution, a value of Tu30% would indicate inter- 1. Using a flap will attempt to increase the velocity over the
mittent flow separation. whole of the upper surface of the main aerofoil;

162 Vol. 124, MARCH 2002 Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


2 The increase in the velocity is largest over the leading edge
of the main aerofoil.
For attached flow, the first trend is consistent with our compara-
tive experimental results for the mean velocity field over a single
aerofoil and two-element aerofoil with a nondeflected flap. For
example, this effect can be observed from a comparison of the
flow field over the single aerofoil case S1 and over the two-
element aerofoil case A1 at an angle of attack of 5 deg. This
demonstrates an enhanced flow field for the two-element case.
The second effect causes the front of the main aerofoil to be
more heavily loaded. This effect works against the attachment of
the flow, and causes the pressure distribution over the main aero-
foil to deteriorate in terms of a more unfavorable pressure gradient
dp/dx. As is increased and the aerofoil becomes more heavily
loaded higher velocities on the front, the unfavourable pressure
gradient will increase faster than for the single aerofoil, resulting
in an earlier separation at the rear of the aerofoil. This is demon-
strated by the results for 15 deg.
Flap Deflection. The above concepts also apply to the de- Fig. 17 Pressure distribution for aerofoils at 10 deg: a
flected flap. Provided that the flow remains attached then the sec- Aerofoil with retracted slat and flap; b three-element aerofoil
ond effect will be dominant when the flap is deflected. The de- with extended and deflected slat and flap from Savory et al.
flection of the flap increases the strength of the assumed vortex, 8
thereby creating higher velocities particularly at the front of the
aerofoil, a general enhancement of the flow field and an increase
in the lift.
On the other hand, if prior to the flap being deflected, a de- of the aerofoil is much smoother for the slatted aerofoil,
tached boundary layer or an intermittent separated flow, however causing a much more favorable boundary layer on the top
small in size exists, then the case of the deflected flap will have a front of the aerofoil.
worse flow field than the situation with a nondeflected flap. The 3 As a consequence, the pressure gradient dp/dx is reduced
already unfavorable pressure distribution over the top surface of considerably for the slatted aerofoil, enabling attached flow
the main aerofoil will deteriorate further when the flap is de- to be maintained for much larger angles of attack than for the
flected, causing a larger low velocity/reverse flow region. For ex- single aerofoil.
ample, the intermittently separated flow field for the three-element This is consistent with the measured higher mean velocities in
aerofoil at 25 deg with a nondeflected flap case B5 will de- the rear part of the aerofoil for the slatted case and it can be
teriorate by setting the flap deflection angle to 25 deg case D5, clearly observed from the comparison of the mean velocity results
thus causing a large scale separated flow stall and a loss of lift. for the two- and three-element aerofoil at an angle of attack of
Therefore, as long as the flow can cope with the change in the 10 deg, shown in Figs. 5b and 9a, respectively.
pressure distribution over the aerofoil surface, the flow will re-
main attached and the deflected flap will enhance the flow field. Lift and Drag. In the current experiment it was not possible to
When the flow over the main aerofoil starts to deteriorate due to directly measure the lift and drag and thereby evaluate the corre-
the stronger adverse pressure gradient, then a pressure modifier is sponding lift and drag coefficients C l and C d .
needed to restore the pressure distribution, particularly over the However, there is sufficient evidence to demonstrate that, pro-
nose of the main aerofoil and to smooth the pressure rise in the vided the flow is attached, then current CFD codes will produce
streamwise direction. This pressure modifier can be a slat and its adequate solutions of the mean velocity field and pressure distri-
main effect is explained in the following section. bution. By integrating the pressure distribution the corresponding
C l and C d coefficients can be derived.
Slat. As pointed out by Smith 17, a common misconception As described in Maddah 13 such a numerical calculation was
of the working principle of the slat is that it blows high energy carried out using the CFX code, for two-element and three-
flow into the boundary layer. However, to a large extent the slat element aerofoils to demonstrate the variation in C l and C d with
has the opposite effect. The effect of the slat can be demonstrated in the range 10 to 25 deg and f 0, 10, and 25 deg. The experi-
by simulating it using a clockwise vortex as shown in Fig. 16b. mental results in this investigation, have shown that the flow is
The resulting pressure distribution over the main aerofoil and the attached for cases B2-B4 and D2-D3, and intermittently separated
related lift are functions of the strength of this vortex which is for cases B5 and D4. Separated flow was identified for case D5
itself a function of: Slat deflection angle, slat gap, slat overlap and and hence the numerical prediction for this case is significantly
slat shape. From Fig. 16b it is observed that the direction of the less accurate.
circulation over the slat as represented by the vortex and over Prior to stall occurring for any of the investigated aerofoil con-
the main aerofoil are in opposite directions in the gap region be- figurations ( 15 deg), the lift for the single- and the two- and
tween the slat and the main aerofoil. This causes lower velocities three-element aerofoils with a nondeflected flap is almost the
over the leading edge of the main aerofoil and hence less suction same. This can be explained by considering the interaction of
and a consequent reduction in local lift. The effect of the slat on upstream and downstream elements. The circulation increases on
the pressure distribution is shown in Fig. 17 for an aerofoil at the upstream element, while the circulation decreases on the
10 deg with and without a deflected slat, Savory et al. 8. This downstream element. Thus, adding a nondeflected flap increases
figures demonstrates three effects: the circulation over the main aerofoil, while the circulation around
the flap is decreased, thus, the overall effect is a slight increase in
1 That the slat causes a substantial reduction in the pressure C l . In addition, there is a considerable increase in the lift due to
suction and hence in the velocities over the nose of the aero- the extra area provided by the flap. Adding a slat to the single
foil. aerofoil, results in higher circulation over the slat and the unload-
2 More importantly, that the pressure recovery on the top front ing of the front of the main aerofoil. The overall effect is a slight

Journal of Fluids Engineering MARCH 2002, Vol. 124 163

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Nomenclature
A1 to
A3 two-element aerofoil configurations with a non-
deflected flap
B2 to
B5 three-element aerofoil configurations with a non-
deflected flap
c chord length of the main aerofoil
C d drag coefficient
C l lift coefficient
C p pressure coefficient
D2 to
D5 three-element aerofoil configurations with a highly
deflected flap
dp/dx streamwise pressure gradient
E 1 ,E 2 output voltage of anemometer
S1 to
S3 single aerofoil configurations
streamwise mean velocity component
U
U max maximum measured streamwise mean velocity com-

ponent
min minimum measured streamwise mean velocity com-
U
ponent
u 2 /U
Tu turbulence intensity defined as u /U
U mean free stream velocity
u 2 streamwise normal Reynolds stress
u v Reynolds shear stress
v 2 cross-stream normal Reynolds stress
y cross-stream distance
y R cross-stream extent of reversed flow region
Fig. 18 Numerical prediction of: a Coefficient of lift C l for angle of attack
three aerofoil configurations: Three-element aerofoil, 10, 15, f flap deflection angle
20, 25 deg: series 1: , f 0 deg; series 2: , f 25 deg; se- s slat deflection angle
ries 3: , Two-element aerofoil, 5, 10, 15, and f 0 deg; b
coefficient of drag C d for three aerofoil configurations: legend Appendix 1-Experimental Uncertainty
as in a
The FHW traveling mechanism enabled the location of the lon-
gitudinal and vertical positions with an accuracy of 0.1 mm.
Overall, taking the stiffness of the FHW system into account the
decrease in C l . For the three-element aerofoil, the overall effect accuracy of the absolute spatial resolution of each point was esti-
on the main aerofoil is a balance between the increase and de- mated to be within 0.25 mm.
crease in circulation due to the flap and slat, respectively, resulting The output from a hot-wire sensor is a time varying signal typi-
in values for C l similar to those observed with the single aerofoil. cally representing a velocity component. For a turbulent flow the
Figure 18a shows that for the nondeflected flap case, C l in- anemometer signal will be of a random nature. When using digital
creases linearly to 25 deg, and the experimental results con- data analysis, the continuous signal is replaced by a digital finite
firmed that the flow remained attached for all values of in this time record. Thus, errors are introduced by a time series analysis
range. The delay in the stall angle is caused by the presence of the of a finite time history record.
slat producing a benificial pressure modification on the main aero- The uncertainty in the evaluated velocity quantities U and u 2
foil. The variation of C l for the three-element aerofoil with a can be estimated as
deflected flap, is similar to the three-element aerofoil with a non-
deflected flap, but with a shift in C l at each angle of attack due to C
U
the strong circulation effect of the deflected flap on the main aero- C
1Z /2 U
foil. However, as the aerofoil is more heavily loaded, the flow
U
field was observed to have deteriorated, with stall occurring at
25 deg and a resulting loss in lift. and
The drag for a multi-element aerofoil is different from the drag
uC2
for a single element aerofoil. In multi-element aerofoil flows the 1Z /2 C
u2
flap and slat are usually extended and often deflected. The de-
U
flected leading and trailing edge devices cause a considerable in-
crease in the drag. Consequently, the resulting drag for a multi- where Z /2 is the standardized variable used in the Gaussian prob-
element aerofoil, even at low angles of attack, will be higher than ability distribution described by Bruun 12. For N independent
the corresponding value for a single element aerofoil as has been samples, we have
reported by several researchers. Deflecting the flap will increase
C d as shown in Fig. 18b. 1 u 2 1/2 1
C
U and C
u2
N
U N
Acknowledgments In the present study, the FHW data was evaluated for N200
The principle author wishes to acknowledge a study grant from statistically independent sweeps with a 10 s dormant period be-
the government of the Islamic Republic of Iran. tween each sweep.

164 Vol. 124, MARCH 2002 Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use


For all flow cases a large proportion of the flow field has low/ 5 Biber, K., and Zumwalt, G. W., 1993, Flowfield measurements of a two-
element airfoil with large separation, AIAA J., 31, No. 3, pp. 459 464.
moderate turbulence intensity. For illustrative purposes consider
6 Nakayama, A., Kreplin, H. P., and Morgan, H. L., 1990, Experimental inves-
flow regimes with a maximum turbulence intensity of 20% this is tigation of flowfield about a multielement airfoil, AIAA J., 28, No. 1, pp.
the case everywhere except near or inside a separated flow re- 14 21.
gion. 7 Braden, J. A., Whipkey, R. R., Jones, G. S., and Lilley, D. E., 1986, Experi-
Specifying Z /22.33 for a 98% confidence level, it therefore mental study of the separating confluent boundary layer, NASA CR-3655.
also AIAA paper 86-0505.
follows that 8 Savory, E., Toy, N., Tahouri, and Dalley, S., 1992, Flow regimes in the cove
C
U 1
regions between a slat and wing and between a wing and a flap of a multiele-
ment aerofoil, Exp. Therm. Fluid Sci., 5, pp. 307316.
12.33 0.213%

U 200 9 Alemdaroglu, N., 1993, Experimental investigation of flow around a multi-
element airfoil, AGARD-CP-515.
10 Maddah, S. R., Gough, T., and Bruun, H. H., 2001, Investigation of slat heel
C
u2 1 effect on flow field over multi-element aerofoils, Proc. ASME FEDSM 2001
12.33 116% Conference, New Orleans, USA, session F-252-03.
u 2
200 11 Jones, J. B. C., 1993, An optimised slat to maximise performance in both
aircraft take-off and landing, PhD thesis, University of Hertfordshire, UK.
12 Bruun, H. H., 1995, Hot-wire anemometry, Oxford University Press, Oxford
References London.
1 Innes, F., Pearcey, H. H., and Sykes, D. M., 1995, Improvements in the 13 Maddah, S. R., 2000, Study of the flow field over the multi-element aero-
performance of a three element high lift system by the application of airjet foils, PhD thesis, Dept. of Mech. and Med. Engineering, Univ. of Bradford.
vortex generators, Proc. Conference High Lift and Separation Control, Royal 14 Coles, D., and Wadcock, A. J., 1979, Flying hot-wire study of flow past an
Aeronautical Society, University of Bath, UK, pp 25.125.11. NACA 4412 airfoil at maximum lift, AIAA J., 17, No. 4, pp. 321329.
2 Moens, F., and Capbern, P., 1995, Design and testing of leading-edge high- 15 Mahmood, Z., Khan, M. K., Seale, W. J., and Bruun, H. H., 1995a, Compari-
lift device for laminar flow wing applications, Proc. Conference on High Lift son of measured and computed velocity fields over a high lift aerofoil, Proc.
and Separation Control, Royal Aeronautical Society, University of Bath, UK, Seventh International Conference on computational methods and experimental
pp. 7.17.13. measurements, Computational Mechanics Publications, Southampton, UK, Ca-
3 Seetharam, H. C., and Wentz, W. H. J., 1977, A low speed two-dimensional pri, pp 203211.
study of flow separation on the GAW-1 airfoil with 30-percent chord Fowler 16 Mahmood, Z., Khan, M. K., and Bruun, H. H., 1995b, Flow over high lift
flap, NASA CR-2844. multiple aerofoils, Proc. of Conference on High Lift and Separation Control,
4 Olson, L. E., and Orlow, K. L., 1981, On the structure of turbulent wakes and Royal Aeronautical Society.
merging shear layers of multielement aerofoils, AIAA paper 81-1238. 17 Smith, A. M. O., 1974, High lift aerodynamics, AIAA paper 74-039.

Journal of Fluids Engineering MARCH 2002, Vol. 124 165

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/14/2015 Terms of Use: http://www.asme.org/about-asme/terms-of-use

También podría gustarte