ftcA- -r-3 31
Liu U I -
NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS
TECHNICAL MEMORANDUM 1331
INVESTIGATIONS OF THE BOUNDARY-LAYER CONTROL ON A FULL
SCALE SWEPT WING WITH AIR BLED OFF
FROM THE TURBOJET*
By P. Rebuffet and Ph. Poisson-Quinton
SThe following account reviews the various stages of a research
program relative to the high-lift devices on a swept wing by combined
suction and blowing (jet action), with ejectors fed by air bled off
(extracted) from the turbojet.
After reviewing the essential principles of the boundary-layer
.control obtained by comparison with theory, the electric analogies and
the wind-tunnel tests as well as the essential elements of ejector
,operations, the writers describe the tests made in the large tunnel at
Chalais-Meudon on a full-scale model of the SO 6020 wing.
They comment on some results relative to take-off and landing and
examine the airplane-turbojet energy balance.
The writers emphasize that this investigation has been carried out
in a remarkable spirit of teamwork and particularly their chief collab-
orators Messrs. Mirande and Ravailhe, Jousserandot and Chevallier.
They have also drawn on the unpublished reports of Messrs. Lebrun,
Legras, Nieviaski, Pontezitre of the O.N.E.R.A. and of Mr. Chiaffiote
of the Hispano-Suiza Company.
S reference area
V0 speed at infinity
*"Recherches sur l'Hypersustentation d'une Aile en Fleche R6elle par
Control de la Couche Limite Utilisant le Prelevement d'Air sur le
Turbo-Reacteur." la Recherche Aeronautique, O.N.E.R.A., No. 14, March-
April, 1950, pp. 39-54.
NACA TM 1331
width of blowing slot
flow volume, flow mass, and flow weight
Cq', Cq", Cq
mean speed of suction flow tube coefficient of flow
Cq = V- (expanded to 150 C, 760 mm of mercury)
flow coefficient of infection suction and blowing
flow coefficient of injection suction and blowing
momentum flow coefficient
area of mixer cross section Sm
area of injector cross section S'
area of diffuser outlet section S
area of mixer section Sm
L length of mixer (circular ejector)
D diameter of mixer
mass flow of suction fluid q m
mass flow of engine fluid qm
momentum flow at blowing slot qm V
S momentum flow at ejector q', V'
e flow of kinetic energy at blowing slot
flow of kinetic energy of engine jet
The airfoils suitable at sonic speeds are generally characterized
by a low Czma which is still further reduced by the sweptback form
of the wing.
qv, qm' qp
NACA TM 1331
These two factors led to the study of high-lift devices designed
to develop high Cz on modern high-speed aircraft, to improve the
landing as well as the take-off.
The control of the boundary layer responsible for the separation
of flow and the limitation of Cz has been studied for a long time on
thick airfoils, without combining the suction or the jet action with a
modification of the airfoil camber. The power involved and the flows
(suction or blowing) were considerable.
Numerous technical studies have been made especially by the
Germans (ref. 1) during the war, on more suitable airfoils and with
partial-span flaps; some were flight-tested.
In France, certain airplane firms included in their research
program on high-lift devices, the study of suction or jet action, and
it was proposed to combine the two methods by utilizing injectors.
The S.N.C.A.S.O. had considered, since 1946, the construction of a
scale model of the large tunnel at Chalais-Meudon, for the SO 6008 bis
The jet airplane affords, moreover, a new possibility of boundary-
layer control by suction or by jet action (blowing).
The first solution has been the subject of several studies
refss. 2, 3), the second, more recently, was tested by the Hispano-
The aerodynamical department of the O.N.E.R.A. has, since 1946,
correlated the research data on boundary-layer control, checked the
wind-tunnel data against those obtained in potential flow by electric
analogy, and made a theoretical study of the effect of sinks (suction).
These two comparisons made it possible to orientate the experimental
investigations, since the purpose of the boundary-layer control is to
assure a potential flow under certain provisions to be specified later
These basic researches had to be made on a Chalais-Meudon mock-up
Operation at a Reynolds number near that at landing, the mock-up
having the dimensions of a flying airplane
Testing of components at full scale
Allowance for certain number of contingencies imposed by a product
similar to that of a real airplane
NACA TM 1331
It was advisable to make this study on a concrete case. The
earlier projects of the S.N.C.A.S.O. were compatible with the program
of the O.N.E.R.A., the control devices were applied to a swept wing
model of the SO 6020 type. Besides, bleeding off air from the turbojet
made it possible to reproduce at Chalais-Meudon, the image of the air-
plane, by dissociating first the airplane from the turbojet, which was
installed outside of the tunnel, thus, assuming a safer operation as
well as a more accurate way of recording the drag, that was to be
determined simultaneously with the lift.
It was, nevertheless, possible to measure the thrust of the jet,
with due regard to the air bleeding from the turbojet.
The study at Chalais-Meudon had to be made in competition with the
The S.T.A., engine section which had put the Nene Hispano-Suiza
turbojet at the disposal of the O.N.E.R.A.
The Arsenal de 1'Aeronautique, which had lent the turbojet support
frame and the corresponding control cabin
The Hispano-Suiza Company that constructed the collector for the
air bleed and assumed the control of the turbojet for the dura-
tion of the tests
The S.N.C.A.S.0. which, manufactured the model at the request of
the O.N.E.R.A. and made the study and perfected it
Lastly, the airplane section of the S.T.A. which underwrote the
About 45 to 50 men were involved in this undertaking, which
514 polars and pitching moments
2000 hinge moments
31500 pressure measurements on the wing (700 pressure gage
speed and temperature measurements
NACA TM 1331
PRINCIPLES AND APPLICATION OF BOUNDARY-LAYER CONTROL
BY HIGH-LIFT DEVICES
Of the multitude of systematic studies by the O.N.E.R.A. on this
subject (ref. 4) simple guiding ideas can be enumerated:
It is known that the maximum lift of an airfoil is limited by the
break away of the boundary layer from the upper surface. This separa-
tion has its origin in the appearance of a pressure gradient and is
accentuated with the incidence or with the deflection of a flap. The
remedy is to reduce the peaks of the speed increase by acting on the
curvature of the center line of the profile or else to prevent or limit
this separation by controlling the boundary layer.
In the first method, the fluid is assumed perfect and yields an
incurved profile center line, in a kind of downward deflection of
several flaps, hinged front and rear.
The Pe6rs-Malavard electric analogy method affords a very fruitful
study of the best distribution of the increase of speed over the upper
surface of an airfoil; for a given lift coefficient, it is possible to
increase the camber of the mean line for distributing the peaks of the
speed increase accompanying each variation incurvative due to the flap
The most dangerous increase, which appears at the leading edge can
be reduced by a deflection of the latter which puts the stagnation point
back at the nose of the airfoil. This deflection must be such that the
increase at the nose and the droop are of the same order of magnitude.
In viscous fluid the maximum lift is limited by the separation
downstream from the peaks; it can be remedied by boundary-layer control
either by suction or blowing. Both methods have a secondary effect on
the circulation around the airfoil: source effect for suction which is
theoretically calculable, or the induction effect for blowing. It has
been shown that the lift increase was proportional, respectively, to
the induced flow and to the momentum flow. This secondary effect,
negligible as long as the fluid is separated, remains apparent only
after elimination of the separation. In other words, plotting the
lift increase ZCz against the flow parameter Cq or the momentum
C,, yields a curve with asymptotic direction from the critical flow
assuming the readherence of the boundary layer. Above this value, the
profile can be regarded as functioning in perfect fluid for which the
laws (XCz, flow) are linear. Experience indicates that it is not
important to exceed the readherence flow from the energy point of view,
because the lift increases only very little.
6 NACA TM 1331
The first wind-tunnel test was made on wings with drooped nose
fitted with flaps and slotted suction or blowing slots in the droop.
Experience has proved the efficiency of the boundary-layer control for
obtaining readherence, the maximum lift being no longer limited except
by the separation at the leading edge.
The applications are difficult owing to the over-all dimensions
of the suction or blowing channels and to the practical impossibility
of mounting auxiliary compressors in the airplane. To make it practi-
cable, it is necessary to utilize a part of the engine power installed
on the airplane and to reduce the dimensions of the air ducts along
This is the reason why our studies on the application of boundary-
layer control were oriented from the beginning toward the employment
of compressed air, making it possible to distribute the blast along
the wing span through a small size duct supplied by air drawn from the
compressor of the turbojet.
The compressed air can be expanded directly in a blowing slot but
this method poses problems difficult to solve: discontinuity of stressed
skin covering at the upper surface, control of the width of a very fine
slot before being rigid, and especially the correct distribution of the
flow along the wing span.
Lastly, the ejector, in spite of its low energy efficiency, has the
advantage of simplifying the wing structure considerably and at the same
time assuming the boundary-layer control by suction followed by blowing
at two points of the profile.
In general, an ejector (fig. 1) consists of:
An injector supplying the "primary air" characterized by the
velocity head p' (dynamic pressure)
A suction chamber and suction cone where the sucked or "secondary"
A mixer whose speed and temperature are equalized
A diffuser terminating at the blowing slot
NACA TM 1331
It is characterized by the geometric parameters X and a; its
aerodynamic functioning is defined by p, g, and e.
Ejectors have been studied in static tests as well as on airfoils
in wind tunnels for a long time. It is conceded that the static pres-
sure existing on the profile at the location of the suction and jet
action are similar, the functioning of ejectors is the same with or
without flow past the profile, this remark justifies the static tests.
The jet velocities have, for the utilization of ejectors, as far
as the boundary-layer control device is concerned, as much significance
as the suction volume; in such ejectors the operating conditions differ
considerably from the classical.
The theoretical study and the tests involved several types of
ejectors: an ejector with flat mixer, without diffuser and a set of
ejectors with circular mixer terminating in a diffuser, the section of
which was developed from a circle to a rectangle (fig. 1).
The parameter X can be varied by changing the section of the
,mixer or by varying the section and the number of injectors; the
parameter a by modifying the mixer or the diffuser. The effect of
the mixer length L and that of the variation of the loss of lead due
to suction was studied at the same time.
The comparison of the two types of ejectors was extended to the
coefficient of discharge p, the speed of the blowing jet Vs, and the
ratio of the momentum g. The results prove the identity of the
functioning of the flat and circular ejectors.
The choice between the two solutions can be guided by considera-
tions of size, weight, and facility of construction.
The circular ejectors favor high values of X; values that give
high discharge P but a low momentum ratio g, with blowing slots of
little thickness, because the section of the mixture does not occupy
the entire span of the wing.
The flat ejectors with flat injectors simplify the design, but
produce only small values of X for these straight slots, hence a
low but a high g.
The best solution therefore seems to lie in a combination of
ejectors with flat mixers and multiple circular injectors, so as to
obtain high X values for straight blowing slots.
8 NACA TM 1331
Arrangements Adopted at Chalais-Meudon
The arrangements and the dimensions of the ejectors for the model
of the SO 6020 airplane used in the large wind tunnel at Chalais-Meudon
was determined in the light of these tests. It was finally decided to
adopt multiple circular ejectors with the following characteristics:
X : 28 a 0.75 L 5
Figure 2 defines the development of the curves g, g, e as
functions of the injection pressure p'. A brief study of this graph
shows that the induced flow is about 21l times that of the injected flow,
that the momentum flow at the injector exit is sensibly maintained, but
that the ratio of the kinetic energies is relatively small, of the order
of 20 percent.
In the range of operations customary at Chalais-Meudon, the injec-
tion pressure p' is such that the speed in the throat of the injectors
The characteristic coefficients P, g, and e depend, to some
extent, on the loss of lead at suction and, consequently, on the setting.
At g = 0.8 for the straight slot corresponding to al = 0,
g reaches in effect, unity for the .a = 15 setting.
It is quite evident that it is important to design the suction slot
correctly in order to obtain correspondingly minimum loss of lead.
Exploration of Blowing Jet on a Flap of the Wing SO 6020
In this study the injectors were supplied by compressed air for
the purpose of ascertaining the velocity distribution in the blowing jet.
In the case of the circular ejectors, figure 3 shows at 100 mm
downstream from the blowing slot a regular distribution in span already,
although with a slight fluctuation attached to each ejector.
This distribution is still more uniform when using an ejector with
flat mixer and circular injectors, of a total section equal to that of
the preceding arrangement but having twice the number of injectors.
Figure 4 shows the velocity profile of the blowing jet in the axis
of an ejector or injector 100 mm downstream from the blowing slot.
NACA TM 1331 9
TESTS ON A METAL MODEL IN TWO-DIMENSIONAL FLOW
Before making the tests on the full-scale model in the large tunnel
at Chalais-Meudon, a very complete study was made on a SO 6008 bis,
metal airfoil of 0.838-m chord, 1-m span, equipped with ejectors
(fig. 5). According to a static test for defining the ejector charac-
teristics, followed by a test between panels in the wind tunnel at
Porte d'Issy, this wing element showed very promising results from the
point of view of high-lift devices for landing and take-off.
The pressure distributions measured on the profile have been
compared with those obtained in the electric tank (fig. 6).
The identity of the curves assumes that the separations are com-
pletely reabsorbed by blowing, while this is not the case for high
settings. Moreover, suction and blowing introduce a secondary effect
of sinks and induction that was not represented in the theoretical
study. It follows a different chordwise load distribution for an
Figure 7 refers to the following configuration:
leading edge setting T = 30
first suction flap al = 250
second blowing flap a2 = 45~
which are marked 30 B-25-450, the letter B indicates that.the leading-
edge slot is plugged.
At constant incidence the readherence of the boundary layer to
the first and second flap is manifested by a sudden increase in Cz,
beyond the flow of readherence the linear law (Cz, Ck) prevails, which
corresponds to the action on the potential flow.
The unit curves (Cz, i) undergo a translation of the ordinates with
increasing CO, but it seems that the incidence of Czmax decreases
as a result of the stalling of the leading edge which proceeds from a
critical value of the maximum speed increase connected with the leading-
A series of tests was run on this wing element with varying leading-
NACA TM 1331
For a given contour, the stall produced for a given Cz is so
increased considerably as the leading-edge radius is increased. The
analysis of the pressure distributions shows the great increase of
admissible critical speed increases, when the radius increases.
The solution of the combined flat mixer and multiple circular
injectors was recently attempted on the metal wing element SO 6008 bis.
The model was tested in the wind tunnel at Cannes. According to
the first results obtained, there is a net improvement of Cn (of the
order cf 30 percent), which is manifested by a more effective control
of the boundary layer.
The practical construction of flat mixers reveals itself, as
predicted, much easier than that of multiple circular ejectors.
THE SO 6020 MODEL AT CHALAIS-MEUDON
The 10-percent thick wings copy the outside form of those of the
SO 6020; they are joined to a fuselage, modified only in the front and
rear parts, the length of the actual fuselage being unsuitable for the
The construction is of wood, but with due regard to the placement
of the essential components of the real wing structure.
The wing has an adjustable leading edge and two flaps (fig. 10),
the setting oal is controlled by electric jets and measured by
Cemented strain gages afforded the over-all measurement of the
hinge moment of the flaps al + a2.
A series of static pressure orifices was installed in the section
AB (fig. 8). Exploratory graphing hooks of the boundary layer are
arranged at both sides of slot al and in the blowing jet o2. The
wool streamers on one wing enabled the flow to be checked, and in
particular to follow the readherence on the different parts.
Compressed Air for Model
The turbojet is housed with its control cabin outside of the wind
tunnel (fig. 11). The air is bled off through special pipes which
deflect a part of the flow feeding the combustion chambers (fig. 12)
and terminate in a double collector.
NACA TM 1331
The air enters the model through a symmetrical circuit (fig. 8)
without introducing momentum, with insertions of flexible elements,
that cause no stress at the wind-tunnel balance.
At entry in the model, a special joint enables variations of the
incidences without vitiating the measurement of the pitching moment.
The characteristics p', T' of the fluid (sensibly at rest)
before being distributed to the injectors spaced along the span, and
also the total flow injected, q'p, are measured.
In all tests the temperature T' is about 500 C, whereas, during
bleeding off it may reach 2000 C. For equal mass flow, the speed of
ejection and hence the momentum flow is therefore much similar. The
experimental results must be corrected for application to the airplane.
Functioning of the Turbojet
Figure 13, made at the Hispano-Suiza test stand, shows the evolu-
tion of the principal characteristics of the turbojet plotted against
the pressure and the flow bleed-off weight.
The tests at Chalais-Meudon were carried out by varying the speed
of the turbojet between 6000 and 11,600 rpm. The experimental points
corresponding to three speeds are shown on the graph, they are situated
on a straight line defined by the total section of the injectors.
The diagram contains the network of constant rotational velocities
and bleed-off temperatures, the isotemperatures in the nozzle (after
the turbine) and, in particular, that which limits the diagram to its
upper part. The isothrust curves of the turbojet also included, involve
the heat balance.
Procedure of Test
For a given geometrical form of the model, at an incidence I and
a tunnel speed VO, the inlet pressure at the injectors (and consequently
the primary flow of the ejectors) were varied by changing the speed of
the turbojet. The airspeed VO was varied between 18 m/sec and 35 m/sec
(3 106 < Re <6 x 106).
To each injected flow q'p, there corresponds a secondary flow
4q'p which is none other than the flow sucked through the slot al;
the total flow (1 + g)q' is ejected by the blowing slot c2. The
momentum flow is decreased and from it the parameter Cu.
12 NACA TM 1331
The balance measures the three components of the resultant; the
elements necessary for the calculation of the pitching moments, hinge
moments, and pressure distributions are indicated; the observation
and the wool tuft photographs are extremely instructive.
For computing the lift and drag coefficients, the wing area
affected by the boundary-layer control was chosen, as it affords a
much easier comparison with the tests in two-dimensional flow.
RESULTS OF TESTS
Representative Parameters of the Aerodynamic Phenomena
Suction and blowing induce flow changes at the wing which are
manifested by lift and drag changes.
Lift.- Boundary-layer control makes it possible to make the flaps
more efficient, and hence to'increase the lift considerably. Moreover,
the suction entails an increase in circulation resulting in a tCZ = K.
Clq", the coefficient K being considerably higher as the suction slot
at the upper surface is nearer to the trailing edge. In our specific
case (slot at 57.5 percent from leading edge), this coefficient, given
by the theory, is very low (K = 2.38) and the gain in lift is practically
negligible for the flows involved (Cq" << -- ) By contrast the rise in
the circulation by the induction of the blowing jet entails a con-
siderable increase in the lift Cz = K' E"C; the coefficient K' like-
wise increases when the jet approaches the trailing edge, but it cannot
be defined theoretically; the effect of blowing being preponderate, it
is reasonable to represent the evolution of Cz as function of the jet
action parameter C,.
Drag.- The suction-blowing combination must be considered: A
certain mass of fluid q", is sucked in at a mean speed Va (it can
be obtained by measuring the velocity distributions above the lip (or
rim) upstream from the slot); a mass qm is ejected through the blowing
slot at a speed Vs.
1.. ..; .':
NACA TM 1331 13
At low incidences for which the speeds Vs and Va can be con-
sidered identical with their projections on the speed at infinity VO,
we obtain a force
-Rx = qmVs q Va = 1 + p)Vs Va q"m
-1 Va qm
in the absence of an exact measurement of Va, it can be estimated equal
to VO with due allowance for the local increase of speed at this
position (fig. 14), on the other hand, the V ~ 4VO and 4 3. The
subtractive term is thus very small and is neglected.
q V 2Cq
-Cx = 1-. 2 = q = Cp
Spsvo2 s/i SV
s/i being the relative width of the blowing slot. The drag curves are
likewise plotted against Cu, which has been computed from Cq' by
C = s/Z
with due amount of the variation of i with the pressure.
Velocity distribution near the suction and blowing slots.- Fig-
ure 14 shows by way of example a velocity distribution upstream and
downstream from the suction slot as well as downstream from the blowing
slot, for several values of CI; the boundary layer upstream from the
first slot is little affected by the suction, and completely suppressed
downstream from it, the pocket of the speeds of the blowing jet indicates
on the other hand, that the induction effect accelerates the fluid well
above the jet itself.
NACA TM 1331
Mechanism of Readherence of Boundary Layer
It is interesting to follow the successive stages of readherence
of the flow over the rear part of a wing for a specific configuration
of flaps: nose not drooped, first slot suction al = 250, second slot
blowing, a2 = 450, incidence I = 120.
The pressure distributions recorded in the section AB are shown
in figure 15, the visualization of the flow by wool tufts in figure 16,
and the lift curve (I = 120) as function of C, in figure 17. For
CP = 0, the two flaps have completely separated (phase a). The first
flap readheres rather quickly when suction begins, while the flow is
still severely agitated over the second flap (100 Ci = 1.3, phase b);
nevertheless, the increment of the lift is appreciable.
In the next phase (c), corresponding to 100 Ci = 5.5, the flow
has completely separated from the two flaps. The orientation of wool
tufts indicates the complete suppression of the oblique flow toward
the tip of the sweptback wing. These different phases are manifested
in the pressure distribution by a marked increase of speed over the
rear part of the profile, which tend toward the theoretical increases
of speed, whereas the correlative increase of circulation entails an
increase in the speed at the leading edge. If the flow is increased
further, there comes a moment where the circulation is such that the
maximum increase at the leading edge exceeds the critical value and
induces separation. In this phase (d), the lift increases is no longer
linear and the curve (Cz, Cp) bends inward. When the separation from
the wing is generalized (I = 150 and over), another lift increase due
to the action on the potential flow and the component of the blowing
momentum for increasing flows, is observed.
The formation of the total drag as function of Cg is also *
represented in figure 17; it increases with CP as a result of the
increased induced drag.
If the latter is curtailed, the curves, for which the flow is
sound, are sensibly reduced to a unique curve the mean slope of which
is that of the theory (-Cx = +C-).
Figure 18 shows that the readherence for the C1 is considerably
greater as the profile curvature is more pronounced; nevertheless, all
curves tend toward one asymptotic direction.
A drooped nose reduces the speed increase at the nose as seen from
the pressure distributions of figure 19. Its effect on Czmax is
NACA TM 1331
The static tests proved that the aerodynamic resultant is equal to
the blowing momentum; this roughly checks with the result obtained in
the study of ejectors.
The same property is again verified with relative wind as seen in
figure 20, where the setting of the polars is sensibly equal to C, at
incidences for which the flow, without boundary-layer control, remains
With boundary-layer control, the polars are sensibly parallel to
the induced polar, although the latter is computed by the classical
method of Prandtl, not applicable rigorously, to a swept wing.
Different configurations with moderate flap settings are designed
for use at take-off. Certain forms are to be discarded for a specific
take-off speed, be it that they do not furnish the corresponding Cz,
or give an incidence too close to separation or give rise to abnormally
high drag. A range of take-off Cz is established for seeking the
best forms chosen by the minimum (Cx Cxi). Figure 21 shows the
corresponding network; it is apparent that high Cz necessitate a more
Figure 22 compares the drag of different forms at constant Cz
and variable a2. The Cx = f(a2) at the right for C, = 0 and
100 Cg = 5 is shown by way of example.
For a2 = 0, the gain in Cx is greater than C, and remains
substantially the same when a2 varies.
The influence of the three elements that define the profile camber
is decomposed, in order to obtain the best Czmax.
In the study of the effect of the drooped nose, the continuity of
the profile was reestablished by an appropriate setting. Figure 23
shows the role of the drooped nose which, by reducing the speed increase
at the nose, delays the separation and consequently lengthens the unit
curves Cz = f(i).
16 NACA TM 1331
The best setting ) lies between 150 and 30. A study was made
also on a drooped nose the suction being assured through the wing by
ejectors in similar manner to the suction on flap al, whose slot was
then closed (fig. 10).
It could be observed at this occasion that the flap al overcame
the separation, even for a setting of 250, by induction of the blowing
The suction in the droop of the nose, by impeding the separation
from the forward part of the wing, postpones the incidence of Czmax
a little too.
Varying only al, the best setting increases with C,, owing to
the prevented separation due to the suction, according to the curves
of figure 24.
The effect of the blowing flap a2 (fig. 25) shows that the
optimum is near 400 or 45, depending on the values of C.
In addition, increases considerably with Cu; this factor is
favorable for assuming the lateral control by blowing aileron.
For a contour of primary interest from the point of view of landing
it is advisable to take the total drag that corresponds to Czm, for
several Cp into consideration.
Figure 26 gives by way of example the polars obtained for the con-
figuration 30 B-25-450. The rise in Cz is substantial, especially
for a swept wing with thin profile. Correlatively, the very significant
drags are due in a large measure, to the induced drag.
Pitching Moments and Hinge Moments
Pitching moments.- One difficulty accompanying the use of high-
lift devices is the correlative appearance of a high pitching moment.
The equilibrium of the airplane necessitates an appreciable elevator
setting which becomes strikingly nonlifting and reduces the Czmax of
the airplane to the same extent.
The tests made at Chalais-Meudon on a model without empennage do
not permit a direct treatment of the problem.
NACA TM 1331 17
The O.N.E.R.A. has recently begun a series of studies in the wind
tunnel at Cannes, on a complete airplane model equipped with drooped
nose and blowing flap. The first results indicated that polars
balanced up to Cz of the order of 4 are obtainable without having to
change the initial setting of the tail.
Incidentally, it is noted that the blowing effects are favorable
on the tail that is in a sound flow and highly deflected; the only
limitation is the stalling over the tail.
Hinge moments.- The flaps al and a2 are constructed in four
elements distributed along the span of a semispan wing (fig. 8). The
hinge-moment measurement of al and a2 results in the following:
For the contour 30 B-25-45 taken as example, the hinge moment
decreases 30 percent from root to tip, when the control of the boundary
The improvement of the flow over the suction and blowing flaps,
multiplies the initial hinge moment by a factor varying from 2 to 3.5,
from root to tip. This confirms the qualitative conclusions drawn
from the wool tuft observations, particularly in the case of swept wing.
The hinge moment does not change much with the incidence, even a
little beyond Czmx, an element favorable for lateral control.
For a nondrooped nose, the hinge moments start to decrease before
Czmax, even with boundary-layer control; this confirms the significance
of adopting a drooped nose.
As to the forces required for flap control, they do not seem to
pose particular structural problems.
BALANCE AIRPLANE TURBOJET
Bleeding off air from the turbojet limits its thrust as a result
of the increased temperature upstream from the turbine. It must be
allowed for in the performance determinations.
By contrast, the decrease in thrust for a given turbojet speed
is negligible as shown in figure 13, while the specific consumption
increases a little during the short period of bleed-off.
18 NACA TM 1331-
The preliminary study of ejectors and the over-all results pointed
out previously have brought into evidence a relationship between the
injected and blown momentum.
Consequently, with the employed notations
q'm' = gqV, = gC SVO2
The aerodynamic study defining Cu, the weight and the Cz defining
V0, the term q'mV' supplied by the turbojet, is deduced. It is
important to locate the corresponding points of operation on the bleeding-
To this end, a net of curves of equal q',V' can be plotted on the
chart from which the equivalent surface of the injectors is then deduced.
By way of example, for a specified landing speed of 45 m/sec and
operation at 100 CP = 10, we get q'mV' = 266; on the extrapolated
diagram of the turbojet, the maximum thrust limited by the temperatures
is 1780 kg equal to Cx = 0.67; if the airplane C, is higher, the
permanent flight is slightly lower; for the usual wing loading, the rate
of descent is not more than 1 to 2 m/sec. The diagram also shows that
the injector exit should be considerably increased.
The adaptation to an optimum configuration is obtained by a similar
process: for a specified Cz = 1.75, the configuration is OB-15-15
(fig. 21); the maximum climb is reached with 100 C1 = 2.5, the corre-
sponding drag is balanced by a thrust P = 1028 kg; the excess thrust
(sP = 2190 1028) makes it possible to obtain the rate of climb.
The diagram shows the ejection section to be much smaller than that
used on the Chalais-Meudon model.
The foregoing results demonstrate the effectiveness of boundary-
layer control on a wing of pronounced sweep and whose basic profile is
NACA TM 1331 19
They show, moreover, that the source of energy used, bleeding off
from a turbojet, results in a significant energy balance since the
thrust loss is negligible at take-off and enables landings in sensibly
horizontal steady state.
The summary numerical applications merely prove that different
numbers of injectors must be used at take-off and landing, such as two
rows of injectors, for example, one for take-off, both for landing.
A study under way on suction at shock wave level, on airfoil
SO 6008 bis, zero incidence, has proved the complete absence of break-
away for extremely small suction flows.
It seems justifiable to elaborate the study of the solution by
ejectors and flat mixers. A much greater control efficiency and simple
design should be attainable, since the mixer is reduced to two flat
The lateral control should be examined correlatively with the
high-lift devices. The efficiency of a differential aileron setting
should be checked, an efficiency which appears attained, since precisely
the flow over the flap ao is correct, even at high incidence.
Lateral control by differential injection might be explored.
Such studies have been scheduled in the supplementary program on
the Chalais-Meudon model.
Although the work at Chalais-Meudon was made on a full-scale wing,
with corresponding limitations, it should be pointed out that only
laboratory studies are involved, and that the application of boundary-
layer control by bleeding off from the turbojet poses technical and
structural problems which are outside the scope of the present research
Translated by J. Vanier
National Advisory Committee
NACA TM 1331
1. Poisson-Quinton: Id6es nouvelles sur le control de la couche limited
applique aux ailes d'avion. Les Cahiers d'Aerodynamique, no. 3,
Deplante: Hypersustentation, commander Transversales. Technique et
Science Aeronautique, no. 2, 1946.
2. Sedille: La propulsion par reaction en combinaison avec 1'aspiration
de la couche limited. ler Congres National de l'Aviation, 1945.
3. Dupin et Morain: L'alimentation en air des turbo-reacteurs par
aspiration de la couche limited. Technique et Science Aeronautiques,
no. 5, 1947.
4. Poisson-Quinton: Recherches theoriques et experimentales sur le
control de la couche limited. VIIe Congr'es International de
M6canique Appliqu6e, Londres, septembre 1948.
NACA TM 1331
Figure 1.- Test setup of ejector.
1000 1500 2000 p mmHg
Parameter of operation of the ejector units used on the S.O. 6020
... ... ..... .........
2 T 13
22 NACA TM 1331
Figure 3.- Spanwise distribution of blowing jet.
NACA TM 1331
Figure 4.- Velocity profile 100 mm downstream from blowing slot.
24 NACA TM 1331
Figure 5.- Metal model; view of inside, cover of flap al has been removed.
NACA TM 1331
Figure 6.- Comparison of theoretical and experimental pressure.
26 NACA.TM 1331
Figure 7.- Example of high-lift device in two-dimensional flow.
NACA TM 1331
Sweepback p=31 2o'
1/4 chord line
Figure 8.- Model S.O. 6020.
NACA TM 1331 ..:;
nACA TM 1331
Figure 9.- Setup in large tunnel at Chalais-Meudon.
NACA TM 1331
Figure 10.- Section of profile: suction from first flap or the leading edge.
NACA TM 1331
Figure 11.- Turbojet and control cabin.
Figure 12.- Bleeding off air from turbojet.
NACA TM 1331
Figure 13.- Characteristics of Hispano-Suiza Nene turbojet operation with
NACA TM 1331.
Figure 14.- Velocity profile adjacent to suction and blowing slots.
* c* .
NACA TM 1331
Figure 15.- Experimental pressure, section AB of wing.
NACA TM 1331
Figure 16.- Visualization: (a) flaps separated, (b)a partially separated,
(c),l and a2 readhering.
NACA TM 1331
Figure 17.- Evolution of lift and drag with C .
ACz ___Separation ot leading edge
(b) -- -
11 3-- / -
( 2Om/s OB-25-45
Vo + 28m/s
5 X 35m/s 10 1 100C.
NACA TM 1331'i
f f 15 f0 100c
Figure 18.- Effect of flap setting on the C4 of readherence.
NACA TM 1331
Figure 19.- Effect of drooped nose.
NACA TM 1331
Figure 20.- Flap polar not set.
NACA TM 1331 39
Figure 21.- Comparison of take-off configurations.
NACA TM 1331;
Figure 22.- Effect of blowing flap setting on drag.
* *' -, *
NACA TM 1331
Figure 23.- Effect of drooped nose n.
NACA TM 1331 -
Figure 24.- Effect of suction flap setting a1.
... .f-: .
SNACA TM 1331
Figure 25.- Effect of suction flap setting a2.
Figure 26.- Family of polars for a landing configuration.
MACA-lUanle 4-23.l2 1000
m m g
? 6 ;'E u C.O
U U m 0. Z aU
4 0 .E 4n -i dB
h" c u
Q:5 & n5-
w RnZ Go
E-. ~ c U
t a, 00 g=0 -
u a ,C 4, l
0 no u OZ4
SM6 Do: a~
00 =1 CL 0 C'
m -' U 0 6 -W
P 0 0j-^^< ^S
a -;u S
;>~O M0; mi-i
-^>a i a! Sb' ra 1" 1"
z z: 5:2EQ LL. = <~-~d
a r 0 3 u I.
0 0 U 0
LoE 3 in iE a
u 'S.s "h
-, U. 5-0
0 C'I L. Qg
a=o 22 w
r3 -0 sT
ug = u3 m'
0 (UP r V~:
Z V r) z
> ap an ^^
?A E a
On E w~" E
m m...0 0
>. -E B .' c
QiE, E =
-0 = 0 e
c =1 E 0'
fu m .0
0 r, nC
< Lr C.
A C4 C4 c 0
.4 0) Io 4
S?~ ,~ KE a 'SS
:4 "I G
*0 c%0 0D c ac j
'B M *i ..U:io .
54--lihfl- MI '0 0 ",
illl$ 1 II~ W
r- be 0-. .
0 N T *=
All u At 9004 *4 0
4 01 0
Q-: ba 0 4)- -1. 0-
4i pi 'U' j cm
esi wS NUo 0.3'
0c i rj~~~
u-- be.-4 ti s''. -l '<^
0 0 B2O
E;! M fl -;1
cl; 1d; ; ; tj
. 0 "! w
,"| Sdnii ni $ d 3
'14 bb A V! cMl a
| a^ |^m ^j <
ci w 0- L.B
E u .o s 3cc i -g D:
5 kU E; 0 d
Ar in u MB C4 ORCia4
i2 cMS 4..,' 1 1' -Z I t 0 WS' o.^ .'0^
^^^S~(D 5~a 5|g=|g
w (-t U14 w
WO 0- 4
cnO.- U2^ _" w
0rl !5 1 2 M
0 0 ) V-
0 ~ P4 4 bit
;L. 0 --
S.*- haZ kn h k- 'ua -.3 u A,* 001 k f
0 0 P4 C a; m Z
4 W F. W. o I 9
>0 0 O
cr1 W -- bb
=" t t QA,
Z, z 9 u Mm c 1 44 I
0 v ~ ~ U
S rl C'! 'WO m c%
0 u w *5
P4 U 1 040. W'
-0h -C U C4 C 4
I 1 st 1 M^
.i O S
L- F n f _
o 1 W2 T o 0 W <
-o o ..' U
E .-: c, F.
H U. a,
i o a o o -
u; 0 :n
Z U- ., t s
0 ,W= I aF
za u :h -PN as !H, 0
^l; "^ g ,j-
'-;. 3 a u^ a o MC T
A" ,b.; C,3 C O
2 l bD C' 0 C :4
0 b DQ -.
o.. o C o l
0 cu0 c
Ai U0 ni m C z az
n^ c Q CO *; _
,a a- 0
0 04 .4 '0
.0 0 U u..
5g Sol -
Z1. g a g .g ) # 040
Z Vw 40
< 5% 01 .
ti C4 CM O r4
4 i V\ cu [ r, C4 o
o S oU BO
F. M WC b. U U ffl i
r -00.9 u M0
4 U U ZA U)
W-6 0 0-
dn t gf b 0
0 1 .v | -
CC 0"a Cdle
0 -o4 .. W ,g i a- -| 2f
NWB '""6 --
" 8 --------------------------------------------
w E 4-
4i |l C 0 ^i~ El
Q4 10 C 0)
c'e0 lgolcu c
ca Q 13 t W b 8.
SO Ek Ws 0 = 0 m W ,
t4 0 0. W 0 wd 0 En, to
S C iilQ w3 o)c:Nga SSS D 'oe'
cd Lq Z L-
cuu a Ja zi
u 0 oia W La 5
4. 1 4)4OEFd
L. S C4 u o 4 ) Om n
^'4O^fu4^ ssbfl* -4 0 ~ ) 2 004 940 OM c
"'iS^ inajraai "^ a c 0ap cj -i
Z, El 0 0
UNIVERSITY OF FLORIDA
3 1262 08106 66 2