Comparison of wind-tunnel and flight measurements of stability and control characteristics of a Douglas A-26 airplane

MISSING IMAGE

Material Information

Title:
Comparison of wind-tunnel and flight measurements of stability and control characteristics of a Douglas A-26 airplane
Alternate Title:
NACA wartime reports
Physical Description:
17, 28 p. : ill. ; 28 cm.
Language:
English
Creator:
Kayten, Gerald G
Koven, William
Langley Aeronautical Laboratory
United States -- National Advisory Committee for Aeronautics
Publisher:
Langley Memorial Aeronautical Laboratory
Place of Publication:
Langley Field, VA
Publication Date:

Subjects

Subjects / Keywords:
Stability of airplanes   ( lcsh )
Bombers   ( lcsh )
Aerodynamics -- Research   ( lcsh )
Genre:
federal government publication   ( marcgt )
bibliography   ( marcgt )
technical report   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
Summary: Stability and control characteristics determined from tests in the Langley 19-foot pressure tunnel of a 0.2375-scale model of the Douglas XA-26 airplane are compared with those measured in flight tests of a Douglas A-26B airplane. Agreement regarding static longitudinal stability as indicated by the elevator-fixed neutral points and by the variation of elevator deflection in both straight and turning flight was found to be good except at speeds approaching the stall. At these low speeds the airplane possessed noticeably improved stability, which was attributed to pronounced stalling at the root of the producion wing. The pronounced root stalling did not occur on the smooth, well-faired model wing. Elevator tab effectiveness determined from model tests agreed well with flight-test tab effectiveness, but control-force variations with speed and acceleration were not in good agreement. Although some discrepancy was introduced by the absence of a seal on the model elevator and by small differences in the determination of elevator deflections, correlation in control-force characteristics was also influenced by the effects of fabric distortion at high speeds and by small construction dissimilarities such as differences in trailing-edge angle. Except for the wave-off condition, in which the tunnel results indicated rudder-force reversal at a higher speed than the flight tests, agreement in both rudder-fixed and rudder-free static directional stability was good. Model and airplane indications of stick-fixed and stick-free dihedral effect were also in good agreement, although some differences in geometric dihedral may have existed because of wing bending in flight.
Bibliography:
Includes bibliographic references (p. 16).
Statement of Responsibility:
Gerald G. Kayten and William Koven.
General Note:
"Report no. L-99."
General Note:
"Originally issued March 1946 as Advance Restricted Report L5H11a."
General Note:
"Report date October 1944."
General Note:
"NACA WARTIME REPORTS are reprints of papers originally issued to provide rapid distribution of advance research results to an authorized group requiring them for the war effort. They were previously held under a security status but are now unclassified. Some of these reports were not technically edited. All have been reproduced without change in order to expedite general distribution."

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003613603
oclc - 71226411
sobekcm - AA00006243_00001
System ID:
AA00006243:00001

Full Text

(Mci I


'it


ARR No. L5Blla


NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS





WARTIME I1 REPORT
ORIGINALLY ISSUED
March 1946 as
Advance Restricted Report L5Hlla

COMPARISON OF WIND-TUTnEL AND FLIGHT MEASUPMENITS
OF STABILITY AND CONTROL CHARACTERISTICS
OF A DOUGLAS A-26 AIRPLANE
By Gerald G. Kayten and William Koven

Langley Memorial Aeronautical Laboratory
Langley Field, Va.


NACA


WASHINGTON
NACA WARTIME REPORTS are reprints of papers originally issued to provide rapid distribution of
advance research results to an authorized group requiring them for the war effort. They were pre-
viously held under a security status but are now unclassified. Some of these reports were not tech-
nically edited. All have been reproduced without change in order to expedite general distribution.


L 99


DOCUMENTS DEPARTMENTT


d; ,.;a~'"
c




































Digitized by the Internet Archive
in 2011 with funding from
University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation


http://www.archive.org/details/comparisonofwind001ang






( '..- 5

NACA ARR No. L5Hlla

NATIONAl ADVISORY CO;.JI1fTEL FOR. AEONAUTICS


ADVANCE IEST.j.ll D REPORT


COElP.::.i.iSC OF WIND-TrUT'L AI~'T FL.PT MEASU'PEJLErT

OF ST:,BILITY ; -D C00TROL CHARACTERISTICS

OF A DOUGLAS A-26 AIRPLANE

By Gerald G. Kayten and WVilliamn oven


SUMI.ARY


Stability and control characteristics determined
from tests in the Langley 19-foot pressure tunnel of
a 0.2375-scale model of the Douglas XA-26 airplane are
compared with those measured in flight tests of a
Douglas A-26B airplane,

Agreement regarding static longitudinal stability
as indicated by the elevator-fixed neutral points and by
the variation of elevator deflection in both straight and
turning flight was found to be good except at speeds
approaching the stall. At these low speeds the airplane
possessed noticeably improved stability, which was
attributed to pronounced stalling at the root of the
production wing. The pronounced root stalling did not
occur on the smooth, well-faired model wing. Elevator
tab effectiveness determined from model tests agreed well
with flight-test tab effectiveness, but control-force
variations with speed and acceleration were not in good
agreement. Although some discrepancy was introduced by
the absence of a seal on the model elevator and by small
differences in the determination of elevator deflections,
correlation in control-force characteristics tas also
influenced by the effects of fabric distortion at high
s-,.-ds and by small construction dissimilarities such as
differences in trailing-edge angle. Except for the wave-
off condition, in tinich the tunnel results indicated
rudder-force reversal at a higher speL'd than the flight
tests, agreement in both rudder-fixed and rudder-free
static directional stability was good. lIodel and airplane
indications of stick-fixed and stick-free dihedral
effect were also in good agreement, although some differ-
ence in geometric dihedral nima have existed because of









2 NACA A-r No. L5HIIa


wing bending in flight. The use of model hinge-moiaent data
obtained at zero sideslip appeared to be satisfactory for
the determination of aileron forces in sideslip. Fairly
good correlation in aileron effectiveness and control forces
was obtained; fabric distortion may have been responsible
to some extent for higher flight values of aileron force
at high speeds. Estimation of sideslip developed in an..
abrupt aileron roll was fair, but determination of the
rudder deflection required to "maintain zero sideslip in a
rapid aileron roll was not entirely satisfactory.


INT.CDjC TION


Although the qualitative reliability of wind-tunnel
stability and control test results is generally accepted,
very few opportunities have arisen for determination of
the quantitative agreement between measured flying qual-
ities of an airplane aJd flying qualities predicted on
the basis of model tests.

In connection with the development of the DouglasA-26
twin-engine attack bomber, a series of investigations has
been conducted at the Langley Laboratory of the National
Advisory Committee for aeronautics. These investigations,
the results of which have not been published, included
tests of a 0.2375-scale powered model of the XA-26 airplane
in the Langley 19-foot pressure tunnel and flight tests
of an A-26B airplane. By use of the unpublished wind-
tunnel data, calculations have been made predicting the
flying qualities of the airplane for correlation with the
characteristics measured in the flight tests. The results
of the correlation are -resented herein; the flying qual-
ities are not discussed except for the purpose of comparison.


OD, L, AIRPLAITE, AND TESTS


Photographs and drawings of the A-26B airplane and
the XA-26 model are shown as figures 1 and 2, respectively.
In table I general dimensions and specifications are shown
for the airplane and the model, as well as for the r-'odel
scaled up to airplane size. Some discrepancies of neg-
ligible ic.ortance are noted in this table but it cun be
seen that, with respect to generall dimensions, the XA-26
and the A-26B are essentially the same airplane. As shown









~JACA A2h a. L~U-illa


in fi-ur-e 1, the model during the stability and control
tests .-~s equipped with a fuselA.e nose which was som~eowhat
dif 'ere.tt frcm that of the airlane. 'i-,e spinners shown
on t-.e modrll propellers were not used on the airplane, and
t'.e a-rr-.lane oil-cooler ducts outboard of the nacelles
:.ere remo-,ved from the model wing during the stability and
control Les.s :-ith the exception of the aileron tests.

Several icre significant differences existed between
the 'odel anSi the airplane. During most of the tunnel
tests the noodel rudder and the elevator, which were of
th;e nl.'..n -v;er.-:ang-balance type, remained unsealed, but
the airr.lane- control surfaces were equipped with rubberized
ca,.vas s-als. rhe control surfaces, all of which were
fa.:ric-c;ove:ed on the airplane, were of rigid metal con-
structi.'n on the model. The airplane ailerons were equipped
withh Ublanci-,"_ tabs arranged so that 80 of aileron deflec-
tion pr.duced ap-.roximately 0o of opposite tab deflection.
On 't:-. -m.odel t-h balancing tab when connected moved 10
for a 10 aillcrn deflection.

Thin 1:i.cal strips were fastened to the upper and lower
surf'a.:-.s of c:he airplane elevator causing small ridges
di -act'l in front of the tab. These ridges were not
reo.rec-snt-ed on the model, but their effect on elevator and
tat 3:a.aracteristics is believed to be negligible.

The wind:-tunnel program included a fairly extensive
sri .s of conventional stability and'control tests. The
nod. l ail:-r'on tests were made at a Reynolds number of
ap.r'oxi.atel' 5.4 x 100. The remaining Irodel tests
aV-re 1ai--. at ia eynolds number of approximately 3.6 x 106
except fo:r t-he tests at high thrust coefficients, which
b:-causr. cf modI-l motor limitations were made at Reynolds
nuimberz r:edc.;d to approximately 2.6 x 106. The portion
of the f.i-:.t tests devoted to stability and control were
of lth t.e. usually conducted by the NAiCA for the purpose
of det-er-.r.ning the flying qualities of an airplane. The
we~ihti. of th: airplane, which varied from 27,000 to 31,000
oouinds in t1 ~ flight tests, was assumed for the analysis
of tih- tu.inne.l data to be 26,000 pounds corresponding to a
'in. loading of 51.8 pounds per square foot. The analysis
vja3 )as2 d on an altitude of 10,000 feet, which represented
an ar.pr.oxima-tc necan of the fli.ht-test altitudes.

Ai.nl-,sis of the tunnel data has been made for condi-
tions representing airplane rated nower and 75-percent
ratsed poo.er at the appropriate airplane weight and altitudes








hC.\ ARR No. L5Hlla


and for a gliding flight condition. In representation of
the gliding flight condition, it has been assumed that
eng.in's-idling and zero-thrust conditions may be considered
identical. any discrepancy in results introduced by the
difference between these power conditions probably will
be small.

In commuting elevatcr, aileron, and rudder control
forces from n...del hinge-moment data, The corresponiing
control linlcaees measured on the airplane were used.

COEFFICIEai S AD 3S':cOLS


60 elevator deflection, degrees
6f flap deflection, de:,rcos

6t tab deflection, degrees

Ch hinge-moment coef"'cient (2
\qbc2/

Vi indicated aircpeed, miles per hour

Fe elevator control force, pounds

Tc thrust coefficient / T-


2 wing-tip helix angle, radians

CL lift coefficient ILift
-qS/

whe re

H hinge moment, foot-pounds

b ing so':r, feet
c root-mean-sluare chord, feet

q dynamic pressure, pounds rpr square foot

(f)









HACA A.R 'o. LHYlla 5


p mass density of air, slu.s per cubic foot

V airspeed, feet per second

T total thrust (two propellers), pounds

D pr_ opller diameter, feet

p rolling velocity, radians per second

S wing area, square feet

a angle of attack, degrees

at tail angle of attack, degrees

g acceleration of gr9avty, feet oer second per
second


RESULTS AND DISCUSSION

Longitudinal Stability and Control


Curves of elevator angle and elevator control force
required for trim in straight flight throughout the speed
range are shown in figure 3. Various flap and power
combinations are considered at three center-of-gravity
locations. For the flaps-retracted conditions, the tunnel
control-force curves were obtained by applying the tab-
effectiveness data of figure 4 to the tab-neutral curves
estimated from the tunnel hinge-mnoment data. The amount
of tab deflection required to adjust the tunnel curve for
trim at the flight-test trim speed was determined for each
nower condition and center-of-gravity location, and this
amount of tab deflection was assumed constant throughout
the s-:eed range. Inasmuch as model trim-tab tests were
not made with flaps deflected, the trimmed control-force
curves for this condition were obtained by means of a
constant adjustment to each original curve of Ch
against CL. This constant hin;e-ui:;ent shift is believed
justified because the data of figure 4 indicate a negli-
gible change in tab effectiveness with change in power
(flaos retracted) and because analysis of stabilizer-
effectiveness data indicates that the variation in
avers-Pe- n:-'.ic-pressure ratio with speed is small for









NACA ARR No. L5Hlla


the flaps-deflected condition. The fl.-.s-deflected
control-force curves for zero trim tab are included in
figure 3.

The sideslip required for straight flight at low
speeds was considered to have a negligible effect on the
longitudinal characteristics of this airplane; hence, the
characteristics determined frov- tunnel data are based on
tests at zero sideslip.

The variation of tab effectiveness witn speed has
been calculated from flaps-retracted "-ind-tunnel tests
made at elevator-tab settings of 35 and -30 with 6e = 0
and is shown in figure 1;. compared with the flight-test
curve.

Elevator deflections and control forces in steady
turning flight are shown in figures 5 to 7 for various
center-of-gravity locations. The calculated results are
based on tunnel tests at the thrust coefficient approxi-
mately corresponding to the appropriate flight-test
conditions.

Although some small differences exist in tne absolute
elevator angles, the slopes of the curves in figures 5, 5,
and 7 show good agreement between tunnel and flight results
for both strir-ht and turning flight, except at speeds
close to the stall. At these low speeds, the flight data
show pronounced increases in the amount of up-elevator
movement required for speed reduction in straight flight.
These marked increases are not apparent in the tunnel data.
This discrepancy in results is believed due largely to the
fact that the production airplane exhibited a decidedly
more definite stall at the wing root than did the smooth,
polished model. Although direct comparison of identical
configurations is not possible, the difference in stalling
characteristics at the wing root is indicated by the dia-
grams of tunnel and flight-test tuft studies shown in
figures 8 and 9. The more pronounced root stalling on
hle airplane would, in all probability, be accompanied by
a reduction in downwash and rate of downwash at the hori-
zontal tail as well as a decrease in wing pitchinFg ;o:nent',
resulting in an inmroverient in stability and requiring
greater up-elevator deflections for trim. At higher air-
speeds the agreement between flight and tunnel results is
reasonably consistent with the ex-erimental accuracy of
both.









p':;.- ARR Io. L5Hlla


The tunnel and flight curves of elevator-fixed neu-
tral point plotted _ainst airsr-eed in figure ]0 for the
flaps-neutral conditions agree to within apr-l:oximately
2 .--rcent of the mean aerod.;iamic chord except at low
speeds .,1th idling power. This difference is practically
within the bounds of the experimental accuracy with which
the flight and the w'ind-tunnel neutral points are deter-
mined. The discrepancy increases with reduced airspeed
as the airplane demonstrates comparatively greater stability.
Because of the difficulty in obtaining consistent neutral-
point results, particularly at very high airspeeds, neutral
points were not determined for these s..:-eds. The curves
of figure 5 serve as a measure of the stability in the
hi:h-speed range and are, in fact, believed more reliable
for comparison throughout the speed ranj" than the neutral-
point curves. Although the curves for the flaps-deflected
conditions are included for completeness, direct conrarison
should not be made inasmuch as the flap settings used in
flight and tunnel tests were not identical.

Examination of the straight-flight control-force
curves of figure 3 reveals comparatively poor agreement
between tunnel and flight results. The force measurements
shown in the tab-effectiveness curves of fi'uie L, however,
are in excellent agreement. Both flight and tunnel control-
force measurements are believed to be accurate to within
approximately 3 pounds. Although some discrepancy in
the elevator control-force curves of figure 3 would be
expected because of the absence of a seal on the model
elevator, analysis based on brief check tests in which the
model elevator was sealed indicated that differences of
the ::.-_nitude shown in figure 3 cannot be attributed to
effects of the elevator seal. In an effort to determine
the cause of the disagreement, the effects of the discrep-
ancies in elevator deflection were investigated. Hypo-
thetical control forces were computed from tunnel hinge-
moment data by using the values of elevator deflection
determined from flight rather than those determined from
tunnel data. For these computations, the wind-tunnel tab-
effectiveness data were used, but the tab deflection was
that employed in the flight tests. The curves obtained
in this manner are shown in figure 11 compared with the
flight-test data. In general, .ree.ient in figure 11
appears considerably improved; for several flight con-
ditions, in fact, agreement is excellent up to spe.-ds
above 200 miles per hour, beyond which the flight-test
curves become noticeably more stable. This difference
r.may be explained to some extent by the observations of









NACA ARR Ho. L5Hlla


elevator-fabric distortion and internal pressures trade
during the flizhL tests. The internal pressures .were
found to be only slightly higher than free-stream static
pressure, causing fabric distortion of the ty':.e illus-
trated in figure 12. As demonstrated in reference 1,
elevator-fabric distortion of this tyne may be expected
to produce increases in the variation of force with
airspeed at high speeds. Inasmucn as the flaps-retracted
flight-test trim speeds of figure are all in this
high-speed range, the trim-tab deflectioDs required to
trim t-e control forces com'uted from tunnel data are
different from the tab angles used in flight, and the
control forces originally cornuted from tunnel data
(by using the aisount of tab deflection required for zero
force at the high-speed flight trim point) could not be
expected to agree well with the flight control forces.
The lack of agreement in the original results was
further -r-r!vated '- the elevator-deflection differ-
e.ces at low speeds, caused by the root stalling effects.

In addition to the effects of elevator-deflection
differences, fabric distortion, and elevator rsp, agree-
ment in the control-force results is believed to be
influenced by small but significant construction discrep-
ancies as, for example, differences in surface condition
and .n trailing-edge ar~-le. At a representative section
the trnilin.-edge enale measured on the model elevator
was 12.70, whereas the corresoDnding anrle measured on
the airline was 110. Kone of these effects would be
expected to influence nrcrreciably the agreement in tab-
effectiveness results.

As seen in figures 6 and 7, tihe flight. tests show
considerably greater variations :" control force with
acceleration, and the values of force er g show con-
siderably greater variation with center-of-gravity
location, although the elevator-free maneuver point

Fe
-e= 0 is approximately the ssae. Because the absence
g
of an elevator seal -;as believed to be more significant
in accelerated flight than in str'.,~ t flight, control
forces were estimated for both tne sealed and the unsealed
elevators by ass, 'l;c' constant pitching-moment and hin_e-
:nocint slopes and using the sealed-elevator -inre-noment
data obtained in the previously mentioned checi tests.









:ACA A'"? .o. L5Hlla


The respective values of 6Ch-/66e and 6Ch/6bt used
in tLese computations were -0.0057 and -0.0018 for the
ui3.:l--i elevator and -0.0050 and -0.0032 for the sealed
ele,.t.cr. The resulting. curves of force oer g against
center-oi-gravity location are shown in figure 15. The
co.i-ve f-,r the unsealed elevator is practically identical
'.': t- ?.t previously determined for the unsealed
el:vstor (fig. 7) by the method of reference 2. For the
se-lej elevator the values of force per g are still
ve-- :;u.-:: lower than the flight-test values, although
t'-e -.'ariation of Fe/g with center-of-gravity location
is ...-re nearly parallel to that determined in flight.
Th-e cowmarison of control forces in accelerated flic-ht
hse beei ,iade at a fairly high speed. Reference 1
i:-.icec.: that fabric distortion of the type experienced
i" t.he A-26B flight tests may be expected to produce
increases in the variation of force with acceleration
in t-.e n.rl'nal center-of-gravity range and in the
..ri tion of force per g with center-of-gravity
icptioln. This comparison as well as that for straight
fli_:t 'iould also be influenced by any differences in
co:rntrol-surface construction.

Arreement in the curves of elevator-free neutral
-oi. t 9s-einst airsoeed (fig. 10(c))is rather poor and
te z :r.- w..orse as the seed increases. The flight-test
el-ev tr.r-free neutral point moves rapidly rearward
with ~- 1creasini: speed, and at high speeds the airplane
9a--, rs .:.ore stable with elevator free than with elevator
fi-,:e. It is believed that this large rearward shift
in 1-e elevator-free neutral point with increasing air-
sn-e i.:si be a result of the fabric distortion.

In general the present correlation indicates that
succ.;ful prediction of elevator control-force charac-
ter-i.tics from wind-tunnel data can be made only if
e:-.trei.e care is used in representing closely the air-
il.nse in its construction form particularly with regard
to rme control surfaces. Agreement with flight
:es-ure,.-nts might also be improved considerably if
eifet-r. s such as fabric distortion could be taken into
account. A more beneficial solution, however, would
be to minimize these effects in the construction of
the alrolane.









NACA ARR "o. LSHlla


Lateral Stability and Control

Steady sideslip characteristics.- Characteristics
of the i-su-In. ni teas~; -i lTsl-lp, which are used
as flight-test measures of directional stability,
directional control, dihedral effect, side-force
characteristics, and pitching moment due to sideslip,
are shown in figure- 4. Although complete hinge-morent
data for the model ailerons and elevator were not
obtained in sideslip, aileron forces in sideslip were
estimated from the tunnel data by taking into account
the change in effective angle of attack due to sideslip
but assuming no direct ch3r:-e in aileron hinge-moment
characteristics with sideslip.

For both idling and rated-power flight with flaps
retracted, figure l1 shows excellent agreement in the
variation of control settings, anglo of bank, and rudder
force with sideslip, although some difference exists
in absolute values. Some of the difference in absolute
values may be due to the fact that model tare tests
were not made in sideslip. It is especially interesting
to note the close agreement in the variation of aileron
angle with sideslip, which serves as a flight-test
indication of dihedral effect. It was found in the
flight tests that the airplane wing in normal flight
a-oPeared to bend upward noticeably with respect to its
position at rest. Despite the wing bending, however, the
amount of effective dihedral determined from flight
tests was also found to be no raterer than that which
would ordinarily be expected for an airplane of this
type with 4.50 of geometric dihedral. Analysis of the
elastic properties of the model wing under load indicates
that the model wing bending was negligible. On the basis
of the agreement between model and airplane results,
it appears that the observed airplane wing bending mar
have had very little effect in increasing the dihedral
effect beyond the normal amount for 4.50 of geometric
dihedral. Further information regarding the elastic
properties of the airplane wing and the effects of
these properties would have been desirable but was
not available. Co.iarison of the flight and tunnel
aileron-force curves appears to indicate that little
error was introduced in determination of the latter by
the assumnotin that aileron hinre-moment characteristics
remained unaffected by sideslip. The sideslip charac-
teristics with flaps deflected Co not agree as closely









1IACA A27 No. LHlla


a; ... ti-e flaps-retracted characteristics, particularly
in t-e cnse of the aileron-deflection and rudder-force
.- t icns. 'I.e flicht-test rudder forces show a tendency
Lo .':ar reversal in figure 14(c) but do not actually
r3-.-rse 9s in the case of the model forces. At an
a"r:'eed slightly lower than that for which the data
.'e 'rcs~nted, however, rudder-force reversal did aosear
in ..e flight tests in this wave-off condition.
r ... 1 effect with flaps deflected and rated power
l- .:.r:.eed appears somewhat lower in the tunnel
-:.r-..:.enLets than in the flight data. rne flap deflection,
.-. -.e, was 50 greater on the model than on the air-


In fi-ure 15, rudder hinge-moment characteristics
esti:.3tQa from flight-test rudder kicks are compared
.':ic,- r'.ier hinge-moment characteristics measured in
c.-Le tunn.--l tests with flaps retracted. Although the
:o.-ie .1 rudder hinge-momient and force results are for an
u-r:e le-'1 rudder and are also subject to effects of small
s'su- r-e _and trailing-edge irregularities as in the case
o' t.e elevator results, agreement in this respect is
c-oo. A.s previously shown in figure 14, the rudder
forces n. steady sideslip are in good agreement for
t..l fl: condition; In regard to rudder hinge moments,
t.- tunnel results, which showed no positive values of
.. i- r~r.eter 6Chj/a for the rudder, indicated that no
r.iAde-r making would occur in flight. This indication
.'as co.-i'rmed in the flight tests.

Alleron characteristics.- Io tunnel tests were made
t, i-r: -'tisste sclr'on c. aircteristics for the 3:8 tab
lin: ,-re .ith which the airplane was tested. If, however,
lit'nr t.sb effectiveness is assumed, these characteristics
f -.- tne flaos-retracted condition can be estimated from
r-.e r-esits of tunnel tests of the plain ailerons and
t-ie ?il-eons with a 1:1 balancing-tab ratio. Estimates
o. co-:itrol force and helix angle made in this manner
:-- c -.-ri Rred with flight measurements in figure 16 for
in-icted' airspeeds of 155 and 358 miles per hour.
A- recoim':ended in reference 2, helix angles were
pb 0.80C
Sst i~ated as where C1 is the total aileron
2V C p
rollin.-:-.oment coefficient and a value of 0.57 was
u-ed a; the damping-moment coefficient Cp Although









NACA ARR No. L5Ella


the angles of attack selected for these estimates
correspond to rated-power flight at the appropriate
seeds, the model aileron data were obtained in power-
off static tests. Inasmuch as the tunnel measurements
were made for right rolls only, the tunnel estimates
are exactly symmetrical for right and left rolls,
whereas the flight results are not. Agreement in the
curves of helix a'--le is excellent in the ra-~ e where
coe -.arison was possible. There is, however, some
indication that the tunnel estimates, based on the
arbitrary 0.3 factor, n' It be slightly optimistic for
hi-h deflections at high speed. At the low airspeed,
agreement in the force curves is good except at the
highest aileron deflections, where the control forces
for given aileron deflections are slightly higherr in
the flight records than in the tunnel estimates. At
the high soeed, the control force required in flight
for a total aileron deflection of 14i is approximately
40 Pounds (or 73 percent) greater then the force
indicated by the estimated curve. The greater dis-
crepancies in the control forces at the high speed are
believed largely due to the effects of aileron fabric
distortion. As in the case of tne elevator, the
aileron fabric was found in the flight tests to undergo
considerable distortion at this high speed. The
distortion was in a direction to produce higher control
forces.

If the assumption of linear tab effectiveness is
not entirely valid, actual wind-tunnel tests with a
5:8 tab linkage would indicate the control forces
somewhat lower than those estimated herein for the
3:8 linkage at the higher deflections.

Sideslip due to aileron deflection.- Curves of
sides!-.'- sn 1. si.i ri.iLing vA..c_ l yr against time in
an abrupt rudder-fixed aileron roll out of a 300 banked
turn are shown in figure 17. In addition to the sn;:jli-
fied sideline estimate of reference 2, the motions have
been calculated by the operational method of reference 3
and also by the tabular-integration method of reference 4,
in which slope variations in the curves of rolling-
moment, yawing-mnoment, and side-force coefficients
n ainst n-le of sideslio are taken into consideration.
This method of tabular inteicration has been shown in
reference 4 to be more reliable for -eneral use than
methods requiring 7 the assumption of constant slopes.









:'ACA A :-o. L5Hlla 13


\r- clhe subject airline, .".: ch exhibited essentially,
crtr:t :-.t 3o1nes, the three i:iethods of cor ut action based
or. In- -tlnnel results re epos1r to L Sve very sir.ilar
r-.. ... respect to rai::r u: sideslir an-"e, all o
..1.. -~:troxirately 4o higher thane: the flish.t-test
I-. A.:.n- tae factors possibly contributing to the
1-: .:_' -erfect agreement is t: i rlence between the
*:r.s a:rt reous control deflection assus:.ed for the cc'I.ou-
t.cT-'.nZ rind the actual control ;:ovemr..nt in the fli 't
t-:t. Another factor influernciv7: t.: results m:ay be the
S.i- n normal acceleration experienced by the airalane
in it: rill out of the turn. Altho.' no fl-ht record
cf i:---i e acceleration was obtained for the test in
q.:.sti'n, similar fli-i '-test results indicate t:*Pt a
c..n:...er. s:e variation :.e-- have occurred during the
.*:--.=. Analysis indicates that the change in normal
-' L-.-r' iLon and, conseaque ntl, lift coefficient may
-:.tr:-,....c- conditions considerably different fror those
c- :-,rel in the theoretical calculations.

A -i le static estimate of the amount, of rudder
.i-l::-..~ i required to maintain zero sideslip in en
nll r. rn roll was clade as su -.:sted in reference 2; that
is, i --s assnued that the desired rudder deflection
w.l i- that required to counterect the combination of
a l: .-.-;, -Iverse yawing rmo :ent and awing moment due to
roll;n-. The estimated value obtained by ti -_s Imethod ':,as
*-.- .i ely 8 for fleas-retraced flight wth levei-
f-Li :', er st an indicated eirs- od of 1>.5 niles per
-.-. al.hot,.: no fli:-ht-test dara were recorded for
'.11-. ilron rolls at chis fli -t condition in wvich
z::;-o s-I*.; lio was maintained by means of varying rudder
defl:-'.icns, flight-test records for constant rudder
-ett:!;:- indicate that the rudder deflection estimated
fr.... !.:sl results would be notic-easl lo :er than that
r:r-: .'i -_ n flight. For severely rolls ith partly
afLc ailerons, however, essentially zero sideslip
n-;- t --nt -ined, and the esti .e te. rudder deflections
'.,r ..,rn to be in fair agreement with the -.: ..
efl-c.:, s required in fli-~-t.


C TCT 'hllTG RE"A- C


Srys~ility and control characteristics determined
fro'r. L .1ley 19-foot-nressure-tunnel tests of a








NACA k2R No. L5T11la


0.2575-scale powered mod-el of the Douglas XA-26 airplane
hav- been compared with results of flight tests of a
Douglas A-26B air. la:e.

The significant results of the comparison may be
suimarized as follows:

1. Good correlation was obtained regarding elevator-
fixed neutral points ar-1 the variation of elevator
deflection in both straight and turning flight except at
speeds approach-'" the stall. At these low speeds
the airplane showed a distinct improvement in stability
not indicated by the model tests. The difference was
attributed-. to the fact that the ,prcno'.inced stlling at
the root of the production airplane wing did not take
place on the Smo?;, well-faired ridel wing.

2. Th--. variations of elevator control force with
airspeed and acceleration wore not in good agreement.
Although some discrepancy was introduced by the absence
of a seal on the model elevator and by small differences
in absolute values of elevator deflection, the corre-
lation in control-force characteristics was also
influenced by the effects of fabric distortion at high
sp -ed and by small construction dissimilarities such
as differences in trailing-eige angle.

3. Elevator tab effectiveness as determined from
tunnel data was in good agreement with flight-test tab
effectiveness.

4. Agreement in both ruC.der-fixed and rudder-free
static directional stability was Cood except in the
wave-off condition, in which the model tests indicated
rudder-force reversal at a higher speed than the flight
tests.

5. Model and airplane indications of stick-
fixed and stick-fr.-e diledral effect were in good
agree.:*..nt, althouc'-h some slight difference in geometric
dihedral may have existed because of wing bending in
flight. The use of model hince-moment data obtained at
zrro sideslip app:-ared to bo satisfacto-r. for the
determination of aileron forces in sideslip.

6. Fairly good correlation inaileron effectiveness
and control forces was obtained. Fabric distortion was












clie. 3d res onsible to some extent for higher flight
v: -t'-:- of aileron force at nigh speeds.

7. Est:'r.,r'tion of sideslipr developed in an abrupt
-i r.-n roll was fair, but deter.-iination of the 2:iaxi;:ur.
r"'i: r deflection required to maintain zero sideslip
-. bru-t roll was not entirely satisfactory.

.i the basis of these findings, it appears that
n: -nt between stability eard control charact.eritics
ti.. ed fro wind-tunnel results ana those measured
.' ht cannot be coinletel:' satisfactory unless
---t-n factors now usually neglected in wind-tunnel
:.t n; can be taken into consideration. These factors
:.-1 :' small differences between the model and the
-ne and include differences in elastic proi:erties,
.-ice finish, and construction accuracy. These factors
s:':u-1j be considered, if possible, in future investi-



TLr.-:. 1y Temoriel Aeronautical Laboratory
national Advisory Coari.ttiee for Aeronautics
Lan"le"- Field, Va.


iA a A- to. LE'rla









IACA ATR No. L5HIll


i. eathews, Cl rules 'r.: An A'alytical Investigetion of
tie Effects of Elevotor-Fabric Distortion on the
Longitudinal Stability and Control of an Air-olane.
FACA ACR No. TLE30, 1) .

2. Kayten, Gerald G.: Analysis of vVind-Tiu;nel otability
and Control Tests in Terms of Fly1ng C-lities of
Full-scale Airplanes. NACA ARR No. 3J22, i~ r.

. Jones, Robert T.: A S:..jlified Application of the
.ethoc' of COerators to the Calculation of Disturbed
Yotions of an Airolane. NACA Reo. Uo. 560, 1956.

'I. Wolowicz, Chestsr I.: Prediction of i.otions of an
Airnlene Riesult:ri :from Abru)t movementt of Lateral
or Directicnal Controls. NACA ARR :TD. L50O2, 1945.













NACA ARR No. L5Hlla


.4-.4


o o* a8 4
* (- 0 &aj o 0 sO 0 U\ m 0o 0



tI O0CoQ0-4 -.4 Nm OON a .CNo N
a a










0o ocu C on no
z Z
0


S0 (I4 '
.0,% 4-NWCi 3 4.0..0 '-4 i(tn"CO CoO >000C 0 0 LrN(UWO 0 CO
2 .ru~f O r0i ON 0T\ ,M( '. r-4\sK\ t'rW s 0 KM4O r


0 r ti> 4








A .-l.. t. -
Ci .1'. 0 0 -4


-44





4 1 LOr%
o CO. 00 Ca .O It
*6 .4'ooao 0 s u 11 va1 -4 uO -t- r' w-1. 0 It. ir- P'
a 0o os34o' r-O.04zn .C P u'oa ooo.-4oo c- o NV (\l
I 1 4'.rl ;\ *O C



0 z




4)
"-


*o O
4, -. v


40 0- 4 a V 0- t


-4 i4t c 14
0' 4'0 0 O1 0*. .4 .a

1 d' 0 t 1 00'-4 a
.2 0. 6P 4 ..W4)


%' 6) r. 0 0 r .P
4 .4 6 40 4)- .4War 4w '0 '0 O0 a .0 0'



.44 0 .. -4 00 '0. 0 `
2 OAI ra 0 C 0 4) a .00 (0 C,
t a, o o0 4Or6 -I P u0 0 <



f 0 c20WH rCO m x 2C.4 Pj
0 *4A 0 0 6. 4- 0 0 6 V 0 C 4- 4 r CIO
0, O0l.4W 4 .4 .O .4 .. .4 W2*4 .44 -


2Q I(Dpc. : 4 4 4) O0 0 r '0'062B0G '0.0 6 4)
.i 4 -1 6 0 '1- 1- 2 a C. r- I ) 6 0 v .-r-46)4O0
.-t04).rf O 2H 2 -1 0 .0 4P 4) 06- 23 I .*4 C.I.0

a 4) 066D .4 .4 P..-' -. 0 A 6C.410 r. 9. f4 4)



G14 0 G 4H u4 0,
0 '-4 C. O. 0
.4 .4 .4 0: 0 P




I









NACA ARR No. L5H11a Fig. la









VC:












o



1 0
L'S











o



01


(' C)C)
S4:D







4 -












bo
I r .
~ ~ ~ r ', .. ,"
,,
L
":"r (E}
e P-I





.,
..,
t, -.,.-.




1










NACA ARR No. L5Hlla Fig. lb


















bo











0




a z



cd








0 a,


I tw


ra
-O
OC



*0 .-











NACA ARR No. L5Hlla Fig. 2a











0






I
z *S 55 .

















Su







-D
'o


^ N



N L,








Fig. 2b NACA ARR No. L5H11a



-- .-- 09

o






O '

-- -





















6.





I s, ,, 0
<---------\ uj--------- /








NACA ARR No. L5Hlla


b6p p '9% ibUt JO.O/al.j7

i <






O
i' < Z 0


3 _t
~

F_ _z z j
E i _
[] j ^


I-
0


o



0I
Fl
n3


0 -




r,


'"7/ ,./oJ /OJ/ 1o/0 JO./Vlc/7


00













- -7o C
_____0_


UMOG;
UoJ
0


Fig. 3a


Q0








4







o2 -e










S-i
0

cr)














&G l

7zz


U,
U;
0:
4
1"8
4'
0t











rO

K)~

KS


'SZ
K


"b
0




0
-S


x,
K


?.Z
0






Izz






!
-t^
oJ






K
0
QJ

0
0








nQ)

K







NACA ARR No. L5Hlla


40ap 'ag 'b/6UlD J/O-//1 lS7


o o









TI




0 :
U_


0


ooc

o0a0


E



oi
i


CI


r-^
rn


Q)
Iq


lin


q9(9/ '. /9JO /O/U60 Jo1/1lAa /


UMOA 0
c c



z






o o
( 0 I























/OOn


0- -Z






4-


Fig. 3b







NACA ARR No. L5Hlla Fig. 3c


dn ap "a9 dg6/a Jo/Ada/3
a on
L/AO16









0 h
cr







I r, I s




II








CCO
II I







00 OOf0
n.1
-_ __ j
c ...- -- -, --4 I

















ql (% '-3"Joj /OJ (o0 Jo/,417 .







NACA ARR No. L5Hlla


Sb&ap W ?'a/buo jo/Aag/_
o n 9 o



















( c3 1q j
So0















-Q



0







ql" 'j/' 9J0og /OJ/ 6uo3 JwOOAI/_7


(.


2


C-)
K'z


Q.








q/sd


Fig. 3d


tz,q
(r)




Q).
S1
O% -I%
% !

vj o
$ r s

^^ z!
6 ^4^ 1
p
^^^S


i~--






NACA ARR No. L5Hlla


28



124













8



o
0





O
3


Rated power, df =Oo Tunnel
--- = o ., 6,-=-0
__0 a __ ____
0 Rotedpower, cf=0 Flight
El Idling 6=0J 0


/








--~~--- 7 ---
_______ ____ __ --- --- ---- --- --- --- --- --- ---
0/











/


NATIONAL ADVISORY
COMMITTEE FOR AERONAUTICS


/40


220


260


300


340


/00


airspeed, V, rnph


In dicated


Fiqure 4


-Variation of elevator trim-tab

effectiveness with ai/rpeed.


/80


Fig. 4





NACA ARR No. L5H11a


07



C
0





ci
U





-c


I '/O U GD J On 0 "O V (
dn umcy7
bap 2 /buz)o/oJ, uo o. ao c7


*s


b
i




oI
o'
i 0.
S0









0
^ Q,








r
4C


OQ
cZ1%.
S^c3







U .L%
K ^)


- '0


o -^




' '0

ci)
ci, 1
K (J O
0% cZ1


Fig. 5






NACA ARR No. L5Hlla


U

ini
OW
'- c
z-




~oo


a
cn o
ISV o ^




O4

U SU
p b o-
Sa C: k






o





CCOL
00









u '
00
K
oo
Q ) o,1^





I 1^?

K? P -~


\ i a,
^ ^
^ ^
E" S
,\ Cr S
G'


c7/ adoJ I/OJJuO4 jo/w4a/fa U/l auLx


Fig. 6







NACA ARR No. L5H11a


4


IN 0







60






120
b





80

- 40


-Tunnel

--- Figqht





























n IL II I '--
N '
N_______


N^
.^^ N


44


Center-of-qravity /ocation, percent MAC.


Figure 7- Variation of elevator control-force and oeflec tion

gradients with center-of-gravity location. V=260,,:/'/es

per hour at /0,000-foot altitude d 6 f-Oraftd power;


steady turning fight.


NAT ONAL ADVISORY
COMMITTEE FOI AERONAUTICS


Fig. 7


43d


%







NACA ARR No. L5Hlla


____ NATIONAL ADVISORY
Surbulent COMMITTEE FOR AERONAUTICS
S tQlled

Figure 8.- Diagrams of stall progression in the gliding condition.
Engines idling; flaps and landing gear up; cowl flaps closed;
oil cooler one-half open ; Douglas A-26B airplane.


Fig. 8

























14



o I '



S.4,

O 4 8 Z /6 M2
Angle of oaHckhcwafe


NACA ARR No. L5H11a


c = 14.4'
c, =/.33


[ Unsfalled e1'm "e -"




C Cross flown Camplefely
t\ e disction 5staaed
of Or fr


\-Unsteady


cc /7.4"
C -1.3


NAtIOMNA AOVISOY
CUMMIUit FOR AtIOIAUIiCS

Figure 3. PTwer- offsfta/l,. y',- for the 0.2375-scale mode/of the XA -56 a/rpolne.
,5 '."- J2' model configure lon with airplane oa-cooler -. a number, 4.5 x ;
Mach number, 0.131; 6f = 0 .


Fig. 9


ac = 6.9"
c, -aO.


aC = 10.2/
C, =/./7


!UnsIeody

cc = 12..3'
C, = I.3/







NACA ARR No. L5Hlla


Q)


o \ CN
Cr) Ct)


Q)







C-

o,






0)





0


0)


Figo 10a


'' VAI -luaDao '1ulod ,o/DJ/naN






NACA ARR No. L5Hlla


-Q








Q)












0%
Qb





) 3 )


K


D'Ji/ 1U.QaJqad ',UIOod ,o.,nan


Fig. 10b







NACA ARR No. L5Hlla


c) 0 C


3'FN 1uaD1ad C'u/od pun1


S )









So
^Q







8 ,
o U



Q)
i3 :


Fig. lO






NACA ARR No. L5Hlla


3 40


O


40

40


0



40

40


0


100


140 /80 220 260


300


Indicated airspeed, ,; mph
(a) /iaps retracted; rated power.


Fiqure //' .-Vr/aton of elevator control/ force with
indicated airspeed. Mode/ elevator and tab de flec -
tlons identical with flight-test settings.


N.


()
r>




K
o


u

K
0


Fig. lla






NACA ARR No. L5H11a


40


0



40

40




40
40


0


40
40


0 ___




t c. g.
(deg) (percent M4C) Tunne/ F/i gh
2.0 up 3/.9 0
1.2 up 278 --
0.3 down 22.9 ---



0-0- -
[ J ... -Er -Q ... ..^
------ ----E-8- -
Eco












NATIONAL ADVISORY
COMMITTEE FOR AERONAUTICS


/40


/80


220


260


300


/00


Indicated a/rspeed, V mph


(b) Flaps retracted; 7,-percent rated power


Figure / Continued.


Fig. 11b






NACA ARR No. L5Hlla


z40 z
% --- -- o- ---"

de e.g.
(deg) (percent H4C.) Tunn
2.3 up 32.0
40 1.7 up 274
/.2 up 23.0
(3 f 3l 3
a 40



O 40-- ---

o 0



Fo _
04






NATIONAL
COMMITTEE F(
40
/00 140 180 220 260

Indicated airspeed, ,n mph

(c) F/aps retracted; T, = 0.

Figure /ZII. -Concluded.


300


Fig. 11e







NACA ARR No. L5H11a Fig. 12


4 =.Vi360 mph



Vl=370 mph
c at 28 percent MAC.
Approxi mate
point of fabric
offochment
ot_ me Vi=300mph



Vi=320 mph
cat t 2 percent MAC.
No-load fabric fenson, 2.7/b





Vi= 270 mph

section under no load
.~--- Section in flight




Scale, in.
2=20 mph o 2 \&






Vi 170 mph
Elevator section 942 in. from center /ne
of airplane
o/ air,'/ane NATIONAL ADVISORY
COMMITTEE FOR AERONAUTICS

Figure 2.-E/evator-fabric o'istorhon at
various indicated airspeeds, Doug/as A-268
air/o/ane with center of qrav/ty at )2 percent
MAC. except where noted.







NACA ARR No. L5Hlla


b6 Jad q/
"Cua/poj0b 3-oJ-IOJ -JUO


U
0


4
4
-Ic

z1
0


Cz

Q)

'U
K)
0O

QS

Bz
(3
0







(3)


Jo_./DCla1


t3





b












o
b k;


o"




8h a`rs,

P^

\e,
c^c
`EJ O


Fig. 13







NACA ARR No. L5Hlla


01



I


(Z

Z,





(Jz
$1
6 1

Q

s*0


20b /0 0 /0 20


F-i ght
0
0
O


Control
Tunnel
Rudder -
Elevator -- -
Alteron
(total)


-- Tunnel
A Flght


















NATIONAL ADVISORY
COMMITTEE FOR AERONAUTICS


Left Sidesl/p ange, deg Rig/h
(a) Flaps retracted rated power; V /4/mr//es per hourl

Figure 4. Jteady rdoeslp characteristics.


/ -- -- -_^ S -







C ---- ----_--





^-----
0-



A< zL


940
U
o



T\ -40^


200


Fig. 14a







NACA ARR No. L5H11a


C-,
0


Q


0






0
c z
aO


-Q


20


/0


0


/0


20


10


0


/0


0
or


\40


Left Sides/lp angql, deg


Flight
0
<


Control


Tunnel


Qudder -
/levator -- -
A .'ron
(total)














-- T jnne,


Right


(t) Flaps retracted; T, 0; V =/33ml/es per hour.
Figure /4-Continued.


Fig. 14b








NACA ARR No.. L5H11a


20


Q 10

OZ
o

Q /









O o




Q
c^1 0



*-/


t-


'Q40


Q i i

Qo g

-M


20C


Control
Flight Tunnel
0 Rudder --
0 Elevator --
o Aileron
(total)















-- Tunnel
A Flight


NATIONAL ADVISORY
COMMITTEE FOR AEIMAUTICS


/0 20


Lelt Sides/lo anq/e, deq Right

(c) F iops deflected; rated power; VIlll miles per hour
Fqiure /4.- Conc/led.


So0


Fig. 14c







Fig. 15 NACA ARR No. L5Hlla




4-


C' U














0 13
/ aI
-- --- -o- -.-


-------------------4------------o Oc' b



70------------ o ^













s C






.Q_
P f







4 --
_I ---z 7



LuaIlNJ'ao) quawuow-abu1( U/ abuoqD







NACA ARR No. L5H11a


2001


iS


0


s:
Q)










Q
0-
e


/60

/20

80


40

0

40

80

120

/60

200
08

.34

0

-"4


I
--------I-----






7-





-L
--- -- --- -- + -----/
-- -- -- -- E ^ -- --/
__ __ __ __ __ __ __ __/ __ _

ZZS^Z^ZZZZZ/
I~
^ -


ZZZZ^ZZ /ZZ
/Y


Tunnel
/35
--- 383


Flight
0
El


40 30 20


/0 0 /0 20 30 40


Left Change in total/ aileron anqle, deg R/ght

Fqure /6. Variation of aoleron wheel force and hel/x
onq'e pb/2V with change in total a/leron angle in
rll/ with rudder fixed, flops retracted, and rated
p I'4 r. NATIONAL ADVISORY
COMMITTEE FOR AERONAUTICS


Fig. 16







NACA ARR No. L5H11a


Reference Calculated from
-- Reference 3 wind-tunnel data
----Reference z


----- light


0 .4


S 1/2 16 20 24 28 32 3.6 40


Time, jec


NATIONAL ADVISORY
COMMITTEE FOR AIONAUTICS


Figure / 7- Poll/ng velocity and J/des//p during aileron
roll out of 30J banked -turn. 6f= 0; 0V =/J45mi//es per
hour atof/ 0O-foot altitude; level 7I/ght power


4,4,

/ F/
/ / I
_ __ t __ _


Fig. 17








UNIVERSITY OF FLORIDA
II 1 l I I III I Tl I I kImn1 1 I
3 1262 08104 970 1


:lE3il-TYC OF FLORIDA
S,_:',JrEfTS DEPARTMENT
1._" i..-.RSTOU SCIENCE UBRARY
.- BOX 117011
.:,I IESVILLE, FL 32611-7011 USA





















I










I,,