AL162036C' It A numerturl method tor the derlgn and anrlyris of

countetrorrting propellem. S. 2. Playle. K. D. Korkan. and E. von Lavante

flexas A 8 M University. College Station) (AIAA, SAE, and ASME Joint Pmwl-
s h Cmfemm, 20th. Cm'nnai;:3H.June 1 1- 13, 1984,AIAA Paper 84- 12CYi.)
Jmrmal of Propulsim and Power (ISSN 0748-4656), vol 2. Jan.-Feb. 1986, p.
57-63. 18 refs. Grant NAGS-354.
A numerical method has been developed using the techniques of Lock and
Thaodonan as dasc&M hy Davidson to design and analyze counter-rotating
propellers. The desrgn method develops ihe optimum prGp9ll9r $3ometry by
calculating the planform and twist distribution for each propeller disk through the
use of s,wif&inputs of engine shaft horsepower,diameter. and disk spacing. The
malysis inethod celculates the performance of a given counter-rotating propeller
system at any flight ccnditron. Urrng the NACA fourdigit airfoil family. the per-
formance of a counter-rotatingpropeller design for a given flight & i was
investgated in the desiin and analysis mode. (Author)

This Abstract Published In

INTERNATIONAL AEROSPACE ABSTRACTS@

The Original Source ~ocurnent

Is On file in the AIAA LIBRARY
AMERICAN INSTITUTE OF AERONAUTICS AND ASTRONAUTICS
TECHNICAL INFORMATION SERVICE
555 WEST 57th STREET
NEW YORK, NY IWIQ
 A Numerical Method for the Design and Analysis
of Counter-Rotating Propellers
S. C. Ylayie,* K. D.Ko:S.sn,t and E. von LavanteS
Texas A&M Chnir:ersrt.y, Coiiege Station, Texas

A nuwericpl method has been developed lasing !he technique of Lock and Theodorsen as described by David-
son to desig~m d ezalyze counter-rcrtPtieg propellers. Tbe design metbod develops the optimum propeller
g e o m r y by cakuhting the planform and twist distribution for each propeller disk through the use of specific
inputs of engine shaft horsepower, diameter, and disk spacing. The analysis method calculates the performance
of a given coenter-rotating propeller system at anv fligbt condition. Using the NACA fourdigit airfoil family,
the performance of a coun;er-rotating propeiler desig~fir, a @OYIB flight cordition was investIgutd in thz design
a d analysis mode.

Nomenciature 0 = solidity
9 = resultant velocity angle
= propeller disk area = advance angle
= interference angle 90
Q = rotational velocity
= Betz's coefficient w =dswnwasl: velccity
= chord

-
= drag coefficient
lift coefficient
= power coefficient
Sdwripts
B = back propeller disk
= torque coefkient
F = tront propeller disk
= thrust coefficient
= diameter
= advznce ratio
= circulation 1. Introduction
HERE has been renewed interest in finding a more effi-
= inverse lift-to-drag ratio
=Mach number
= revolutions per second
T cient replacement for current aircraft propulsion systems,
with one such approach utilizing a counter-rotating propeller
= shaft horsepower config~ration.An initial theory regarding the mechanics of
=ambient pressure, constant of calculation counter-rotating propellers was developed by Locks in 1941.
=torque Since then, many investigations into the advantages of
= freestream dynamic pressure, constant of counter-rotating propellers have been conducted. For exam-
calculation ple, Biermann and Gray2 conducted full scale wind tunnel
= radius t&ts on counter-rotating propellers in both tractor and pusher
=radial distance, constant of calculation configurations. It was found that an 8 to 16% increase in pro-
=constant of calculation peller efficiency could be gained depending u p n installation
=thrust position. In a later test, Biermann and Hartman3 found that
= axial interference velocity the performance of counter-rotating propellers was
= freestream velocity significantly improved at lower advance ratios. McHugh and
= radial interference vclocity Pepper4 have shown that the counter-rotating propeller con-
= resultant velocity figuration is highly receptive to the use of aerodynamically im-
= rearward helical displacement velocity proved airfoil designs. Other investigations into the perfor-
=tip loss factor mance of counter-rotating propellers conducted by Gray5
= percent radial location have indicated that the overall efficiency of a counter-rotating
=angle of attack propeller is not seriously affected by changes in rotational
=blade twist angle speed or small changes in blade angle of the aft propeller disk.
= angles defined in Figs. I and '2 These changes did, however, have a moderate effect when the
=angles defined in Fig. 2 propeller was operated at peak efficiency. In an experimental
= propeller efficiency study, MillerVound that the vibration of counter-rotating
= ambient air density propellers caused by mutual blade passage or by blade passage
through the wake of a wing was not significant. Bartlett7 has
Presented as Paper 84-1205 a? the A I M / S A E / A S M E 213th Joint shown that locking or windmilling one of the propeller com-
Ptopuiskm Conference, Cincinnati, OH, June 11-13, 1984; received ponents of a counter-rotating configuration has a detrimental
A p i l IS, 1985: revision received Sep. 4, 1985. Copyright O effec? on totd propeller efficiency. For example, a counter-
American Institute of Aeronautics and Astronautics, Inc.. 1985. All
rights reserved. rotating propeller with one propeller disk disabled results in a
*GraduateRescuch Assistant. Aerospace Engineering Dcpartmmt. totd propeller efficiency that is lower than the individual dfi-
Studat h k A I M . ciency of the rotacing propeller.
t Associate Prlfcsror, Aerospace Engineering Department. This brief literature surwy has indicated that counter-
Associate Fellow Ar \A. rotating propellers exhibit many advantages over single-
$Assistan! Profca: ;, Aerospacs Enginwing Department. Member rotation propellers such as higher peak efficiency, better off-
&r* A
deipn nerformance. and a reduced total torque of the svstem.
 PLAYLE, KORKAN AND VON LAVANTE J. PROL ilLSION

The increa5e in peak efficiencjsand improved off-design per- development. The results of the comparison of the two
f~--~,~ ofccouncer-rotating
e iystems allow far smaller pro- methods by Playle16 have shown that the Lock and
ph.sion units to be installed on the aircraft. The disadvantages Thedorsen method, given by Davidson, was more accurate.
of counter-rotating propeEer configwations include gearbox Davidson's approach yielded realistic designs of counter-
complexity and an increased vibrational state caused by the rotating propeller configurations in terms of planform and
periodic blade passage. Research conducted by Strack et. al.' performance values and was adaptable to further develop-
has shown that .with current improvements of present day ment. In addition, inclusion of drag and compressibility
technology, lightweight and reliable counter-rotating propeller shown in this work were the most useful modifications using
gearboxes can be built. Thus, it is evident that counter- this theoretical approach. Davidson's method was also found
rotating propellers can indeed offer a more efficient means of to be adaptable to include off-design analysis, disk spacing,
propulsion. blade weep, and different angular velocities on each propeller
The work described previously has been limited in the disk.I6
analysis of counter-rotating propeller configurations.
Therefore, it is necessgly to develop a condstent methodology 111. Theoretical Design Development
to obtain a design appoach as well as detailed analyses of The main advantage of counter-rotating propellers stems
counter-rotating propeller systems. In the present study, ex- from the swirl velocity losses of the front propeller disk being
isting theoretical models for the design and analysis of recovered by the aft propeller disk. The front disk imparts a
counter-rotating propellers were investigated and compared to tangential velocity to the air as it passes through the front pro-
determine a method that could be developed for w e as a peller disk plane. This swirl velocity acts as an additional
design and ar&h method. 11,pular velocity for the aft disk, without the power plant hav-
ing t o drive the aft disk at a higher angular velocity.'Figures 1
11. SelecQonof Design Method and 2 show the velocity diagrams for the front and aft disk,
The majority of the work descri'bed in the previous section respectively. It may be noted that a tangential interference
was conducted with existing single-rotation propeller blades in velocity is recognized by the front disk, but is typically an
a cou~ter-rs;z;iugconfigurativn. Therefore, a requirement to order-of-magnitude smaller than other inttrferenx velocities
develop a theoretkal means by whizh counter-roiating pro- and therefore neglected, resuking in a first-order theory f o ~
pellers covld be dt5igned and analyzed was initially in- the design and analysis of counter-rotating propellers.
vestigated. Three previouly developed theoretical models for Airfoil data, such as lift, drag, and angle of attack are
the design of counter-rotatuig propellers were considered and specified for each radial location along the propelier blade.
examined for feasibility of further development. This is done through the use of airfoil data banks, utiliziag
The first method, developed by G i n ~ e lmakes
, ~ no presump- either tabulated data or empirical formulations that yield air-
tion about the distribution of spanwise circulation across the foil lift and drag as functions of mgle of attack and Mach
blade. It is restricted to the use of aerodynamics that result number.'' The airfoil data used in the calculations during the
from propellers, which were built in accordance witk the struc- design prccess was selected to maxirnize the lift-to-drag ratio
tural constraints that existed at the time the theory was of khe airfoil used at each radial location dong the blade. In
developed. This structural restriction was applied in 1943 and regions near the hub, where the propeller blade quickly transi-
is not pertinent wiih respect to the current state of technolog;. tions from an airfoil to a right circular cylinder, a different ap-
For example, with the advent of composite materials and high proach has been taken. To accommodate this transition, the
strength alloy metals, the structural assumptions that Ginzel chord length is linearly interpolated from that calculated to a
based his method on are no longer valid. structurally feasible cylinder at the hub. The design procedure
Other methods for the design of counter-rotating propellers involves calculations, which include division by the lift coeffi-
include the SBAC methodloand a theoretical model developed cient, to determine chord length. Since the lift coefficient of a
at Un&ed Technologis Research Center (UTRC)." The circular cylinder is zero, it may be noted in the ciwelopment of
SBAC method is computationally very cumbersome as it is a the theoretical model that it is necessary to maintain a finite-
design-by-analysis method and requires &ta interpolation and lift coefficient to insure arriving at positive values of the blade
cross referencing, and the UTRC method requires extensive chord.
computer time and meznory. Once :he airfoil data have been specified, a comparison
In another method, Naiman12 uses a modified strip theory scheme is initiated whereby the power input to the propeller
in which approximations for the interference relations from and the power absorbed by the propeller are matched. This
the front and aft propeller disks are assumed such that the in- iteration scheme begins with an assumed value for the pro-
terference calculations are functions of each propeller disk in- peller efficiency, which need not be accurate since q will be
dependently. Naiman uses sectional aerodynamic characteris- changed by the iteration process. From the assumed propeller
tics as a basis for his derivations, and hence allows the use s f 'efficiency Betz's coefficient is calculated, which is the
various families of airfoils and different structural con- parameter through which the iteration process is conducted.
straints. Sectional circulation to determine the interference Davidson has derived Betz's coefficient using calculus of op-
relations is also utilized by Naiman, and therefore the method timization as:
is adaptable to changing technology.
Another counter-rotating propeller calculation method in-
~ . ddescribed
vestigated was that of Lock1 and T h e ~ d o r s e n ' as ~~
by Da\idson.Is This method uses an interpolation scheme be- Calculation of the rearward helical displacement velocityi4
tween both propetler disks for the interference of one disk may be accompiished by the expression
upon the other, and requires the use of the circulation of the
entire configuration as an input for calculation of the ill-
terference relations. Similar to Nairnan's method, Davidson's
apprgach is easily expanded to off-design analysis. Using the experimental values of circulation by ~heodorsenl~
Since Ginzcl's method was considered unacceptable, as were as a function of r/R and (V+ w ) / n D , ihe circulation K ( x ) can
the SBAC and UTRC methods for ease of use and applicabil- be found for each radial locati~nas noted in Fig. 3.
ity t o design, the nethods of Nairnan and Davidson were The present method for d&.gning and analyzing counter-
chosen for further .development and investigation. The results rotating propeller configurations is two-dimensional by virtue
of computer-generated designs for these two theoretical of the strip-analysis approach. However, the propeller is
kT'iod& Were anpared to cieierrnar~er.iur;G i--Gi G- s ' i ~ . ~ ; =
:&rc-&s
~;r;z&$ &.m.md ZZL! $k::f?~n
Zi 1. ~k211"1.~32.
fOr re;rsom of accuracrj, computation:l efficiency. and funher a method to include the differen& between two- and three-
*
 Fig. 1 Velocity dhgnar for froat prapelltr disk.

Fig. 3 Circulation functifs*Tor dud-rotation p r o ~ c n four

; blades,
froat and back propeller disks (Ref. 13).

Equation (6) is integrated to find the total power ckfficient

for the power absorbed by both the front and aft propeller
% disk. The power absorbed IS compared to the power input to
Fig. 2 Velocity diagram for back propeller disk. the propeller by the power plant, which is dictated by the
available horsepower of the power plant specified in power
coefficient form by

dimensional flow is required. Davidson derivzd a version of

I f the two values of the power coefficient do not match within
Lock's tip-loss factor15 for specific kpplication to counter-
rotating propellers, which is a well behaved function that is a given toleranc~,Betz's coefficient is adjusted and the power
different from a single-rotation propeller. A reformulation for absorbed is recalculated following the outlined procedure. The
adjustmerrt made to Betz's coefficient represents alterations in
Lock's tiploss factor that includes drag has been derived15J6 thp planform of the propeller blade, and thus propeller blade
and takes the form
loading. Should the power absorbed be lower or higher than
xo= qs/(2p-am) (31 the pow& input, Betz's coefficient is increased OF dmeased,
where respectively. The iteration process is continued unti: the power
absorbed by the propeller is within a predetermined percent-
age of the power input to the propeller.
The condition of the power absorbed by the propeller being
eqwl to the power input, defines the optimum propeller for
the given flight conditions since the configuration was de-
signed using tbe maximum lift-to-drag ratio for the chosen air-
foil at each radial location.Also, since the form of Beu's coef-
ficient has been determined through calculus of optimization
The solidity-lift coefficient for each radial location on each and used as the iteration parameter, the resulting design is
disk is then given by considered to be the o p t i m u ~counter-rotating propeller for
the input condition.
The physical geometry of the propeller is calculated axe the
solidity-lift coefficient radial distribution has been defined
through the above iteration process. By iterating between the
resultant interference velocities and the resultant total veloc-
ity, both the interference and total velocities can be found by
the following expressions:
with the value of CD/CLarrived at through use of the airfoil
data banks. Having calmbated the solidity-lift coefficient, the
differcathi gower coefficients, which are measures of the
power absorbed by each propeller disk, can be computed by
 PLAYLE, KOKKAN AND VON LAVANTE J. PROPULSleiN

where

The differentialthrust and torque are functions of i!x resul-

tant velocities and calculated by the following expressions:

Integration of Eqs. (14) and (15) results in values of thrust and

torque, hence the thrust coefficient, torque coefficient, and ef-
ficiency, i.e.,
-

Fig. 4 Assumed values of maximum thickness to chord ratio and

design lift coefficient distribution for NACA 4 digit design, both pro-
peller disks.

At this point, the radial values of the blade planform and

twist distribution for each propeller disk can be cJculated
from

where

It is noted that the values for angle of attack were selected

through the iteration process. in the present method, which
uses the airfoil data banks, ,?;: numerical routine has used
values for angle of attack that correspond to maximum lift to
drag ratios at each radial location for each propeller disk.
As previously discussed, there exists a difficulty in the
calculation c;f chord lengths in the region near the hub. In this
segment of the blade, the structural requirements force the air-
foil to take on the characteristics of a circular cylinder. As a
result, the solidity-lift coefficient may take on negative values
as the shmk region of the propeller is approached because of
large values of C , and relatively small values of CL.AS noted 0 l I 1 I I I
in Eq. (51, for large values of C,/C,,the numerator becomes
negative. Examination of Eq. (20) indicates that for negative
. 0 Q.2 0.4
RADIAL
0.6
LOCATION. r/R
0.8 1.0

solidity-lift coefficients, the chord length would also result in a Fig. 5 Resultant chord to diameter distribution using the NACA
negative value. In the present study, it was determined that the four-digit airfoil family, front and back propdct disks.
most acceptable means to calculate chord length in the vicinity
of the hub region would be through a linear interpolation of
the chord length from the radial location, where the slope of foil family to determine the optimum counter-rotating pro-
the planform goes positive to a value of zero at r / R =O. peller design with respect to performance.
Once the chord length has been calculated, a new value for
the solidity-lift coefficient must be found to complete the
calculations for the radial variation of thrust and torque, i-e.,
IV . Application of Counter-Rotating
Propeller Design Method
Using the NACA four-digit airfoil family, a design for a
aCL= (No. Blades)(Chord)CL/Dxx (24) counter-rotating prppelIer system was performed. The
specified input flight conditions corresponded to a typical
By this present approach, any possibility of negative chord commuter class aircraft, i-e., freestream velocity of 625 ft/s,
lengths in the region of the hub during the numerical design altitude of 28,000 ft, shaft horsepower engine of 900 with the
procedure is eliminated. condition that one engine is available for each propeller disk,
Once the theoretical mrvlel had been hlly developed, the engine rpm of 1500, and a blade diameter of 7.50 ft. The -air-
design method was-appiieciuriii~ngik S A G :uui-di& sir- '-=I -c.v;-..M +hiz-Gnoce
I..M.II.U... ." -
C...r....-.,rn ohnrr( rl~crrmn~~ftnn
--.-..-----.--,
-
snwr! !n _r!g.
JAN.-FEB. 1986 COUNTER-ROTATING ?ROPELLE KS 61

4, was dictated by propeller structural considerations. The ent design and result in 1, i- 0.798 and a, = 1.093, hencc, a
assumed design lift coefficient radial distribution, shown in total propeller efficiency of 0.946. It may be noted that a pra-
Fig. 4, results in a well accepted integratbd design lift coeffi- p e k r efficierlcy greater than 100Vo was found for the pro-
dent of 0.52. pelle; back disk. This is a direct result of the swirl velocity
In a study conducted bv Korkan and Playle,'* it was shown component effect and using the conventional expression for
that for similar flight conditions, an efficient counter-rotating calculating propeller efficiency.
propeller configuration consisted of six blades per disk. The theoretical model previously described, designs a
However, it was also shown that there is a small chan~ein counter-rotating propeller configuration for a given operating
total progclier performance between the fdur-, five-, and six- condition and specified geometrical constraints. The analysis
b b e ptx disk configurations. Based upon this result, a of the counter-rotating propeller design in an off-design
counter-rotating propeller system of four blades per disk was operating condition is also required to determine thc full ex-
chosen for the present investigation. The spacing between the tent of the performance envelope.
d i s h was assumed to be zero, in addition to identical rpm for
each propeller disk.
Utilization of the counter-rotating tharetical design model
for the conditions previously mentioned results in the
developed blade planforms shown in Fig. 5. It may be noted
t b t there is little difference between the front and back pro-
pelier disk beyond the radial location of approximately 40%.
Before this poiut, tht: back propeller chord becomes larger
than the front propeller chord. The maximum chords of both
diks r e d in approximately nine inches, which is considered
acceptable. The propeller design b!alle twist distribution,
shown in Fig. 6, indicates a significant difference between the
front and back propener disks. This difference in twist comes
directly from the effect of the swirl velocity component from
the front propelkr disk impinging upon the back prGpeller
disk, with the back propeller disk requiring less twist i o main-
tain the angle of attack corresponding to the maximum lift-to-
drag ratio.
The differential thrust distributions for the present design,
shown in Fig. 7, indicate that the back propeller disk will
result in a higher integrated thrust coefficient than the front
propeller disk. Fcx 'his case, C,=0.458 as compared to a
CT=0.330, respectively, resulting in a total thrust coefficient
of 0.788. It may also be noted that the maximum dCT/dxdoes
not occur at the same radial location for the front and back
propel!er disks. For example, the front propeller disk max-
imum dC,/dx occurs at approximately the 70% radial loca-
tion, which is considered optimum, whereas the back propeller
disk maximum dCT/dx occurs further inboard at approx-
imately the 60% radial location. The differential power coeffi-
cient distribution, shown in Fig. 8, yields identical distribu- RADIAL LOCAT 1ON, f/R
tions for the front and back propeller disks due to the condi- Fig. 7 Resultant differea:ial thrust coeffiint distribution using
tion of equal horsepower being supplied and absorbed by each NACA four-diit airfoil family, front and back propeller disks.
propeller disk. Tbe integrated power coefficients, i.e.,
C, = 1.378 and C , = 1.395 with the difference resulting from
the iteration tolerance, yieids a total power coefficient of
2.773. With the thrust and power coefficient values defined,
the resulting propeller efficiencies may be found for the pres-

I I I 1 I
0 0.2 0.0 0.6 0.8 1.0
0 0.2 0.4 0.6 0.8 1.0
Rmlk b c A f l # r r/R RAP~AL LOCATSOW, r/R
wii-6 pmpdk b M e twist tliotribetioa asiq NACA four- Fi. 8 Resultant differential power coeffir3ent ruing the NACA
Sg2=%S-,-.s---a: F-+gi! M d ! fear$?. +??? red w ~ F y e e ?nj*c-
I I - ----- --7 *-- (C -- eWvL-- ,, - - --- -
"* -" -

REPRODUqlB lL ITY OF THE OR1 G I ML PAGE

-- I S PO@R
 PLAYLE, KORKAN AND VON LAVANTE J. PROPULSION

- - - -- BACK DlSK EFFICIENCY

_---
, - FRONT DISK EFFICIEHEY
TOTAL EFFICIENCY

0@
/c---
--+-----
DES~GH POINT
-I

. /'

Fig. 9 Off-design efficiency of NACA f@nr-digitcounter-rotating 7 !.G 2.9 3.0 G.0 5.1 CSrj
propeller configuration. ;,: !:,'.CE PLTIc, !
Eig. 11 Off-design power coefficient of NACA fourdigit counter-
rotating propeller configuration.

BACU DIS.: THRUST COEFFICIENT

-
---0-

- FRONT DISK THRUST CMfFICIEWT

TOTAL THRUST COFFFICIEWT
result in the actual airfoil data. From these initial airfoil ap-
proximations, a value for the solidity may be found,
o =(No. Biades)(Chord)/Dxx (25)

Once the value of the solidity has been obtained, values for
the interference velocity angles can be computed. Since these
angles are functions of the aerodynamics of both propeller
disks, the following expressions yield values for BF and BB:

The jnterference angles define the angle of attack for each

radial location since

These angle-of-attack values yield airfoil characteristics from

the airfoil data bank and, therefore, lead to an iteration pro-
cess by which the final airfoil data used in the performance
calculation may be defined. This iteration scheme continues
FLg. 10 Off-deign thrust coefficient of NACA fourdigit counter-
until the airfoil data used at the start of the process results in
rotating propeller conriuration. the interference conditions tha-t yield identical airfoil data.
Once the iteration process has converged, the performance
of the counter-rotating propeller is found through the use of
V. Theoretical Analysis Development Eqs. (8-19) as in the design method. There is no difficulty in
A major reason for the selection of the Lock and calculating values of the circulation function K ( x ) for the off-
Theodoisen design method as given by Cavidson, is because design case once the interference values of the velocity are
of its adaptability to the propeller analysis mode. The analysis found. Using the above analysis method, a counter-rotating
method described in the following section uses an iteration propeller can be analyzed at any given operating flight
scheme to obtain the final results, which converges on a+foil condition.
characteristics. To begin the ar;.a!ysis iteration scheme, the
propeller blade geometry, such as radial distribution of chord M. Application of Coanter-Rotating
length, twist distribution, and flight condition are specified. PropeUet Analysis Method
7-L- :-:*:-I -.-a --
A s i r r z a ~ ~ . a Wa Q 1 U S
C,.->L-
I v r
.:-A'--.
tIlC r u a --..-.- 2-*:--*- - m - -
L i I ~ i u b c i=
rzs.>-- , --#---A
,grJv .%S%CLLU, - ;hc ah&><i-i
&i~&g &c%& &3<r:fYE2 111 T;;~=.~z.zs
=;-PC
and need only be approximate since the iteration method will tion, an investigation was conducted tor the oftdesign perrbr-
--"--* * --* -
PAGE I S*P ~ O R
 COUNTER-ROTATING PROPELLERS

5 ~ c of e the counter-rotating fied-pitch propeller configura- Acknowledgment

tion, designed in Sec. IV, usirrg the NACA four-digit airfoil This research was supported by the National Aeronautics
family. The variation of the advance ratio, J, was obtained by and Space Administration Lewis Research Center, Grant
holding the frcestream velocity and twist distribution constant NAG 3-354.
a d varying the rpm of both the front and back propeller
disks, keeping equal values of rpm for both disks. References
The rcipultina, values of propeller efficictrcy are given in Fig. Lock, C. N. H., "Interference Velocity for a Clow Pair of Contra-
9 an8 indicate a very hrasd rand flat efficiency map across a Rotating Airscrews, A &. M 2034, July 1941.
b e variance in advance ratio. This result shows that a ~icrrnann,D. and Gray, W. H., "Wind Tunnel Tests of Single-
counter-rotating propeller designed for cruise, will also be ef- and Dual-Rotating Pusher Propellers Having from 3 to 8 Blades,"
ficient in off-design flight ccnditions such as take-off and N?,CA ARR (WR L-359). rebruary 1942.
climb. Figure 10 indicates a large ixmease in the thrust coeffi- Biermann, D. and Hartman, E. P., "Wind Tunnel Tests of Four
cient at the lower advance ratios, which is also a dcfinite ad- and Six Bldde Single and Dual-Rotating Tractor Propellers," NACA
vantage for the takeoff and climb conditions. The power coef- Rept. 747, 1942.
ficient decreases a t the off-design conditions as given in Fig. 4 ~ c ~ u gJ.h ,G. and Pepper, E., "The Characteristics of Two
11, but does so in a ratio with the thrust coefficient to yield the Model c -Blade Counterrotating Pusher Propellers of Conventional
and 1mp;oved Aerodynamic Design," NACA ARR (WR L-404). June
relatively flat efficiency curve. These performance calcula - 1942.
tions have been accomplished at fixed pitch conditions, and it ' ~ r a W.
~ , H., "Wind Tunnel Tests of hal-Rotating Propellers
may be assumed that the perforn~anceenvelope would be fur- with Systematic Differences in Number of Blades, Blade Settifig, and
ther enhanced by utilization of a variable pitch a r constant Rotational Speed of Front and Rear Propellers," NACA ARR L4E22
speed mechanism. This series of calculations has indicated the (WR2-80), May 1944.
full extent of the advantages of counter-rotating propellers 6~iller.M. F., "Wind Tuqnel Vibration Tests of Dual-Rotating
when analyzed in an off-design perfomance situation. Fur- Propellers," NACA ARR 31 1 1 (WR L-405), Sept. 1943.
ther, this off-design analysis shows clearIy that counter- 'Bartlett, W. A., "Wind-Turmel Tests of a Dual-Rotating Propeller
rotating propellers are not only efficient at the design point, Having One Component Locked or Windmilling," NACA ARR
but are also extremely efficient at operating conditions other L5Al3a (WR L-214).January 1945.
%track, W. C 'hip, G., Woesbrich, A. I.., Gotston, J., and
thar; the design point. Bradley, E., "Technology and Benefits of Aircraft Couaterrotatigg
Propellers," NASA TM 82983, October 1982.
9~inzel,F., "Calculation of Counterrotatin3 Propellers," MACA
TM 1208, March 1949.
VII. Conclusions
The theories of Lock and Theodorsen, as described by
Davidson, were determined to be acceptable theories for the
'O~fandat
af British Aircrft Constructors. Ltd., London.
"~ober,L. J., Private communication, NASA Lewis Research
-
d Method of Propeller Petformance &timotion, Society

design and analysis of counter-rotating propelIers. Using this Cer ter, 1983.
theoretical model, the full advantages of counter-rotating pro- "Nairnan, I., "Method of Calculating Performance of Dual-
Rotating Propellers from Airfoil Characteristics," NASA ARR 3E24
pellers have been shown, and indicate approximately, the WR L-330). May 1943.
same efficiency over a large spectlum of operating conditions '3~heodorsen,T., "The Theory of Propellers-Determination of
other than the design point. However, with regards to the the Circulation and the Mass Coefficient for Dual-Rotating Pro- '

limitations of the current model, the radial interference on the pellers," hACA Repr. 775, 1944.
front propeller disk is neglected, which is a n assumption that 14Theodorsen,T., Theory of Propellers, McGraw-Hill Book Co.,
may not be totally valid, especially in the hub region. The cir- New York, 1948.
culation functions of Theodorsen are calculated for optimum lS~avidson, R. E., "Optimization and PerformanceCalculation of
propellers with equal angular velocity on each propeller disk. Dual-Rntat~onPropellers," NASA TP 1948, Dec. 1981.
Although jt has been suggested that these values for circula- '6~Iayle,S. C., "A Numerical Method for the Design and Analysis
tion are still valid for nonoptimum propellers, a question as to of Counter-rotating Propellers," M. Sci. Thesis, Texas A&ii Univ.,
June 1984.
their validity for the cases of different angular velocities on I7Korkan,K. D., Private communication,Texas -3 Univ.. 1984.
each propeller disk may be posed. These cited areas are left for "Korka& K. D. and Playle, S. C., "On the Design of Counter-
future studies. rotating Propellers," SAE Paper 830773, April 19153.