PROCUREMENT EXECUTIVE.

MINISTRY OF DEFENCE
AERONAUTICAL CURRENT RESEARCH PAPERS COUNCIL

Some

Remarks on the Induced Velocity\$ Field of a Lifting Rotor and on Glauerts Formula by
A.R.S. The City Bromwelf London University,

LONDON:

HER MAJESTYS STATIONERY 1974 PRICE 65p NET

OFFICE

C.P. N0.1301~ October 1971

SOMEREMARKS THE INDUCED VELOCITY FIELD OF A ON LIFTING ROTORAND ON GLAUERT'S FORMULA

A. R. S. BRAMWELL The City University, London

The induced velocity field of a lifting rotor is discussed in relation to the pressure field rather than the vortex wake in an attempt to obtain a clearer understanding of the relationship between the induced velocity and rotor forces. A number of results are are compared with those obtained from derived and, where appropriate, the theory of the vortex wake.

An investigation into the validity of Glauert's formula indicates that it appears to be true for all rotor loadings for the linearised, "high speed", case. Calculations of the induced power show that for typical rotor

loadings the power in hovering flight is about 10 per cent greater than the "ideal" induced power, rising to about 15 to 20 per cent greater in forward flight.

*Replaces A-R-C.34 822

-21. INTRODUCTION

A knowledge of the induced velocity at a helicopter rotor is essential for the calculation of rotor blade forces and moments. For many applications a detailed knowledge of the complicated induced velocity field is unnecessary and Glauert's formula9 for the mean induced velocity often gives acceptable accuracy. Glauert's formula has sometimes been misinterpreted, however, and used wrongly as a basis for detailed calculations of the induced velocity field. This note considers the pressure field of the lifting rotor, rather

than the vortex wake, in an attempt the relationship between the induced Certain symmetry relations, obtained are used to investigate the validity throughout is based on a linearized

to obtain a clearer understanding of velocity and the rotor forces. from the theory of the vortex wake, The discussion of Glauert's formula. analysis. no practical application but it is into the development of the induced

The work of this note has probably hoped that it leads to a better insight velocity 2. field.

EQUATIONSOF MOTION Euler's equation of motion for a fluid is

If the motion is regarded as being that of a small disturbance superimposed on a uniform ilow of velocity v , i.e. 91 = VtG that squares and products equation becomes of the components of

7 such Euler's

V are negligible,

8t

3. + 5.Q;

= -$V~I
and remembering that is constant, V is a constant vector

Taking the divergence of both sides of (2), V.ij = VS = 0 by continuity, we have, if p - $ V'p i.e. = V (7. v2p PV = 0 ) = 0 since

Thus, for thelinearized problem, the pressure field satisfies Laplace's For incompressible flow, (2) equation, even when the flow is unsteady.

-3-

can also be written

giving potential" since, as can be seen from (3) @ is known as the "acceleration the gradient of @ g ives the fluid acceleration. first introduced by Frandtl in 1936, has The "acceleration potential", wings of been used by Krienes' and tinner2 to find the loading of elliptic arbitrary aspect ratio and circular wings respectively. The latter formed the basis of Mangler's work3 in which he calculated the induced velocity distribution of a disc carrying a load similar to that of a helicopter rotor. We illustrate the calculation of the velocity Then, if components by considering

the particular case in which 3 q VT. (2) becomes, for the steady case,

V = bz

t vd + uk

Assuming the disturbance

velocity

to be zero at a great distance

ahead

of the disturbing load, the velocity components anywhere in the flow field For example, equation (6) gives can be obtained by integration.

(7)
is to construct solutions to Laplace's equation boundary conditions and then to obtain the distIn Mangler's work urbance velocities from integrations such as (7). Laplace's equation was found to be satisfied by a sequence of associated Legendre functions which, in addition,gave the required pressure jump across the disc. These could be regarded as a series of pressure "mode shapes" gave an and it was found that the first two, in a suitable combination, acceptable representation of a helicopter rotor loading. 3. THE UNIFORMLYLAODEDDISC The uniformly solutions loaded disc is one which leads to comparatively insight into simple The problem, briefly, satisfying the appropriate

and gives a useful

the mechanism of the development

-4-

of the induced velocity

field. that a function which satisfies acnxs a surface element

It is well-known from potential theory Laplace's equation and gives a discontinuity is

' where n is the unit normal to the surface andvf) our problem, the pressure difference across the rotor the field due to a three dimensional distribution doublet For a uniform pressure

is, in the terms of disc. Equation (8) is

or dipole4.

where the integration is taken over the whole surface, which need not be necessarily plane or closed. which the surface subtends at any point Now the solid angle (p can be expressed' as

so that

(9) can be written Q,= OI-

as -npq w

p = afq/4r

. py

that is, the pressure at any point in the field can be expressed very simply A sketch of the in terms of the solid angle subtended by the surface. pressure field in a dimetral plane is shown in Fig.1. Further, the acceleration ii= field -V@ a can be obtained from ( 3 ) as

= -2-&t! vSg&v($)dS = -$$j$.vv (+)ds

By an extension of,Stoke's theorem4.

(II)

where th& suffix C denotes the boundary of the surface Hence, (11) becomes an element of the boundary.

and

di

is

-5i *AJl

&VPsc

OS xdi
0

The integral "actor we note coinciding this result and ring, by them field that due it with

on the to is the

right

hand carrying in form

side

of

(12)

gives pressure

the

acceleration distribution Of a vortex and ring

a surface identical bounding that,

a uniform to the velocity

field

edge since

c , fig.2. the pressure

Katzoff field velocity

and others' has to satisfy

obtained Laplace's of a was

by arguing the the to

equation vortex used load.

same boundary two fields determine

conditions can be regarded

as the

potential The result to carry

as equivalent. wings designed

camber

lines

of

a uniform

4. CALCULATION Since determined pendent velocity velocity those are fig.3, the of the once the

OF THE NORMAL INDUCED VELOCITY pressure the disc field satisfies distribution To find to the to along to the rotor the flight the disc, Laplace's is equation It velocity", that (4), the (5), Thus, to by the flight if it is completely indethe

pressure incidence. normal found

prescribed. "induced we note direction.

is also i.e.

components components, along

appropriate (6), will Lt.' be and W'

by integrating

equations

and perpendicular components normal w-

velocity the velocity

and perpendicular disc x f flight (4) and is given x

direction,

&Qs between equations the

w'strl

where rotor

x disc.

is

the Then,

angle from

direction (6)

and

the

normal

to

the

The pressure the along tables the given axis

gradient

normal and to the

to

the

plane

of

the

disc along fig.4.

can be found a diameter

from and

by Kuchemann perpendicular

Weber6. disc are

The values shown in

-6-

From equation

(12) and fig.4

we easily

find

that

this

component is

. (4)
In particular, along the axis of the disc, f = 0 we have

(5)
Integrating, we get
13-p = -+p 1 It(+) I ) @ j0 .

. (6) (7)

This variation

is shown in fig.5 (propeller) case is found by putting

The induced velocity for the axial any f, x= 0 in (Xi), giving,for

i a_F!d: = - PV s ad ~

=
that u = 0 as *aI

(8)

Then, integrating (15) and ensuring i.e. upstream of the disc,we have

&

The axial

variation

of induced velocity

that the slipstream velocity mate value within a diakter's

and we see has reached nearly 90 per cent of its ultidistance of the disc.

is shown in fig.6

The integration of (4), which corresponds to the linearized form of Bernoulli's equation, shows that the axial velocity component depends only on the local pressure, i.e. & = Thus justifies, for f PJ' + constant. the linear theory at any rate, the idea of the"independence of blade elements" of elementary propeller theory.

For points downstream of the disc, and for all points for which r>!? the constant in the above expression for u is zero and the axial velocity

-7component earlier by the uous fig.1, and that since there for of is that disc, everywhere. that it the simply the local proportional pressure upstream also of to minus is the local to axial of pressure. the solid Now, we saw angle must pressure the disc subtended be continfield, ( 1"> R 1, However, increases, at f theory =R

proportional disc the in the

so that It axial the

the from

velocity the of disc

follows is

sketch the of i.e. of plane the

velocity becomes velocity

zero

actually

negative

as we proceed

downstream. continually wake, the vortex

we know that must points

downstream of as is velocity, assumed,

be a discontinuity behind the disc,

a definite course, in

propellers. Koning7 has also disc, ahead simplify x= of investigated but of the 90 is the "high speed" case, x = 90. From (18) the the the impression is false radial given component in the of work velocity that for a disconof the assumpa

uniformly tinuity tions

loaded extends made to

disc

and is

a consequence

analysis.

5. FORWARD FLIGHT, Another the induced case velocity

interest is

In the wings, bounding the in the the

method integration lifting of

used

by Katzoff was first

and others' performed 12). point

to

calculate of

the the

camber contour

of

on an element Thus, X' , 3' with and

surface the element gave

(equation dL' at

dx,
X, 7

dy'denoting the point

components the field,

integration

Expressing integrating little over manipulation

these the

quantities circular

in contour

terms of

of the

polar Potor

rotor disc,

coordinates we find after

and a

that

where

jJ,

T/R

-8Along rotor, y the longitudinal = T , and axis (21) of symmetry, and ahead of the centre of the

becomes

where k(b)
Similarly, along

is

the the

complete rear

for

p > 1
integral axis

bE(bj
of the y) first : 0 kind. , we have

longitudinal

The induced Along the lateral

velocity axis of

along the

the rotor,

longitudinal +I$ = r/z,we

axis have

is

shown

in

fig.7.

Now

Hence & within 6. the disc,

has the

constant =

value *&/&!ord 61 outside the

the disc.

lateral

axis

and -%{I-

MOMENTUM CONSIDERATIONS At this stage at it is useful to consider disc in the forward the relationship flight of between and the the the

pressure velocity the induced

jump there.

a position In axial U;:

on the flight, the

induced

assuming well-known in the form

rotation

slipstream the
thrust

to be negligible, velocity

we have

relationship

between

and

-I- =
and, for an arbitrary

qd(vtu;)v;
the differential

.
form

(w

loading,

df

zp

(vtv;)v,hl

.w

-9is assumed to hold. The validity and the this in the is to in linear result of this relation, paper8it as has been from the as stated is shown shown fact general that by Glauert7, to be untrue earlier. the flow, induced and that FOP at (26). rotor in

has necer?been general. It

established is true for of is the it order Glauert', is

a recent case arises

The simplicity velocity the these flow increment behind

same direction confined apply

as the

disc easy to

to a well-defined momentum principles velocity the formula

slipstream. and arrive for calculating

reasons Now, in

obtain in

a mean induced 1926, proposed

performance,

T=
where Although for this formula no proof is that is v' of it = (27)

2pAh
da has been to the given, (25) induced in the the velocity justification hovering of case,

(27)

claimed

reduces

= 0

and assumes the correct loaded wing when V An interpretation stream, speed v havjng the and being component

form for large. the

an elliptically

of

formula as the downwards .

(28)

is by it

to

imagine it

a cylindrical upon the rotor (27) is ultimately

slipat acquires

s3me diameter deflected of 2U;

rotor,

impinging so that momentum

a downwash readily that it it

By applying of the such

principles

obtained. leads easily from

The only to (27)

virtue but

a fictitious has often and, indeed,

representation been that gained

is that

impression

is"obvious"

momentum

considerations

a differential

form,

C/J- = 2p v v; dA
analogous Now, rate means velocity as in solved of this of to (26), is the also lift applicable. of of is the the rotor air, must although of

.
be accountable considerations flight terms case, of x= path the in terms are and the local 90 of

(26)

change

momentum rotor

momentum to simply the in the

by no induced

simple

when the

inclined

can no longer (26). To see this

be expressed we refer and

loading, which we

back

to

analytically, integration.

(22)

(231,and being we have air

consider integrated seen

a physical is the

interpretation pressure to it of develops pressure becoming gradient the under i.e. rotor's some gradient twice

The function which

shown in fig.4, QPlag local vertical acceleration. the influence in edge of front the the pressure of pressure the leading the an upwash, leading distance again

to be proportional the rotor

As the field disc. gradient edge.

approaches imparts air moves

comes the

which As the reverses At the

an upwards downstream

velocity,

and a downwash trailing decreases edge finally the

behind reverses

and the

downwash

gradually

- 10 the value at the centre of the rotor. In this special case the slipstream

boundary, or vortex wake, degenerates into a flat strip; there is no "flow" through it, the momentum changes which occur being associated with changes of the direction of the air in the neighbourhood of the rotor, that is, in the conversion of an upwash into a downwash. In the case of inclined flow it can be seen physically, and also from (X3), that the induced velocity at a particular point of the rotor depends on the variation of the pressure gradient along the path of integration and not simply on the local discontinuity, as happens to be true in the axial case. Thus (28) cannot be valid in general and attempts to use such a relationship to connect the local rotor loading with the induced velocity require a special are fallacious. imterpretation Such attempts, of the increment in any of mass flow case, usually

pVdA.
Another point worth noiing is that, although there is a pressure discontinuity at the disc, the gradient, and hence, the flow acceleration, is continuous< as is also the flow velocity. There is no sudden deflection of the flow field as is sometimes thought to be the case in supposing (28) to be true. Further, since any axially symmetrical loading can be build up of an assembly of elementary concentric circular loadings, the centre of the rotor occupies the same geometric position in every such elementary loading. Hence, the induced velocity at the rotor centre depends only on the total must local loading; if it is zero, as would be expected, the induced velocity be zero there also. This does not hold for any other point on the rotor, i.e. the local induced velocity, generally, will depend on the shape of the loading 7. distribution as well as the magnitude.

GENERALFORWARD FLIGHT CASE be inclined at an angle x to the

In the general case the flow will rotor disc and we have the relationships x' I&X

+ psx

(13) then becomes

But along the path of integration, dz = - da t&Xand so' s-0 3

i.e.

along a path parallel

I to the k-axis

Y~~~d~

@9)

11

where, in the integrand, Now (14) can be written

X and

are related

by the path of integration.

and along the path of integration x where and w= = = rwsl+ f.hE, rs,; 'y = contint. p-in = rtbsrY - -h*o

"4 Hence for a point

on the disc

Apart from a slight difference of notation and definition of the a&e corresponding to l/J' the above integral is identical to that of ref.10 in which the induced velocity of a uniformly loaded rotor had been calculated by integrating the effect of an inclined cylindrical vortex wake. Another difference is that the induced velocity given here is expressed directly in terms of the pressure jump instead of the circulation of a unit slice of the vortex wake. Unfortunately, the integral cannot be evaluated analytically for a general point on the rotor disc except for points on the longitudinal axis of symmetry, but even then, as shown in ref.10, the result can only be expressed of elliptic integrals of the first and third kinds. The special case, F = 900, agrees with the results given in (2'2) and (23). 8. THE SYMMETRY RELATIONS

A circular disc carrying a uniform load generates an elliptical vortex one, a twowake. By superimposing a skew-symmetric wake on the original dimensional elliptic wake is created by means of which Katzoff 11 obtained certain relationships between the induced velocity components. They are (i) If P and Q are two points on the rotor disc syrmnetrically located about the lateral axis, the sum of the induced velocity components UG), and ti, is equal to the vertical component of the induced velocity field within the two-dimensional wake. Since this is constant, it follows that is also constant and the induced velocity distributjon up+ wa is skew-symmetrical with respect to the lateral axis. (ii) disc If f' and and lie are symmetrically outside it we have

located

about the lateral

axis of the

- 12 where plane the the flow at the v I point wake air. elliptical the been circulation shown to, v; in ref.10 ultimate rotor is the longitudinal<: to the component P (or it this velocity a circuit 0 wake centre; is of velocity in Now, itself gives the into the the ellipse flow relative rise wake and out the to axis, of the u say. the about to

corresponding is due to

Q ), fig.8.
induces motion upon which to

elliptical surrounding about the

motion of the

The component wake is round

normal threaded

By considering wake, within, is twice it has

that

L&q p/L? times this is latter that

velocity component

and parallel the value

the at the

and that

Ref.ll

gives

the

complex

potential

of

the

ellipse

flow

as

where coordinates major Ir terms axes

t by of

= <++ n't;y' theellipse giving becomes

C

are * 8' and

q

elliptical

coordinates ,,," . with

related 1;:

to

the

cartesian semi3

;~SC.O~~~~S Rsiq X so that,

E,aeT=th;

= RCQSF of a'

(33),

equation

(34)

in

Q,ticy Differentiating (35)

= with

24
respect

CGdtx x to

[ 1' -

/pi&

(35) as

gives

the

velocity

components

Now, from (32)

on the

lateral

axis
5 q

w, ,b&

a, and

&? si?x=

say ii

and

i'

%'

. have

Then,

and writing

, we finally

Equation induced rotor velocity centre.

(361,

shown the

plotted lateral

in

fig.(q),

gives

the of

distribution the value together at

of the with

along These

axis also

as a fraction been given in

values'have

ref.12,

* Here, i.e. minor the

as in d

ref.11, axis Thus,

we adopt corresponds we have <

the to =

conventional the 1 major and axis '8'

axes and =

system the I

for axis

the to

ellipse, the

axis.

- 13 other values in the lateral plane, but the calculations were obtained a numerical evaluation of an integral derived from (31) with I//
q

by
/z

However, for the special case of the lateral ible to rearrange the integral and solve it as (36). 9. VERIFICATION OF GLALERT'S FORMULA

axis, it has been found possexactly, giving the .same result

Let x be coordinates in the rotor plane and n ',I' coordinates 4 in the plane perpendicular to wake axis, fig.(lO). The cross-section of the wake parallel to the rotor plane is, of course, a circle with radius e Consider now an annulus concentric fig.(lla). The mean induced velocity with & the rotor and lying outside it,

over the annulus is

where as in the previous section, 0~ and ti, are the induced velocities at points equally spared about the lateral axis and A is the area of the annulus. From (32) equation (37) can be written as

If elliptical

and w wake, v'

are the velocity

and stream functions

of the flow about

* in the-rotor and ellipse plane are related by

The coordinates

so that

and

can be written CJ z

as &lx w SJ ( 0w

A(wFC)

- 14 so that, from (39),

where the suffix

C denotes the contour of the annulus,

section, we have

fig.Cl1.a).

Now, from the previous

The contour consists of two circles of radius R and /+R say, where ,& > 1. Although the function in the integrand has branch points, they both lie within the inner circle so that no discontinuity of the function occws between the straight portions AD and 8c . Hence, the integrals along these portions cancel and we need only consider integration round the circular arcs T q R and 7" = ,& . If "d : r&"' we have *

P, Jc(GV)~~ - I sc 0

;y, )&$dq

- p

s0

(cp,- h$iydy

where the suffixes 1 and 2 denote the values on the inner and outer circles respectively. Gn the inner circle x' = R&svand 1' =e.?l~(uWsXand, therefore, q-4, =
2;

uu;

coscc;(

On the circle

of radius

Considering

the

first

term,

the

contribution

to

the

total

integral

is

of which

the

real

part

is

easily

found

to be

The second unit that radius) to

integral Cartesian

can be evaluated coordinates in

by reverting complex form

back by

(in

terms

of

where we find

the

contour

is

the

unit

circle

about

the

origin.

On rearranging

The only root sign

pole

is

a simple find

pole the

at residue

8 q o .
of the

Expanding integrated

the to

term

under

the
Hence

we easily

be unity.

r = np (ffCoS%)
and the This part term of contribution cancels S((P-4dj velocity load is with due to the is over zero. of velocity this result is, as follows:on the rotor is We saw in section 8 with the first zero; ten one in in I is found on the words, with the to be outer

27rp2Q2u; cd @I
circle, of the the real disc carrying

so that, other

mean value

induced a uniform

any annulus

concentric

a circular

The implication that the induced

distribution

skew-symmetric

- 16 respect to the lateral axis. Hence, the mean value of the induced velocity over the rotor must be the same as the (constant) value 'V,' on the lateral axis. To find this value we put v = 0 in (31) and, by means of standard tables, we perform the integration with respect to This is easily found i*

ButA P

is the thrust u;

loading I J

T/4

so
. . .

ZpA\/

I411

This is Glauert's formula for Vi when v is large; that is, Glauert's formula is true for a uniformly loaded disc. This result has also been obtained by Shaidako:3who related the thrust and induced velocity to the impulse and circulation of the vortex rings shed by the disc. Now suppose we have a uniformly loaded rotor of radius R a thrust T,,say. Then, by (41) the mean induced velocity is carrying

If a concentric

uniform

loading

is added over a circular Viz over the disc

disc of radius is

the mean induced velocity

But the mean induced velocity over the annulus between? = R, = R has b een found to be zero;therefore, over the whole rotor,

and

= J/i w
i.e. the mean induced velocity vi for the combined loading, is given by where

is the total

thrust.

- 17 circular loadings the mean induced velocity, depends only on the total thrust and not on the radius of the smaller loading. And since any axially symmetrical loading can be built up from an arrangement of such concentric loadings, Glauert's formula holds for axially symmetric loadings also. Another case which can be examined fairly easily is the one for which the area outside the load is a circle which just touches the circular load, fig.(llb). Taking the origin as the point of contact of the circles, let pR be the distance of the centre of the larger circle from the origin and the distance of a point on this circle. Then, as is well known from coordinate geometry, ? is given by I-= where 8 and a let p
2pRcos (0-d)

Thus, for two concentric taken over the larger radius,

are defined
q

in fig.(llb so that

). (42') can be written

As before,

PR

h= Ignoring late

2pc0s (e-d)
in front of the integral (40), we wish to calcu-

(43)

the constant

where the integration

is taken round the ccntour

made up of the rotor

circumon

ference, as before, and the circle (43). On the outer circle longer constant but a function of the angular coordinate of the point the outer circle. me value of found to be zero. Then, if and "a
a de

J
.
l

t ak round the rotor en

circumference

has already

been

p.f~?c~

da

With these values we easily find that and, by the same means as for the previous

case, we find

that ,als0.

em J,/m

= 2sQ(itwrF)

- 18 Hence, since these quantities cancel, the mean induced velocity taken over the lune (lib) is also zero and Glauert's formula holds for any rotor load distribution made up of such basic loadings. For the more general case, represented complicated by fig.(llc), it was found that solution was

the algebra became prohibitively not attempted. But,

and an analytical

(38,

by Green's theorem,

Equation

(34) gives (Paand, in terms of elliptic coordinates,

(40)

Using (40) and (41), the numerical computation of (39) was found to be quite simple for the regions lla, b, c and confirmed the results obtained The computation also showed that, analytically for .regions lla and lib. within the numerical accuracy of the calculations, the mean value of the induced velocity OV~P the region llc was also zero. Thus, we-have the rather remarkable result that if a circle is drawn u

round a circular area carrying a uniform load, the mean induced velocity taken over the region between these two circles is zero. It can easili be verified that w does not vanish for every region

- 19 taken is outside the with circular the elliptical ellipse load. For example, if we take a" ellipse it as the onto outer the which rotor contour. to of two

confocal

wake boundary which of

and project

plane, Since constant for this It means loading loadings

we obtain this ellipse, values case, would that the

another

can be regarded the rotor itself,

and the of 5 in

boundary the elliptical quite the

correspond the integration

coordinates, easily result over the and is for the rotor from

can be performed appear, mea" then, induced that

found general is

to be non-zero. circle, fig.(llc), of of the

velocity

independent

since of

any loading varying radii. formula the

can be constructed

a" arrangement

circular

Thus, to be true for low velocity, 10.

Glauert's whatever cannot

for load

the speed

linearized distribution. is of theory. the

"high

speed" it

case is as the

appears true induced

rotor forward

Whether same order

still

speeds,

when the

be checked

by linear

THE INDUCED POWER IN FORWARD FLIGHT The fact that flight the the is slipstream a distinct of a uniformly of loaded moving air This the thrust loading power rotor. power rotor enables leads the to in hovering the the kinetic well by the of

and vertical energy know" induced the quite of result

cylinder

slipstream that the For in the vertical

to be calculated induced power is

easily. simply rotor the the the that

multiplied symmetry

velocity. easily of

a" axially flight

symmetric enables over

slipslream

induced whole induced

to be calculated It for can be shown a given the It is The

by evaluating calculus when the is uniform. that of

SVi dT
variations velocity is induced This when the power is

by means thrust rotor is

least

distribution, the induced 13% higher flight is "ideal" velocity than often the

and therefore induced power.

loading,

called

can easily triangular induced the of

be show" the power

radial about

distribution "ideal"value.

induced in vertical and

and forward increasing velocity bearing in are valid rotor being consists of the it into mind very for in

estimated take the

by calculating non-uniformities

"ideal" loading

power

by about account. that different all flight the

15% to

and induced is;

The question butions is for a given of

induced in vertical

velocity and forward To answer and calculate this the In the

distriflight, we rate ulti-

loading 1.15 loaded

a factor

about

cases? flight

consider at which mate

a uniformly kinetic

forward to two flow

energy the energy wake. energy it. E,

is

supplied of

the

slipstream.

slipstream (i) the vortex (ii) the

energy E,

contributions within a unit length of the

uniform

inclined

of

the

air

outside

the

wake

due to

the

movement

of

the

wake through

-..zoIn unit that the power time the length is of the ultimate wake increases by v units so expended cross-sectional mass of value within the wake of air the in

V(E,+ E,),
area a unit of the length velocity to moving the wake the with is wake wake is is r Q2c8sX x . axis Also of the and if rotor in prR2cos lateral But, induced on the is

Now the therefore Vi the the Ls the

velocity 8, axis

and relative is the energy E, itself absoulte in =

2V;.

as discussed

section to its

velocity of the air therefore

2v; tan,z/2
in the wake is

normal

so that

velocity

2q set 7Llz. e Th
To calculate we use the formula

2p1r3rQ~cos;Csu~11/2
energy -;fjhy, and stream On the tih
T-f2

the 14

kinetic

o f the

flow

outside

the

elliptic

wake

where evaluated E.

and

v,

are

the

E2 = potential

functions E

of has the

equation constant

34 value

on the given

wake boundary. - L by e 9'

boundary , hence

-2u, R ta&

S&I-/

Y =
giving

2u;R tn.&i

~0s7

The induced

power

is

therefore

= 2pre2q2v =
where Thus as in It tical u; for is the induced a uniformly loaded

TV;
velocity rotor flight. on the the lateral for axis. induced power is the same formula

hovering would rotor

and vertical be useful loadings. to the arbitrary induced

to be able Unfortunately loadings velocity closely'

to

calculate the induced

the

induced

power distributions

for

?~a=-

velocity in

corresponding has whose calculated radial

cannot field resembles

be found for

general

but

Mangler3 loading If

an axially those occuring

symmetric in

distribution

practice.

- 21 X = "/R 1s t h e non-dimensional . radial coordinate of the disc Mangler's

where in

T is

the It

total should

thrust. be noted to at X

The shape that

of

this (42)

loading implies

distribution that the

is blade

shown

fig.12.

equation

loading

is proportional The induced velocity

the

rotor

corresponding

to

(42)

can be expressed

Where are

~iv;, by

is

the

mean "momentum"

induced

velocity.

The coefficients

(2,

given

where For

$ even

Ivalues

2 of

and fl3

I-

CosF

lfCbS~

Mangler rotor induced However culated induced component is the but the power the quite drag in induced

also

gives

the are

induced

velocity

field

far for

downstream calculating considered

of the

the

expressions in special easily in the terms far drag the

much too the

complicated uniform flight, formula 15 the loading

same way for case of wake the of the i.e. "high by making

above. can be cal-

speed"

= 90deg,

use of spanwise in is the

circulation so-called

which expresses the and the normal velocity plane. If D;

Trefftz

formula

te

D; giving the induced

$p
power

rwdr
c -

-e

- 22 where r is the local circulation and &' the induced velocity. In the case of the uniformly loaded rotor the "circulation" is simply proportional to the "chord" and the distribution is therefore elliptic. It can easily be 'verified that the circulation corresponding to Mangler's loading is

Although Mangler gave a fomula.for the induced velocity in the far wake the integral (45) becomes too complicated to be evaluated analytically. However, the table of values given in his original paper 16 , together with equation easily. (461, enables the integral to be.evaluated The induced power can be expressed as numerically quite

where

1 -

and W = @w/v. are the values given by Mangler. ,. of 1 gives

$$G *

(vzz)(it3;t2)dn
Numericalintegration

= r.r7+ i$*
where 6, = TV;, is the induced power of a uniformly loaded rotor. One should also be able to calculate the induced power by considering the backward tilt of the local blade thrust vector due to the induced velocity , as in classical blade element theory. The thrust carried on an annulus of width df is and this elementary thrust, dT of coume,

,%rApdr
The

is shared by the b blades of the rotor.

induced torque is therefore

where v; is the local induced velocity The induced power contribution is

at the blade.

= 2crrdp J;dr
since A/J
is a function of X only and v; is periodic with respect to azimuth the mean value of q Hence from (42) and (43) we find depends only on the first that term of (43).

P;

= 7

TV;, f$(,.

x2)dx

- 23 c !fr @ vi, = 1. 172 TV;, numerical difference, is the same answer as

which, apart from a negligible before. Now for axial flight

at velocity

b,

the momentum theory

gives

dT

= 2wApdr

Using Mangler's

= 2pvv;kdf or v; = -4 W loading, equation (42) gives

T
we2 . $t/z

v.

* =

2p

and this is identical to the first term of the series (43). Since similar results are true of the other (elliptical) loading considered by Mangler and for the uniformly loaded rotor, it is reasonable to assume that x axially symmetric loading leads to a series of the form (43) of which the first term gives the induced velocity for the same loading in axial flight. Hence, the induced power of a symmetrically loaded rotor in forward flight is the same as that in axial flight fast enough for linearization to be valid. For the non-linear relationship of hovering flight it can easily be shown that the induced power for Mangler's loading is about 1.11 times that of a uniform load. Thus the induced power factor for Mangler's load rises from 1.11 in hovering flight to 1.17 in forward flight. If the radial loading of the disc is of the form = the induced power factor f+ k f+k = I9 can be shown to be given by (f+p

Itin
for hovering flight, and J+k= for forward flight. w ftrr It should be Numerical values of these factors are given in fig.12. remarked again that if the disc loading is proportional to X', the corresIfi ponding blade loading is proportional to n . For a given loading tor in forward flight it appears, therefore, that the induced power facflight.

is somewhat higher

than in hovering

SYMBOLS

fluid acceleration vector pressure induced power ideal induced power fluid velocity vector rotor radius disturbance velocity components in rotor disturbance velocity vector induced velocity in plane of rotor

plane

longitudinal velocity in plane perpendicular flight speed of helicopter coordinates in rotor plane rotor disc incidence elliptic coordinates in plane perpendicular air density

to vortex

wake

to vortex

wake

potential function acceleration potential wake angle azimuth angle in rotor

plane;

stream function

REFERENCE.5 -No. 1

Author(s) K. Krienes W. Kinner

W.

Title,

etc.

Die elliptische Tragfl&he auf potentieltheoretischer Grundlage. Z.A.M.M. 20 (1940) Die Kreisformige Tragflgche auf potentialtheoretischer Grundlege. Z.A.M.M. 18 (1937) The induced velocity ARC R & M 2642. field of a mtor.

3 4

Mangler and H. B. Squire

C. E. Weatherburn s. Katzoff, Frances M. Faison and Hugh D. DuBose D. Kichemann and J. Weber Durand (Ed.) P. M. Goorjian

Advanced vector analysis. Chap.IX. G. Bell and Sons Ltd., London. Determination of mean camber surfaces for wings having uniform chordwise loading end arbitrary spanwise loading in subsonic flow. NASA Report 1176. Aerodynamics McGraw Hill of Propulsion (Appendix) Pub.Co.Ltd., London.

Aerodynamic Theory Vol.IV Div.M.Sect.3. Dover Pub.Inc. New York. An invalid equation in the general momentum theory of the actuator disc. AIAA Journal Vol.10 No.4. April 1972. A General Theory of the Autogyro. ARC R & M 1111. Evaluation of the induced velocity an idealined helicopter rotor. NACA ARR No. L5ElO. field of

9 10

H. Glauert R. P. Coleman, A. M. Feingold and C. W. Stempin Ii. H. Heyson and s. Katzoff W. Castles, Jr. J. H. Deleeuw V. I. Shaidakov and

11 12

Induced velocities near a lifting rotor with non-uniform disc loading. NACA Rpt. 1319. The normal component of the induced velocity in the vicinity of a lifting rotor, and some examples of its application. NACA Rpt. 1184. Vortex ring method applied to the aerodynamic design of lifting systems. Soviet Aeronautics Vol.9 No.3, (1966). A Treatise on Hydromechenics. Part II Hydrodynamics. G. Bell & Sons, London. The ABC Helicoper. AIAA Paper No.69-217. of a rotor.
(1948).

13

14

A. S. Ramsey
M. C. Cheney, Jr.

15 16

W. Mangler

The induced velocity field RAE Report No. Aero.2247,

34 822

Fig.1

Sketch

of

pressure

field

for uniform

load

Fig.

Sketch

of vortex

ring

flow

34 822

Fig.3

Rotor

Coordinates

34 822 P Along rotor axis

-2

iI

! r/R

-1

--5 0 i

1 I/-

Along diameter

Fig.4

Pressure

gradient

normal

to rotor

34 822

Fig.5

Pressure

variation

along

rotor

axis

-2

-1

I 0

'44

Fig.6

Velocity

variation

along

rotor

axis

34 822

0 1.
-7

Fig.7

Induced

Velocity

along longitudinal

axis

Fig.8

Symmetrical

points

in wake

34 822

Fig.9

Induced

Velocity

along

lateral

axis

Y Yf V V

.Y ,I --. x Y

X ,,I ! .\

>

m-..y

Fig.10

Wake

Coordinates

lla

llb

Fig.11

Regions

of

Integration

0.2

0.4

0.6

0.8

Fig.

t2

kkMgk&

Pressure

Dihbution

Form-d

Flight

rtk

Hoverlng

7)

Fig. 13

Induced

Power when

Rotor Loading i prcp~&nai

propartiallal

to rn

ARC CP No.1301 October, 1973 Bramwell, A. R. S. SOME REMARKSON THE INDUCEDVELCCITY FIELD OF A LIFTING ROTQR AND ON GIAUERT'S FORMUIA The induced velocity field of a liftiw rotor is discussed in relation to the pressure field rather than the vortex wake in an attempt to obtain a clearer understanding of the relationship between the induced velocity and rotor forces. A number of results are derived and, where appmpriate, are compared with those obtained from the theory of the vortex wake. An/ ARC CP No.1301 October, 1973 Bramwell, A. R. S. SOME REMARK.5 THE INDUCEDVELCYZITY ON FIELD OF A LIFTING ROTOR AND ON GIAUERT'S FORMUIA The induced velocity field of a lifting rotor is discussed in relation to the pressure field rather than the vortex wake in an attempt to obtain a clearer understanding of the relationship between the induced velocity and rotor forces. A number of results are derived and, where appropriate, are compared with those obtained from the theory of the vortex wake. Ad

ARC CP No.1301 October, 1973 Bramwell, A. R. S. SOMEREMRKS ON THE INDUCED VELCX2ITY FIELD OF A LIFTING ROTOR AND ON GIAUERT'S FORMUIA The induced velocity field of a lifting rotor is discussed in relation to the pressure field rather than the vortex wake in an attempt to obtain a clearer understanding of the relationship between the induced velocity and ro++r forces. A number of results are derived and, where appropriate, are compared with those obtained from the theory of the vortex wake. Ad ARC CP No.1301 October, 1973 Branwell, A. R. S. SOMEREMARKSON TBE INDUCEDVELOCITY FIELD OF A LIFTING ROTOR AND ON GIAUERT'S FORMUIA The induced velocity field of a lifting rotor is discussed in relation to the pressure field rather than the vortex wake in an attempt to obtain a clearer understanding of the relationship between the induced velocity and rotor forces. A number of results are derived and, where appropriate, are compared with those obtained from the theory of the vortex wake. An/

An investigation into the validity of Glauert's :ormula indicates that it appears to be true for all .otor loadings for the linearised, "high speed", case. Calculations of the induced power in forward flight shows that for typical rotor loadings the power is &I to t5 per cent greater than the "ideal" induced power.

An investigation into the validity of Glauert's formula indicates that it appears to be true for all rotor loadings for the linearised, "high speed", case. Calculations of the induced power in forward flight shows' that for typical rotor loadings the power is @to 45 per cent greater than the "ideal" induced power.

:ormula

An investigation into the validity of Glauert's indicates that it appears to be true for all :otor loadings for the linearised, "high speed", case.

An investigation into the validity of Glauert's formula indicates that it appears to be true for all rotor loadings for the linearized, "high speed", case. Calculations of the induced power in forward flight shows that for typical rotor loadings the power is kC to 45 per cent greater than the "ideal" induced power.

Calculations of the induced power in forward flight shows that for typical rotor loadings the power is kCl to t.5 per cent greater than the "ideal." induced power.

.- _ _

___- . - - -

_, _ _ _ .

- - - ___..

- .-,..--

-.

__._____.

~. ..^.

----..---

_ ^ _. _ _

__

,.

. -. - - -

-. .

_-

C.P. No. 1301

C.P. No. 1301 IhBN 470885 0I, I