NATION^ADVISORY ~ O M M ~ ~ E

FOR AER~NAUTICS
TECHNICAL NUTE 4304
NATIONAL ADVISOFE COMMITTEE FOR AERONAUTICS
TECHNI CAL NOTE 4304
MATRIX METHOD FOR OBTAINING SPANWISE MOMENTS AND
DEFLECTIONS OF TORSIONALLY RI GI D ROTOR BLADES
WI T H A RBI " Y LOADINGS
By Alton P. Mayo
SUMMARY
A matrix sol uti on for the spanwise bending moments and defl ecti ons
of a torsi onal l y ri gi d rotor blade subjected to an arbi trary loading i s
presented. The method includes the canti l ever, teeteri ng, and hinged
blades i n hovering and i n steady forward f l i ght. The method is compara-
ti vel y short, involves only standard matrix procedures, and does not
requi re that the mode shapes or natural frequencies be known.
INTRODUCTION
Numerous methods are avai l abl e f or cal cul ati ng the uncoupled span-
wise bending moments and defl ecti ons of rotor blades i n steady unaccel-
erated fl i ght. Some of these methods (refs. 1 and 2) involve tabul ar
solutions of the di f f erenti al equation of blade deformation. These
tabul ar procedures are very lengthy, parti cul arl y when numerous loading
conditions and rotor speeds are being investigated. Other methods
(refs. 3 and 4) require the cal cul ati on of the natural-mode shapes and
frequencies of the blade. The cal cul ati ons of mode shape and frequency
are laborious and ei ther must be repeated f or each rotor speed or an
approximate correction must be applied to account f or the effects of
rotor speed. A matrix method which avoids some of the di f f i cul ti es of
the tabul ar and modal sol uti ons i s presented i n reference 5; however,
tabul ar sol uti ons of the equation f or blade bending moment under quasi-
stati c conditions are required before the method can be applied.
I n the present method the di f f erenti al equation f or blade bending
moment i s solved enti rel y by matrix procedures. The method permits the
determination of the bending moments and defl ecti ons di rectl y without
preliminary quasi -stati c, mode-shape, or natural-frequency cal cul ati ons
and, as a resul t, the method is shorter and much work i s eliminated.
The method i s adaptable to cantilever, semirigid (teeteri ng), and f ul l y
2 NACA TN 4304
arti cul ated (hinged) blades i n hovering and i n steady forward fl i ght and
is applicable to design calculations and experimental data analyses.
The di fferenti al equation for the blade bending moment i s composed
of matrix expressions for the centrifugal i nerti a, verti cal i nerti a, and
damping loadings, and the contribution of each term to the blade loads
and moments may be determined. The effect of blade structural damping,
which has been omitted i n other methods, is included i n the present
method. Comparisons are presented between the bending moments obtained
by thi s method and the method of reference 3 for the first harmonic
loading on a ful l y arti cul ated blade and the second harmonic loading on
a cantilever blade.
SYMBOLS
Some of the symbols l i sted i n thi s section are i l l ustrated i n
figure 1.
A
a
C
E1
F
blade-element average aerodynamic damping coefficient,
- - 1 r, lb-sec2
2 i n.
blade -element l i f t -curve slope , per radian
blade chord (at center of blade element), i n.
blade bending sti ffness, lb-in.2
centrifugal force acting on mass of blade element divided
by $, lb-sec2
structural damping coefficient
perpendicular distance from axis of rotati on to flapping
I
hinge, in.
imaginary component of complex number,
blade running load, lb/in.
J -1
load on blade element, lb
bending moment, lb-in.
NACA TN 4304
3
m
ne
R
r
clr
v
ZM
Z
B
P
Subscripts:
a
C
d
H
i
mass of blade element, 1b-sec2
i n.
number of blade elements for extended matrices
distance from center of hub to blade ti p, i n.
spanwise distance from center of hub (measured along
undeflected blade) , in.
wi dth of blade element, i n.
forward velocity, in./sec
elements of an influence coefficient matrix (see table 111)
blade deflection, measured from a plane of rotati on per-
pendicular to shaft axis, i n.
rigid-blade flapping angle, measured froma plane of
rotation perpendicular to shaft axis, radians
angle of fl exi bl e blade at hinge, measured from a plane
of rotati on perpendicular to shaft axis, radians
blade slope due to blade bending, dz/dr, radians
rotor angular velocity, radians/sec
angular frequency of applied loading, radians/sec
l b - se c2
4
i n.
ais density,
blade azimuth angle (measured i n direction of blade
ti on from downwind position), radians
aerodynamic
cosine component
damping
due to forces paral l el to plane of rotati on
i nerti a
rota-
4 NACA TN 4304
L pertai ni ng to a load stati on
r
S
pertaining to a spanwise stati on
structural
S si ne component
1,2,. . .n
v
W blade weight
Z defl ecti on
f centri fugal loading
stati on number or number of harmonic
due to forces perpendicular to the plane of rotati on
A dot over a symbol i ndi cates fi rst deri vati ve with respect to ti me.
Two dots i ndi cate second deri vati ve wi t h respect to time.
Matrix notation:
c1 rectangular or square matrix
c l i diagonal matrix
column matrix
inverse of a square matrix
{ }
L J row matrix
METHOD
The di fferenti al equation for the structural bending moments of
hinged, teeteri ng, and canti l ever rotor blades i s set - ~p i n matrix form.
Matrix expressions for the de moment due to the various blade loadings
are derived and added to f the blade-moment equation. The di fferen-
ti al equation i s fi rst derived i n a generalized form and i s later adapted
to (1) the case where the rigid-blade aerodynamic loads are known from
design cal cul ati ons and the fl esi bl e-bl ade structural moments or deflec-
ti ons, or both, are desired; (2) the case where the blade structural
moments are known from strain-gage tests and the total aerodynamic
NACA TN 4304
5
moment, including aerodynamic damping, i s desired; and ( 3 ) the case
where the aerodynamic loads have been measured and the structural moments
are desired. The procedure for obtaining the rotating-blade character-
i sti c matrix i s al so presented so that the natural modes and frequencies
may be determined.
Blade Loadings
I n the calculation of spanwise bending moments, a torsi onal l y ri gi d
rotor blade or i ts equivalent (an el asti c blade f or which the secti on
moments are zero and the aerodynamic center, el asti c axis, and center of
gravi ty coincide) may be treated as a simple beam subject to numerous
superimposed loadings. If chordwise effects are excluded, the blade i n
the forward f l i ght condition i s subjected to the following di stri buted
loads :
(1) The aerodynamic load La (excluding the aerodynamic damping
effect which i s di rectl y proportional to blade flexure plus flapping
deflection velocity; see i t em 5 )
(2) The centri fugal load F
( 3 ) The blade weight L,
(4) A normal i nerti a loading due to verti cal accelerations
L i v
( 5 ) An aerodynamic damping load
L a
(defined by eq. ( 5 ) ) , di rectl y
proportional to blade defl ecti on vel oci ty due to both flapping and
fl exure
(6) The blade structural damping load
Lds
The combined moments of these si x loadings are balanced by the blade
i nternal mament due to blade bending resi stance.
Derivation of Blade Moment Equations
The blade geometry, the forces acti ng on an i nfi ni tesi mal ele-
ment of the blade under forward-flight conditions, and the resul ti ng
force system are shown i n fi gure 1. From fi gure 1 it i s seen that the
bending moment Mrn at any stati on rn due to the loads acti ng on the
blade is given by:
6 NACA TN 4304
The matrix equivalent of equation (1) may be wri tten i n the form
Equation (2) corresponds to a step-diagram representation of the
2v and 2H loadings of equation (1). The [r] matrix i s an i ntegrati ng
matrix with values that depend on the stati ons and i ntegrati on method
used. The LFJ matrix incorporates the values of the horizontal loads
{LiH}
(dbensi onal with respect to rotor speed) as wel l as i ntegrati ng
numbers. The [r] and [F] matrices are derived i n the appendix.
Substituting the fi ve components of {Lv} enumerated previously
i nto equation (2) yi el ds the following bending-moment equation of the
combined loadings:
Equation (3) is the general form of the equation f or blade bending
moment used i n thi s study and pertai ns to hinged and canti l ever blades
i n hovering and forward f l i ght. I n order to obtain the blade bending
moment for any speci fi c loading condition, the matrix expressions f or
the various loads acti ng are substi tuted i nto equation ( 3 ) .
Loading matrices.- The expressions for the various component loads
i n equation ( 3) become qui te simple i f equation (3) is restri cted to the
expression of the moments due to loadings of a single frequency.
total aerodynamic loads on the blade may be broken down i nto components
which are of a sinusoidal nature, and the standard procedure is to
develop equation (3) for the moment due to a si ngl e frequency.
posi ti on may then be used to obtain the total moments due to al l of the
harmonic components on the blade. The loading matrices which appear i n
equation ( 3 ) are given by the three following equations.
The
Super-
For a parti cul ar frequency 03 of the applied sinusoidal load, the
verti cal i nerti a loads {Liv} are given by
NACA TN 4304
biv}=u.9 Llm J I4
where the elements of the r mJ matrix are the masses of the blade
elements .
The aerodynamic damping loads
{Laa} =i Sh[ I A J
are given by
{zl
7
(4)
( 5 )
where the diagonal matrix A] is equal to [-$ pacr
The elements of the
f A3 matrix can be deduced fromsteady aero-
dynamics f or the l i f t on a section of chord c and length & rotati ng
at uni t angular vel oci ty and with a uni t normal velocity.
The structural damping moment [r](Lds) is given by
Moments and defl ecti ons at a si ngl e frequency.- The blade Zoadings
at a parti cul ar frequency u) (eqs. (4) , ( 5 ) , and (6)) may be substi tuted
i nto equation ( 3 ) i n order to obtain the equation f or the blade moment i n
terms of the structural moment, the aerodynamic moment, and the blade
defl ecti ons as
The deflection (2) i n equation (7) i s composed of the fl exural
The fl exural bending bending and the effects of blade-root rotati on.
i s defined i n the appendix as
The total blade defl ecti on including root rotati on is, therefore,
given by
a NACA TN 4304
For small angles
based on assumptions of small angles, ph i s assumed equal to tan ph.
Equation (9) , therefore, becomes
tan ph =ph, and i n the equations to follow, which are
Substi tuti ng equation (10) i nto equation (7) yi el ds the following
rel ati on between the aerodynamic loads and the structural moments for
the fl exi bl e blade:
where
[Mzv] = [lrImZ1
The parti ti on l i nes of equation (11) effecti vel y separate structural
moment effects from root slope effects, both of which act to balance the
aerodynamic moment. If the root slope ph i s set equal to zero, the
terms pertaining to the root slape drop out and the resul ti ng equation
i s the equation f or a cantilever-blade bending moment.
blade equations are given i n detai l i n a subsequent section.
The canti l ever-
Equation (11) may be inverted i n order to obtain the reci procal
matrix rel ati on
NACA TN 4304
where
{Ma} = 1r1{La}
9
Equation (12) gives the structural moment {M.}
and root slope ph
on the-rotati ng hinged blade i n terms of the applied aerodynamic moment
{Ma}.
is proportional to the total defl ecti on velocity.
The equation includes the effect of the aerodynamic damping, which
When calculations are to be made for many rotor speeds, it is advan-
tageous to put equation (12) i n the form
where the matrix [N] i s evaluated for each harmonic, and
The matrix CN] need be evaluated only for each harmonic number.
Hovering condition. - Equation (12), which was derived for the
forward-flight condition, would al so apply to the hovering condition i f
~~ .
the terms which pertai n to the blade verti cal vel oci ty and acceleration
are omitted.
g =0
and u) =0; and the equation for hovering becomes
Because the hovering aerodynamic loads axe steady,
Cantilever-blade equations.- If the blade-root angle p, i s zero
i n equations (12) and (l 5), the result i s the equation f or the canti l ever
blade i n forward fl i ght:
10 NACA TN 4304
or
I f the cal cul ati ons are to be made for,numerous rotor speeds, equation (17)
may be put i n the same form as equation ( 13) .
The cantilever-blade equation f or the hovering condition is:
C A L U T I O N PROCEDURES
Matrix Evaluation
The cal cul ati on procedures presented are i l l ustrated f or the hinged
The detai l s given pertai n to the ten-segment breakdown of the blade.
blade load, weight, and sti ffness di stri buti on shown i n fi gure 2. The
procedure f or ei ther a canti l ever or a teeteri ng blade i s a si mpl i fi ca-
ti on of the hinged-blade case. The parti cul ar blade-section breakdown
i n the matrix tabl es (see fi gs. 2 and 3 ) was sel ected to provide more
stati ons near the root of the blade where rapi d changes i n blade moment
and sti ffness occur. The nuniber of blade elements used may be increased
by the procedures given i n the appendix f or increasing the order of each
matrix.
suffi ci ent.
However, it i s believed that f or most blades ten stati ons are
I n order to apply the hinged-blade equations, it is first necessary
to evaluate the various matrices i n the equations. The hinged-blade
equation, previously given as equation (12), i s
NACA TN 4304
11
~~~ = ~1. 0 +i g) i-----+---] 1.01 i (0) +Q2[F] [[%]I {r - h)] -
LoJ : o I
The structural damping coeffi ci ent g i s a function of the fre-
quency, and the value used should correspond to the value of LD for
which the equation i s being evaluated. The matrix ------4--- i s an
i denti ty matrix where the zero, as the last element on the diagonal,
imposes the condition of zero structural moment at the blade hinge.
This matrix i s given i n tabl e I. The matrix [F] expresses the moment
on the blade caused by the centri fugal forces acti ng on the defl ected
blade. The [F] matrix i s given i n tabl e 11. The matrix [[ZM]! {r - h>l
expresses the blade defl ecti on caused by the structural moment and the
root slope. This matrix i s given i n tabl e 111. Thematrix [Mzd
expresses the verti cal i nerti a moment i n terms of the blade deflection.
The cMzv] matrix is given i n tabl e I V . The [r] matrix i s an i nte-
grati ng matrix which expresses the moment of a loading and i s given i n
table V. The matrix [Ad expresses the aerodynamic damping load i n
t er ms of the blade deflection. The matrix [A] is given i n tabl e VI .
1.0] I(o}
[Lo, I 01
The number of di gi ts needed i n the evaluation of the various matrices
and i n the succeeding operations for the most accurate resul ts has not
been rigorously established. However, l i mi ted experience has shown that
carrying the basi c masses, distances, sti ffnesses, and so f orth to four
di gi ts and, i n succeeding operations, allowing the number of di gi ts to
accumulate to a maxi mum of ten gives good resul ts.
f i l l i ng out the given matrix tabl es is as follows:
The procedure-for
(1) Set up the i denti ty matrix as shown i n tabl e I, which i s
already completed i n i ts enti rety and may be used as is.
(2) Obtain the weight di stri buti on f or the blade and break it down
i nto ten elements of equal length. Determine the mass of each element.
Use the product mr for each element as described i n the appendix, and
f i l l out tabl e 11.
( 3 ) Obtain the sti ffness di stri buti on for the blade and break it
down i nto the smaller elements shown i n fi gure 2. Use these sti ffness
values and the values of @ and f i l l out tabl e 111.
12 NACA TN 4304
(4) Use the mass values of i t em 2 and f i l l out tabl e I V .
( 5 ) Substi tute the value of Lr i nto tabl e V.
(6) Use the values of the chord of each element, the l i ft-curve
slope of each element, the radius to the center of each element, and
the values of p and &, and f i l l out tabl e VI .
Application to the Hinged Blade
The equations for the hinged blade are applied t o those problems
which are believed to be the most general. The cases discussed include
the determination of (1) the flexible-blade structural moment due to
the rigid-blade aerodynamic loads; (2) the structural moment due to the
measured aerodynamic loads; (3) the moment of the total aerodynamic
loading as determined from the measured structural moment; and (4) the
aerodynamic damping load.
modes and frequencies of the blade i s al so discussed since, i n some
cases, thi s information may be desired.
The procedure f or obtaining the natural
Structural moment due to rigid-blade aerodynamic loads.- The stand-
ard design practi ce i s to obtain the rigid-blade total loading from
- -
rigid-blade aerodynamic load and i nerti a balance equations. The ri gi d-
blade aerodynamic and i nerti a loads are then applied to the fl exi bl e
blade which is considered to f l ex about the rigid-blade position. This
procedure i s equivalent to the appl i cati on of the rigid-blade aerodynamic
load only (minus an aerodynamic damping component which is proportional
to 5) to the equation f or the fl exi bl e blade which was derived as equa-
ti on (12) and i n which the defl ecti ons are referenced to the plane of rota-
ti on perpendicular to the shaft. The equivalence of the standard design
procedure and the procedures of the present method i s proved by the
following:
The rigid-blade total load (aerodynamic plus i nerti a) which is nor-
mally applied to the fl exi bl e blade (considered to flex about the ri gi d-
blade posi ti on) i s given by
2
Lri gi d - - (La)rigid - mrij - mu rg
total
total
where the terms on the right-hand si de of the equation represent the
rigid-blade aerodynamic, verti cal -i nerti a, an& centri fugal -i nerti a loads,
respectively.
terms of moment, it i s useful to convert equation (19) to i ts moment
equivalent.
the matrix form.
Since the equations of the present method are wri tten i n
Also, it i s useful t o express each term i n equation (19) i n
Thus, equation (19) becomes
NACA TN 4304 13
where the terms on the right-hand side of the equation are the ri gi d-
blade aerodynamic, verti cal -i nerti a, and centri fugal -i nerti a moments,
respectively.
If the rigid-blade moments given by equation (20) are applied to the
fl exi bl e blade, which i s cansidered to f l ex about the rigid-blade posi-
ti on, the following equation resul ts:
Moment of rigid-blade loading
/-
A
,
+ 2 [qv] (Zrigid} - "*LF1{'rigid} =
Er] (La) ri gi d
{ total^
Moment of flexible-blade loading
/-
A
The terms on the right-hand si de of the equation give the moments
associated with the blade defl ecti on {zflex) about the rigid-blade
posi ti on. These terms, from l ef t to ri ght, represent the blade-bending,
the centri fugal -i nerti a, the verti cal -i nerti a, and the aerodynamic-
damping moment, respectively. Equation (21) implies that the blade
moments resul ti ng from the flexible-blade defl ecti ons must balance the
applied rigid-blade moment. The fl exi bl e- and rigid-blade defl ecti ons
I f it i s recognized that
'flex
zflex plus Zr i gi d is equal to the defl ecti on z (measured with
respect to the plane of rotati on) and i f the terms of equation (21) are
collected, then the following equation resul ts:
and 'rigid are shown i n fi gure l (b).
crl (La)rigid =(1.0 +ig){Ms) +n2[F] {z) - m 2 [ i d {z) -
{ totad
Adding the term -isbu[r][AJ {Zrigid } to both si des of equation (221,
col l ecti ng terms, and rearranging yi el ds
14 MCA TN 4304
where
Equation (23) is recognizable as equation (7) where (La> i s
defined as the total rigid-blade aerodynamic load minus a component
which is proportional to the rigid-blade defl ecti on velocity.
equation (7).
The matrix
is equal to the matrix multiplication [rjrrn] shown i n
l M4
Substituting {zrigid) ={r - h}B i n equation (24) yi el ds
The flexible-blade moments, therefore, may be determined by applying
equation (25) (the rigid-blade total aerodynamic load minus a rigid-blade
deflection-velocity load) to the flexible-blade equation (eq. (12)) which
is referenced to the plane of rotati on.
I n order to obtain the flexible-blade structural moment due to
rigid-blade airloads:
(1) Obtain the rigid-blade total aerodynamic load f or each harmonic
and remove the flapping effects by using equation (25).
(2) Use the loading j ust obtained along with the previously deter-
mined matrices, and obtain the moment by means of equation (12).
The reason the rigid-blade loads were handled i n thi s manner was to
keep the blade equations referenced to the plane of rotati on i n order to
provide for more general applications.
Structural moment due to measured ai rl oads.- The procedure for the
cese of measured ai rl oads i s as follows: Since measured loads include
al l aerodynamic damping effects, the aerodynamic damping term i s dropped
from equation (12) and the aerodynamic moment {Ma} i s given by
NACA TN 4304 15
{Ma} = [r] (Ltneasured}' Equation (12) is then applied by use of the pre-
determined matrices to obtain the structural moment-
If the hinged-blade equations are used with measured airloads where
o =fl and the equations involve no aerodynamic damping, the last row
and last column automatically drop out of equations (ll) and (12) when
the matrix multiplications are performed, and
i n the solution. This behavior of the equations i s consistent with the
physical concept of blade rotation i n the plane of no flapping or ri gi d
mode resonance.
Bh i s no longer involved
Moment of total aerodynamic loading as determined from measured
structural moment.- A requirement for determining the total aerodynamic
moment i s that the blade flapping angle at the root of the fl exi bl e
blade Ph be measured as well as the structural moment. The procedure,
then, is to apply equation (11) and drop the aerodynamic damping term
from the equation because the structural moments are a resul t of the
total loads applied to the blade.
The aerodynamic moment obtained is thus the total aerodynamic
moment and includes the moment due to al l of the aerodynamic damping
loads.
Aerodynamic damping loads.- I n some of the foregoing applications
it my be desired to determine the aerodynamic damping load or moment
which is proportional to the blade total deflection velocity 5
order to remove these effects from the measured data.
moment l i nearl y rel ated to the deflection velocity are given by
i n
The load and
\
and
The values of { F: } may be ei ther from measurements or from eal-
culations for the lower harmonics, but
calculated for the higher harmonics because of i ts extremely small
magnitude.
Ph
would probably have to be
16 NACA TN 4304
Determination of blade natural modes and frequencies.- If the blade
rotati ng natural modes and frequencies are desired, the procedure f or
thei r determination i s as follows:
(1) Solve equation (15) for uni t loads applied at the center of
each of the ten blade elements.
(2) Use equation (10) and the resul ts of step (1) and determine the
deflections due to the uni t loads.
( 3 ) Set up the numbers from step (2) i n matrix form, postmultiply
r ml , and follow standard i terati on proce- by the diagonal mass matrix
dures. (See ref. 6.)
The ten-element breakdown of the blade al l ows ten degrees of free-
domand thus permits the calculation of ten natural modes and frequen-
cies; however, the use of the standard i terati on procedure f or obtaining
accurate resul ts above the fourth natural frequency i s rather di f f i cul t
because the i terati ons converge slowly at these higher frequencies.
Application to Cantilever and Teetering Blades
The foregoing procedures were based on the hinged blade. When the
problem i s for a canti l ever blade, the only change i s to use the corre-
sponding cantilever-blade equation and matrices. The cantilever-blade
equivalents of the hinged-blade equations (10) and (25) are not given
but are obtained by making ph and p equal to zero i n these equations.
The cantilever-blade matrices are given i n tabl es I through V as the
el pents above and to the l eft of the dashed parti ti on l i nes.
I n an anal ysi s of a teeteri ng blade, the even harmonic loadings are
considered as applied to a cantilever blade; the odd harmonic loadings
are considered as applied to the hinged blade. The reasons for the di f-
ferent considerations f or di fferent harmonics i s that even harmonic
loadings are symmetrical loadings, that is, of equal si gn on both si des
of the hub; the odd harmonic loadings are unsyrmnetric.
NUMERICAL EXAMPLES
I n order to i l l ustrate the method, it i s applied to a canti l ever
and to a zero-offset hinged blade.
and sti ffness, with the exception that the sti ffness of the hinged blade
was zero at the hinge. The weight and sti ffness di stri buti ons are given
i n figure 4.
Both blades were of the sameweight
NACA TN 4304 17
The aerodynamic loading applied t o the hinged blade i n forward
f l i ght is shown i n fi gure 5(a).
harmonic of measured blade ai rl oads as defined by
The loadings shown are pl ds agai nst
coeffi ci ents f or the first
spanwise stati on of assumed Z1,c and 21, s
By using the loading of fi gure 5(a) and the weight and sti ffness of the
blade shown i n fi gure 4, the structural moment coeffi ci ent shown i n
fi gure 5(b) was calculated by means of equation (12).
damping term of equation (12) was dropped because measured loads already
included aerodynamic damping effects. The effects of structural damping
were not included, although the inclusion of structural damping i n the
equations should improve the accuracy of the calculations f or the case
where the blade i s near resonance. However, near blade resonance, the
effects of smal l errors i n the calculated blade weight and sti ffness on
the calculated resul ts are l i kel y to be large, and the problem requires
extreme accuracy i n al l components.
The aerodynamic
The structural moments calculated by the modal method of reference 3
for the loading of fi gure 5(a) are al so shown i n fi gure 5(b).
calculations the first four natural rotati ng modes and frequencies were
used i n equations of reference 3.
dropped from the equations. The mode shapes and natural frequencies
were obtained by the method previously outlined i n the cal cul ati on pro-
cedures.
obtained by the two methods is very good.
For these
The aerodynamic damping terms were
A s can be seen i n fi gure 5(b), the agreement of the resul ts
The aerodynamic loading applied to the canti l ever blade was the
forward-flight second-harmonic loading shown i n fi gure 6(a).
loadings shown are pl ots against spanwise stati on of assumed
The
22,c and
coeffi ci ents for the second harmonic of measured ai rl oads as
22,s
defined by
This loading and the weight and sti ffness di stri buti on of fi gure 4 were
used i n equation (17) to cal cul ate the structural moments shown i n
fi gure 6(b). Al so shown i n figure 6(b) are the resul ts obtained by the
method of reference 3 by using the same aerodynamic load (from f i g. 6(a))
and four rotati ng symmetrical modes and natural frequencies. The resul ts
calculated by the present method and by the method of reference 3 are
seen i n the fi gure to be i n good agreement.
The structural moments on the rotati ng hinged and canti l ever blades
caused by 100-pound loads applied at various spanwise stati ons are given
18 NACA TN 4304
i n figures 7 and 8.
and (18). By using the curves for the manent on the canti l ever blade
caused by 100-pound loads at various stati ons and the aerodynamic loads
measured at various stati ons i n the hovering test of reference 7, the
structural moments i n hovering were calculated. The calculated moment
and the blade defl ecti ons f or the hovering condition are shown i n fi gure 9.
The moments were obtained by the use of equations (15)
The various matrices used i n the example cal cul ati ons are shown i n
tables VI1 to X.
l i nes pertai n to the canti l ever blade.
elements pertai ns to the hinged-blade case.
The elements above and to the l eft of the parti ti on
The matrix of al l of the given
CONCLUDING RESIARKS
A matrix method has been derived f or determining the structural
moment and deflections of hinged, teeteri ng, and canti l ever blades. The
method avoids any preliminary mode shape, natural frequency, or quasi-
stati c cal cul ati ons and, as a resul t, is comparatively short and involves
only standard matrix procedures.
The equations are well sui ted to the determination of the blade
bending moment at various rotor speeds and f or numerous loadings. The
method puts no restri cti on on the blade defl ecti on shape and thus permits
the blade to assumemore complex shapes that could be simulated by the
superposition of four natural modes, as i s usually employed i n a modal
method of sol uti on of the di fferenti al equation for blade bending moment.
Thus, the present method should give more accurate resul ts than the modal
sol uti on f or the higher harmonics, where the blade defl ecti on and load
di stri buti on are of a more complex shape.
The method includes the effects of the primary i nerti a and damping
loads on the blade and al so incorporates mainly the following assumptions:
(1) small angle consideration, (2) step diagram of the blade mass and
sti ffness, ( 3 ) step i ntegrati on procedures, and (4) steady average aero-
dynamic considerations i n the damping terms. I n addition, the method
does not include the effects of:
with blade deflection, (2) vari ati on of rotor speed with azimuth angle,
and ( 3 ) any torsi onal considerations.
di stri buti ons may be removed by increasing the number of blade elements;
however, f or most blades ten elements should be suffi ci ent.
(1) radial change of mass posi ti on
Some of the effects of the step
The inclusion of structural darnping i n the equations should improve
the accuracy of the calculations for the case where the blade i s near
resonance. However, near blade resonance, the effects of small errors
i n the calculated blade weight and sti ffness on the calculated resul ts
are l i kel y to be large, and the problem requires extreme accuracy i n al l
NACA TN 4304 19
components. These unfavorable factors, i n some cases, could be somewhat
offset by experimental determination of the elements of the influence
coeffi ci ent mtri x and the mass of the blade elements and by some
experimental estimate of the structural damping.
blade i s not near resonance with one of the applied load frequencies,
the structural damping term i s rel ati vel y unimportant and may be dropped
from the equations f or blade bending moment.
LZM]
I n cases where the
Langley Aeronautical Laboratory,
National Advisory Comi ttee f or Aeronautics,
Langley Field, Va., May 7, 1958.
20
APPENDIX
INCA TIi 4304
BASIC INTEGRATING MATRICES
Moment of an Arbi trary Load.ing
The general bending-moment equation (eq. (1)) i s given as
I n this section the matrix equivalent of the expression
is derived.
The i ntegrati on of equation ( A l ) may be accomplished by using a
procedure based on step, trapezoidal, or parabolic representation of the
loading.
is considered to act at the center of each step.
I n the present method the step diagram is used, and the l oad
The procedure f or obtaining the moment of a loading i s as follows:
(1) The loading i s replaced by a system of equivalent concentrated
l oads as shown i n the following sketch:
Arbitrary l oad. stati on arrangement
NACA TN 4304 21
M1 M2 M3 M4
I I I 1
4 9 m n
'M3
"-' a4.I
(2) The moment due to the loading i s desired at the stati ons shown
i n the following sketch:
-
I-
r Ml rMl r Ml r Ml
rM2 rM2 rM2 rM2
rM3 'M3 rM3 rM3
r M4 ' M4 'M4 rM4
L
The loads are centered for 10 equal stati ons at 0.05(R - h), O.l ?(R - h),
0.25(R - h), and so forth, whereas the moments are taken at stati ons
0 . 8 5 ( ~ - h), 0.75(R - h), and so forth except near the root where stati ons
at 0.0125(R - h) and 0.0623(~ - h) are used to include possible hub
effects.
With the arrangement just shown the moment equation i s given as
'L2
%2
rL2
L2
r
I n equation (A2) any elements of negative si gn resul ti ng from the
subtraction are t o be replaced by zeroes since the moment contribution
of the loads inboard of a moment stati on is zero.
22 NACA TN 4304
/ \
M. 85
M. 75
M.65
M. 55
M. 45
M- 35
M. 25
M. 15
M. 0625
M. 0125
------
Mo
c-
I f the rectangular matrix of equation (A2) is symbolized by
[rJ,
then equation (A2) may be wri tten as
1
2
3
4
5
6
7
8
? =
8.87
9- 37
----
9. 5
, r
I.> = (1.1{.)
-
0
0
0
0
0
O 0
0
0
0
0 -
which i s the matrix equivalent of
r
\
L. 95
L. 85
L. 75
L. 65
L. 55
Fljt:::
L. 25
L. 15
L. 0625
L.05
L.0125
where the i ntegral is to be evaluated at a sel ected number of stati ons.
The
[r]
matrix is given i n tabl e V for the ten load and moment
stati ons of fi gures 2 and 3 . The matrix of tabl e V may be extended to
include more stati ons by expanding the equation {M} = [r] {L} as shown
i n equation (Ah), where the rectangular matrix postmultiplied by the
diagonal matrix i s equal to the [r] matrix.
5
6
6. 87
7. 37
4
5
5. 87
6- 57
3
4
4. 87
5. 37
----
5.5
0
1
2
3
3.87
4. 37
1
2
2- 87
3.37
0
1
1. a7
2- 37
0
0
0
0
0
0.87
1- 37
0
0
0
0
0
0
0
0
0
0.37
0.5
NACA TN 4304 23
To extend the matrix to more stations, continue the given number
sequence of the rectqul ar matrix and obtain more rows at the top and
more columns on the l ef t. Keep unity numbers on the diagonal and zeroes
on the ri ght. The moment stations are then the given stati ons (shown i n
tabl e V) multiplied by l o/%, (where
elements), and the obtained sequence is continued columnwise in order to
provide additional stations. For example, the elements of the f i rst
column of a 20-station arrangement would be 1, 2, 3, 4, 5, 6, . . . 16,
17, 18.87, 19.37, and l g.5(r - h)/(R - h); the load stati ons would be
0.025, 0.075, 0.125, 0.175, . . . 0.925, and O.g75(r - h)/(R - h); and
the moment stati ons would be 0, 0.00625, 0.03125, 0.0725, 0.125, 0.175,
. . . 0.875, and O.g25(r - h)/(R - h).
is the new number of blade
Moment Due to Horizontal Forces Acting on a Deflected Blade
The general bending-moment equation of the text (eq. (1)) is
I n thi s section the matrix equivalent of the expression
i s derived.
The system of equivalent horizontal concentrated loads, as shown i n
the following sketch, is again used, although some other representation
might produce sl i ghtl y more accurate results at the expense of simplicity.
24 NACA TN 4304
A t stati on
' 1
the moment due to horizontal centri fugal forces may
be wri tten
Collecting
as
terms and writing equation (A6) i n matrix form yi el ds
I f thi s procedure i s now applied and extended t o obtain the moments due
to the centri fugal i nerti a forces at ten sel ected moment stati ons on the
blade, the following matrix equation results:
NACA TN 4304
where
25
k 9
n2
F =mgr g=-
9
h
LH1
il2
F1 =mlrl =-
If the rectangulas matrix of' equation ( A8) is symbolized by [F]
(given as table 11), then equation ( A8) may be wri tten i n shorter form
as
which is the equation
Mrn = -
ZH(Z - %)dr i n matrix form and is
J rn
t o be evaluated at sel ected stati ons. The moments MiH are the moments
on the blade caused by the centri fugal forces acting on the defl ected
blade.
I n order to extend the matrix equation (A8) to n stati ons, make
al l of the negative summations on the shown diagonal go from Fn t o Fkr
where k i s the subscript of the F values i n the preceding column as
i l l ustrated. The elements off the diagonal of summations are obtained
by extending the indicated sequence to Fn. The new moment and deflec-
ti on stati ons are
sequence is extended columnwise f or addi ti onal stati ons. The new blade-
element arrangement has to have equal-length increments
10/ne
times the given stati ons, and the obtained
&.
26 NACA TN 4304
Blade Deflection Due to Structural Moments
For the bl ade the general bending-moment equation of the text
(eq. (7)) is given as
+I
I n subsequent derivations, a matrix expression for the defl ecti on
is required. The defl ecti on (z} is composed of a component due to
the flexible-blade root slope and a component due to blade bending.
thi s secti on a matrix expression f or the defl ecti on component due to
blade bending i s derived. The deri vati on involves establ i shi ng the
matrix equivalent of the anal yti cal expression
I n
The deri vati on of the required matrix expression is i l l ustrated herewith.
The i ntegrati on is first i l l ustrated with four stati ons and then extended
to more stati ons.
I The blade sti ffness and structural moment di stri buti ons are repre-
sented by the following step diagrams:
Bl ade sti ffness di stri buti on
NACA TN 4304
27
I Ar I A r I
M
.025 .1 .2 - 3
Bl ade moment di stri buti on
The segmental arrangement just noted could be made differently, but
(1) it
the particular arrangement shown was chosen for the reasons that:
permits a large change of blade sti ffness at the blade hub, and (2) it
gives a moment stati on where there are large changes i n moments when the
methods are applied to the hinged blade.
The blade angles /3r resulting fromthe bending deflections are
r
0 E1
the eq. Pr =>& &) giyen by the equation
Pr =
and can be deduced, for example, from the foregoing distributions as
follows :
A t (. - h, = 0.025,
(R - h)
28
NACA TN 4304
These di stri buti ons may be wri tten i n matrix form as follows:
B. 025
B.10
P.20
P.30
0
0
Ar
-
E13
E13
Dr
-
0
0
0
&
M 1
M2
M3
M4
NACA TN 4304 29
I n order to extend the procedure to obtain the defl ecti ons i n
terms of the moments, the following sketch proves helpful:
30 NACA TN 4304
Extending and col l ecti ng terms and writing i n mtri x form yi el ds the
following:
2.05
z.15
2.25
2*35
-
0 0
- 0 3$
-2
g z 2 2 &
--
4EI2 E13
Symbolizing the square matrix by [Q] gives equation ( ~12) as
Rearranging equation (Al2) to have the bl ade ti p values i n the
upper rows and extending to obtain defl ecti on values at the twelve
sel ected stati ons resul ts i n the following matrix.
and 3 for stati on location and the defi ni ti on of the elements.)
(See fi gs. 1, 2,
WCA TN 4304
2.95
z.85
2.75
2.65
".55
2.45
2.35
=.25
2.15
2.0625
z.05
'. 0x25
8
-
:I10
0
0
0
0
0
0
0
0
0
0
0
3 3 4 s
E I ~ E I ~
-2
cur
0 -
E17
0 0
0 0
0 0
0 0
-2
crr 2E2
- -
E I ~ E I ~
-2
O &
0 0
0 0
0 0 0
0 0 0
0 0 0
0 0 0
0
0
0
0
-2
O L V l
E13
0 0
0 0
0 0
0 0
32
Equation (Al 4) may be written as follows:
<
NACA mJ 4304
r
=.95
?85
2.75
2.65
z.55
2.45
4
2.35
z.25
?15
=. 0625
?05
2.0125
L 3
E
4
4
3
2
1
0
0
0
0
0
0
0
0
6 8
24
4
-
7
21
4
- 6 0 1
0
0
0
0
0
0
0
0
0
5
4
3
2
1
0
0
0
0
0
18
4
-
0
5
4
12
4
-
3
e
4
2 0
6
4
-
1 0 0
- 3
4
0 0 0
3
32
-
0 0 0 0
0 0 0
0 0 0
NACA TN 4304 33
To extend to more stati ons, simply extend the establ i shed number
Keep uni t values on the extended sequence i n the rectangular matrix.
diagonal and zeroes below.
and defl ecti on stati ons is as previously discussed i n the
The procedure for obtaining the new moment
[r] and I - -
LF] matrix derivations.
moment stat ions .
The values of E1 must pertai n to the new
The resul ts obtained by using ei ther equation (Al4) or (Al?) have
been compared with resul ts obtained by graphical numerical double
i ntegrati on of a known M and E1 di stri buti on, and the resul ts
agree very wel l .
could be determined experimentally from the actual blade.
If desired, the elements of equation (Al4) or (Al5)
34 NACA TN 4304
1. Flax, A. H., and Goland, I,.: Dynamic Effect i n Rotor Blade Bending.
J our. Aero. Sci ., vol. 18, no. 12, Dec. 1931, pp. 813-829.
2. J ohnson, W. C., and Mayne, R.: Effect of Second-Harmonic Flapping
on the Stresses of a Hinged Rotor Blade.
Part I11 (Contract W33-038 ac-11439), Goodyear Ai rcraft Corp .,
May 2, 1946.
Rep. No. R-107-4,
3. Beman, Alex: Response Matrix Method of Rotor Bl ade Analysis. J our.
Aero. Sci ., vol. 23, no. 2, Feb. 1956, pp. 153-161, 172.
Calculation of the Bending Stresses i n Heli- 4. Guillenchmidt, P. de:
copter Rotor Blades. NACA TM 1312, 1951.
5. Berman, Alex: A Parametric Analysis Method for Rotor Blade Struc-
tural Design. Proc. Thirteenth Annual National Forum, Am. Heli-
copter SOC., I nc., May 8-11, 1957, pp. 111-126.
6. Scanlan, Robert H., and Rosenbaum, Robert: I ntroduction to the
Study of Ai rcraft Vibration and Fl utter. The Macmillan Co., 1951.
7. Rabbott, J ohn P., J r.: Static-Thrust Measurements of the Aerodynamic
Loading on a Helicopter Rotor Bl ade. NACA TN 3688, 1956.
NACA TN 4304
Moment station, (r - h)/(R - h) I
' @h
0.85 0.75 0.65 0.55 0.45 0.35 0.25 0.15 0.0625 0.0125 I
I
1.0 0 0 0 0 0 0 0 0 0 10
0 1.0 0 0 0 0 0 0 0 0 10
- -
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0 0 1.0 0 0 0 0 0 0 0 10
0 0 0 1.0 0 0 0 0 0 0 10
0 0 0 0 1.0 0 0 0 0 0 10
0 0 0 0 0 1.0 0 0 0 0 10
0 0 0 0 0 0 1.0 0 0 0 10
0 0 0 0 0 0 0 1.0 0 0 10
0 0 0 0 0 0 0 0 1.0 0 10
0 0 0 0 0 0 0 0 0 1.0 I 0
-r-
-----_ -_ -_ -_ -_ -_ --_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
0 0 0 0 0 0 0 0 0 0 10
L I ,
35
Moment
stat ion,
r - h
-
R - h
0. a5
.75
.65
- 55
.45
.35
- 25
.15
.0625
.ox5
----
0
36
NACA TN 4304
TABLE 11. - CENTRIFUGALFORCE MATRM
[F]
pn =Wn]
Deflection station,
(I - h)/(R - h)
0.95 0.85 0.75 0.65 0.55 0.45 0.35 0.25 0.15 0.0625 0.05 0.0125
O l
FI.0 -%o 0 0 0 0 0 0 0 0 0
F10 "9 - X I Fn 0 0 0 0 0 0 0 0 0
F9 Fa -E F, o 0 0 0 0 0 0 0
F10
-10
-9
n=10
d 3
Moment
station,
r - h
R - h
-
0.85
* 75
.65
.55
-45
.35
- 25
.15
.0625
.0125
_ _ _ _
0
NACA TN 4304
37
~
Mrment station, (P - h)/(R - h) .
0.85 0.75 0.65 0.55 0.45 0.35 0.25 0.15 0.0625 0.0125 I
Lk f l e Ct i On
station,
a
R - h
0.55
.e
.75
.6
.55
.45
.35
2 5
.I5
. e 5
.05
.a25
NACA TN 4304
I I '
0 0 0 0 0 0 0 0 0 0 1 0
I
I
I
~ o o o o o o o o o o l o I
I
I
NACA TN 4304
39
-
3
8
8
In
In
3
9
0
In
0
I
In
0
c\!
In
0
M
In
2
In
0
'"
ln
0
\9
0
?
0
m
0
0'
-
I
0
0
0
0
0
0
0
0
0
0
0
' a
0
0
0
0
0
0
0
0
0
0
a
a
cu
0
0
0
0
0
0
0
0
0
a
b
cu
a
M
0
0
0
0
0
0
0
0
a
a
cu
a
M
a
f
0
0
0
0
0
0
0
a
b
cu
a
M
a
f
a
In
0
0
0
0
0
0
b
b
cu
b
M
b
f
b
In
a
W
0
0
0
0
0
a
a
cu
a
M
b
f
a
In
b
v)
a
t-
0
0
0
0
4
4
cu
b
M
a
f
a
In
4
W
a
P-
a
co
0
0
0
a
a
5
5
5
%
5
r i
a
cu
a
M
a
f
b
"9
a
?
a
t-
In
P-
W
P-
2
a
P-
co
I I
a3
I 1
0 1 0
I
I
I
I
0 1 0
I
I
cr; I & I
40
NACA TN 4304 '
.~
r a
'4
w
i
I I
0 0 0 0 0 0 0 0 0 0 0 0
r(
0 0 0 0 0 0 0 0 0 0 3 0
i3
0 0 0 0 0 0 0 0 0 0 0 0
h""
B
0 0 0 0 0 0 0 0 w o o 0
M
0 0 0 0 0 0 0 k o o o o
B
f
0 0 0 0 0 0 % o o 0 0 0
B
ln
0 0 0 0 0 ? ! n o 0 0 0 0 0
B
v)
o o o o b o o o o o o o
B
t-
0 0 0 g o o 0 0 0 0 0 0
B
Q,
o o % o o o o o o o o o
B
0 g o 0 0 0 0 0 0 0 0 0
2
0
4
p o o o o o o o o o o o
I B I
NACA TN 4304 41
-
0
I
o o m i 0
a 1
? I
I
0 0 0 0 0
0 0 0 0 0 0
r l I
3 I
0 0" 0 1 0
I
I
4 I
0 0 0 0 0 0
0 0 0 0 0 0
Y?
9
9'
O O O N 0 0
t
0
0
I
0
42
NACA TN 4304
I I I
0 0 0 0 0 0 0 0 0 0 1 0
I
I
$ I %
d l 8
g I W
cu. I M
0 0 0 0 0 0 0 0 0
I
I
I
I
0 0 0 0 0 0 0 0 0 0 1 0
NACA TN 4304
43
s
a
- -
In
$1
9
0
In
cu
9
0
9
In
0
I
In
0
c\!
In
0
?
In
0
f.
In
0

In
0
4
In
0
c-
In
0
9
-
P-
?
W
0
0
4
4
3
P
9.
M
M
cu
cn
t-
In
?
E
%
0
a:
cn
a3
2
?
t-
W
s:
I
M
W
W
cu
0
In
r l
W
9
P
0
cn
f
M
9
cn
cu
In
?
W
W
0
cu
9
19
c-
t-
f
cn
W
W
cu
cu
In
K\
M
W
t
%
0
0
3
o?
In
0
In
a3
9
In
cn
f
cu
r!
a3
f
M
In
I
3
v!
W
f
a3
cn
s
W
f
3
r!
u3
cu
0
0
0
cn
0
K\
9
8
t
In
P-
4
f
d
9
s
9
W
cu
In
s
3
0
0
0
0
0
cu
IC
o?
f
a3
t-
rn
\9
ti7
o?
3
0
0
0
0
0
0
8
cu
8
9
s
9
cu
3
0
0
0
0
0
0
0
M
0
r l
o?
In
cu
r l
s
0
0
0
0
0
0
0
0
4
8
a)
t-
t
0
0
0
0
0
0
0
0
d
0
3
0
0
0
0
0
0
0
0
0
0
9
0
0
0
0
0
0
0
0
0
44 NACA TN 4304
0
0
0
0
0
0
0
0
0
0
0
8
?
cn
0
0
0
0
0
0
0
0
0
0
8
I
8
K!
cn
CO
d
0
0
0
0
0
0
0
0
0
0
?
cn
8
9
8
f
a
d
P-
cu
0
0
0
0
0
0
0
0
0
I"
9
cn
0
a
d
8
s
t-
cu
0
0
W
K\
'9
0
0
0
0
0
0
0
8
I
cn
0
0
a
d
K!
si
5
cu
8
'9
8
'r
v)
K\
In
f
0
0
0
0
0
0
8
I
OI
0
0
a
d
K!
8
2
cu
0
0
v)
K\
'9
8
'r
In
f
0
0
f
I n
o!
0
0
0
0
0
8
I
8
K!
OI
03
d
8
f cu
8
'9
v)
K\
8
'r
In
f
0
0
f
In
o!
5:
9
f
v)
0
0
0
0
0
I n
cn
I
0
0
a3
d
K!
0
I n
2
cu
0
0
v)
K\
'9
8
'r
Ln
f
8
iA
8
9
f
v)
0
0
cu.
P
r;
3
l-
P-
In
+
NACA TN 4304
(a) Force system.
(c) Geometry.
45
- 90
(b) Deflections.
Figure 1.- Definitions of force and geometric symbols.
46
NACA TN 4304
Spanwi se st at i on, r- h
R- h
10
1
Figure 2.- Hinged-blade segmental arrangement for the matrix tables.
F1O
0
Spanwi se st zt i on, r/R
Figure 3. - Cantilever-blade segmental arrangement for the matrix tabl es
NACA TN 4304
Y
N
47
48 NACA TN 4304
r-h
Spanwi se st at i on, -
R-h
(a) Spanwise vari ati on of fi rst harmonic coeffi ci ents of blade ai rl oad.
(b) Calculated spanwise vari ati on of first harmonic structural moment
coeffi ci ents due to first harmonic ai rl oad coeffi ci ents.
Figure 5.- Airloads and structural moments for hinged blade used i n
numerical examples.
NACA TN 4304
49
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0
Spanwise stati on, r/R
(a) Spanwise variation of second harmonic coefficients of blade ai rl oad.
Spanwise stat i on, r/R
(b) Calculated spanwise variation of second harmonic structural moment
coefficients due to second harmonic airload coefficients.
Figure 6.- Airloads and structural moments for cantilever blade used in
numerical examples.
50 NACA TN 4304
r-h
Spanwise stati on, -
R-h
Figure 7.- Structural moment on zero-offset rotati ng hinged blade due to
100-pound concentrated loads at various stati ons.
.7 .8 .9 1.0
Spanwise stati on, r/R
Figure 8.- Structural moment on rotati ng canti l ever blade due to 100-
pound concentrated loads at various stati ons.
WCA TN 4304
NACA - Langley FIeId, Va.