NINETEENTH EUROPEAN ROTORCRAFT FORUM
Paper
n"G4
ROLLER VIBRATION ABSORBER FOR HELICOPTER MAIN ROTOR HUB
by
W.HAWRYLECKI, J.KLIMKOWSKI WSK PZL SWIDNIK CO. (POLISH AVIAT.WORKS) September 14-16, 1993 CERNOBBIO (COMO) ITALYASSOCIAZIONE INDUSTRIE AEROSPAZIALI
ROLLER VIBRATION ABSORBER FOR HELICOPTER MAIN ROTOR HUB
W.HAWRYLECKl, J.KltMKOWSKl WSK PZL SWIDNIK CO. (POLISH AVIAT.WORKS) 1.ABSTRACT
During the first flight of PZL-Sokot helicopter equipped with a main rotor composed of 4 epoxy-glass lades,lt
was found out,that
the level of loads on the rotor head anns, as well as the level of vibration Inside the helicopter cabin, were alarmingly high. Modification made In the blade construction Itunnlng natural frequency reduced the loadson
the rotor head anns but the vibration level Inside cabin remained exceptionally too high and Increased with 1\y\ng speed.Through analysis II was found that use of a dynamic vibration absorber In the rotor head should reduce the vibration to the level not exceeding the relevant standards and regulations. \! was decided to add to the main rotor head a dynamic roller vibration absorber of the Solomon type so far It seems to be the only construction of this type which works succesfully on the serial helicopter. The absorber works successfully within enUre range
of
operating conditions. This paper covers the progress In development of absorber. The attention Is paid mainly to tests and measurements.2.NOTATION
R - distance from center of rotation to center of ann bushing, {m}. 0 diameter of arm bushing, [m].
d - diameter of roller, [m). m mass of roller, [kg). <f) - hub rotational angle, (red).
a. - roller degree of freedom, [rad). 10 - roller moment
of
Inertia, [kgm2]. xh,yh- hub coordlnates,[m).o> - rotor angular speed, [rad/sec].
1!. - measured angle of roller position,
I
red].r
-
roller spin, [radj. e=
(D -d) I 2 jm}.N - number of rotor blades. P - per one revolution. 3. INTRODUCTION
In Poland, the first dynamic wlbratlon absorber was Installed on the helicopter main rotor hub in 1950. II was of blfilar type and fulfilled
Its
rote successfully.This paper relates to PZL -sokol helicopter equipped with one 4-blade fully articulated main rotor with blades made entirely of epoxy-glass composites.
The maximum helicopter take-off Is 6400 kg. The helicopter ls designed, constructed and tested In Transportation Equipment Factory PZL-Swldnlk.
During the first flight of the helicopter, It was found out, thai In hovering, vibration level inside cabin as well as ln other points of the fuselage was within expectallons. Vibrallon started however to increasing with forward speed and reached hazardous level at high speed. The diagram In /Fig.1/ illustrates the problem.
One the basic of numertcal analysis II was found that vibrations were excited by forces perpendicular to the rotor shaft axis. The results of the computallon were verffied by means of strain gauge measurement of share forces and bending moments along the rotor shaft lenght. Special attention was paid to the determination of natural frequencies and the corresponding modes of helicopter vibration. One frequency was almost equal to
the /4P/. The resonanse of the frequency /4P/ appeared. The change of rotor shaft angular speed gave positive but sUlllnsulflclent results. However, sUch a Change may cause the worsening of the helicopter CharactertsUc. Therfore, most attention was devoted to the application of the rotor head vibration absorber. To solve the problem, It was decided to use the absoroer of the roller type, /Flg.2/ so far never used on the other helicopters. II was proved, quite property. It has a number of advantages when comparing it to absorbers of
the
blfllar type used commonly.4. MATHEMATICAL MODEL OF ROLLER ABSORBER
Flg.3a Illustrates scheme of the dynamic model thai was applied to describe the roller diceplacemenl. Ills convenient to describe the roller movement using the Lagrange's secund order equations. When the roller rotates wlthaut slip, position of the roller center ~nertla coordinate system (x,y) Is determined by the relations:
~
=
R
cos~ +e
cos(~+a)y
=
R sinm + e sin!m+al
The roller klneUc energy for the lnerUal system (x,y), Is expressed as+
y
... J:L .. --- .. -- --- ---· -~ -~--- ··--~ ~ ~ ~ ~ (1) _ 1 ~ ~ 1 r-~~ + ry~ ~J E!;= ;
m C;: b .:" 'f.b] + ; I 0 ·.:--: . . . . :-:d_~...
... (2)By substituting (2) to Lagrange's equation, the equation decribing movement of the roller and rotor hub center, is obtined.
I +
I
+yhe cos(¢ +al
+le
2+ R e cosa
+~
02
~~
¢
=
0
Assuming constant angular velocity of the rotor shaft
>:
=
0 h - · - - - · ·y h :;:;:: (} . ¢
=
w=
canst.,
slmplll!ed linear equauon of the roller movement Is obtained:Hance, natural frequency of the roller movement Is: 0
= "".
J
~..
J
sincx ~ 01 1 1+
4 Io m d2 (3) (4) (5) (6) In the equations (3) and (5), damping Is neglected. It may have essentlallnfluance on the roller movement, and hance, efficiency of the absorber. In Ref.[1 ,2,3,4) one can lind more information about mathematical models of absorbers,both roller and bifilar type. The nature of the damping forces Is difficult to evaluate In ·theoretical way, especially for the absorber of such type. Therefore the attention was paid mainly to laboratoryexperiments and flight tests.
5.DETERMINING OF THE ABSORBER DIMENSIONS.
The main dimensions of the absorber were detennlned on the basis of the formula (6), assuming that the natural frequency equals to (3P).The Inertia force produced by rollers must compensate the hannful force that exdtes vlbraUon. The maximum value of this force was determined during flight with V max and during landing approach. Durtng the absort>er wO!l<., the roller must roll without slipping Inside the crank bushing . .The roller rotatloo was detennlned by value of a /Fig.3a/.
The force produced by the absorber Increases with angle a .. However, the angle a cannot exceed some value because slipping appears between the roller and bushing and the absorber loses Its features. By means of many exsperlrnents It was detennlned, that the angle a should not exceed 30".
6.MEASUREMENT OF KINEMATIC PARAMETERS
Measurement of the roller dlsplecement was carrtOO
out
by means of the system whose simplified scheme Is Illustrated In Flg.3b. By means of two mlcropolenliemeleftl·Aaf!Ci.B<XlllllOOiedtogetiJer wlth telescopic ann, the roller spin angle Ymeas and the roller displacement angle 11 were measured. WHh maintaining the condition where no slip occurs during rolling, the angle Ytlleor and 11 are related to each other, according to geometrical relatlooYtheor
=
11[(2e/d)+ 11 + (2e/d)arcsln[(2rld)slntl.j (7) The theoretlcal angle of the rolle(s spin Ytheor can be determined measuring the tlme hlstOI)' of 11. The 1 angle Is measured by potenciometer B. Formula (8) exspress theslip= 'ttheor- Ymeas
(8)
value and sign of roller slip. 7 .ABSORBER TUNNING
Four sets of vartous diameter rollers were made. They ensured the absorber resonanse frequency within the limits of (3P)• 1%. As a criterion for tunnlng the absorber, the amplitude of verllcal acceleratloo In the cockpit of the /4P/ frequency was assumed. Additlooal cl1terton was the phase shltllng between the rotor shaft bending . moment (3P) and angle of roller displacement a. Fig 4. shows the measurement results for the sets used. On the basts of this and similar dlagmms, the best roller set
was
selected. During flight tests, It seemed that at strong exdtatlng force, when helicopter was approaching landing, the angle of roller displacement a exceeded30"
which resulted In roller slip. flg.5 shows the diagram which Illustrates the roller slip. To Increase the absorber reaction force while maintaining small angle a, the absorber construcllon was changed by Increasing length ofIts
anns R. Respectlvely, other dimensions of the absorber were changed. The absorber having the following main dimensions R = 450 mm.,d = 108.5 mm.,m = 7.89 kg.,D = 168 mm. 10 =0.139 kgm
2 worked
wlthBut slip In all the flight conditions, which was con1irrned by experiments. Fig.
'1-
Ulustrates baste quantities characterising absorber effldency.8.FUNCTIONAL TESTS
The aim of the tests was examining the vibration damper for Hs race wear interwlly and !hate wear effect on the vibration damper operatloo effldency. The wear measure were differences In the race geometry from the drawing dimensions and the efficiency operatloo measure were vibrations and loads In a selected helicopter places measured on the helicopter In flight after the determined vlbrallon damper operatloo durallon on a special stand. The stand drtven with elec.motor was designed to simulate the vibration damper operation (rprn,vlbraUon amplitude In the plane of revolutloos, angular displacement of the rollers). Measurements of the vlbratloo damper for effectiveness have been carrt00
out
for the same helicopter. The tests for three pieceS of the vlbrallon dampers were carriedout.
Each of the vlbratloo dampers was operated on the test stand for 1500 hours and every 500 hours has been tested for appropriate operatloo In flight. The test results were pasitive.9.FATIGUE TESTS
The purpose of the fatigue tests was establishing the fatigue life of the vibration damper components and the attachment of the vibration damper to the main rotor hub. Three types of the samples were used for the tests. The samples consisted of the orlglnal vibration damper components. Totally nine samples (three samples of each type) hare been tests. Loading spectrum was established bavlng on the lnfllght test results. Component of the lowest fatigue life was the disc (Fig.6).
10.PROPERTIES OF THE ROLLER ABSORBER
For comparlason within the scope of research program, similar test were canied out for the blfllar absorber. Theoretical comparison of both types
Is
desCI1bed In Ref.[3J. Main dimensions of the both types differed slightly -weight of the roller type was 50 kg., and the blfllar one 56 kg. Operating efllclency of the either absorber was similar. However, production cost of the roller absorber was four. Urnes smaller then this for· theblfllar one. The· overaH·welght of the roller· absorber;· indudlng·eovertng·etements·, Is less than 0.8% of the
helicopter take-off weight.
The separate problem to be solved, was theselectlon of materials for the absorber rollers and crank bushing. By experiments the steel ensuring assumed fatigue life of this parts, was chosen. Proper operation of the absorber is influenced by some construction details of the rollers and crank bushing. These details are unique design which Is patented.
The roller absorber Is sensitive to the external conditions, partlculary to water,oll,etc. Therefore, the rollers and crank bushing should be well protected from the envlronement.
11.CONCLUSIONS
1. According
to
experience gathered for several years, the roller absorber can be used on helicopter.2. Production and repair costs of the roller absorber are lover then analogous costs related to the blfllar absorber one.
3. The roller absorber proves Its efficiency comparable to that of btfllar absorber. 4. Weight of the roller absorber should not exceed 1% of helicopter take-off weight.
5. The functional and fatigue tests have confirmed the usefulness of the vibration damper choice for vibrations damping In the plane of rotation on the PZL-Sokol helicopter main rotor Installed.
References
1. Zurakowski,B.,"A VIbration Absorber for a Hinged-Blade Helicopter Rotor", Prace lnstytutu Lotnlctwa, 93(211983) pp.3-78
2. Mouzakls,T.,"The Monofilar Dual Frequency Rotor Hub Absorber" AHS Journal, OctOber 1983,pp.22-28. 3. Pasley,P.R,and Sllbar,A.,"Optimale Auslegung von Salomon Schwingungestifgem•,lngenleur Archlv.XXIV Band 1956,pp.182-187
4.Marthy,V.A., and Hammond,C.E.,"VIbration Analysis of Rotor Blades with Pendulum Absorbers", AHS Journal, January 1981, pp.23-29.
I .
.
I
II
uI
·:g,
6•
/
/
...
,...
4•
I
.. 2I
0_o. __
~o. _
lo __
ltL lo_lOO~ 140 16Q_lBQ_200.22.Jorwo.rq ~peecl Cl<m/o]
Fig.l Level of vertical acceleration
vs forward speed.
__ f'ig. 2
Ro11 er type absorber of PZL-So}se>l. rotor hub.
a).
~
I
b).
Fig.3
a) Dynamic model of absorber.
b) Scheme of system for measurement of
roll_~r: -~i_:>P
l
aceme~n,t=·===d•O. 1092 m
/
/
/
/
/
d•O. 1 /v
/
.8 .6/
//
Y.
/
. /
__.
~
v
...
...,
'"'
d-0. 1085 m EO 100 120 140 160 - ... 180 200 220 2402~---
-- ···---·---
...-Fig.4Level of
cockpit____:::_~Tt:_ica,l_'":.£_<:;el _
_eration
(4P),..,
•
'-' 4 ~ ~ £ )o - - - · - - · - · - -··--····-· ...•• - . . . . · · · · · · · · · -A3S A20 1'1I
n'
11
AOSv
vgo
1\ ~ ! ! I .7'5Go
v
v
v'
v
v
v
v
.
I
. . . 0 1Fig. 5
F_r.<'~m~.!'!t__
2.LL~.Q.rd . of th_f'_I.Q)l e_r measured rotatt.Cl.D.
slip.
~i~.
6.
angler
illustrating
the
meas.
Vib•a.iion
o.bsorber model
u~et:l}o,..
she:;!>
ca. leu lo.
t
•'on.
I
.. ··---··- ---·- --- -- ---___:
_____
..,1
with
... to al amp! put ab901 b.
00111 l. of (4 ) with ut 0090 b. oto
.~~ n~~~l
w th 0090 b. ·~ ~)•
lt.h 8 uv
...
...
/
4 ,......
v
/
...
...
~
0 2~
.
¥
•
~9
...
.
.
·• 0lo
lo
0 120 160 200Fig·"' Am£1 itude.s
_
_<:>L_y_~rt;jcal
coc~pit
5_<:;~
l era_t;)on with a.J1d without
absorbe1· vs forward sp_eed.
G4-8 '