c .

C.P. No. 32 (12738) A.R.C. Technical Report

MlNlSiRY
AERONAUTICAL

OF SUPPLY
RESEARCH COUNCIL PAPERS

CURRENT

Air

Bags as Flexible Supports in Ground Resonance Testing ~of Aircraft

BY R.L.Moffat and D.R.Gaukroger, M.A.

Crown Copyright Reserved

LONDON:zHIS

MAJESTYS STATIONERY 1950 (

O/WE

Price 1s. 3d. net

C.P.32.

Technical

Note No. Structures September,

47

1949

ROYALAIRCRAFT ESTABLI-T. Air Bags as Flexible Supports 'in Ground Resonance Testing of Rlrcraft . ' bY

R.L. ivioffat ana l3.R: Gaukroger;

X.8.

Stlf'lness and natural frequency tests have been made on an air bag designed for the ground resonance testi% of a medium S'LZC arrcraft. Results arc compared with those obtained from an existing theory and from a new theory put forward in this report. Agreement between the experrmental results and the theory of this report is very good. The results show that the present design of aw bags does not provzde the low frquenoy anticipated.

LIST OF CONTENTS Introduction Typ of azr bag testerl Method of test Test results

4.1 Stir'lhess tests 4.2 Frequency tests

Comparison of ~xperimentzit Discussion Conclv sions of' results and theoretical results

1.

LIST OF ILLUSTRATIONS u. Type of air bag tested Test rig for nir bag Height-Pressure curvee for ?W'ferent loads-expermental Height-Load curve - oxpe3montal Height-Load o~rves for diffevcnt oxperimenta1 initial heights loads - experimental 1 2 3 4 5

Frequency-Hexght curves for different

Frequency-Load curves - experimental and theoretxdl Height-Pressure curve - theoretical

Introduction

A consideration1 of various types of flexible support for airoraft during ground resonance tests indicated that one of the most promising methods of providing the desirable low support frequency was by the use of air bags. No detailed invostigntian of the behaviour of air bags under vs.rious conditions of lonkng has been madeup to the present and the method of support has not been widely used in the'resonanoe testing of aircraft. Recently, however, sir bags were specially &signed for uso with a large aircraft rind the results obtained indicated that the support frequency WAS higher than tho frequency predicted theoretioally.1 A detailed investigation was therefore made of the characteristics of on air bag of similar design to those used in the above tests, and the original theoretical treatment1 was examined in detail. This report gives the results of stiffness and frequenoy tests made on the air bag, and shows that the fonnuln for the bag frequency of Ref.1 is lncorreot. A new formula is established end shown to agree well with the test results, It is concluded that the present Zesign of air bagsdoesnotprovid- L the required low frequency support for verticnlmotion of the aircraft for ground resonance testing and that further development IS necessary.

The air bag on which the tests were made was constructed of three The ply dinghy fnbrlc cud the main dimonsions nre given in Pig.1. sides of the bag consist of five lobes, the junctions of which are held in position by 30 cwt. cable. During design experiments the bag was found to be unstable in both shear and roll, To overcome this in- _/ stability, vortiocl diaphragms wore provided internally, and a surrounding apron nttachod to the lobes was fittod externally. The dinphragms allow from exp,%sion of the b&g up to a height of thirty inches; this height is subsequently referred to as the tdesrgn height' and bag heights Apressure of 2poundsper square are expwssed as percentages of it. inch above atmospheric is regsrded as 'design pressure' end this gives a 'dosign lo-d' of 8150 pounds, ether-loads being expressed as percentagos of design load. 3 Method of test

3.1 The test rig shown in Fig.2 was used In both stiffness end ' frequency measurements, A 61 x 6' x $' steel plate "A" for&d a platform to which additional weight could be directly added; IDa to a fixed point at one end and whi.le c besm "B" hinged vertically "x to the loading plate at,its centre point hoted as a stabiliser. For the stiffness tests increments of load were added and the variation In bag height measured. Three sets of initial conditions wore exsmined- heights of 30" 30.5" md Jl undw a load of 24.6 of' The difficulty cf loading above 73.7$ design load (i.e. 2000 lb). of design load (&CC lb) by means of dead weight prevented a direct Loading up to the full dessgn load. For a fourth stiffness test therefore a lover system was used and the starting conditions were:bag height 31" snd load 614 (5900 lbs). The load was inoreased Considerable friction was jnorementslly to llO.@ (SC00 lbs). apparent in the lever system, and for this reason dir-eat loading was used in all other tests. The varlntion of pressure with bag hoight under constant load was measured in order to determine the effect of the internal-diaphragms, and tho docreasc of supporting area with hoi&t.

3.

3.2 TO obtain the bag frequency under load the beam "Ei" was excited vert1cauy at Its free end; -xt ivas convenient to excite by hand since damping was so lorr that only at the resonant frequency was it possible to obtau any apfireciable amplltu'de. The frequency was measured by , tlmlng a number of cycles. 4 4.1 IT_est.results Stlfrness -. tests .

Curves of bag height against au pressure Two points are notewokthy.shown xn Flg.3.

for various

loads are

(1) (11)

that the stiffening effect of the,lnternal becomes apparent between 100 nnd~'l03$,of

diaphragms the design height, of

that belox the dcslgn height there is a slight vanntion pressure with height, lndicotlng that the bag supportIng this is due to area decreases with uxrense of height; the rounded edges of the top and bottom of the,bcg.

The stiffness test curves arc shown U-I Figs.4 and 5; these lndlcate a.dlaphragm effect between 102 and 103s of design height, but omng to the 'r&Sk of damage to the air bag, heights above 103.3% (31 inches) were not lnvestlgated, and the crltxal hei.#t for dlaphra@ effect is not>accurately established from the static test _I results. 4.2 Frequency tests,

'

x6t.b load and height IS shovn by the L The bag frequency variation k cross plot of these results IS shown in set of curves 2.n Fxg.6. Flg.7 which gives varxatlon of frequency against load (a) as measured, measurements, and (c) as calculated (b) as calculated from the stiffness from the formula gxven III pora.5.1 of this report. The frequencxs obtalned from stiffness me?surements have been corrected nssumlng MOthermal conditions for stlifness and adlabbutic conditions for frequency measurements. Due to dlnphrcgm~effect a crltx$ol height IS apparent in hg.6. This is at 101% of,deslgn height and indicates the upper llmlt of bog height permuslble for eff-ective operntlon. /< ,' > The comparison of experimental and theoretical results 5 From the conslderntion of a simple cw column of height S,-unit p and applied loading Vo, cross-sectional area, internal pressure such a colums~~s given by:that the stiffness oF Molyneuxl establishes K J-c+ 1 ". therefore . . . . . (1) . . . . . (2) *.... (3)

where Y 1s the adiabatic constant for nw, he concludes that the frcguency of the applied load on the column is.-

f = 1 2x
Using the relationship he then derives

YPo g SW,

w. = Po

I+.

Relationship (2) is, however, inconsistent wrth the use of Wo in (l), where it represents the applied structural loading (inertia of the ambient air being ignored). On this interpretation, (2) should be replaced by v& = PO - PA pA bY being the atmospheric prossuro, and oxpressron . . . . . . (4) (3) is then replaced

. ... ..

(5)

where

Ap = excess bag pressure

above atmcspheric.

Using formula (5) a theoretical curve is shown in Fig.7 and shows good agreement with the measured frequcnoias. There will inevitably be some difference due to the bag expandin& laterally under load; this should reduce measured frequencies compnrcd with theoretical and Fig.7 confu.ms that this is so. Wg.7 also shovs the error which arises from the use of the frequency f'onula

The linearity of the stiffness ~ux-ves (Figs./+ and 5) seems ;It first sight to be erroneous; the curves should follow the pV curve for the mass of air in the bag. Rut, as 1s shown in Pig.8, the section of tho pV curve in which the pressure and volume are varied is small by comparison with the rest of the curve, and the stxffness ourves <are assumed to be linear within the limits of experimental error. Similarly, with dif'fzxng initial condition in the stiffness tests, different stiff'nesses would be expcctcd, ance the mass of air, and But tho pV curves for the hence the isothermal constant is varied. small air mass variation which occurs in the bag are nearly parallel, and again the otiffness variation cannot bc measured within the limits of expermental error. 6 Discussion of results

In designing a flexible support for ground resonance tests it is shov,n that a sufficiently close estimate of air bag frequency can be obtained from the fonnula:f, where f -$ 1 +sJ

= vertical frequency in c.p.s. v PA = atmospheric pressure Ap = excess air pressure in bag above atmospheric s - bag height, 5.

For all but very large aircraft the lower limit of the lowest fundamental symmetric frequency can be taken as 3 c.p.s.1 and the corresponding desirable support frequency as 4 of this valuel: viz. For the latter value and taking an air bag 30 ins. in height 1 0.p.s. the internal pressure necessary is 12.4 lbs/sq.in. At this pressure a bag diameter of only 14.5 inche s would be required to support the design load of the bag already tested. Generally, it would be impractical to USC air bags more than six feet in height, and if the bags are in direct contact with the under surface of an aircraft a loading ~11excess of 2 pounds per square inch is prohibitive. A bag of this height and at this pressure would have a frequency of 1.3 c.*.s. (approx.). This value is in excess of the desirable frequency of 1 o.p.s. but might be reduced if special means were used to distribute the load of the aircraft. 7 Conolusions

The experiments and theory of this report show that the present design and application of air bags do not provide the desirable lm frequency support rcqmred in the ground resonanoe testing of medium and large size aircraft and that further development is necessary.

LIST OF REFERENCES Title, 1 W.G. iMolyncux etc.

Flexible supports for the ground resonance testing of aircraft. R.A.E. Report No. Structures 32. September, 1948. ARC. 11.964.

kt. .?l78.Ci32.d3.

Prrnred

1% heat

BIstctn.

6.

FIG. I.

i
I/ 1 +----/ I >L+----/ . +---, \q+-I I i I

W-0

ACROSS

CORNERS.

SCALE :- r:

FULL SIZE.

MATERIAL

:- 3 PLY

DlNcjHY

FABRIC.

FIG.I.

TYPE

0~

AIR BAG 0

TESTED

C~FTDIAMETER)

FIG.2.

POINT

OF

%:TAToN FREQUENCY TESTS. LOADlbdq PLATFORM A. HlNCp. SEAM 8. \

I I

1 A,

I I

FIG.2. TEST RIG FOR AIR BAG.

FIG. 3.

AIR PUMPED 1~70 BAG, TO INCREASE N~lc,HT AT GIVEN LOAD.

O-5.

-x

= POOOLB. LOAD 0 = 3000 LB. LOAD A= El = 5000L5. LOAD

'24.6
= 36.0 61.4

% 0
20

94

95

96 es

97

DESlqN 98 ,

HClcjHTj 99 100 I 101 I 102

103 31

30 S4q IWcjHT (INS3

FIG.3.

BAG HEIGHT - PRESSURE CURVES DIFFERENT LOADS EXPiRlMENTAL.

FOR

FIG.4.
AMOUNT OF AIR I,N BAG KEPT CONSTANT.

00%
000

LOAD

LB

6001

STIFFNESS

= I800

LB/IN.

SOOI

as

96

% DESIGN 97 98 HCI qHT

HEIGHT 93 (IN)

100 30

101

IO2

FIG.4.

BAG

HEIGHT - LOAD

CURVE EXPERIMENTAL

4.

_I

-3

6000 LOAD L0. 5000-

CONDITIONS X LOADING COMMENCED AT 31 2000 LB. @ LOADING COMMENCED AT 30-S 2000 LB. 2000 LB. q LOADING COMMENCED AT 30 AMOUNT OF AIR IN BAG CONSTANT FOR EACH CONDITION, BUT DIFFERENT FOR THE VARIOUS CONDITIONS.

3000 -

HEIGHT. 2QOo. 27 29 HEIGHT IN. 30

FIG.5

BAG

HEIGHT-

LOAD

CURVES FOR EXPERIMENTAL

DIFFERENT

INITIAL

HEIGHTS

FIG. 6.
AIR PUMPED INTO BAC-, TO INCREASC HtlqtIT AT GIVEN LOAD.

- 24.6 -36.6

-l
I

PERCENT&q #es
OF DESlG,N

A -49. El -61.4

LOAD.

l-

96

98

% OF DESIGN HElCjtiT.

1 3

IO2

FIG.6.

BAG FREQUENCY -HEIGHT VARYING LOAD EXPERIMENTAL.

CURVES FOR

FROM MEASURED

STIFFNESS.

MEASURED

VALUES.

BAG HElG$iT = DESl+ FOR ALL CURVES.

% 08

OESlqN

LOAD.

FIG. 7. BAG FREQUENCY - LOAD CURVES. EXPERIMENTAL & THEORETICAL.

FIG.8.

bv

DIAGRAM

c z T -z 2 2 I? In it? LL

NORMAL OPERATING

40.

CONDITIONS.

dd 20. \ I 0

IO
BAq

20
HEQHT (IN.) =w

30

40

FIG.8.

BAG PRESSURE - HEIGHT THEORETICAL.

CURVE.

C.P. No. 32 (12738) A.R.C. Technical Report

Kmgsway, LONDON, c 2, 429 Oxford Street, LONDON, I, w w PO Box 569, U)NDON. sel, 13a Castle street, EDMBURG, 2 1 St Andrews Crescent, CARDFF 39 King street, MANCHESIER, 2 1 Tower Lane, BRISTOL,1 2 Edmund Street, BIRMINOHAM,3 80 Chichester Streef, BELFAST, or from any Bookseller

York How,

1950
Prtce Is. 3d. net

S.O. Code

No. 23-9006-32