e~r. tp jb ~;r, ur-o ,. ~-f~~i - dticsignificant amount of wind tunnel data pertaining to...
TRANSCRIPT
ARCHIVE COpy DO NOT LOAN
A LOW-LOAD THREE-COMPONENT FORCE BALANCE FOR MEASUREMENTS IN A LOW-DENSITY WIND TUNNEL
roved fO!j.pub\ic re\e~R. -11 ~ do$;~ment has b~en app ". riA-' b~o" TP 0 (
1 s oo, j"b I'on ,', 1I:,,\l11111"! d. J\ D 1LrI II '1 17 ~;"r, ur-o ,. ~-f~~I"-
George D. Arney, Jr. and Wade T. Harter
von Karman Gas Dynamics Facility ARO, Inc.
TECHNICAL DOCUMENTARY REPORT NO. AEDC-TDR-64-280
December 1964
Program Element 62405334/8950, Task 895004
(Prepared under Contract No. AF 40(600).1000 by ARO, Inc., contract operator of AEDC, Arnold Air Force Station, Tenn.)
__ I ARNOLD ENGINEERING DEVELOPMENT CENTER
AIR FORCE SYSTEMS COMMAND
UNITED STATES AIR FORCE
NorlCBS Qualified requesters may obtain copies of this report from DOC, Cameron Station, Alexandria, Ya, Orders will be expedited if placed through the librarian or other staff member designated to request and receive documents (rom DOC.
When Government drawings, specifications or other data are used (or any purpose other than in connection with a definitely related Government procurement operation, the United States Government thereby incurs no responsibility nor any obligation whatsoever; and the fact that the Government may have formulated, furnished, or in any way supplied the said drawings, specifications, or other data, is not to be regarded by implication or otherwise as in any manner licensing the holder or any other person or corporation, or conveying any rights or permission to manufacture, use, or sell any patented invention that may in any way be related thereto.
AEDC·TDR·64-280
A LOW -LOAD THREE -COMPONENT FORCE BALANCE FOR
MEASUREMENTS IN A LOW-DENSITY WIND TUNNEL
d f ublic. release 1r1 r Tms document hasbeen 3Ppr~v~ 0prPv D D~ JTf'- ~ • .' its distribution is unlimited. ~. 1) f+-o II '"1 0 ()1f
\+11 ~r'~
By
George D. Arney, Jr. and Wade T. Harter I I
von Karman Gas Dynamics Facility
ARO, Inc.
a subsidiary of Sverdrup and Parcel, Inc.
December 1964
ARO Project No. VL2517
AF - AEOC Arn01d AFS 'penn
References to named commercial products in this report are not to be considered in any sense as an endorsement of the product by the United States Air Force or the Government.
A E 0 C· T 0 R·64-280
FOREWORD
The authors are grateful to the several persons who have made helpful suggestions throughout this work. Special acknowledgment is due Mr. A. B. Bailey, Mr. J. F. Roy, Mr. J. C. Potter, Mr. R. P. Van Hook and Mr. D. E. Boylan.
AEDC-TDR-64-280
ABSTRACT
This paper describes a three-component force balance which has been developed for measurement of lift. drag. and pitching moment acting on models in a low-density. hypervelocity wind tunnel. Although designed for a particular wind tunnel. the general arrangement and the principles involved should be applicable to similar situations.
The balance is of the external type operating on the nulling principle. Nulling of each component is accomplished automatically through the use of closed loop control systems. Calibration data and sample wind tunnel measurements are presented which demonstrate that the balance is suitable for the intended application.
PUBUCATION REVIEW
This report has been reviewed and publication is approved.
Ll R(;u..t~ R. Cureton
'uMltf}tl/AJean A. Jack
Captain, USAF '~Colonel. USAF Gas Dynamics Division DeS/Test DCS/ Research
v
AEDC-TDR-64-280
CONTENTS
ABSTRACT. . . . . . . . . . . v 1. 0 INTRODUCTION. . . . . . . . . 1 2. 0 DESCRIPTION OF THE BALANCE 2 3. 0 STATIC CALIBRATION . . . . . . . . . 3 4.0 SAMPLE WIND TUNNEL FORCE MEASUREMENTS 4 5.0 CONCLUDING REMARKS. . . ... 5
REFERENCES. . . . . . . . . . . . . . . . . . . 6
ILLUSTRATIONS
Figure
1. Mechanical Arrangement of Balance. 7
2. Schematic of Nulling and Measurement Systems a. Lift System (2 Required) . 8 b. Drag System (2 Required) . 8
3. Photograph of Balance. . . . . . 9
4. Force Diagram of Balance and Calibration Loads 10
5. Drag Output vs Load. . . 11
6. Lift Components Output vs Load a. Component Ll. . 12 b. Component L2. . 13
7. Lift Components Lj and L2 Counts Output per Unit Load vs Station . . . . . . . . . 14
8. Lift Load Measurement Error 15
9. Moment Measurement Error . . 16
10. Interaction of Drag Load on Lift Components 17
11. Example Wind Tunnel Lift, Drag, and Pitching-Moment Measurements on Blunted Cone (Shaded Regions Lie between Extremes of Free-Molecular and Newtonian Theoriesfor Lift and Moment) . . . . . . . . . . 18
vii
A E DC·TDR·64·280
1.0 INTRODUCTION
Noticeably lacking in the reported literature on gas dynamics is a significant amount of wind tunnel data pertaining to aerodynamic forces acting on bodies in high -speed, rarefied flow fields. This is partially due to a lack of sufficiently sensitive force-measuring balances capable of obtaining data in the typically small, low-density wind tunnel of today, wherein limited model size combines with the rarefied flow to result in extremely small forces. To the authors I knowledge, little has been done to correct this deficiency beyond the development of simple, singlecomponent, drag balances.
The purpose of this report is to document the development of a three-component balance which is capable of resolving lift, drag, and pitching moment in the relatively typical low -density wind tunnel (Gas Dynamic Wind Tunnel, Hypersonic (L) )(Ref. 1), at the Arnold Engineering Development Center (AEDC), Air Force Systems Command (AFSC). Briefly, this tunnel is a continuous type, arc-heated, ejector-pumped, hypersonic system. Based on typical tunnel flow conditions, initial performance requirements established for the balance were as follows:
Maximum Load Accuracy (Ibj' or in. -lbf) (lbf or in. -Ibj)
Drag 2 x 10-2 ±2 x 10-4
Lift ±2 x 10-3 ±4 x 10-5
Moment* ±4 x 10-3 ±4 x 10-5
Additional requirements originally set forth having a governing infl uence on the balance design were:
1. The size obstruction allowable in the flow field is unusually restricted because of small tunnel size, i , e., models are roughly 0.5 in. in maximum diameter.
2. The balance must be insensitive to the near vacuum environment on the order of 10 microns Hg and to heat transfer from the relatively high -enthalpy flow fields wherein total temperatures are up to 72000R.
*For the purpose of defining pitching-moment range and accuracy, the moment center is defined as a point on the end of the model sting at its centerline.
Manuscript received December 1964.
1
AEDC-TDR-64-280
2.0 DESCRIPTION OF THE BALANCE
The balance is of the external type and is composed of one drag and two lift components, pitching moment being determined from the latter two. All components operate on the nulling principle. The mechanical arrangement of the balance can best be explained by reference to Fig. l. Note first that the model supporting sting is rigidly attached to a yoke which is supported by the two lift components through vertical flexures. In turn. each lift component is supported through the use of two sets of cross flexures and is braced in the vertical position by a horizontal stiffened flexure. Both the cross flexures and stiffened flexures are made of stainless steel shim stock of sufficient width to give lateral stability. A static weight is used to counterbalance the lift component. yoke. sting, test model. etc. In order to reduce the mass of the system.: all dynamic components were made of the lighter metals, aluminum and magnesium.
Nulling and measurement of the force acting through each lift component is accomplished automatically through a closed loop control system as shown schematically in Fig. 2a. Any error in the vertical position of a particular component is sensed by a differential transformer which transmits this information to a servoamplifier. This amplifier in turn sends a current through the driving coil of a magnet assembly which. through a series of flexures. applies the necessary force through the component to maintain a null position. Finally. the magnitude of the current is recorded on a strip chart and is a measure of the restoring force acting through the component. One obtains the output for an applied load simply by subtracting the reading before or after the load is applied from the reading obtained during load. Depending on the weight of the test model attached to the sting, it is. of course. usually necessary to position the counterweight so that the required driving current for nulling the balance will be within range of the servo.
Drag is meas ured by applying restoring forces to each side of the yoke at the points shown in Fig. 1. The drag measurement is obtained by summing these two forces. This was done as a matter of convenience and is not intended as another component in the balance. However-, it is conceivable that with minor changes in the flexures connecting the drag servos to the yoke. it would be possible to measure yawing moment.
The drag nulling and measurement systems are similar to those for lift. the main difference being in the linkages and the fact that Consolidated Electrodynamics Corp. Type 1-124 servoamplifiers were used in lieu of their type 1-126 (see Fig. 2b). Also. it may be noted that no
2
AEDC·TDR-64·280
dashpots were required for damping as was the case with the lift components.
A photograph of the balance is shown in Fig. 3. This is included primarily to show the aerodynamic shielding of the components in or near the flow field. The horizontal shield for the rear of the yoke assembly and the top of the balance housing are water cooled. Not shown in the photograph are the side plates for the balance housing.
3.0 STATIC CALIBRATION
The technique used for static calibration of the balance is illustrated by Fig. 4. Briefly, drag loading was accomplished by use of standard weights suspended over a pulley, while lift and moment loading was accomplished by suspending standard weights from the model supporting structure at eight different load stations.
Calibration of the drag component in counts output versus applied load is shown in Fig. 5. Data points obtained with the balance operating both at atmospheric pressure and near vacuum are shown. Also, several points were repeated with artificial lift load applied. No interaction of lift on drag measurements could be detected. Linearity is seen to be very satisfactory with the best straight line through the data yielding a calibration constant of 1. 488 x 10-4 lbf/count. Since none of the calibration points deviate from this straight line by more than a count, it is concluded that the accuracy of the drag component is within the bounds of the original design requirement of ±2 x 10-4 Ibj throughout the required range.
For lift components L1 and L2, the counts output versus applied lift load at each of the eight load stations (see Fig. 4) is shown in Fig. 6. Attention is called to the fact that the data include points obtained under vacuum as well as at atmospheric pressure. Also, points are included where an artificial drag load was applied continuously. * As desired, all the data for each station form a straight line for both L1 and L2. By next reading the slope of these straight lines, the output counts per unit load were plotted versus load station location as shown in Fig. 7. Again,
*The reader should not conclude from these points that there is no interaction of drag on Lj and L2. As a matter of fact there is a small scale shift which will be discussed later. These data are included to show that there is no change in sensitivity when drag load is applied.
3
AEDC·TDR·64·280
straight lines fit the plotted points. Finally, calibration constants for L1 and L2 were both determined to be 3.33 x 10- 5 lbf/count.
The calibration data were next analyzed to determine extent of error in lift and moment measurements resulting from use of this calibration factor. Applied lift load should equal, within specified tolerances, the sum of L1 and L2 measurements. Similarly, about the balance moment center, the moment introduced by the applied lift load, or the product of the applied lift load and its moment arm, should be equal to the summation of restraining moments obtained from measured L1 and L2 and respective moment arms (see Fig. 4). The measurement errors found by comparison in this manner are shown in Figs. 8 and 9 for lift and moment, respectively. Again, it is noted that the balance performs within original design requirements throughout the required load range.
As stated previously, the addition of drag loads results in a zero or scale shift in L1 and L2 output. The magnitude of this shift versus the applied drag load is shown plotted in Fig. 10. Fortunately, the shift is linear and of small magnitude.
4.0 SAMPLE WIND TUNNEL FORCE MEASUREMENTS
As a preliminary to making force measurements on lifting bodies, the interaction of drag on L1 and L2 was again calibrated, with the drag load being applied by mounting spheres of assorted sizes on the balance in the actual wind tunnel flow. Since spheres should cause zero lift and moment, this is a satisfactory method of accounting for any slight misalignment of the balance and the flow field. The result of this calibration is shown in Fig. 10. As was the case with the static calibration, the interaction is linear and of small magnitude. This latter calibration with spheres was used to correct for interactions during subsequent wind tunnel meas urements.
For the purpose of obtaining sample lift, drag, and pitching-moment data, a blunted 10-deg cone with a O. 5-in. base diameter was chosen as the test model. This particular model was chosen because drag data had been obtained for it using a single-component balance in the same flow field being used to test the present balance (see Ref. 2). The flow conditions used for the tests were as follows:
Working Gas Nitrogen Total Temperature 39600R
Total Pressure 20 psia Mach Number 9.9 Unit Reynolds Number 750/in.
4
A EDC·T DR-64.280
In making wind tunnel measurements, the procedure was to first obtain zero readings on all components with no flow, but with the tunnel under vacuum. Then flow was established and readings were obtained with flow over the test model. Finally, flow was stopped and zero readings again obtained. The entire procedure consumed less than 45 sec on the average, allowing little time for heat transfer to affect the balance mechanism.
Resulting measurements on the blunted cone are shown in Fig. 11. Note that the drag data fall within the band of scatter of the previous data obtained with the single-component balance. Unfortunately, there are no other data for equally low -density flow conditions with which to compare the lift and pitching-moment measurements. As would be expected, however, the results do lie between the values obtained from the extreme cases of free-molecular and Newtonian theories.
Since the rear of the sting shield (see Fig. 3) is exposed to a pressure not necessarily equal to the natural model base pressure, the question may arise regarding an influence on the balance measurements. Because of the hypersonic flow condition, it can be shown by the ratio of impact pressure to the pressures involved at the base that there should be a negligible influence. Moreover, the present drag measurements compare very well with the single-component balance wherein the same problem did not exist. As a further check on base pressure effects on the present balance, the pressure to which the rear of the sting shield was exposed was varied from approximately 60 to 160 microns Hg. Also, the spacing between the rear of the test model and the end of the sting shield was varied from approximately 0.05 to 0.5 in. No significant change in the balance measurements was detected in either case.
5.0 CONCLUDING REMARKS
From the s tandpornt of producing minimal obstruction in the flow field and also from the standpoint of minimizing effects of heat transfer, the choice of an external balance design has been shown to be satisfactory. A utomatic nulling and force measurement through the use of closed loop control systems on each of the three components proved to be adequate for obtaining the required sensitivity. It should be noted that the balance met the original design specifications, i , e., measurement of lift within ±4 x 10-5 lbj, drag within ±2 x 10-4 Ibf, and moment within 4 x 10-5 in. -Ibj ,
As was shown in the data presented, the operating environment did not have any adverse effects on balance performance. The bench calibration is precisely the same as that under tunnel conditions. During
5
AEDC·TDR-64·280
the usual test interval, approximately 45 sec, there is no significant zero shift attributable to varying model temperature.
The balance, although designed for a specific tunnel, could be easily modified to operate in tunnels of different configurations. The load range of the balance is also quite flexible and can be increased or possibly decreased by use of different driving coils. It is also feasible to modify the balance so as to provide a five-component balance--lift, drag, side force, pitching moment, and yawing moment. This could be accomplished by modifying the two drag components to read yawing moment as well as drag force and adding a component to sense side force.
REFERENCES
1. Potter, J. L., Kinslow, M., Arney, G. D., Jr., and Bailey, A. B. "Initial Results from a Low-Density Hypervelocity Wind Tunnel. II Hypersonic Flow Research, Ed. by F. R. Riddell, Academic Press, 1962, pp. 599-624.
2. Boylan, David E. and Sims, William H. "Experimental Determination of Aerodynamic Drag on a Blunted 10-deg Cone at Angles of Attack in Hypersonic, Rarefied Flow." AEDC-TDR-64-60 (AD 435860), April 1964.
6
/ Drag Flexure No. 01
~ t"=- To Nulling and Measurement ~::i:i======H.. . 0z~ System (Drag Flexure No.
Similar)
Shielding Not Show
Test Model
String (Typ. )
-J
Cross Flexure
Note: Components not shown of LZ unit similar to L1.
Fig. 1 Mechanical Arrangement af Balance
~ Flexure to L1 Nulling and Measurement System (LZ Similar)
11025751
Oashpot }>
m o n·-i o Al· 0.... · IV en o
A EDC- T DR-64·280
Carrier Voltage
Servoamplifier (Consolidated Electrodynamics Corporation Type 1-1261Dashpot
Ctr. Weight Driving Current
t=:==r~~~;=r--Drivi ng Coi I Magnet
a. Lift System (2 Required)
Carrier Voltage
~ :::x:====:>ef1 Servoampl ifier (Consolidated Electrodynamics Corporation
Magnet-4~h1 Type 1-124) Signal Error
Drivi ng Coil Differential
Current Shunt Transformer
Driving Current
To Strip Chart Recorder
Differential Transformer Error Si nal
b. Drag System (2 Required)
Fig.2 Schematic of Nulling and Measurement ~'~Iems
AEDC·TDR·64·28D
Phofograph of Balance
9
AEDC·TDR·64·280
Moment Arm Schedule
Station No. X, in.
1 2 3 4 5 6 7 8
1. 74 1. 24 0.74 0.24
-0.26 -0.76 -1.26 -1. 76
Lift and Moment Moment Center Calibration Load
Restoring Forces Model Supporting Structure _~ (Dl + O2)--/-, •
I
.. I. 1,0-...-.tDrag Calibration Load L I. D
Restoring Force L1
All Dimensions in Inches
Restoring~
Fig.4 Force Diagram of Balance and Calibration Loads
10
AEDC·TDR-64·280
180 ---------------------, Flagged Symbols Indicate Poi nts Repeated under Vacuum
160
140
120 /0 VI
-I-' 0c ::J 100 0
U -I-'
::J c..
-I-' 80::J 0
60
I o
40
20
o o 5 10 15 20
Drag Load, Ibf x 103
Fig.5 Drag Output vs Load
11
100
E 40 ~
8-::::J Q.
'S 20 o ~
t-:l
o
-20
-40 '
Solid Symbols Taken with 0.02 Ib Drag Load Flagged Symbols Taken under Vacuum
SOl ./'" I
oo~ ~ »> I
, , , ,
o 5.0 10.0 15.0
Lift Load, lilt x 104
a. Component L 1
Fig.6 Lift Components Output vs Load
Sta. m>
1 n0 . -t 0 ;u. 0~ ,....,2 a> 0
3
4
5
6
7
8
I
20.0 1-10-258-0I
\.
Sta.100 I I
80
60
VI- 40 1
c: ::::l 0
U-::::l c.. I ..... -::::l 20 CN 0
I
-20
»I I I I I I m
-40 0 o5.0 10.0 15.0 20.0 n
Lift Load, Ibf x 104 [ii2581J o
;0, ob. Component L 2 .,. ,...., ex>Fig.6 Concluded o
6
7
8Solid Symbols Taken with 0.02 Ib Drag Load Flagged Symbols Taken under Vacuum
~ ~ ~ ..........,Q 5
~~ ~ ~ ~ 4
~~~ ~ \I IA
3
°rwz& r;f
:= : II
I1 : i 0
--i
AE DC·T DR·64·280
5 x 104 r--------------------------,
4 x 104
3 x 104
.s:'"c: -::J
8 Ix 104
Ol------+----------~i-------I
-1 x 104 I I L2 L1
-2x 104 8 7 6 5 4 3 2 1
Station
Fig.7 Lift Components L 1 and L 2 Counts Output per Unit Load vs Station
14
11025821
Symbol
o 'V o o Ll a o C)
4
3
2
I a 4
I 0
0 0
f-' ~ CJ1
)( .:l
Station
1 2 3 4 Flagged Symbols Taken under Vacuum5 Solid Symbols Taken with 0.02 Ib Drag Load6 7 8
! <:>
- /
<:>
o Limits of Permissible Error
6 • 0 a J 0 o
•.10
%e
011 a 0 e .:l
:E 0 ~ ,00 "V / ff -<I
"V 0 0! -1 ~ <:><:> / "V • 0
-2 <:>
0 "V
<:> "V
J> m o::l
"V
n-i·o0 5 10 IS 20 25 ;0
Applied Load, Ibf x 104 o-l>.~2~ · · CD '" Fig.8 Lift Load Measurement Error o
--l
Flagged Symbols Taken in Vacuum Solid Symbols Taken with 0.02 Ib Drag load
4
~ c••3 1.-....-limits of Permissible Error I
o c
0 I o2 1
0
• •00 OOI~.1· Q
0"b • !-x E
I 0C...... v W c
en QL. V v W- ~
0 L........, -1 c c Symbol Station
0 0 I 0 1c "V 2
-2
<> 4 ·0 I
c 0 3
LI 5 -3 a 6
0 7 0 8
-4
o 0
v
» m o n,
o AI. 0..... a:>'" o
-40 -30 -20 -ill 0 ill 20 30 40 Applied Moment, in. -Ibr x 104 11025841
Fig. 9 Moment Measurement Error
---
AEDC-TDR-64-280
Symbol
o Component L1
~ Component L2
2
N .....J ~/"0
1c -ro
~ .....J ~ -o-e-C / _e_
0
C _e~O 0 0 u--- ~ ro A L •a.> C
---c V'l
-1 .... Note: Closed symbols obtai ned by static ::J 0 loadi ng. Open symbols obtai ned by
U drag on spheres in tunnel flow Held.
I-2 1 2
Drag Load, Ibf x 102
Fig. 10 Interaction of Drag Load on Lift Components
17
11[ 18
~ 10 160 ....-l
X 14.is"
6..-- 12:tr:::f, 10
:tc L..
sI- 10
4 8
3 oq- 6 0 ....-loq-
2 40 x ....-l
X .s: ~ .....
I .I -:=i
2
..... c 0 0 co .
.....c -II-- -2 Q,)
E 0 -21- -4:E:
-6-3 t -4 -8
-5 I-- -10
-6L -12
~_ ..............._.....
........0. .......... ......
-40 -30
» m(Band of Scatter in Drag Data from Single- _ o
Component Balance _-~-......... ()
.;- ~..... · ~__ __....- Ll_ o -- --,,- -- :;0
--._~---- ~ 0~---P:. ....· ,. ----"----_ .....6_--- · IV 00
Symbol o
l:! Drag o Lift o Pitching
Momen
-20 -10 0 10 20 30 40 Angle of Attack, deg 11025861
Fig. 11 Example Wind Tunnel Lift, Drag, and Pitching-Moment Measurements on Blunted Cone (Shaded Regions Lie between Extremes of Free-Molecular and Newtonian Theories for Lift and Moment)