You are on page 1of 12

*'

b

A SKIN FRICTION BALANCE

-

*' b A SKIN FRICTION BALANCE - Eugene S. McVey' James W. Moore' Prof e s

Eugene S.

McVey'

James W.

Moore'

Prof essor

Associate Professor

Non-Member

ASE

Member AHE

INTRODUCTION

,

The measurement bodies requires that

friction meter may

conditions.

be subjected to a rapidly changing temperature environment which varies from a

few degrees Rankine

friction meter is expected to operate in a vehicle traveling at relatively high

Mach numbers and

the vehicle skin is usually at a premium for measurement instrumentation so it

of

in-flight

skin

friction drag on hypersonic

subjected

field,

a

aerodynamic

measurement

instruments be

to extreme environmental

In addition to a

severe vibration

skin

(due to liquid hydrogen

lines) to 2000°F or more.

100,000 feet.

A skin

at altitudes in excess of

Mounting space near

is desired to keep package dimensions near the skin to two inches in thickness

or less.

available from the literature to guide the design

Little

information

is

of

a

skin friction meter

for high temperature environments,

although Kistler

Instrument Incorporated' claim operation to 1500'F for a skin friction meter

they have developed. This performance is obtained by enclosing the meter in

a cooling jacket to overcome the high temperature problem.

information indicates that Kistler uses a closed loop null balance system to measure drag un a small section of skin which is free to move parallel to the

Available

skin

surface.

Fenter

and Lyon?

have reported

on

a

skin friction meter In a comprehensive report

developed to make measurements on Arobee-Hi Rockets.

they describe an open loop system which uses a flexure to constrain a test segment of the missile skin. A linear variable dirrerential transformer (LVDT) is used

"Both

with the School of

Engineering

and Applied

Science,

University

of

Virginia, Charlottesville, Virginia.

was supgorted by the Force Measurements Section of the Instrument Research

The research

on which

this paper

is based

Division of NASA, Langley,

Virginia,

3 N4SA Grant NGR 47-005-026.

(NASA cn OR +NIX OR ~d NL'MBERI

<CATEGORY,

I

I

2

to mea& e flexu

is proportional to friction drag.

3 displacement

and provide

a read-out

signal

since displacement

system is described in this paper which operates satisfactorily

room temperature and is believed to be capable of operating to 2000°F with the

A basic

at

substitution of

a few critical high

temperature

have been taken to 900°F

on Components developed

components.

for high

Preliminary data temperature operation,

and continuous

operation of

these components in a 2000°F

environment

is anticipated

by substituting special materials in the basic construction.

motor are located in the hot envirodment while the amplifier segment of the control loop is placed in a remote, cool location. Basic theory and low temperature experimental results Hill be presented here even though a high temperature capability has been the main concern in the general approach to

the problem and in the design because this information should be of value to

and

The sensor

others working in the area in view of the fact that

information

be reported when the data are complete.

little or no related

is available in the literature,

High temperature performance will

SYSTEbl OPERATION AND ANALYSIS

A closed loop system configuration was

which works on a null

balance principle. The system is shown schematically in Fig. 1 and its block

diagram is contained in Fig. 2.

flexure is displaced. The displacement is sensed by a transducer (e.g. an LVDT for translational motion) ivhich has an output voltage proportional to displacement

The sensor output is amplified and used to drive a motor which produces a force

in opposition to the skin friction force.

When a skin friction

force,

is present the

FX

Motor

armature current

is measured

and recorded because it is proportional to motor force.

With proper design

the measurements can be made essentially independent of both the motor flexure

constant and changes in system gain due to other parameter changes as will be demonstrated .

The linear differential inertia is

dF

X

+ CFm -

Kfe

= J8

equation for the flexure and

its associated

where:

X is skin friction drag;

F

K

J

d

f is the flexure constant;

inertia

is the rotational is the distance

of

all moving parts;

from the tangent

to the flexure;

C is the distance from the servo motor to the

flexure;

(1)

3

8 is flexure displacement

angle.

K is the motor

m

It follows that

gain constant

The motor

force is

Fm=K i

ma

where Km is the motor

force constant.

ea =

iaRa

i ma where Km is the motor force constant. ea = iaRa + Kv4e The motor

+ Kv4e

The motor

armature voltage

is

(2)

(3)

for

are very

,0005 radian

small

deflection

good here,

or less.

angles and

For example, the total motor deflection is expected to be

negligible armature inductance.

Both

assumptions

The distance

from the flexure to the motor

is 4,

and the motor

back

emf

constant

is Kv,

and

B

a

is the armature circuit resistance.

From Eq.

(4)

ea(s) - 4Kvs8(s)

i (SI =

a

Ra

(5)

The armature voltage will be the output

of

constant Ks

of

the

type transducer with

sensor multiplied

must

by the gain sensitivity introduce a Thus

the amplifier,

phase

Ka.

For

a translational

output

is

4,K

0.

S

the transducer

shift

so that

The amplifier

180'

the system will have negative feedback.

motor

e

a

=

- Ks48Ka

From Fig.

2 the closed

armature current

-

loop transfer

(6)

function in terms of drag force and

ity)

is

JIZ,

JRa

To demonstrate the relationship

JRa

between

the drag force and the motor

4

current in the steady state, assume that a step of force Fx/s

system. In the steady state

by the final value theorem.

If

Eq.

(8) may be written

i

Rom Eq.

-

a,ss

-

Fx

Km

d

‘-

-

4

(3) the value of Kmir

d

Fm = -F -

x.c

may be inserted to show that

is

applied

to the

(10)

(11)

Thus, by making the amplifier gain constant Ka sufficiently large the motor

force nulls out the applied force and the motor current is a direct indication

of drag force.

From Eq. (8) and Eq.

(9) it

is apparent that

changes in

system gain

and changes in the flexure constant Ks do not affect system accuracy if the

inequality of Eq. (9) is satisfied.

Thus, it

is desirable to make the amplifier

gain, Ka, relatively large consistent with system stability.

results have demonstrated that K

relationship of Eq. (9) valid if stability compensation is added.

Experimental

a can be made sufficiently

large to make the

The greatest

sources

of

error at high temperatures

is

anticipated

to be due to variation of the motor

zero position.

assembly have been designed

type construction has been used to avoid temperature dependent magnetic materi 91s

force constant K

m and the position

sensor

The high

temperature model motor,

position transducer and flexure

effects.

Air

core

to minimize errors due to thermal

5

and to make Kmand Ks independent of temperature except for dimensional stability

problems.

temperature instead of a permanent magnet field as in the low temperature model.

This means

that

an electromagneti-c motor

field must be used

at high

In

order to assure operation

of the

system in the amhient vibration

and

acceleration field as well as operation over a wide temperdture range, it is necessary to pay particular attention to the form of pivots used for the moving parts. The pivots must hold the parts completely rigid against all motion, except for rotation about the measurement axis. To satisfy the requirements of the control system,backlash, stiction and hysteresis must be zero or very small. It would be preferable if the pivot exerted no rotational forces of its own, such as the spring constant Kf Eq. (1) ff, used in deriving the control equations.

The preceding requirements

indicated

that

the best

pivot would

probably be either an air bearing or a flexural pivot bearing.

chosen for initial study because it is self-contained and because of its high rigidity about the non-rotational axes. To meet temperature requirements, it is necessary to use materials which have high strength at 2000' F and which are not

brittle at extremely low temperatures. Haynes Alloy 25 as the flexure members.

The latter was

Initial studies are being made using

The extremely

small forces to be measured require that

the

spring

constant of the flexure be very small relative to the control loop gain as

shown by Eq.

in Fig. 3a, By adjustment of the tension in the flexures the rotational spring

constant can be made positive, negative or zero.

and rotational

(91,

Therefore a three member flexure will be used

as diagrammed

The environmental

force field

includes both

translational

vibrations and translational and rotational accelerations, In order for the control system motor to not need to overcome the effects of this field, the moving members

are tounterbalanced RS shown in Fig. 3b. counterbalanced about their own axis.

flexure system so that they tend to rotate in opposite directions when an

angular acceleration is applied to their common base. and rotational inertia effects are cancelled.

The two rotating members

are each

by

a

They are then

coupled together

Thuqboth

translational

.

6

EXPERIMENTAL RESULTS

An experimental model was built and tested at room temperatures. A

Collins Manufacturing Co. LVDT 55-203with a sensitivity of 1 volt per mil was used to measure flexure displacement. A Bendix, Inc. flexure Model No.

6016-800 was used (the experimental system was not counterbalanced) as the

pivot.

of Alnico V mounted on the sting for

wire wound in a coil 1.5 inches long to form the armature of a solenoid

motor.

The motor consisted of a 0,125 inch diameter, 1.0 jnch long piece

An

operational

the field and 2000 turns of No.

amplifier

followed by a unity

gain power

30

amplifier was

used to obtain amplification and motor drive,

The

system parameters are:

Ka

= 100

v/v

Kv

=

negligible

Km =

43

lb/amp

K = 0.5 v/mil

S

Kf

= 0.8

in.

lb,/rad

Ba =

99

J

= 5.8

ohms

X

aft lb ses

10

.c

=

3.3 mJ

Uncompensated open loop frequency response data are shown in Fig. 4. As can be seen from the plot, the open loop gain constant is 225, a value

which is high enough to allow relatively large variations in eny system parameter

except Km without affecting calibration more than a fraction of a per cent.

example, the flexure constant varied 5 per cent when heated to 500' F. change had negligible effect on the force reading.

This

For

necessary to use lag compensation

such a manner that the resonant peak frequency is nQll above the crossover

The mechanical resonance makes it

in

frequency.

not import ant.

LGg

compensation was

used is this case because

speed of

response was

A linearity curve is

shown in Fig.

5.

It is apparent that the system

is linear to within 1 per cent,

1 per cent at room temperature,

The repeatability was

also found to be within

REFERENCES

1.

Kistler,

Group,

Inc.,

"Sin

Instrument

Friction Drag

Research Division,

Sensor,"

Files of

Force Measurements

Virginia,

NASA, Langley Field,

2.

Fenter

and Lyons,

"The Direct Measurement of Local

in Free

Skin Friction

Flight ," DRL-391, CM-877

Texas,

1957.

on

Aerobee-Hi

Laboratory,

Rockets

Defense Research

University

of Texas,

3. Seemuller, W.

W.,

"The Design of

a Skin Friction Meter

Environmental Conditions,

IS Thesis,

University

of

for Use in Extreme

Virginia,

1966.

4.

Moore,

J.

W.

and E.

S. McVey,

"Investigation

of

Systems and Techniques

for Multicomponent Microforce Measurements on Wind Tunnel Models,"

Semiannual Status Report EME-4029-102-66U,

March,

1966.

Air

Flow

t

/

L- Ramjet

-

Motor

-

\\\\\\\\\\\\'A\

Skin

1

Flexure1

\

LFloating

Element

( target 1

Displacement

Transducer

Control

EIements

Fig. 2

Moving

Mernber

Flexur

Pivot

\ b

c. w.

i

>

C.W.?

C.W.

A

Flexures

.FIex ural

Pivot

Fig. 30

Fig. 3b

,.

lo3

IO2

IO

b

-

225

0 I

IO-'

I

I

5.4

FREQUENCY -

IO

Hr

IO

6

5

4

3

APPLIED

FORCE -

GRAMS