waterfall rains

8/16/2019 Waterfall Rains

1/18

A0 A061

901

Pt*DUE

UNIV LAFAttflt IPW PRO

~

fCT

SQUID

IEAOQUARIflS

F/S 20/I

A WAU.—FLOflDIRECTION PROeE FOR (95€

IN

S€ PARATING NC REATTACNI—fl c (U)

NOV

78 J K

EATON.

A

H

JEANS

.

.J ASI4JAEE

N000II—7S—C—1t13

UNCLASSIFIED SQUID

SU

3aU

_

~

__

_

u

_ fl

~~~~~~~

nyC

_


2/18

i r

~

L.

I .

I J L

L.

~~~

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I

~~~~~

~

‘

~

I I

[ 2 5

L4

~

iIi~


3/18

“

T

~~

T

1II

i

~~

LEVEV

PROJECT

SQUID

TECHNICAL REPORT S U 3 P U

A

WALL FLOW DIRECTION PROBE FOR

USE

I

N

SEPARATI

N

G AND

R

EATTACHING

F L O W S

BY

~

‘

JOHN

K. EATON ,

ALBERT

H.

JEANS

,

JALAL ASHJAEE and JAMES P. JOHNSTON

LLJ

STANFORD

UNIVERSITY

STANFORD

,

CALIFORN

IA

94305

PROJECT SQUIDHEADQUARTERS

D 0 C

r

~

r?1

~

r ?n

~

rIEr

W E S T

L A F A Y E T T E

,

INDIANA 47907

U

DEC

7

1918

NOVEMBER

1978

~

l

~

t.bU

U

L

~

Project

SQUID

is

a cooperative program

of

basic research relating

to Jet

Propul

sion. It is

sponsored by the Office of Naval Research and

is

administered

by

Purdue

University

through

Contract NOOO14

~

75.C-1

143,NR-098-038 .

T his

document

has been approved for public release and

sale;

its

distribution

is

unlimited.

78

12

04 1 .1 0 8


4/18


5/18

A WALL-FLOW-DIRECTION

PROBE

FOR

USE

IN

SEPARAT I

N

G

AND

REATTACHING FLOWS

by

John

K. Eat

on

*

,

Al bert H.

Jeuns

*

,

Jalal Ashjaee

*

and

James P. Johnston

ABS

T

RA

C

T

The measuranent

of

the

exceedingly

unsteady

behavior of

nominally

two-dimensional , turbulent

flow

s

in regions

of separation and

reattacl’inent

is facilitated by the development

of

a

new instrument;

the wall— flow-direction probe

which

determines

the

instantaneous flow

direction

(upstream

or

downstream

) n

a

thin

layer

of

fluid

very close

to

the wall. The

probe

was tested in

low-speed

,

unsteady,

separating

and reattaching air

flows

. It

appears

to offer considerably more

accuracy

than

other methods for the

determination of

time mean separa-

tion and

reattachment

points.

In

addition

, the probe

and its

control

circuits

are

relatively i nexpensive and easy to

construct

.

ESSION

fo r

______

NTIS

\ h I ’

:

~

r

~

Infl

DOC

Li O n

0

UNANN OI

~~

’

~

0

JUSIfI

CA ’

~

BY

—

~

STRIBU

~

H

~

VAt

~

BIUF

tOOtS

~~~

id

or

SPECIAL

-I

~~~

*

Research Assistants

,

Dept.

of

Mechanical Engineering

,

Stanford Univers i ty

.

**

Professor of Mechanical Engineering

,

Stanford

Univers

ity

.


6/18

1

.

INTROD

UC

T

ION

As part of

a

continu i ng program of

research

on

internal flows we

are

currently studying nominally two-dimensiona l

separating and reat-

taching turbulent flows which contain regions where

instantaneou s

flow

reversal frequently

occurs .

The

most notable example of such behavior

occurs

in the transitory stall

regime

of two-dimensiona

l diffuser

flows [1]. In cases

with

unsteady flow reversals

,

the

use

of most

conventional

instrumentation

is

precluded .

A

notable

exception is the

laser

doppler velocimeter

with frequency shifting ,

an expensive and

difficult

technique.

To

facilitate

our research

,

a

simple and

inexpensive

wall-flow -

direction

probe

has

been

developed

which

may be

used

to

determine

the

instantaneous flow direction

(upstream

or downstream

) n a thin

layer

of fluid very near a wall.

Although

the basic idea for this probe is

not new

, we

feel

tha

t other

investigators of

unsteady,

reversing flows

may

find the probe

’

s design

very

useful especially

in

light

of

its

low

cost

,

its

simplicity and

its

relative

ease

of

application in low-speed

air flow experiments

.

2.

BASIC DESIGN

The probe (Figure 1)

consists

of three parallel wires mounted

on

short prongs

which

penetrate

the wall and

are

perpendicular to the flow

.

The larger

,

center

wire

is heated

with a

D.C.

current

of

about

1.5 amperes

to create a

hea

t

ed

wake

in

the layer

of

air close to the wall. The two

sensor w i res

,

operated

as

resistance thermometers

,

detect

the

wake

,

and

through a

simple electronic

circuit

give the instantaneou s sign of

the

fluid

velocity In the layer

of

air which

is

within 1 nm

of

the

wall.

J

~


7/18

3. DESIGN

DETAILS

The three

wires

are

mounted

on six copper

plated

sewing needles

(.3

nm diameter),

which

are

rigidl y mounted in

a plexiglas

cylinder.

The

cylinder

fits inside of the outer shell which

itself

fits

into the

wind tunnel Instrument

ports.

Once

installed ,

the surface

of

the probe

appears to be an integ ra l part

of

the test

surface

to

±0.03 mm

.

Th e

needles protrude

throug h clearance holes

in

the

face

of the outer shell.

With this arrangement

,

the needles may be extended about 5 nm from the

probe surface to facilitate

wire

mounting

.

After

the

wires are

mounted

the

needles are

retracted

so that the wires are about

1 mm above

the

surface.

An earlier

version

of

this probe used

a

center

wire

mounted

flush

to

the

surface. Data sets using both

probes

show tha t the two

configurations give identical results to

wi

thin

the

experimenta l uncer-

tainty of

either type

of

probe.

The

sensor

wires

are 5 micron

platinum

plated tungsten and

are

soldered

to

the ends

of

the needles.

The heater

wire

is

125 micron

mone l

,

and it

is soldered

to

the

side of

the

central pair of

needles

just below

their tips.

The solder

joints on the

heater

wire

are coated

with epoxy to prevent oxidation

and

to provide

mechanical

strength.

The wires

are

connected throug h the

needles

to a five

pin connector

mounted in the outer shell. Only

five pins

are needed since

the

sensor

wires

share a

common

ground.

A block diagram

of

the

electronic

circuitry

is

shown

in

Figure

2.

The two

sensor wires

are

operated in

a bridge which

is

balanced

when

both

wi

res are

at

the same temperature

.

If

the temperature

of

sensor

wire 1 is increased ,

its

resistance

w i l l

Increase

and

a

positive voltage

-2-


8/18

w i l l

appear

at amplifier A l .

This voltage

is

amplified and

then applied

to

a

comparator (C) which controls switch S The switch

gives an

output of 5 volts or zero volts

depending

on the

flow

direction

.

The

final

amp

lifier

(A2) is used to accurately adjust the

outpu

t

zero

.

The circuit

diagram

is shown in Figure

3.

Power supplies are

needed

to supply

~

15 volts

and

V

ref

which may be any

voltage

from 0

to 12 volts. We chose 5

volts

in order

to facilitate

computer inter-

facing. An additional

power

supply

is needed

to supply

current

to

the

center wire

.

Excluding the power supplies , the

total cost of components

for this

control

circuitry did

not

exceed $40.00.

Before using

the

system , the

probe

should

be

placed in

the

flow and

the i nput offset

adjusted with the heater turned

off. If

the

offset

is

adjusted correctly,

the

output

w i l l dither

between

0 volts and

V

ref

This

adjustment

is

qu i te

stabl e

but

should

be

checked

every

hour or

so

in the

course of

a

run.

4.

APPLICAT

I

ONS

The

wall-flow-d

i

rection

probe

can

be used

for several different

purposes. The

fraction of

the time

that

the local flow

near the

wall is

forward

or backward

can be measured

by time-averaging

the probe

’

s outpu t

using an i ntegrati ng voltmeter

.

When

the

fraction

of

time

that the

flow

is

in

the downstream

direction

henceforth called percent downstream

flow

) Is measured for

a number

of streamwise locations

in a

nominally

two-dimensional separating or reattaching flow

,

the

location of

the

mean

*

detachment or reattacisuent

point can

be determined

.

In addition

to

this

*

We

have defined

the

mean

detachment

or reattachment point

to be the

point

where the

flow very close

to

the wall

is

reversed 50 percent

of

the

time.

This point does

not necessarily

coincide

with the

point of zero mean

shear st

ress

,

although

the two

points

are probably quite

close

together.


9/18

~~

••

measurement,

spatial

and

temporal

correlations of

the

wall

flow

direction can

be made

usin g

two or more probes

at

the same time

.

Finally,

the wall-flow-d i

rection

probe can

be used

in conjunction with

hot-wire

anem

ometers

and

other

instruments for the purpose of condi-

tional sampl i ng of the anemometer

output.

The probe

has

been

tested in both

a

separating and

a

reattaching

flow.

In

both of these flows

,

the measured value of

percent

downstream

flow was a weak function

of

the

heater current for currents of less than

1 .5

amperes.

Higher

currents

lead to

shorter

heater wire lifetimes

so

a

current

of 1.5

amperes was used for both cases

presented here.

The probe

was first used

in

a

reattaching

flow behind

a

backward

fac i ng step 5.08

cm

h i g

’ ’

with free

stream

air

speed

s

of from 5

to

20

rn/sec. A

typical curve of percent downstream

flow plotted against

streamwise

location

(Figure

4)

shows

the same genera l

features prev i-

ously observ ed

using

flow visualization

,

(e.g.,

Kim

et al. [2]). These

features

include (i)

an

unsteady

reattacIi

~

e

nt

’egion

approximately

3-4

step

heights

l ong , (ii) a

region

of fairl

y steady backilow located close

to

x

=

4

step

heights

,

and ( i i i )

a

resepa ration point located

about 1

step

height downstream of

the

step

.

The

uncertainty

in each

value of percent downstream flow

is

esti-

mated to be

±1

at 20:1 odds.

The

uncertainty

in the location of the

mean reattachment

point

is

then

less than

‘

.1

step

height (Kline and

Mc

Clintock

[3])

which

is

an

order

of

magnitude

better

than

can be

*

obtained wi th most known methods

of

flow visualization

*

The only exception

is

the method of Rothe [4]

which

requires

tedious frame-by-frame

examination of motion picture

film.

-4-


10/18

The probe

has

also been used in

a

stra i

ght-walled diffuser operating

in

the

transitory stall

regime.

The

throat

width

of the diffuser

was

7.62

cm

and the total

included angle

between the diverg i

ng walls was

12

degrees

.

In t h i s

case

,

the

i n l e t a i r

speed

was 50 ni/sec

.

The percent downstream flow is plotted against streamwise location

in

Figure 5.

Here

the probe

showed

a s m a l l amount

of

asymmetry

. The

measured

percent downstream flow was

s l i g h t l y

different

when the probe

was rotated 180 degrees thus reversing

the

positions of the sensor wires

.

The

average

of the two measurements was used to determine the mean

detaclmnent

point.

5. CONCLUSION

The

wall-flow-d

i rection probe is prov i ng to be

a

useful

and

easily

applied tool

for studying unsteady

revers ng

flows

It

enables one

to

accurately measure

the

location

of two-dimensional

detachment and reat-

tachment points in a consistent

manner

.

Adoption

of the

method by

future

researchers could enhance our ability to

compare

results

of

different

studies.

For

example

,

current

uncertainties

on distance to reattac

~

wnent

for separated flow

behind

steps

and

fences

are

roughly plus

or

minu s one

step height (H).

Use

of the wall-flow-direction probe can

reduce this

uncertainty to plus or minus

0.1 H

6. ACKNOWLEDGEMENT

We thank Mr. Michael Eaton

who

contributed his time

to design

the

electronic circuitry and

Mr. Robin Birch who

’

s

fine

craftsmanship lent

much to

the success of t h i s project.

Also,

we gratefully acknowl

edge

-5-


11/18

—.

~~~~~

-

—

.— ._—‘ ~~

5-.

the support of

Proj

ect SQUID

,

and the National Sc

i ence

Foundation

which

supports the first

two

authors through graduate

fellowshi ps.

REFERENCES

1 .

Fox ,

R.

W.

and

Kline,

S.

J

. (1962).

“Flow

Regime Data and

Design

Method

s

for

Curved Subsonic

Diffusers

,

” TASME J . Bas

~~~~~

in

~~

i

~

a,

Vol. 84,

Series 0,

Sept

.

l96.

~

. pp.

303-312.

2. Kim

,

J.,

K l i n e

,

S.

J., nd

Johns ton

,

J.

P. (1978),

“Investi gation

of

Separation and Reattachuent of a Turbu

l ent Shear

Layer:

Flow

over

a

Backward-Facing

Step

,

”

Report

MD-37 , Thetmiosciences

Div.,

Dept. of Mechanical Engineerin g,

Stanford University.

3.

Kl i

ne,

S. J. and Mc C lintock ,

F

.

A.

(1953),

“Describing Uncertainties

in

Single-Sample Experiments, ” Mechanical

En

~~

neer

i

n

~

, Vol. 75

, pp.

3-8.

4.

Rothe ,

P.

H.

and

Johnston

, J. P

.

(1975),

“The

Effects

of

System

Rotation

on

Separation

,

Reattachment and Performance i n Two-

Dimensional

Diffusers ,” Report

PD-17

,

Thermosciences Div ., Dept.

of

Mechanical Engineering

,

Stanford University .


12/18

l-

~~~~~~~~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

_____

125w

M

O

NEL

HEATED W IRE

FLOW 6

~~

TUN

G

STEN

_ _ _ _ _ _ _ _ _ _ _ SENSOR WIRES

. 4

1

1

THIS

PART

FITS

THIS SURFACE

I

WIND TUNNEL

FLUSH WITH

WIND-

~~~~~~~

—

PORTS .

TUNNEL WALL

.

O

UTER

~

—

SHELL

Figure

1. Wall-Flow-Di

rectIon Probe

(

di

mens i

ons

i

n mm

7


13/18

•

—

~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~

-.•

_

I

_

l l


14/18


15/18

M

O

1

J WV 3

~

.LS

NMOQ

IN]DU3d

10


16/18

•

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

•

u I u I I I I l II I I J l 1

~~

I t I l I

~~

I I I J I I J l J I I 1 I I J I I I I I J I I I I J J

C_

i)

- 0

1

+

1

_

iC

1-

I I I L I I I I I I I I I I I I I

t l i l t

I L l i l i l i 1 1 1 1 1 I l l I l t I l l i l l t

o

o

0 0 0 0

0 0 0 0 0

0

0

~~

~~~

-

CD

iC)

~~

C’)

N

—

—

MOld WVJ

~

1

SN

MOU

1

N33

~

3d

11

-

.

~~~~~~~~~~~~

•

•

~~~ ~~~~

•.


17/18

~

ICu O , 1 v

C L A S S I F I C A T I O N

OF

1

4 1

P A G E

WP,..,

Dat a

EnIs,sd)

READ

IN STRUC T ION S

REPORT DOCU

M

ENTATION

PAGE

BEFORE COMPLETING

FORM

I

~

1Po

~~

? N

UMS(

~~

SU—

3—

PU

G O V T

AC C ISI ION

P40. 1

~

CIPI(N T

’ S

C A T A L O G

NUMa(A

4 T iT L E an d

Su bilil.)

S

T Y P E O P

~

EPO

~~

T

k

P

~~

IOO

C O v ( A f D

A

WALL-FLOW-DIRECT

I

ON PROBE FOR USE

IN

Technica

l

Report

S

EPARATIN G

AND REATTAC

HING

FLOW

S

P

FO

~

MINO O

~

O.

~~

t PO PT

NUN

~~

~~

7.

A u T H O

~~

(a)

b

CON?

~~

A CT O

~

OMAN ?

NUM

UA I)

Jalal

Ashjaee

, John K. Eaton

,

Albert H

.

~

1eans

N000 l4

7 5 - C - 1 l 4 3

and James P.

Johnston

NR-098-038

I. PI

~~

FO NMING O R G A N I Z A T I O N

NA M

E

AND ADO

~~

( S

TO.

PROO

~~

AM

EL.IMEP4 T.

P

~

OJ t CT

,

T A SK

A

~~

IA

O

~~

K UNIT

NUMSE

~~

Stanford Un i versity

Stanford

,

Califo rnia 90007

II COP 4T

~~

O L L I P i G

O F F I C S N A M E A N D ADD

~~

ISS

~ ~~

PQ

~~~

O A T S

Project SQUID Headquarters

November

1978

Chaffee H a l l ,

Purdue Un

i

vers

i ty “

NuMSE

~~

0F PA0 ES

West Lafayette

,

Indiana

47907

13

F4 MO NIt O

~~

INO

A G E N C Y N A M E S

AD0

~~

ESS(((

dflf. ,.nt

from Coni,oSUn

~

OIlS,.)

II S E C U R I T Y C L A S S . .1 t A t .

r.po r )

Offi ce of Naval

Research

,

Powe

r Program ,

Code

473

Unclassif Ied

Department of the

Navy

800 No. Quincy Street

l DECL.

A SSIF i CA T I0 N DOWNO

~~

A DING

— -

SCHEDULE

Arlin gton ,

VA

22217

___________________________

1 .

O I T

~~

I

S U T IO N

S T A T E M E N T

(of this

Rsporf)

This document

has

been

approved

for public

release

and

sale;

its

distribution

is u n l i m i

ted.

li oIs T

~~

IS U T I 0 N S T A T E M E N T of A.

bstr ct

.t,t.r. d

in

85o’

k

~

O,

i dIII.ran i

ire.,

R.por()

Same

‘S

SUP PL EMEP4TA

~~

Y N O T E S

IS KEY WO

~

OS Continua

os

r .v .ra.

aid. II

n c I a ty

and

Id

nhII

~

. by

block

n,anb•r)

Probes

Separation

Re

attachment

Unsteady

Flow

20. A S S T

~~

A

C T (C.nttn u.

an

,.v .r. .

aid.

It

ns c asaafl

—

~

id .nHiy

by

block

nu

b.r)

‘

—

~

he

measurement

of the exceedi

ngl

y unsteady behavior of

n tm lna ll y two

-

dimensional

,

turbulent flows In

reg

ions of separation and

rea

‘

achment

Is

facilitated

by

the

development of a new

Instrument;

the wall-flow -direction

probe

which

determines

the Instantaneous

flow

direction upstream or d

own

s

tream

In

a

thin

l ayer of fluid very

close

to

the

wall. The probe was tested

In l

ow-

speed

,

unsteady

,

separating and reattaching air flows

.

It

appears

to

offer

considerably

more

accuracy

than

other

methods for

the

determination of the

t ime

mean separation

and

reattachment

points.

In

addition ,

the

probe

ari

d Its

~

‘

FOAM

~

DD

J A N 7

t4i.i

4

EDITION

OP

NOV SI J

O S S O L ( T I

Unclassified

5/

~~~~

S’N

01 02 .L.F .O (4 .66O1

SICUNITY

CLASSIFICATION

OP

THIS

P A G E

~

an

Dais

*RtIe.d)


18/18

•

-

.

5--.-

- - -

Unclass i

fled

S C U N I I Y

C L A S S , F R T A T I O N

O P T H IS

P A G E

(W )

~

•n

f l a t a Fnla,.d)

‘

/

con Lrol circuits

are

relatively Inexpens ive and easy to

construct.

waterfall rains

Documents