rr— - -

- ‘
A D—AO 66 371 NAVAL POSTGRADUATE SCHOOL MONTEREY CALIF F/G 20/4
THE APPLICATION OF A LASER DOPPLER VELO CIMETRY SYSTEM TO FLOW Vs—ETC (U)
D~ C 78 J A HARRISON
UN CLASSIFIED NI.

_____
I

_
_

I__ __ _

END
______ ______
 1~ ! K1~’
_~~
• NAVAL POSTGRA DUAT E SCHOOL
Monterey, California

‘ U-

THESIS
~~~ APPLICATION OF ~ A LASER DOPPLER
VELOCII€TRY SYSTEM TO FLOW VELOCITY
~~ASURE~~ NTS IN A TRA NSONIC COMPRESSOR
by

Jeffrey Alan Harrison

December 1978

Thesis Advisor: D. J. Collins

Approved for public release; dIstribution unlimited

L i IT 1-I: L _ _ _
 -

SICU~ ITV CLAUIFI CAT IOW OP 1141$ PAG E (~~ e* D~~a L,I.,.d)

p WI I kA ~~ U ? A ? I~~~II U Aa E
D I~~~ V f
I A D INSTRU CT OP4S
fl~~~
~~r~~~I~~ w ~~~~ —~~ l’ I ~~ II~~ ~ ~~~~~~ B~ ?O*E COMPLETfl’ ~
~G FOAM
1. T 3. ~ CClFI tNT $ CATALOG NUMSt ~
~~~PO~ NU ~~SE ~ 2. GOVT ACCESSION NO

(ji
T I T LE ( Suernl.) T T
~~~~
uP A(P&~~1 I
~~~~~~~~ ~

~~~~~~~~
Measurements in a Transonic Compressor ,
~~~~~~~~~~~~~~~~~~ ~~ ~~~~~~~~~~
S. CONT ~~AC YOR GIA NT NUMSI ~ (I)

eff r e y Al an !Harr 1so 7

~ ~ ______________________
S. •E~~~OSM,NQ OSQ NI ZA?ION NAME AGO £005151 IS. PSQGSAM £L&Mg N y PSOJ ICT . TASK
~ ASEA I WOSIC UNIT ~NUM SI R S

Naval Postgradua te School /

Monterey, Ca l i f o r n ia 93940
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

II. C0N ?MOL. I..ING OFFICE NAME ADOSEU ia. 5S~~SST S I

~~~

Nava l Postgradua te School ‘I _______________________

Monterey, Ca lifornia 93940 70

Il MONITOSINO AGENCY NAM E & A005E$$(IE dStI.i.. *f
t
. C.nf vollisd Offic.) IS. SEC USITY CLASS. (.~ ISIS. ‘~~ .ES)

Nava l Postgradua te School Unclassified

Monterey, Cal iforn ia 9394 0 ii
. 1 I C A TION/D O W14G5 A OING
~~~~k~~~ ~~

iS. OISTSIIU ?ION STATEMENT (• ha. R.e.)

i
~

A roved for public release; distribution unlimited

I?. OlSi SISUflON 5TA 1 EMCN1 (oi Sb. AMi e5 .,S.,.d Ia ISk
. 20 SI dU.
fS,
. f
t .
.
i
J IS. $U
~~~
LEM(NTANY NOTES

Thesis research performed in connection with Nava l Air Sys tems

Comma n d , Code 310A , Dr. H . J . Mueller

1*. NEY WOSOS (cAM . j j4 ~~~~~~~~~~ S~~ MSS~~ Ip &S. a..&Ur ~

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

Laser Doppler Velocirne try

Transonic Compressor
Ve loc ity Measuremen ts

20. 1~~ S1i$AC ? (C..dS.u . _ ,UrUV • ~~~~~ St R 1U•U T d S~~ ult4~’ S, AMSI

F
A laser doppler velocimeter (LDV ) system was developed a rid
utilized to measure the flow velocities in the axial and tangen-
tia l directions behind the rotors of a single stage transonic
compressor . The bac kacatter mode was used to acquire the sca ttez~~
light signals from particles that were seeded into the airflow
upstream of the compressor . The light signals were genera ted as
the particles tra versed a fringe pattern produced by the inter-
section of two laser beams .
j
~ ~~54 7$ $/N O iO 2-01 4 ~ S40 * I
~~~~~~~~ ~~~~~~
sccust? ~ CLAI~~~ ICATIO•~ OF TNS$ PAGE 0 . SINUrId) .
~~~~~

-
____

_ _ -- ~~
~

fic t)qus? ’CLASSSPICA?IOM OP ,U I $ PIO(V~~~~~ IS..

~~~~~~~

I~ V system proved tha t it was possible to make velocity

measurements in the transonic compressor . The results revea led
a varia tion of the flow velocity in both the axia l and tangential
directions due . to the forma tion of a wake caused by the pa ssa ge
of rotor blades , The velocity measurements were comparable to
those acquired through pressure probe measurements .

.‘
~:~~ :s

~~~~~ ~~~~~

~on~3
DD 1473
S/
~1‘t ~%~— fl 14—6$O1 2 ~~~~~~~~~ CLASSI PICA VI ON 5F
TillS P *OES ~~~.lI DIN. 1aIINIdJ

.: . __ .. . — —._-.
~~~~~~~~~ .*_ .- .— - ~~~ - .- -~~~~-.-.— — .-. -- -
. .-
— —.—

~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~ _________________ . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ —, . .
 ~~~~~
- -— -
~~~~~
, —-
———— . .-, .--,- ---- -
-
~
.
-
:
~~~
- -
~~~~~~~~ ~~~
-—-—
~~~~~~~~~~~~
.
— - - —— -
.

Approved for public release; distribution unlimited

The App lication of a Laser Doppler

Veloc imetry Sy stem to Flow Veloc ity
Measurements in a Transonic Compressor

by

Jeffrey A lan Harrison

L ieutenant , Un ite d Sta te s Navy
B.S ., Nort h Caro lina Sta te Un ivers ity , 1972

Submitted in partial fulfillment of the

re qu irements for the de gree of

MASTER OF SCIENCE IN AERONAUTICAL ENGINE ERING

from the

NAVAL POSTGRA DUATE SCHOOL

December 1978

Author
~! ~~~~~~~~~~~~~~~~~~

~~~‘
Approved by: ~~~~~~~~~~~
,
1
) Thesis Advisor

-f :
Chairma n, Depa r~~.ent of Aeronautics
1 .

bean of Sc i rice and engineering

1
_ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _
 1~— - - - - - -~ - •. .-- . -—- — -- —
- .- ,_ _
_ .;_____ .,.__._ - _ _
--
_
~~~~~~ ~ ~~

ABSTRACT

A laser doppler velocimeter (LDV) system was developed and

utilized to measure the flow velocities in the axia l and tan-
gential directions behind the rotors of a single stage transorilc
compressor . The ba ckseatter mode was used to acquire the
scattered light signals from particle s tha t were seeded into
the airflow upstream of the compressor . The ligh t s ignals were
genera ted as the particles traversed a fringe pattern produced
by the intersection of two laser beams .
The LDV system proved tha t it was possible to make velocity
measurements in the tra risonic compressor . The results revealed
a varia tion of the flow velocity in both the axial and tangential

L. directions due to the f o r m a t i o n of a wake caused by the passage

of rotor blades. The velocity measurements were compa rable to
those acquired through pressure probe measurements .

4
 —~ _ ..
- ~~~~

TABLE OF CONTENTS

I. INTRODUCTION 10
II . THEORY 12
A. LASER DOPPLER VELOCIMETRY 12
B. COUNTER PROCESSOR OPERATION 1~4
III. INSTRUMENTATION 16
A. OPTICS 16
B. S~~DING APPARA TUS 20
C. LASER SUPPORT APPARATUS 20
D. DATA ACQUISITION SYSTEM DESCRIPTION 21
IV . METHOD OF APPROACH 23
A. BEAN MODE SELECTION 23
B. LASER SYSTEM ALIGNMENT PR0C~~ URE 25.
C. DATA ACQUISITION SYSTEM OPERA TION 27
V. RESULTS AND DISCUSSION 31
A. DATA RESUL TS 31
B. DATA COMPARISON 34
VI . CONCLUSIONS 35
APPENDIX A - HP-9825 Da ta Acquisition Program 57
Al. Main Program 57
A2. Subroutine “Da ta ” 61

A3. Subroutine “Plot ” 62

APPENDIX B - Development of the HP-9825 Program 63

_ :
~~ .
_ _ _ _ _ _
~~~~~~~~~
APPENDIX C— Abrasion Tests - 65
APPENDIX D — Window Re fraction Considerations 67
LIST OF REFERENCES 69
INI TIAL DISTRIBUTION LIST 70

I I

- -
 LIST OF FIGURES

1. Transonic Compressor 36
2. LDV Counter Principle 37
L 3. Laser Transducer in Alignment and Opera ting
Position 38
4. Atomizer Air Supply System 39
5. Laser A lignment Geome try 40
6. Laser System in Alignment and Operating Position --—— 41
7. Data Acquisition System 42
8. HP— 9825 Computer , Plo tter , Osc illosco pe , and
Counter Arrangement 43
9. Pacer System 44
10. Classic Doppler Burst arid Interference Pattern
11. Ax ia l Ve loc ity Versus Bin Num b er ~46
12. Tangential Velocity Versus Bin Number 147

13. To ta l Veloc ity Versus Bin Num b er 148

14. Yaw Ang le Versus Bin Number 49

L~ -- - —
_ _
—
.
_ _ _ _
- ~ ~~~~~~~~~~~~~~~~~~~~~~~~
.. - -l
~~~
L. — -. ..- - -
-

- ~~~~~~~~~

---~--~~~~~~~ --—
_
_ _
_ _
_ _
_ _
____
_ _____ _______
_ _ _ _ _ _ _
r

LIST OF TABLES

I. Experimenta l Operating Conditions 50

II. HP—9825 Da ta Acquisition and Plotting Program 52
III. Velocity Versus Bin Number - Gra ph Routine 54
IV . Data Acqu isition Control Program 56

I
8

-I _ _ _ _ _ _ _ _
____________ _____________ ____________ _
~~~~~~ -
~~~~~~~~~~~~~~~~~~~~~~~~~
 — -2~~~ ___ .. ~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_ _ _ _ -- -— ..-~~~~~~~~~ --

AC~~JOWLEDGENENT

I would like to convey my sincerest gra titude to Professor

D.J. Collins for his pa tient assistance and expert direction
- toward the completion of the experimenta l research which
culmina ted into this thesis report. Professor Col lins ’ close
supervision during the opera tion of the laser doppler ve].oci-
metry system proved to be not only helpful but also educa tional.
-

Additiona l thanks are extended to Mr . Jim Hammer and Mr.

-

Ted Dunton for their technica l expertise and ingenuity which

helped the research proceed in a smooth and timely ma nner .

II ~~~

I
I

I 1!
- -- - -s ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -a— ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -~~~~ r
n ~~~~~~~~~~~~~~~~~~~~~~~ . .
 _‘ - — -——.---
~~~~~ ~~ ~~~~~~~~~~ ~

I. INTRODUCTION

Traditiona lly, the task of determining flow velocities in

turbomachines has been reserved for meta l probe s of finite
dimensions placed physically in the vicinity of the required
flow cha racteristics. The actua l flow velocity could not be
mea sured directly but was kduced from the relationship be tween
the sta gnation arid static properties of the flow . The accura c y
of the measuremen ts was de pen den t upon t he degree o f accurac y
at which the sta gna tion and static properties could be
determined.
Even thou gh t he use o f a laser d o ppl er ve loc imeter is no t
without inaccuracy, it has become a useful and accura te tool
in the field of fluid flow velocity measurements . As discussed
in Ref. 1, the advantages of using a laser doppler system are
tha t the me thod results in no flow disturbance or pressure loss
as would be present with a me tal pressure probe , and tha t it
offers a very high accuracy without the necessity for extensive
calibration . One disadva ntage of this procedure is tha t an
optical access to the interna l volume of the turboma chine is
necessary without limiting the operation of the ma chine . It is
also required tha t light— sca ttering particles tha t are small
enough to dependably follow the streamlines of the flow be
either present in or injected into the flow .
In th is researc h effor t, a laser do ppler ve loc ime ter was
used to mea sure the flow veloc ity b e hi nd a ro ta ting transon ic

10

. _ . . —— .- - .
—.-— —
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~
_ _ _ _ _ _ _ _ _ _ _ -
_
-~~~~
_ _
--
----_
 - -

1~
—~~~~~~~~

compressor rotor section by seeding the upstrea m flow with

particles smaller than a micron in such a way tha t they would
exit the rotor section in the vicinity of a LDV measuring
volume (Fig. 1). By counting the time t hat eac h par ticle too k
to cross the measuring volume ’ s interference pa ttern , the
instanta neous velocity at a point in the flow was determined .
By measuring the axial and the tangentia l flow components , the
yaw angle of the flow was calcula ted by de termining the angle
formed by the two velocity components as compared to the axia l
direction . Any pitch angle measurements were unobtainable due
to equipment limi ta tions .
The laser doppler velocimeter proved to be an effective
instrument which produced results tha t compa red with the da ta
tha t had previously been obtained by other means .

~l ~~~~

- — - - -
-
L - - ~~~~~~~~~~~~~~~~~~
.:
~~
 —
~~~~ ~~~~~~~~~~~~~~~~~~~~~
-“-——
~~~
—
~~~~~ —= ~~~~~~~~~~~~~~~ —- -
~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — - ~~~~~~~~~~~~~~~

• II. THEORY

A. LASER DOPPLER VELOCIMETRY

Laser doppler velocimet r y derives its basis of opera tion
from the doppler frequency shift of the scattered light caused
-
‘
when laser light comes in contact with a particle in the flow
with some velocity component. The doppler frequency shift is
measured by an electronic counter directly.
The laser and its optical accessorie s were arranged so as
to focus the two laser beams on the point in the flow to be
investiga ted. At this intersection , it is necessar y to arra nge
-

the fringe pa ttern such tha t ea ch fringe was perpendicula r to I

the direction of the velocity component desired. To m easure

the flow velocity in a particula r direction , either axially or
tangentially, particles tha t were smaller than a micron were
passed through the fringe pattern . The particles were small
enough to follow the streamlines of the flow accura tely, but
large enough to sca tter enough laser light to be acquired by
the photomultiplier as a signal.
Normally, high speed compressed flow does not contain
adequa te numbers of contaminant particles tha t have diameters
which are no larger than the fringe spacing and provide a good
signal—to—noise ratio. To accomoda te this lack of particles ,
a process known as “ seeding ” was employed to inject particles
into t he flow up stream of the measuring vo~~ime .

12

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
:T1i: i: ~~~~ ~~~~ ~~~~~~~~~~~~~~~~~~~~~
 ~~IL ~~~~~~~~~~~~~~~~~~~~ __ . —

The velocity mea surements were made based on the f a c t tha t

as each particle passed through the fringe pattern , it caused
H an optica l signa l of scattered light to be produced as it passed
through the bright fringe while no signa l was produced as it

- -
traveled past the da rk fringe . Knowing the spacing between
.

bright fringes which was proportiona l to the wa velength of the

laser light and inversely proportiona l to the half-a ngle between
the two beams and measuring the frequency of the light signa l,

H one o b ta ins the veloc it y by :

V D~ . f D~ =

w here laser wa ve len gth

la ser beams ’ half—a ngle
D~ fringe spacing
f = frequenc y
V = particle velocity
The ba cksoatter lens focuses an image o f the intersect ion
point or focal point on the photomultiplier. The backsoatter —

mode was used in this study due mainly to the presence of the
compressor ’ s hub which prevented the passage of the laser beams
from one 3ide of the c ompressor ’ s casing wall to the other side .
Due to the f a c t tha t the intensity of the light scattered back
towa rd the laser transducer was a small percentage of tha t which
would ha ve been received in the forwa rd scatter mode , it was
also necessary to use an increased laser power leve l anc ~ to
employ a beam expa nder to e f f e c t i v e l y increase the Intensity
of the scattered l i g h t .

---s
~~~~~ - - - .-- -~~~~~~ --‘ ---~ - .- - - - - —- - ——...---------
 _ _ _ _ _ _ _ _ _ _ _ -
’ T ’ — - — -- — - • - --—
-- ----- - - - - - -. - - . .— - .
--————
T
~~~~~ ~~~~~~~~~~~~~~~~~~~~~ ~~~~~ ~~ ~~ ~~~~~~~~~~~~~~

B. COUNTER PROCESSOR OPERATION

The 55L90 Counter Processor incorpora ted various functions
with its 250 MHZ clock to determine the doppler frequency of
the sca ttered light. Once a particle crossed the initia l
• fringe of the interference pa ttern, two counters were triggered ,
each being timed using the output of the 250MHZ clock. One
counter measured the time interva l for the particle to cross
5, or NL, fringes and ha lted while reta ining its time measure-
ment . The second counter continued with its counting until a
tota l of 8, or NH, fringe spa ces had been crossed by the particle.
The time measurement from the second counter, TH, was com pare d
in the compara tor to the time interva l measured by the first
coun ter , TL, if the time from the second counter was 8/5 of
the time measured by the first counter within the accura cy, E,
established by the COMP ACCURACY switch , then the signa l was
considered to be a valid reading and a “ da ta ready ” comma nd
was outputted from the compara tor . This LDV counter principle
can be visualized in Fig . 2. The e q uiva len t fre q uenc y was t he
outpu t of t he coun ter processor an d was trans ferre d to a
computer where it was processed to determine the particle ’ s
velocity .
Included in the counter processor were low—pass and high-pass

filters . By means of t hese f ilters , t he hig h fre q uenc y no ise

component of the signa l and the low frequency pedesta l were
elimina ted. These filters were used in conjunction with
threshold window adjus tments which set the upper amplitude
limit. If this limit or threshold was exceeded , the signals
114

_
-:--:- -
~ i~~~~~~~
- - . :- -
_

- - .~~_ ~ __ _
. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~ — —-- .---- ----~
~ ~~~~~~~~~~
 p -~~~~~~~~ iL~~ ~~~~~~~~~~~~~~~
—. - •—-- --
~~~~ ~
—
~~~~
-- — — —- —--—.-- -—. — —
-— -. •—
~ ~~~

were rejected since this would represent a signa l from large

particles. The amplifier gain control performed a simila r
-
function in tha t it attenua ted the input signa l to reduce the
noise signal below the 50 millivolt level and to bring the
scattered light signals above the same level . These validation
•

-
~~ circuits were required to provide adequa te and reliable da ta .

H 15

_
_ ___
- -
- - ..-
• — • • •
-
_ .•-_
~~~~~ ~
- - g--_ _
..- --’ ,- --- _
~
- -_
-—-——
_________________
, - . .. -.. —.

— .— , . -
~~~~~~~~~~~~ -- ~~~~~~~~~~~~~~~~~~~~~~~ —~~~ ~~~~~~~~~~ — -~~~~~ --— —— . —---- ~--~~~~~~~ -——- — -. - — --—--- — -- ~~ - ---— - -5- - ---- --- -~~~~~~~~
 I
~ -
~~~ ~ -- ~~~~~~~~~~~~~~~~
- - -- - - -,— — -—--. •- - - -I
~~~~~~~~~~~~ ~~~~~

III. INSTRUMENTATION

A. OPTICS
The optics requirements for the laser appara tus except
for the 55L77 Bea m Expa nder were provided by the 55L88 Trans-
ducer. Figure 3 shows the laser transducer with beam expander
in its alignment and opera ting position. The small window in
he right end permits the laser beam to enter the transducer.
t
The photomultiplier is the cylindrical attachment on its upper
surface . The transducer housing has the capability of being
rota ted 360 degrees to accomoda te the measurement of any flow
velocity component in a plane orthogona l to the transducer ’s
axis. For the single channel backscatter mode , the fol low ing
• sections of the transducer were employed: the 55L84 Beam
Splitter , the 55L86 Back Scatter Section , and the 55L87 Beam
Sepa ra tor , Lens , and Mounting Section.
The Beam S pli tter Sect ion perm its a s ingle laser b eam to
enter the transducer through a small window . Using a beam
spl it ting pr ism , the one beam is split into two beams . Addi-
tiona l prisms separa ted the two beams and redirected them so
tha t they propa ga ted p a r a l l e l to the centerline of the trans-
ducer. An adjustment on the beam splitter provided the ability
to adjust one of the beams in a plane orthogona l to the pla ne
formed by both of the laser beams . This adjustment was used in
he alignment process to check the forma tion of the interference
t
pattern and to insure tha t the two beams were intersecting one
another a t the foca l point.

16

- -i
~~~ u~ r ~~~~~~~~ -J
. - . .
- - ..~~~

—.
- ~~~~~~~~~~~~~~~ - - --- ~~~~~ -
---—
~~~~

— .- -- - - -. - - — -- -- - --
. — -- .. - - - -.-- — - —-- ,- -- . .-- - -- -- - --- --•- -.
~ ~~~~~
 -~~ —- •

~~~~~~~~~~
•

Ad justments on the front end of the laser optica l bench

provided a means by which the optima l beam spot intensity was
• achieved . The ad justments altered the orientation of the
transducer chassis In order to align the input beam with the
ent’-’ance window so tha t it entered the beam—splitting prism
properly. This insured tha t the beams were of equa l intensity
and of optima l quality . The adjustments provided limited move -
me.it of the transducer in pitch , yaw , and roll.
The beams from the B~ am Splitter passed through the Back-
sca tter Section and into the Beam Sepa ra tor Section which
contained pairs of mirrors tha t by selection determined the
separa tion distance of the two beams as they exit this section .
• A beam sepa ra tion knob provided beam separations of 20mm , 4Omm ,
and 80mm . There were three foca l lenses which ga ve the laser
a ppara tus the capability of using focal lengths of 120mm , 300mm ,
or 600mm . The combina tion of foca l length and beam separation
de termined the angle be tween the two beams as they converged
towards the focal point . This angle was important due to its
direct rela tionship to the number of fringes in the interference
pa ttern and the fringe spacing for a particula r laser wa ve-
length . The Beam Expa nder a ttached to the front of the Beam
Separator Section where the foca l lens would usually be placed.
The f o c a l lens was the n screwed into the open end of the Beam
Expander .
As sta ted previously, the Backscatter Section was placed
be tween the Beam Separator and the Beam Splitter Sections .
This section accepted the backsca ttered ligh t a fter it had

17

—- -.— —••——•—— -. .— ~ _ ~ J •.•- -•• - ,,• • tM’ ,..-. & * r -

- ___ __ - --———_— .•- •5••_. _- — —&,-•..• •—-•‘—•-••-~

• ‘_ a_ — . ~~~~~~~~~~~~~~ ~~~~~~~~~~
- •- —-- --—-. . ~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~ •
 I
- ••-• -.
- :--‘-
— . .
~~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~ — —
~~~~~~~~~~~~~~~ ~~~~~~

passed through the focal lens , the Beam Expa nder , and the
Beam Sepa ra tor . The foca l lens focused the scattered light
onto a mirror in the Ba ckscatter Section to re flect the light
towards the pinhole in the side of the section where the photo-
multiplier tube was secured . When the light was properly
focused and aligned on the pinhole , the only light allowed to
pass through the pinhole and into the aperture of the photo-
multiplier was the scattered light from the particles traversing
the Interference pa ttern.
Us ing me tr ic A llen wrenc hes , a djustments were ma de to ref ine
the positioning and focusing of the backscatte r focal point with
respect to the position of the pinhole . Two Allen screws affect
the positioning of the backscatter mirror perpendicular to the
direc tion of the b ac kscattere d ligh t; t hi s in turn a ffects t he
position of the foca l point . One other Allen screw adjusts the
vertica l position of the pinhole providing the ability to -

-
correc t the focus of the foca l point at the pinhole . A side
view port is provided in the casing of the Backsca tter Section
to provide viewing of the pinhole and the backscattered light’s
foca l point while adjus tments are being per formed.
Used in conjunction with the 55L86 Backscatter Section was
• the 55L77 Beam Expa nder . The Beam Expa nder utilized the
rela tion be tween beam waist diame ter and the diame ter of the
beam at the foca l lens to optimize the sca ttered light signals
derived from backscatter measurements .

18

-
LL
• • —
~~~~~~~~~~~~~
----- 5-- - ------ - --.- -- — - --- --- - • • -•— - - -~
-
 - -
_ _ _ _ _ _ _ _ _ _ _
- - ••--- -- - — -• —,
•
~ ~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

dw a i s t 2 )iiaser. f’
- _ _ _ _ _ _ _ _ _ _ _ _ _

• dwaist beam w a i s t diame ter (equa l to the diameter

of the measuring volume )

• X laser wa velength
~ laser =

• f = focal dista nce

Db -
diame ter of laser beam on foca l lens
The Beam Expa nder was capable of decreasing the diame te r
of the measuring volume by a fa c t o r of 1.9. This reduced size
of t he measur ing volume was p r Im a r i ly the r e s u l t of art increase
in the diame ter of the laser beam prior to Its exiting the
transducer and par tly due to the increase in the beam sepa ra tion
between the coincident beams where they exited the focal lens .
he e ffect o f th is decrease in s ize of the measur ing volume
T
was an increase in its light intensity as well as a better
spa tia l resolution within the measuring volume .
The Beam Expa nder also effectively increased the collection
aperture of the Backsca tter Section. This was accomplished
by accepting the scattered light signals through the larger
lens and reducing the volume of the signals to a point where
the signals were accepted by the backscatter aperture . Both
of t hese e ffects affec te d th e intens it y of t he bac ksca t ter
signals by increasing It by a factor of 13 (Ref. 3). This
signa l increase significantly enhanced the measuring potentia l
of the laser doppler velocimeter.

19

5- - - - - ~~~~~ ~~~~~~~~~ •-
~ --5--- ••-
--
5-— -.—— — .•- =--• — - •- —5- — .5—
_~~~~~~~~ .
. —
- — - — — - — ~~~~ ~~~~~~~~~~ —

B. SEEDING APPARA TUS

The Thermo-Systems Model 3075 Constant Output Atomizer was
• the device used to produce the seeding particles from olive
oil (Fig . 24). The a tomizer was capable of genera ting seeding
particles which varied in size from 0.02 to O.3,Mm through the
technique of solvent eva p o r a t i o n . So lvent eva p orat ion invo lved
the i n j e c t i n g of t h e s eed ing f luid into a stream of dr y com presse d
a i r and expa nding the aerosol in the atomizer assembly cavity .
For this experiment , the Model 3075 was used in the non-
recircula ting mode ; once the aerosol entered the a tomizer
assembly cavity, the particles tha t were too large to be carried
by the air flow out through the upper exit were deflected by the

-
-

impact pla te .
A ke y com ponent o f the seed ing ap para tus was a Harva r d
A ppara tus syringe pump . The syringe pump provided the means by
which the seeding fluid was injected into the a tomizer . It
also provided the f l e x i b i l i t y of having 30 d i f f e r e n t flow ra te s
which could be selected ins ta n ta neously. 24 she; the Figure
~
syringe pump in the foreground , an d the a ir pressure reg u la tors
and dehumidifier are positioned to the left.

C. LASER SUPPORT APPARATUS

The two main components of the laser device , t he laser b eam
genera tor and the transducer , were moun ted on an aluminum
optical bench which provided rigidity to the system . This
rigidity mainta ined the alignment between the laser and the
transducer and provided suppor t for the position adjustments

20

:
. -- - : . 5- —.- - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
--

TT _ _ _ _ _ _ _ _ _ ~~~~~~~~~~~~~~~ _ _ _ _ _ _ _

associa ted with the transducer. Attached to the optica l bench

was a transla tiona l microme ter which provided the calibra ted
movement of the laser system during the positioning of’ the
laser beams ’ foca l point for da ta measurement . The microme ter
along with an aluminum stiffening bar provided additiona l
support to the optica l bench to keep it from bending under the
weight of the laser components .
The transla tiona l micrometer provided an attachment point
be tween the optical bench and laser components and the laser
support table . The woode n support table was s p e c i f i c a l l y
designed to support the laser system in the close confines of
the transonic compressor test cell as seen in Figures 5 and 6.
The table provided the abilities to incline the laser system
from 00 to 25° and to raise the upper portion of the table a
tota l of eight inches. Figure 6 shows the complete laser
system consisting of the laser g e n e r a t o r to the r i g h t , the
transducer to the left, and the laser support table undernea th .

D. DATA ACQUISITION SYSTEM DESCRIPTION

The da ta acquisition system consisted of an assortment of
computers and peripheral devices tha t were joined together as
shown in Figures 7, 8, and 9. The 55L90 Counter Processor
provided the high voltage source for the photomultiplier. The
photomul tiplier transferred the scattered light signals to the
I
Counter Processor . An oscilloscope received the scattered light
signals and system noise from the Counter Processor on one of
Its channels; the other channel sampled the output of the

2].
 - - -
_ _ _ _
_____________________________ - - - -
~~~
---

Intersta te P24 Pulse Genera tor prior to its being received by

the Counter Processor. This oscilloscope was also connected
-

to the pulse generator v ia a line from the “SYNC OUTPUT” of t he

•

pulse generator to the “TRIG IN-EXT HOR” input . The pulse

genera tor was connected to the PACER c i r c u i t r y to take advanta ge
of its bin position signa l output . The PACER , in turn , was
controlled by the HP-21MX computer with the HP-26244A Termina l
handling its progra m input and output functions . Ano ther
oscilloscope was use d to s ynchron ize an d mon itor t he PA CER
output and the one —per -bla de signa l being t r a n f e r r e d to the
PACER by the HP-21r~1X computer . The HP-21MX system was inter-
faced with the HP-9825A computer by an HPIB (an interface bus).
• To comple te the circuit, the HP-9825A computer was linked to the
Counter Processor and art HP-9862 Plotter .
Figure 8 is a photogra ph showing the counter processor
(
lower left), the HP—9825 computer (lower right), an osc illo-
scope (upper rIght), and the HP-9862 plotter . Figure 9 shows
t he area in the Tur b oma c hine La b o ra tor y con tro l room w hic h
houses the PACER system. The PACER circuitry and HP-21MX
compu ter are loca ted in t he vert ica l conso le a t th e rear w ith
the HP—262424 termina l in front of it.

22

-5
r _ _ ,i . T
- --
_
•
_
~~ ~~~ .. • •

•~~~~~~ 5_5~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~ _

~~~~~~~~~~~~~~~~~~~ .5- L~~~~~~~~~~~
 -— -- 5--—-

IV. 1~
’EThOD OF APPROACH

A. BEA M MODE SELECTION

Accord ing to Ref . 6, the 55L88 Beam Transducer , without
the installation of the beam expander , had the ca pability for
beam separa tions of 20, 40 , and 80mm and for focal lengths of
120, 300, and 600mm . To determ ine wha t the best mode of opera -
tion was to be used in the laser velocimeter survey, an ana ly s is
involving the geometric characteristics of the compressor
section and the mea suring volume properties was conducted . The
analysis was conducted assuming the existence of the worst
operating condition which existed when the measuring volume was
coincident with the compressor hub .
For eac h measuring volume length available , t he an gle from
the cen ter of the measur ing vo lume to the ou ts ide b oun dar y of
the casing window was determined . For any mode to be useful ,
the half angle be tween the incident beams would need ,to be less
than the angle measured above ; this would insure tha t both
beams would be able to enter the outside fa ce of the casing
w indow . Two of the modes failed to provide half angles small
enough to meet these physical requirements . These modes
consisted of foca . lengths of 120mm and 300mm and beam separations
of 240mm and 30mm respectively . ‘

The analysis was continued by determining the beam separa tion

of eac h mode at the casing window entra nce. A com par ison of the
beam separa tions revea led tha t the rela tive differences between

23

- .

j
 -- ,T-
_______________________

~~~:~~~
-
- -
~~~~~

them at the casing window entrance were very small. The

decision of which mode to choose was based primarily upon the
number of fringes available in the measuring volume for each
combina tion of foca l length and beam separa tion . As the resu lt
his decision, the opera t ing mo des to be use d in the ve loc it y
of t
survey were comprised of focal lengths of 300mm and 600mm ,
respective beam separa tions of 14Omm and 8Omm , and respective
numbers o f’ fringes of 614 and 127. The anticipa ted applica tion

of’ these modes was the basis upon which the laser support
appara tus was designed.
With the addition of the beam expa nder to the LDV trans-
ducer , the length of the laser system was extended by 5 inche s
which elimina ted the use of any mode employing a 600mm focal
length because the combination of the foca l length and the
length of’ the laser system exceeded the space available in the
compressor test cell. The beam expa nder increased the trans-
ducer beam sepa rations of 20mm and 140mm to separations of 38mm
and 76mm respectively . The beam separation angle consequently
increased as a result of the increased beam separation thereby
elimina ting the 300/76 mode due to its half-angle being too
large to permit the beam pair from entering the casing window .
This further elimina tion indica ted tha t the beam mode which
was comprised of a focal length of 300mm and a beam sepa ra tion
of 38mm was the only mode available to be used in the experiment.
It provided thirty fringes in the measuring volume .

24

-- - - -~~~~~~~~~~~~ -rn- ~~ — .
~~~~~~~~~~~~~~~~~~ - ,- 5 _ • . - - -_ -
 -_ _ _ _ _ _ _ _ _ _ _ _ _ _
______________________________________ - —

B. LA SER SYSTEM ALIGNMENT PROCEDURE —

~

-
The geometr ical combination of laser inclina tion angle

( o’(), table height (H), and the distance of the transducer ’s

-

foca l lens from the compressor outer casing wall (D), provided
the means by which the laser focal point was positioned at the
-
,
-

desired loca tion within the compressor (Figs. 5 & 6). It was
necessar y to inc line t he laser s y s tem a t the same acu te an gle
which was formed between a horizonta l axis and the radial line
from the center of the compressor hub through the center of the
port opening in the casing wall. The proper combination of
table heigh t and the distance from the focal lens to the
compressor outer casing not only aligned the longitudina l axis
of the laser transducer with the radial line that passed from
the compressor hu b’ s center through the center of the casing
port opening but also placed the focal point on the hub .
To ascertain whether the focal point was in fact positioned
on the hub , laser sa fety goggles were used to observe the -

-
reflection of’ the laser beams on the hub ’ s sur face . If two
s p o ts were seen , more adJus tments were required until only one
ypIca l va lues use d for the laser
spot could be observed . 1’
inclina tion angle , table height, and focal lens distance from
the casing wall were 19 0 , 2 inche s , and 8.2 inches r e s p e c t i v e l y.
Care was taken to ensure tha t the longitudina l axis of the
laser system was perpendicula r to the longitudina l axis of the
compressor section. This was done by aligning the two incident
l ser beams such tha t they were loca ted on the relative horizonta l

25
 _____
5 .5 - - --r ---~~~~~~~~~ -

j
. 5 . 5 - —~~~~~~~~~~ _:_
— 5- - 5— -- ~~~ ~ ~~

axis of the port opening and tha t the beams were aligned such
that they were on opposite sides of the rela tIve vertica l axis
and at the same distance from It.
With the la ser focal point positioned on the hub , t he beam
spot was used to align the pinhole in the photomultiplier so
that the reflected focal point was in focus and positioned .
Once this had been accomplished , the laser system could be
moved by using the calibra ted transla tiona l micrometer to
position the focal point at the desired radia l point in the
compressor for da ta measurements . In t hi s ex per iment , the H
14] inches
focal point was positioned at a radia l distance of ].•
from the compressor hub . The microme ter had the ca pability of
traversing only one inch without surpassing the calibrated H
sca le. To traverse l.~4l inches accura tely, the m icrome ter was
traversed 0.5 inch using its calibra ted scale and then a 0.91
inch me tal plug was inserted into the micrometer to attain the

-
-
tota l required distance . With the completion of the physica l
alignment of the laser system , the foca l volume was positioned
at its desired loca tion behind the rotor trailing edge at an
axial distance of 0.241 inc h , At this position , a blade pair
was 3.249 inches in width .
To test the alignment prior to running the compressor , the
tube carrying the seeding particles from the atomizer was placed
at thci port opening such tha t it directed the particles at low
veloc ities into the compressor while the laser was turned on .
The particles would not traverse the interference pa ttern as

• they entered the compressor but circula ted turbulently a fter

26

__ .55-

_ _ _ .5-— — - - —5 .
5 _ _ _ _ _ _
 - —

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

striking the compressor hub . This turbulent motion allowed

some of’ the particles to tra verse the measuring volume across
the fringe pattern thereby producing scattered light signals .
By adjusting the ba ndpasses and amplitude on the counter
processor, the classic doppler burst was observed on an
oscilloscope (Fig . 10). The doppler burst patterns were only
atta ined if a well-defined fringe pattern was produced and the
image of the foca l point had been positioned and focused properly
on the pinhole at the photomultiplier a perture . Another indica —
tion included a high validation ra te from the counter processor
on the order of 8 o particles/b OO or better.
~

C. DATA AC QUISITION SYSTEM OPERA TION

The da ta acquisition system was developed , for t he most
part , specifically for the purposes of this study . The system
provided the method by which flow velocities in the compressor
could be measured and processed to study the velocity flow field
at the same position behind each of the blade pairs . The part
of the system which was already well established in its operation
was the PACER system (Ref. 7) installed in the control room of
the Nava l Postgradua te School Turbomachine Labora tory (Fig. 9).
Knowledge of’ the da ta flow was necessary to understa nd the
overa ll problem solution (Fig . 7). The sca ttered light,
reflec ted off a particle transiting the fringe pattern , was
detected by the photomultiplier, This photomultiplier output
was observed on an oscilloscope for the purpos~e of viewing the
signal characteristics and intensity level as well as the

27

- 5-.5~~~~~-- - - - _- -5 -5 .
 _ _ _
_ _ _ _
_ _ _ _ _ _ _ _ _ _ _
-
_

magnitude of the system noise level after the signa l had passed
through the a m p l i f i e r section of the counter processor . The
oscilloscope was used ifl conjunction with the amplifier gain
and threshold setting switche s to optimize the signal—to—noise
ratio. The amplifier ga in was used to lower the average noise
level below 50 millivolts and simultaneously placed the signals
above 50—millivolt levels . The thres h o ld sett ing sw itches were
used to vary the threshold level until a high validation rate
was achieved.
The osc illosco pe was a lso use d to o b serve on its o ther
channel the da ta acquisition window which was generated by the
PACER equipment and the pulse generator . The oscilloscope
displayed the window width , the stability of the pattern, and
the time be tween windows . A typica l window width for this study
u~sec s to 24 .0~.t secs.
was 2.O,
The pulse genera tor generated the da ta acquisition window .
The pulse genera tor was interconnected with the PACER such tha t
the one-per-blade signa l was used to trigger the pulse generator
which, in turn, sent the pulse of the desired width to the counter
processor . Wh enever t he pu lse was no t presen t in t he cou n ter
processor , it was inhibited from transferring the raw doppler
frequency da ta to the HP-9825. This process permitted the
measurement of da ta a t prescribed bin locations behind each of
the blade pairs.
The HP—9 825 accepted the da ta from the counter processor
when da ta was a v a i l a b l e for t r a n s f e r and stored it in a data
b u f f e r until a desired number of doppler frequencies had been

28

-~~~~~~~~~~~~ - - . 5-s - -- - ._ ~~~~~~~~ -.-5-5 .5-—

5.. ~~~~~~~~~~~~~~~ --- ~.5
 .5 - .5 5-
— —- ,i
~~~ ~~~ - -
~~ ~~

accumula ted . Eac h of the raw doppler frequencies was processed

by the HP-9825, and the average of a l l of’ the frequencies was

• computed and converted to a flow velocity component. The

combination of flow velocity magnitude in feet per second and
• bi n position was transferred to the HP-9862 Plotter to maintain
a graphica l representation of the acquired da ta .
The HP-9825 was interfaced with the HP-21MX computer which
acted as the controller for the PACER system . The HP—9825
transferred its resulting doppler frequency as each was c omputed
to the HP_26)414A Termina l for output on its display and its
peripheral teletype . Once it had received this informa tion, the
control program arid the HP-21MX computed the next bin position
to be used in the acquisition of the da ta and re— initiated the
process.
Table IV is a listing of the controlling program for the
da ta acquisition system . The program was used in conJunction
with the “BASIC” program designed for the HP-21NX computer .
Its purpose was to i n i t i a te the PACER sequence where the PACER
triggered the pulse generator which uninhibited the counter
processor . After the HP—9825 (logical unit #10) had interfaced
with the HP-21NX declaring tha t it had finished acquiring and
processing da ta for a particula r bin location , the program
upda ted the PACER to take da ta at the next bin position. The
program would continue in a loop until the last bin position
requested In line 20 had been 5.nvestiga ted . After the HP-9825
had completed processing its da ta , it trans ferred its resulting
average doppler frequenc~ to the HP-21MX; the HP-2].MX displayed

29

- -.----5 — —— - ..- .
— ~~~‘-——- -. — --—
—5 - .5——- .—- ~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~

____________ .5.55
 ——
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _- -, - _, -
-‘ - -- - -
- -~~~~ ~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-
- - - --~~~~~~~~~
-5 - - - ~~~~~~~~ -

-
— - — - —5 - ~~~~~
- - - ,- -- — .
-- - ~~~~~~——•~~ - 5— - - - -- -

this frequency along with its corresponding flow velocity

component, the bin position , and the compressor rpm indica tor
on the console display of the HP-2624J4A Termina l and the
teletype (logica l unit # 14).

1:1

30

-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --- - - 5 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 — -‘ - -- - - - --
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -

V. RESULTS AND DISCUSSION

A. DA TA RESULTS
• The experiments for obtaining the da ta for the a x i a l and
tangential components of the flow behind the rota ting rotor
stage of the transonic compressor were conducted under the
conditions of Table I. This table lists the settings for the
counter processor and the HP-9625 as well as the compressor
performance da ta . The experimenta l operating conditions were
essentially steady—sta te conditions due to their being repro—
ducible for each experiment.
The ax ia l f low veloc it y com pone n t was o b serve d to have a
higher velocity magnitude and a weaker wa ke structure as compared
• to the results of the tangential velocity component measurements .
-
— ;- Figure 11 shows the axial velocity measurements as they varied
with bin number . The non-wa ke struc ture exhibits a mean velocity
ma gnitude of 328 ft/sec with a standa rd deviation of ~ 18.9
ft/sec. The two wa ke s truc ture s are c haracter ize d by the
decrease in velocity ma gnitude . The f irst wa ke r eac hed a
minimum velocity of 2514.249 ft/sec at bin position 77 while the
second wa ke attained a minimum velocity of 254.67 ft/sec at bin
position 201. From this da ta , it was determined tha t the first
wake was 124 bins in width and the second wa ke was 10 bins in
width . Using the end of the first wake and the beginning of
the second wake as re ference points, it can be seen tha t the
wa kes were separa ted by 116 bin widths .

31

5-—- 5- -•~- ,- 5——- —5— .—-——— - -en-F — -- - -- ~ 1 -

_.
- — - 5- •- ~~~~~ .-‘-
“ ‘ ‘

~~~~~ —.5-

I
.
-

~ .5.5 5- _ _ - .s~~~a~~~ ,,-,- . —,.-—-— —

_ _ _ _ _
- - 5- ~~-5 _ _ 5 ,
~~~~ ~~~~ .55
 ___________
-,
— -
~~~~~~~~~~~~~~~~~~~
-
:
-
:- -- -r
~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~ — - 5- - -~~~~~~~

The tangentia l flow velocity component measurements as

seen in Fig . 12 show not only an increase in velocity in the
wake but also a more well-de fined wa ke structure . The mean
velocity magnitude of the non—wake regions was 175 ft/sec with -
a sta ndard devia tion of ~ 6.9 ft/ s e c . A sharp increase in
velocity ma gnitude i d e n t i f i e s the two wake s t r u c t u r e s . The
first wake reached a maximum velocity of 335. 248 ft/ sec a t bi n
pos it ion 7]. while the second wake achIeved a maximum velocity
of 338. 16 ft/sec at bin posItion 201. The widths of the first
and second wakes were observed to be 114 and 12 bin widths,
respectively. Using similar references as with the axial
veloc it y measurements , it was found tha t these wakes were
separa ted by 120 bin widths .
Figure 13 represents the tota l flow velocity magni tudes
over a range of bin positions. The va lue for the magni tude of
each total velocity was calcula ted from the axial and tangentia l
veloc ity measurements tha t corresponded to a common bin position .
The tota l ve loc ity was de term ined us ing the fo llow ing equa tion:

VTCTAL V YJAL # V ANG

~~~~AL
a
~ ~
~
For the non—wake regions of Fig. 13, the mean tota l velocity
was 372 ft/ sec . The wake structure is not as well defined in
this gra ph as it appeared in Figs . II and 12 and lacks any
significant change in velocity . S ince the tota l ve loc it y vector
is defined not only by the tota l velocity ma gnitude but also
the flow ~ s yaw angle , this indica tes tha t the passage of the
rotor blade affects to a higher degree the direction of the
flow ra ther than changing the tota l velocity of the flow .
32
- - -

5-- .5~~~ .5- —— - -” - --5 - - - - ---

_ _ _ _ _ ~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~ - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _ _ _ _

This e f f e c t is observed in Fig. 114 where the yaw angle of

the total flow as referenced to the axia l direction is plotted
over a ra nge of bin positions . This gra ph depicts well-defined
wakes with maximum yaw angles for the first and second wa kes
of 50 .20 ° a t bin position 69 and 53.02° a t bin position 201,
respectively . The non-wa ke regions exhibit a mean yaw angle
of 28.06° wIth a yaw standard devia tion of ±1.80°. Therefore,
the passage of a blade causes a maximum cha nge in yaw angle of
22.124° for the first wa ke and 24.96° for the second wake . As
w ith t he ma gn itude o f the tota l ve loc it y , the yaw ang le was
derived from the axia l and tangential velocity measurements
tha t corresponded to a common bin position. The yaw angle was

• ca lcula ted using the following vector diagram and equa tion:

V
)
)TANGENTIAL
~AXIAL
V
9 -
- T~N 4 TANGENTI ~~
VA AL
~~

Turbulence intensities were de termined by the HP-9825

computer program (see Appendix A and Table II) for the axia l
veloc ity component measurements and the ta ngentia l velocity
component measurements . Eve n t houg h onl y four data va lues were
use d to determ ine t he ve loc it y measurement a t eac h bin pos it ion ,
an increase in the tur b ulence in tens it ies of the wa ke re gi ons
as compared to those of the non-wa ke region was evidenced for
the tangentia l velocity component measurements .

33

—
 5- _.
-—5.--- - - - - - - ,- - - -
—. ~~~~~~ 5-
’ ——- ~~~~~~~~~~ — — — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.--
~~~~~~ ~~~~~
—-5-- —— -- - — -5- — — -

B. DATA COMPARISON
Re ference 8 contains da ta obtained by pressure probe
measurements representing the wake structure behind the
•
-

I
rota ting rotor section of the transonic compressor . The
results of these experiments can only be qualita tively com-
pared to those of Ref. 8 due to the da ta being obta ined under
different compressor opera ting conditions . In c omparing the
-

-
-
yaw angles measured in these experiments with those of Ref. 8,
the present measurements indica te narrower wa kes, a grea ter
cha nge in yaw angle , no evidence of an aperiodic wa ve structure
-

in the non—wake regions , and a larger average yaw angle for the
non-wa ke regions .

It

324

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 - — -- - -5- — 5- - -
— ~ --5—-- —

-
~~~~~~~~~~~~~~ _ _ _ _

5- -

VI. CONCLUSIONS

The laser doppler velocime try system was shown to be

ca pable of measuring the axia l and ta ngentia l velocity compo—
nents of the flow behind the rota ting rotor blade s of’ the tran —
sonic compressor . The da ta was characterized by the presence
of two well—de fined wake structures which represented the
passage of a pair of rotor blades. The wake structures of the
axial and tangentia l velocity component da ta were located at
the same bin positions and were of comparable widths . From
t hese measurements , the flow was further analyzed to determine
t he tota l veloc it y vector , its ma gnitude and direction , over
the range of bin positions . Turbulence intensitie s were also

• determined to study the turbulent na ture of the flow in the

wake and non-wa ke regions for both the axia l and tangentia l
directions . A qualita tive comparison of this da ta with da ta
obtained from pressure probe measurements was performed.
To fur ther ver if y the data from t hese invest iga t ions ,
additiona l laser doppler velocimetry da ta should be acquired ,
particularly da ta tha t can be q u a n t i t a t i v e l y compared to da ta
obta ined previously by other methods . I t is a lso recommended -
~

-
that similar experiments be performed by taking da ta behind a
particula r pa ir of blades as was accomplished in Ref. 8.

35

.5-- - - -
—— --- - -5 - - -
~
:
:: ~—
-
~~~~~~~ ~~~~~~ ‘!
~~~~~~~
- --- 5-- - 5 - -- ---- — -

— 5 - --- — -- -

/
/ -I
/
/
I
‘

— _ _ _ _
 - 5-- - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
— — -,

OOPPL ~ R SC HMIT T
mODE
_ _ _ _ _ _ _

N H. N
L
SWI T ~~
L
_ _ _ _ _ _ _ _ _ _ _ _

FRIN G E
N H aKN COUNTER N
L L

HIGH COUNT L 250 m HZ L0~ COU N T

ID l ER CLOCK TIME R
—

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

[DATA 1
I R (AO Y 1 -

OUTPU T N
rH
I

[FREQUENCY

LDV COU N TER PRI N CIPLE

FIGURE 2

-i 37 1-
— — - - - - -5 —~~~ --- ~~~~~~ -— — ~~~~~~~~~~
— ---
p. . -5- - ----—-5-5-- 5-

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

_
- y4 _ - _ - ;
,
- -

~~~ ~~ ~~ ~~ ~~
—

w-
- - ~~~ ~~
-
.:~.
~~~
- - A ~~~~~~~~~~~
~ . ~~ ~~~~~~~ ~~~ ~
-‘ ~
- ‘<-
- ~~ ~~~ ~ -

-- ~~~ ~~~~ --
-
- 5 , “ ‘ ,-~~~ ‘ IIØ5-~
,7l<
~~~~~~~~~ ~

-5 ~~~~~
- ‘
-
, —-
-
~~~ - ~~~~~~~~~
-‘ -
I’ ’- -
‘
-

~~~~~ I
~~
~
- - ,:

5- _ _ _

It
H z
--
- -
-
~~~~ ~~~~~~~~~~~~~~ , ,
— , -_ i —
-

_
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~ ~~

I .

_ I
i;!-
_

-T- —
~ -L
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
‘ Is~~~~~
-- -,
- -
-~ -- ~
~~ ~~~~~~
‘I
~~~~ ~~~ -
5

-
-5-
- -:- ~~~~~~~~~~~~~ -
- ~,
-~~~~~-~~
~~~~~~~~~~~~~
- - - - - —~~~ _ ~a~_
5- - ;5- ,
- ~~ ~~~ —~~~~~ ~~~~~~~~~~~~~— ~
~~~~ ~~~~~~~~~~~~~~~~~~~~
_ _ _ _
h~~_ ~~~

- ~~~~~~~~
_ _ -
~~
)_
5 5
—
I
5-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~ ~~~~~~
-
~~
 - -- - - ‘- - -5- - - -5 ---—-- - ~~~~~~~~~ --- - 5 55- 5--

-
r~~~ ~~ ~~~~~ ~~~~~~~
~~~

- -
-
--5 - - -
“ -
“

:
‘ I , ~~~~~~~ ~~ ~~
S
,~ ~ ~ - ~~~~~~
-
-
-
— / -

~-
~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~
_ ~
,
- ~~~~~~~~
,
-
_
-- -

;i
, , -~ :, ~
~~~ ~~~

-:
;- - - ‘

:~ k- -
~~~ ~~~~~ ~~ ~~ ‘
‘ ~~ ~~~
- ~~

5-
~~~~~~~~~
- c --
_~~~~ I
~ ~ - -
-
< -
~~~~~ ~~~~~~~~ ~~~~~ ~~~ ,- -
~~~~~~~~~~ - ,
- ~~~~~~~~~
< -
- ~ ~~~~~~~~~~
~~~ ~~ ‘
, -- -
,-
,
~ - x - - / * ~~~~~~~~
~~
~~~~~~~~~~~~~~~~
‘
~ - -~ ~~ ~~
— , - , > , :-
-
-
~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~
- -
~~~~~~~~~ ‘- : -

~~~
- - ~~ ~~~~~
~/
, -
~~
/ ,
~~~~~~~ - , 5- -
~
-
- /

5
~~~ ~~~ ’ ~
j - - -
~~~~~
- ~~ ~ ~~
~~
:-
~~~~~~~~~~~~
J
~~
:- .~~~~ - --
-,
: ,
~~~~~~
-
-
-

,,
‘5- ,
-- -
,
, -
,
.

,
;
-
~~~~~~~~~~ -
-- - ~~~~~~
-

-
-5
- ~~~- -
~~~~~~ - 5-~
,
-, :
~
- ~7 ~~~
- rn4,
, ,
. -
- -
- -
-
- —
- -
~~~ ~~~ -
_ ‘/‘_ ,
,~~~~ _~~~ ‘

~~~~
~~
5-
~~~~ -

/ / /-~- ~~~
~
-
/~~
__ ~~
- / ~~ - 7 - -~~~~~~~~~~~ ~~~

H- 5- - -

- -
-

-
-~~~A
- -
- ;
~~~~ ~ ~~~~ ,5~~~ _ ~~~~~~~ ~~ - ~~~~~~~~~~~~~~~~~~~~~~~

39

L. - -- -- ~~~~~~~ - 5---5 - 5_--_ _ -

_ -_
-- -_ _ _ _ _
 -— -
- — - - 5 - -— - 5--
--
—5
ii

t I I
‘I I I

_ _ _ _ _ _ _ _ _

I I

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
L ’ ~
~
A~~~
-
\ -r
~~ ~
—— ~
—1

= (_
_
*-
~~~~~~~~~
_
~~~~~~
=
= ~
t ~I ____

‘I
‘I

——
C,,
C,,
I&1

C.3

40

: , —
5- - - - - - . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.: — 5-- --- 5---,

-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -~~~~-~~~~~~~~~~ —~~~~~~~~~~ —5-
- - -~~~~~~~~ -5 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- -~~~~
r 5

- _
_ _ __
—
-5-- -- ,

--
~~~~

~~~~ -
~~~~~~~~
-
-
- 5--
?~~~~~~“
-~~ ~~ ~ -- -~
0
nil
-
-
--
-
- -
~~~~~~~~~~~

~~:7~4~
(
~~

~ ü

/
/

$
/
--
--

;_ -- ~~~~~
-
~ ~~~~~~~~~~~~~~~~~~~~~~~

/ ~~~~~

~~

~ -
~~ ~~~ 4I~

-‘ -- -
5-5-—
~~, ~
-
- -
~~~~~~~~~~~~ ~ ~ ~
-
-

-
_p
~~~~~~~~~~~~
‘5- -

,
, -
-

-
- ’
.5-:

* -

- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

::I
iT iiz.
-

_ _ _ _ i
 -
—5- - - - -
5- 5-—— -U- ~~~~~~~~~~~~~~~~~~~~~~~~ —

-
i’
I
- •

IP H OT OMULT I PL IE R 1
__________________ _________________________________________

I
LASE R

HI VOLTAGE
~m SIGNAL IN mONITOR OUT

AUX
—
CUUN
.1 R PROC SSOR
DIGITAL
PUT

INPUT
4,
I OSCILLOSCOPE I
t
TRI G IN—
(XT HOR
S YNC OUT PUT

‘P U LSE GE N ERATOR
[

I I HP— 9862 1
I OSCILLOSCOPE it I TELE TY PE I
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

1 PLOTTER]
It H
I PACE R I
_ _ _ _ _ _

I HP~.2 644 I ..~ HP-.9825A I

_ _ _ _

f I T ERMINAj J I comP UT E ~ j
~
fi~P~ 21mx It HPIB I
¶c ~OMPUT ER
I
I
j~~

I PER BLA C SIG NA L

~
I PER REV S IGNAL

DATA AC Q UISITION SYSTEM

FIGURE 7

242

_ _ _ _ _ _ _
_ _ _

-- - 5--— -- - -
~~ ~~~~~~~~~~~ —~~~~~~~~~~~~~~~~~~ - - “ ~~ 5-
 —
-- - — - - —-- -5 - - - — 5- - —-5-
-
—-- ~~~~~ -— •— -
-
-

P5- -5 — 5 - —-— - — 5- -

/_
-
~~~~

~~~~~
- - -

~~~~~~
I ~~~~~~~~~~~

/
~~~~

-‘
$ ; //

1/
a
~~~~~~~ :-
__ _/
:- ,- -5--
~~ , ~~~ _ ,

—ia_i
~ ~~~~~~
5-/
- ~‘/~ ’I’ a - ~~~
~
/
~ -
-
~~~~ -
,-
‘
,*/~
-
-
-- I’- ~~~~
-

~~-th
:
- 5-

- -

- -
C
--
-

’ i. ’ I;
C
I -
- - -
- - / - - -~ ‘
:

4
‘4 4
-

I-
-
~ ~~
_
-
- * ~_~
~~

~~~~~~~~~ ~~~~~~~~~~~
!
- -

//

~/
- ,~~~ i#/ ~~ -” - I- ‘ -- - I-
-

4~7~~~~’ - -
‘-
-- ,
_ 1—
5-
-
~ -
- - ~~~
r *
‘ -
- ,—
/- ‘ .
~ ~ %~ :‘ ‘
~~
“ ‘
~
-

143

_ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 — - -.5 --- - - -
-

-:
II
I 5-5-5-

-1 - -
-

I!
-
5-- ~~~

__ _ _ _ _ A
 I

CL A SSIC DOPPLE R BURST

-
5-—

5--

I
— —

INTERFERENCE PATTERN
F IGURE 10

- -
245

_ _ _ _ _ _ _ _

~~~~~~~~~~~~~~ 5-— ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - — --

 --
-— - - — - ‘ _ -, - --— - —•— -~I—-— --— — ————
— -- --— —-- S 5-5-’~~’~’- ~~~ _ r n _ __~~ 5-,— - --

. LO
H .
.
Lu
,I~~
. . a

.
H .

.
P4
•
. P4
.
.
• .
Li
.

I
•
•
z
z
—
I•
. ~~~ Li —
LU Lfl
— ~
E >
Li
z >-
I- =
I
• z — LU

— ~~~
— .
iLi
• I
. >
.
•
H - . • LU
01 —

.
I

• LU

(‘4
• fy i

.•

•• I I I I I I
I I t I ~~~~ i
LI
-
~~ ~
~~~
-
m
~~
N
hi~~
r’i
N
LI
— —
LI ~~
IA
OJS/th Afl)D13A 1WXH

246

_ 5-_
__ __ _ 1 ii:_
~~ ~ ~~
- 5 _ _ _ __ ‘t :
_
~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~ ~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~
~ —‘-—.5 -

‘1~~~~~~~~~~~~~~~~~~~~~
— ,----- --- .5— ---5- ---— ---•-~~ -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~ 5- — -
- -—-- -— —5-

. in
I LI
I N
,
H
Lii
N
N =
z
•
•
.•
S z
—
•
. 1~14
• .

.
.
~ —
U,
• >5

>..
H ~
Lii (‘~
I
~~~ ~-
LU ~~~~~~~~
— -
• — I
‘-
4 =
z
.
~~~~~
H ~~~~
:
m Z > _

I i •
•
~ _ b

: L
U LU
ci z
:•
I-

I
I
I LU

•
N
• fyi
S

t I I -I 1 1 1 I

LI
~
~
LI
P1
D3S/U) AI1)U 13A 1HIiN3 NB1
— —
~~ ~
P1
Lii
N ~~
P1
LI
Lfl
~ -

247

H _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

-- -- -- -~~~~~~~~~~
- _ _ _ _ _
_ _ _ _ _
_ _ _
 —5--’-- ---~ -•--• _ _
~~~~~~ ~~~~~~~~~~~~
5-5-
~~~~~~~
5-
- _
~~~~~~~~~~~ , _ “
5.
I

LU
•
•
Lii
P4

S
• N
• N
S
• Li
.
E
4 =
.
4

z
5
.
z
• Lii
LU ~~~~~~~~~~~
—
~~~ >5

•
z>.
LU
,
,
I Z — —
• — V L.
-J
• Lii
• >~
• •
S
LU
• Ui I
-
I.
.
p I..
•

H
S
.
•
LU

• (‘4
fy i
1—
I 1 I I I I I I I

iiI Lii Lii

~ P1 ~
P1 (‘4 ~~
(‘4 — ~
— L
ii

~ ]5/U) A1I)013A 1HIBJ.

248

-_ _ _ _ -
- -
_ ;___
~~
—
~~~~~~
•_ — — - -5- —, - - - — — ,—
~~~
--5- - -5 -— -5 —---5- ~~~~~~~~~~ -- - -

5- -
_ _ _ _ _ _ _ _ _ _ --5- 5-
 -- 5 -’ —- -
‘~~~~ ‘ --— -—---- -- - -
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~ ‘
_ _ _ _ _

LU
Lii
• N

.
P-J
- : •, N
• L
•
S
•
.
S
.

1~
•
Lii
S z
-

.
LU
Ez
- - I
—

= — ~~~~~~

Z U 1—
• Lii
• _j
LU
z
LU
• ci
•
.
S

-
.
•.

1 LU

•
N
fy i

I I I I I I I

LU Lii P1 N —

C93 ) 31 NH NBA
~~~~ ~~~

149

— -5-- -5— 5-—~~~ - 5--— - — --

- -- - - - -5- - - 5--- ~~~
- .~~~~ - — - -- — 5 - —- - - -- 5 - - -5 -- -- — • - ——- - . ---—----—--——- — .- 5 - - - - 5 - — ~~~~~~~~~~~~~~~~~ - — ~~ I

A
-
~~~~~~~ ~~~~~~~~~~~~~~~ ~~~
-5- —‘-
-~~~ — - _ _ _ _ -~~~
5-5
~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~
 -i=•-~:--~~~ :--=-—
— — ——-- -
—
——
~~ - — ~~~~~~~~~~~~~
~~~~

TABLE I
EXPERII€NTAL OPERATING CONDITIONS

Counter Processor
Threshold window : 16db

• Amplifier gain: -7db

Photomultiplier voltage : 1.3—1 .14kv
Anode current: l0,
MA
Compara tor accuracy: 12%
Axia l Tangential
Val ida tion rates: 938 888
Low pass filter: 100X106 100X106
High pass filter: i6Xio6 8X1o 6
• HP-9825 Com puter
Ax ia l Tan gent ial
Low pass filter : lOOXlO 6 100X106
High pass filter : 17X106 8X106
Laser power setting : 0.3 wa tts
Seeding probe position: 3200/6.5 inches
Syringe pump position: l~4
Air flow ra te : 55 f t 3/liR
Os ci 1108 C op~
Delay: 1,~t s ec
Vertica l scale: 0.05 Volts
• Da ta acquisition window : 14~~ sec

50

— ‘l —5—- --— —
- -
----5 — --— 5-
-- - - -~~~~~~~~~ -—5----~~~~- 5 - 5 ------
 - - 5-
-_ - - - - :~~~~~~~_ l — — __________________

T
—s-’

~~~~~~

-1 TABLE I (continued )

Compressor Performance Da ta
-
•
RPM - 15330
Pressure r a t i o ( t o t a l— t o t a l ) : 1.1242
Re ferred flow ra te : 10.2609 lb/sec
-
I Total temperature in: 68.12°F
Estima ted average tota l
velocity and yaw angle : 2400 f t /sec and 27 .514°
Pulse Genera tor
Trigger mode selected
Repetition mode : 1HZ
Delay: 10 nanosecon d s
I

Pulse w i d t h : l fr sec (a d j u s t e d to 2,
~ sec)
5-! Rise/fa ll t ime : 5 na nosecon d s
-

Amp litude : 5 vo lts

Output polarity :

51

_ _ _ _ _ _ _ J
 ____________________________________ - -5----— — -—‘5-
-~~~~~~ I ~ -•--~~~ ‘•‘5-
— —5
5--
~~

I
TABLE II
HP-9825 DATA ACQUISITION AND
PLOTTING PROGRA M

ent ”low pass filter ” , H;~~~~~ I; pr t “ low pass ” ,H

~
2: ent “ high pass filter ” ,L; pr t “high pass ” ,L
-

3: erit “how many blocks of points ” ,r6;~ — W V

-

~4: buf “DAT ~~~~~~~~~~~~~~~~~~~~~~~~~ buf “DAT 2” ,4,4;

gsb “DATA ”
I+1— ~ I;moct;if rl.1;rdb(”DAT l ”)
5: ~~~ D
6: if rl l;rdb(”DAT 2” )— ~-D
• cmpD -_~ D;shf(D,l4)_4 D;ba nd (D,17)_
7: ~ ~.r2
8: shf(D,4)— ~ D;otdD2totdr2 —~ D; m dec
9: Dl~ t9/13369314)4-+ D
10:if D~H; ~ —~ D
11: if D’L; ~‘ —~ D
12: if D4 ; gto 14
13 : l/D+r24—~~r14;r5-q.1-—~’r5;D.V--~~V
114: if 1.14; gto 16
-
-
15: jmp — 1~’
16: r5/r4—~-M;V/r4--~ R;~ ’—~.I; r
(R-MN)/M -.
~ R; wrt 731,M
17: gsb “PLOT”
-

18: prt ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ gsb “DATA ”

-
19: Jmp — 124
20: “DATA ” :
21: if r ds(”DAT l ” ) — ~ ;t f r 2 , “DAT 1” , 24

- TT I~J
 - -------5- ---- — - •••I-••- -~ . s_• - ••~~~~~ — - 5 ‘-~~~
- - -- - - 5--- -
-
_ — _ _ _ _ _ _
~~~~~~~~~~~~~~~

TABLE II (continued )

22: dsp “BUFFER 1”

-
23: if rds ( ”DAT l” ) . ,— l ; j mp —1
24 : gto 3 ’
~
25: if rds(”DAT 2”)—Ø ;tfr 2, “DAT 2 ” ,4; gto 31
‘ —I 26 : dsp “BUFFER 2 ”
27: spc
-

~~~

F ’ 28: gto 29
29: ret

30: if rds (”DAT 1”) 4;2—~’rl; gto 25

31: if rds (”DAT 2”)a— l; jmp ~~
‘

32: If rds (”DAT 2”) 14;l—~ r1; ret

- 33: gto 30 4
3)4: “ PLOT” :

35: sd -32,272,-100,650; csiz 2,2.5,1,0

36: pit N,M*1.32401427e_5

37: N +l— ~-N

38: pen
39: ret

53

5-— -5-- _ -_ _
— 5-— —- ~~~~~~~~~~~~~~~~~~~~~~~
-- -- ——5--- ~~~~~~~~~~~~ ~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- - —— •—— -— --5- -
 5 - 5
- ’ 5-
___________
~~‘T
~ — ‘
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TA BLE III
VELOCITY VERSUS BIN NTJMBER
GRAPH ROUTINE

0: sd — 32,2 72 , -100, 650; csiz 2 , 2.5, 1,0

1: pit 100,-100,1
2: ibi “VELOCITY VS BIN NUMBER ”
3: cplt —16 ,2
4: lbl “ BIN NUMBER ”
5: csiz 2 , 2.5, 1,90
6: pit —25, 250,1

7: ibi “VELOCITY (FT/SEC)”

8: axe 0,0,32,50
9: csiz 1.5, 1.7,7/ 10, 0
10: Ø— ’Y
ii: pit ~ ,Y , 1
12: cplt -7, -- .03
13: ib]. Y
124: Ye50— Y
~
15: if Y <i 600; gto ii
-
‘
16: c~,iz 1.7,1.7,7/10,0
17: 0— ~ X
18: pi t X ,~ ’, l
19: if X-0; gto 22
20 : epit —3 ,— i
21: gto 23
22: cpl t .2,— 1

5)4

-
-5- -5- --
~~~~
-5
5- — — — -— 5 - - ---5 5- 5- -5-
~~~~~~~~~ ~~~~~~
- — -
~~~~
—— ~~
— -
~~~~
 5- 5-—.

TABLE III (continued)

23: ibi X
2~4: X+32 —+ X
25: if X ~ ‘256; gto 18
26: pen
27: end
 - -- -- -- —--——--— - - - - -
- ~~~ _ _ L~~ TT :
T : : - ~~~~~~~ ~~~~~~ --~~- -_ -~
_ _ _ _ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-
-

TABLE IV
-

DATA ACQUISITION CONTROL PROGRA M

10 DIM M$ [
12]
, A$~ 3~
11 CALL HPIB (
7,0,0)
20 FOR Xl - 0 TO 255 STEP2

30 CALL RPACE (12,R1, x2)

READ #1O; M
-

40
-

145 PRINT #124 ; Xi,R1,M,1.3)4043E_05*M

50 PRINT Xl,R1,M,1.314043E_05*M
60 NEXT Xi

-
70 END

-
NOTE: If it is desired to take da ta at a particular blade
pa i r , the fo llow ing s tatemen t should b e inserte d in to
the pro gram:
25 LET X2 -(32678-xl )

Ji ~~
56

--
 _ _ _ _ _ —
=—
_ _ _ _

-,
L

APPENDIX A
HP-9825 DATA ACQUISITION PROGRAM

The program found In Table II is a listing of the HP-9825A

program which handled the da ta acquisition process . The
HP-9825A uses a high level programming language known as HPL
which is designed specifically for scientists and engineers.
HPL provide s adequa te power and efficiency for handling
equa tions , da ta manipulation , and input/output operations with
many of its opera tions simila r to corresponding BASIC comma nds .
The following is an explana tion of the comma nds and the flow
of program execut ion wh ich were incor pora ted by the da ta
ac quisition program.

A. MAIN PROGRAM
The ma in program initially assigned a value (normally i) to -
-

he coun ter “N” which corresponded to the num b er of the first

t
bin tha t was to be investiga ted in the wake survey. A va lue for
the low pass f i l t e r , the uppe r limit of the program ’ s band pass ,
was entered and stored in “H” . At this point, the coun ter “I”
was initialized to zero; “I” was a counter whi ch mon itore d how
many doppler frequencies had been accepted into storage prior to
- -

—
being filtered by the established bandpass . The high pass f i l t e r ,
the lower limit of the program ’ s b an dpass,was next entered and
stored in “L” . The ba ndpass of the computer program did not
necessarily ha ve to be e x a c t l y the same lim its as th ose tha t
were programmed on the counter processor; however , the two

s-

- HI ::,1~,,:
1: ~~
:
~ ‘5- 1 5- _
~

ba ndpasses must i n t e r s e c t one ano ther or no da ta flow from the

counter processor to the HP-9825 would ha ve occurred . A dummy
quantity of “ r6” was filled with the “ number of blocks of points ”
de sired , and the q u a n t i ty “V” was given a value of zero where
• “V” was to be the summa tion of doppler frequencies a t t a i n e d
tha t were not filtered out by the c omputer ba ndpass.
The “ buf” comma nd s in step 4 established “DAT 1” and “DAT
2” as read/write buffers tha t were used for buffered DMA (direct
memory a c c e s s ) operation and assIgned its da ta source as the
counter processor . Initia li zation was performe d for the
quantities of “r4” , “r5 ”,and i’M” . The subroutine “DATA ” was
ca lled to accept da ta into the two buffers for la ter processing .
Once the “DATA ” subroutine returned program execution to
step 5 of the main program , the da ta processing commenced . The
“I” , b u f f e r da ta counter , was i n i t i a t e d and the octa l mode was
established . When the computer was operating in the octa l mode ,
all 16—bit parame ters were assumed to be expressed as octa l
numbers . Such was the case with the da ta stored in buffers
“DAT 1” and “DAT 2 ” . For opera tiona l eff icienc y pur poses,only
the da ta in “DAT 2” was to be processed . Using a “ rdb ” comma nd,
one 16—b it bina ry character code was retrieved from “DAT 2” and
stored as the q u a n t i t y “D” . The logical c omplement of “D” was
determined. Due to the fact tha t the four least significant
• digits were of no significance , the da ta string was shifted
four bits to the right to discard the worthless da ta . As
arra nged, the least four bits represented the exponent to which
the number 2 was raised to finally formulate the dopple r

frequency in decimal.

--- - -5 - — —-— -5-- —- —— -- - - — —

~~~~ ~~
- - - - -- — 5 - - - — —-— 5--- — — - - -— — — 5 - - --
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
5- -— .5 — 5-——- - ‘-55-5-— ~~~~~~~~~~ •‘ 5-5-• 5-5-5- ” ••5-5-5- ‘ _
~~~~ ~ ~~ ~ ~~~
_
~~~ .
~~~
_•___ 5- - ---——5---- - 5-5- 5- ”- -~~~~~~~~~~~ ’ 5 - 5 - - —•~ •—- —•-‘-——-— - -
 ~rV~
- -- - ---5
-- - —-
~~~~~~~~~~ — —-5-- -
_ _ _ _ _ _ _ _ _ _ _

To isola te the exponential da ta from the remainder of the

da ta string, the cha racter code was compared to octa l number
17 in a logical “ and ” operation which in e f f e c t elimina ted the
uppe r twelve bits and saved the lower four bits which were
stored in the quantity “ r2 ” . The origina l da ta s t r i n g was n o t

destroyed by this “ ba nd ” comma nd and was used for further

processing . Now tha t the exponential term had been duplicated
and stored , the da ta string was then s h i f t e d again four bits
to the right to elimina te the exponent da ta from the da ta
string .
The low eight bits of this raw da ta string represented the
mantissa tha t corresponded to the exponent retrieved previously .
Both the octa l value s of the mantissa and the exponent were
converted to their decima l equivalents using the “ otd ” comma nd.
As spec ified in step 8, the decima l equiva lent of the mantissa
was multiplied by 2 raised to a power which was the decima l
equivalent of the exponential term . Once this had been completed ,
the decima l mode was established for the execution of the
rema inder of the program. The resulting number of the mantissa-
exponent recombina -tiori was further processed by a number described
in chapter 6 of Ref. 5 to convert the raw da ta to a doppler
frequency in hertz at which time it was also printed for
reference on the pa per tape .
• Up to this point, it was possible for both the noise compo—
nents and the actua l doppler frequency da ta to ha ve been acquired
in the buffers , man ipula ted and processed to obtain the resulting
frequenc ies. However , the next step was to compare the frequency

59

5-- -
--

~~~~~~~
— -- --5-—
——
— - —- 5- — - ---- -—-- -—---—-- -—- — -—---5— ---- - -- _ _ _ _ _ _ --5 _ -_ _ _ _
 — --—-5--—- — — - - - - - -- - ---- -—— - ‘ - - - - 5 - - - - -
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ —
~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-
with the bandpass; if the frequency did not f a l l w i t h i n the
limits of the ba ndpass , it was reset to zero so as to ignore
it with regards to further da ta reduction . Otherwise the
-
-

desired doppler frequency da ta was allowed to proceed and to

be accumula ted in q u a n t i t i e s “ r 4” and “V” . “ r)4” was de fined
as the summa tion of the inverse of each doppler frequency and
“V” was the summa tion of all doppler frequencies tha t were not
elimina ted by the ba ndpass up to a maximum number of 4. A loop
was set up between steps 5 and 15 which continued the above
acquisition , manipulat ion, an d process ing of da ta until “DA T
2 ” had relinguished all four of its 16-bit binary character
codes.
O nce all four da ta strings had b een r e t r i eve d an d p rocesse d ,

the program continued by de termining the average doppler

—
frequency “N” an d the sq uare of the var iance “ R” , otherwise
called the squa re of the turbulent intensity . The average
doppler frequency was trans ferred from the HP-9825 to the
HP—26~-1J4A Termina l for visua l display on its console and for
printing on its peripheral teletype . The “PLOT” sub rout ine was
then called to gra phically plot the flow velocity and bin number
on the HP-9862 plotter . When program execution was reassigned
to the ma in program , the program pr inted the avera ge d opp ler
frequency, the corresponding flow velocity, the squared
var iance multiplied by 100, and the number of valid doppler
frequenc ies received . A ll the parame ters “r 4” , “ rS ” , “M” , and
~
“V” were then in i t i a l i z e d and the subroutine “DA TA ” was called
to acquire new da ta . Main program execution then proceeded

60

----- — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -~~~~~ ~~~~ —~~~ •

-- -
-
— - --5-— -~~~~~~~~~~~~~~~~ -~~~~ —-5 -— - ~~~~~~~~~~~~~~~~ - -~~~~~~~~~~~~~~~~~~~ - -~~~~~~~~~ 5-~~~~~~ s-~~~~~~~~~ — - - ~~~ 5--— --- — -
 T
-
_ -- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
~~~~~

ba ck to step 5. The program continued until it was comma nded

to cease either by the opera tor or the HP-2].MX controlling
computer.

-
-
A2. SUBROUTINE “DATA ”
-

The “DATA ” subroutine controlled the actua l acquisition for

the program. The subroutine initially checked the sta tus
cond itions o f “DAT 1” or b u f f e r 1. If the sta tus of “DAT 1”
was zero , meaning tha t the b u f f e r was ready to accept new
data , 2 bytes of da ta were transferred from the counter to.
b u f f e r “DAT 1” . “BUFFER 1” was displayed on the visua l readout
of the HP—9825 to indica te tha t Buffer 1 or “DAT 1” had just
rece ived some new da ta . The sta tus condi tions of b u f fe r “DA T
1” were checked to determine if the b u f f e r was busy which was
signified by a — 1. If this were the case , a loop was set up
by which the program execution jumped to the line prior to this
sta tus check sta te’ient and returned to recheck the sta tus of
“DAT 1” . This continued uninterrupted unti
l the sta tus o f
“DAT 1” changed to some value other than -1. Once this occurre d,
the sta tus condition was checked again , b ut thi s time a sta tus
for “DAT 1” of 24 was being sought to indica te tha t the buffer
now held four pieces of da ta and was ready to have the da ta
processed using “ ri ” with a value 2 to indica te tha t it was
buffer 1 tha t was full vice buffer 2. Once buffer 1 or “DAT 1”
was full , then b u f f e r 2 ~r “DAT 2 ” was filled using the same
procedure and using the same sta tus checks as with buffe r 1 to
de termine the condition of b u f fe r 2. Once it had been confirmed

6i

L
— 5- - - -- —
— - - - ~~~~~
— - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — rn-- ~~~~~~~~ - -5 5-5- --- _______
 - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_ _ _ _ _

tha t “DAT 2 ” contained its four pieces of da ’ta , ” r l” was given

a value of’ 1; and program execution was returned to the main
program .

• A3 . SUBROUTINE “PLOT”
The “ PLOT” subroutine was utilized to plot the points on a
graph of velocity and bin number as they were being computed .
5-

It initially scaled the 8k” by 11” piece of pa per to the same

scale as the axis drawing program used (Table III) and selected
the size of labelling cha ra cters to be used even though
labelling was not to be accomplished during this sequence .
The subroutine then plotted the da ta point with the bin number
as its abscissa and the flow velocity as its ordina te . The
velocity was determined by multiplying the doppler frequency
by a constan t whi ch concurrently conver ted it to a ve loc i ty o f
dimensions me ters/sec and then to a velocity in feet/sec. The
bin number was upda ted to the next point to be ta ken e i t h e r by
-

-
~ adding 2 or 8 to its va lue depending on whe ther da ta was being
take n in the expected wa ke region or out of the wake region
respectively.
~tter ’ s pen was executed
The raising of the p].-
next and program execu tion was returned to the line foll ow ing
the g8b “PLOT” sta tement.

62

.
--5- - --

~~~~~~~~~~~~
 ________________________ - _ _ _ _ _ _ _

APPENDIX B
DEVELOPMENT OF TI~ HP-9825 PROGRAM

• Knowing tha t the doppler frequency tha t was indicative of

a particle as it traversed the laser fringe pattern was propor-
tiona l to its velocity , a proportionality constant or scale
f a c t o r was de termined by determining the components of the
following expression:

K -

2sin Q
_ _

where is the wa velength of the laser light, is half of the

conver gence ang le be tween the two b eams , and K is the scale
factor. The K was a property of the particular laser genera tor
being used. The angle was measured by projecting the laser
-
-
beams on a wa ll at some know n di stance from the foca l p oin t of
the two beams . The distance be tween the two laser projections
on the wa ll was measured and halved for e a c h beam separa tion.
This va lue was divided by the distance from the foca l point to
the wall and the inverse ta ngent of this r e s u l t was the h a l f -
angle, of the angle formed by the two beams .
Once the scale f a c t o r had been computed , the t h e o r e t ica l
• dopp ler frequency which corresponded to a particula r particle ’s
veloc ity was determined . This proce dure was used to de term ine
the expected doppler frequencies for the range of velocities
a ttainable by using the DISA rota ting disc which contained

63

5 - 5 - _
- ~~~- ~~~~~~~~~~~~~ — ~~

IT — ‘5-
- - —
~~~~~~~~~~~~~
 T
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~ T~ ~~~ -5-

- - - -
-
~~~~~~~

embedded particles of the size comp arable to those tha t would

be used in the seeding process. By focusing the two beams on
the disc a t d i f f e r e n t radia l positions , different tangentia l
velocities were measured as the ‘disc rota ted. The disc had

H the property of simulating seeding p a r t i c l e s a t velocities of

10 cm/sec to 70 cm/ sec.
This me thod was utilized to test the performance of the
counter in its a b i l i t y to a t t a i n the required raw da ta and the If-
HP-9825 software in its ability to convert the raw da ta into
doppler f r e q u e nc ie s . The expected doppler frequencies for -
-

various velocities and beam s e p a r a t i o n settings were compa red

with the actua l frequencies obta ined from the output of the
•
HP—9-825. Within the accura cy of placing the beams ’ foca l point
on the desired radial position of the disc tha t would resul t in
a particular velocity and of the disc rota ting at itS design
a n g u l a r v e l o c i t y , the HP-9825 program output proved to be

effec tive and accurate .

-5
~~~~~”
 -
5-5-
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~
•1 I

APPENDIX C
ABRASION TESTS
‘1
• Prior to performing the LDV experiments involving the
seeding of minute particles into the transonic compressor , the
question of whe ther the seeded particles would have any a brasive

i e ff e c t s on the aluminum compressor rotors needed to be examined.

-

This was accomplished by using the blow-down free jet ir. the
Turbine Lab a t the NPGS annex with the 24-inch diame ter converging
nozzle a ttached to attain art exit flow velocity a t or near a
Mach number of 1. The exit flow was used to simula te the flow
that could be expected at compressor rpm ’ s h igh enough to
create transonic flow on the rotor blades. In the wa ke of the
Jet , a polished plexiglass circular rod of 0.5 inch diame ter
was positioned to simula te an aluminum compressor rotor blade .
A soft ma terial such as plexiglass was used so tha t in the case
where a b ras ion was detec ted from the tests , then the relative
hardness of plexiglass and aluminum could be used to determine —

—
if any appreciable abrasion would also occur on the aluminum
blade edges and sur faces .
Three thirty minute runs were completed in the following
configura tions : 1) a plexiglass rod positioned five inches
from the nozzle exit with no seeding and no obstructions upstream ,
2) a plexiglass rod seven inches from the nozzle exit with
seeding added a t the center of the exit flow administered by a
0 .125 inch diame ter me ta l probe extending into the exit flow ,

- _ _
 and 3) a plexiglass rod seven Inches from the nozzle exit with
the meta l probe in place but with no seeding applied. The
resul ts of the tests were conclusive tha t rio observable abrasion
was ca used by the seeding . Some abrasion was noticed but since
it was common to a ll three tests , it was determined tha t this
abrasion was a t t r i b u t e d to the flow contaminants cha ra c t e r i s t i c
of the blow-down free jet appa ra tus . It was concluded tha t
seeding the flow a hea d o f the transon ic com pressor woul d no t
present an a b r a s i o n problem for the compressor b l a d e s .

~~~i
I

-
-
66
1 - -5—
_ _ _ _
— -——— ——-—-—,-- - —- —-5—
_ _ _ _ _ _ _ _
- ~~~~~~~~~~~~~~~~~~~~~~~~~
-•-—-
~~~~
_____________
—5-—5-———5-—--••--— - -- - — - 5 - - —
•? 4” ~~ - - -
_
-

-- -
~~~~~~~~~~~~~ ~~~~~~~~~~~~
-— — — — - - -— — -5-— -5-~~~~~~~~ -— • — — ~~~~~5-— -~~~~-- 5- — - ~~~~~~~~~~~~~~~~~ - - - — ~~~~~~~
—5--——•—
-
 - - --
‘ 5-5- — - - -- -- - r p . _ -

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

APPENDIX D
WINDOW REFRACTION CONSIDERA TIONS

Even though a plexiglass window was not used in the data

-

ac quisition phase of the experiment , its refract ive effect

upon the laser beams as they propa ga ted through it was inves-
tiga ted. To investiga te the flow velocities behind the
compressor rotor sect ion a t highe r rpm ’s than were allowed for
his exper imen t, it would be necessary to install a window to
t
prevent the intermixing of the compressor ’ s Interna l flow
characteristics and the atmospheric conditions externa l of the
compressor casing a rid still m a i n t a i n the a b i l i ty to ta ke flow
veloc ity measurements . In the investiga tion of the practicality
o f us ing a window, a small c y l i n d r i c a l plexiglass window was
utilized to provide the passagewa y to allow the laser light to
enter the turboma chin e and to permit the exiting of the back —
sca tter light signals. To be able to ma ke accura te a rid useful
ve loc ity measuremen ts, it was necessary to de termine where the
laser measuring volume was loca ted within the flow field . By
passing the converging laser beams through the plexiglass , the
resulting re fraction of the la ser beams was measured . It was
determined tha t for an estima ted index of refraction of 1.50,
the measuring volume should be loca ted at a point 0.16 inch
farther awa y from the inner casing w a l l.
This 0.16 inch d i f f e r e n c e was confirmed to be the actua l
foca l length d i f f e r e n c e due to refraction by setting up the

67

-- -
-- —----- --------—---— -5-—•————•-——~~~~~
-------— — ‘— ‘ - — 5 -5 - 5 —5-— ——--5- -- - -5- - — -
-
~ -5 - --
. ’,
-
~~~
-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~
— -- 5-’

laser a ppara tus with and w i t h o u t the window installed . ml -

tially, the measuring volume was positioned on the hub of the
compressor without the window installed. W hen the w ind ow was
installed, the two beam spots were observed to be sepa ra ted .
Using the calibra ted tra versing mechanism the laser system
was moved farther from the compressor at intervals of 0.04
inch. After eac h interva l ha d been traverse d , the position of
the beam spots was checked. A fter 0.16 inch had been traversed ,
the beam spots were observed , as accurate ly as p oss ib le , to
ha ve merged into one spot thereby confirming the c a l c u l a t i o n s
performed previously .

-
-
68

_ . 5 -
‘ T IIII I TTT
~
_ _ _ _ _

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 — 5- 5- -
“ ‘ ‘ 5 - ’ - - -- -
~~~ ~~~ ~~~~~~~~~ ~~~~~~~~~~~~~~~ ~~
- _ ~~~~~~~~~~ -
~~~~~~~~~~~~~~~~~~~~ ~~ ~~~ ~~~~~~ ~~--—-—-—

REFERENCES

1. Durat , F., Melling , A. and Whitelaw , J. H., Principles

and Practice of Laser Doppler Anemome try, Academic Press,
Inc., New York, 1976.
2. Compact Syringe Pump O pera ting Instructions , Harva rd
A ppara tus Company , Inc .

3. Type 55L77 Beam Expa nde r Instruc tion Sheet , DISA Informa tion
Department .
24. Instruction Manua l Mode l 3 q75 C o n s t a n t Output A tomizer , - -

Thermo-systems , Inc. ~

-
5. DISA 55L90 Counter Processor Instruction Ma nual, DISA
Informa tion Department .
- 6. DISA Type 556 Laser Doppler Anemome ter Mark II Instruction I
Ma nua l , DISA Informa tion De p a r t m e n t .
~

• 7. West, J. C., Digita l Programmable Timing Device for Fast

-
~

Resp onse Ins trumen ta tion in Rota ting Mac hi nes , Nava l
-

Postgradua te School M.S. Thesis, June 1976.

8. Winters , K. A . , Development of a Me thod for Measuring -

Velocity a t the Exit of a Compressor Rotor Using Kulite

Probes with Synchronized Sampling , Na va l Postgradua te -

School M.S. Thesis , March 1975.

69

, - -‘-n——- - :T— .
~~~~~~~~~~~~~~
-

1
- - -- - S - - -- - -

-- - 5 — - - — - 5 _
~~~~
_ - s~~~~ - —‘--5--- — ~~~~ &. S a. - a .4
-
 INITIAL DISTRIBUTION LIST
No. Copies

~~ .
Defense Documenta tion Center 2
Cameron Stat ion
Alexandria , Virginia 223l~4

2. i4brary, Code 0142 2

Nava l Postgradua te School
Monterey , California 939240
3. Department Chairma n, Code 67 1
Department of A e r on a u tI c s
Na va l Postgradua te School
Monterey, C a l i f o r n i a 939240
24. Professor D . J. Co llins , Code 67 Co 6
Depa r tment of A e r o n a u t i c s
Nava l Postgradua te School
Monterey, C a l i f o r n i a 93940

5. LT Jeffrey A. Harr ison , USN 1

110 Brownell C ircle
Monterey , Ca li fornia 93940
6. Mr . J. E. Hammer , Co de 67 1
Department of Aeronautics
Nava l Postgradua te School
Monterey, C a l i f o r n i a 93940
7. Turbo—Propulsion L a b o r a t o r y , Code 67 5
Nava l Pos tgra dua te Sc h oo l
Mon terey , California 93940
8. Dr . H. J. Mueller I
Na va l Air Systems Comma nd , Code 3lOA
Na vy Depar tment
Wash ington , D . C. 20361

70

- - _ _ _ _ _ _ - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ,——- --— —-—- 5-

- - - --