Ni-.
^.f
.'
MOKTE:.:
SGHOQTj
.;:A9394S-500S
J
NAVAL POSTGRADUATE SCHOOL
Monterey, California
THESIS
DEVELOPMENT OF A BOUNDARY LAYER
CONTROL DEVICE FOR TIP CLEARANCE EXPERIMENTS
IN AN AXIAL COMPRESSOR
1
by
Muliukur Tarigan
March 1988
Thesis Advisor
Approved for public release;
R
P.
Shreeve
distribution is unlimited.
T239283
REPORT DOCUMENTATION PAGE
RESTRICTIVE
MARKINGS
UNCLASSIFIED
'
c\ A^'-o=t
-
DISTRIBUTION AVAILABILITY OF REPORT
3
Approved for public release;
distribution is unlimited.
30.V\GRAD \G 5C-ED^.E
'
CNi
^E^CR' \^V3ER(S)
MONITORING ORGANIZATION REPORT NUMBER(S)
5
7a
Naval Postgraduate School
C
-2i;-E55
sraZPCoae)
Stare
ty
Monterey, California
7b
93943-5000
ADDRESS (Cfy
dude
andZlPCoae)
Stare
Security
'ERSC\A, A.--C=
State,
(C/fy,
and ZIP Code)
93943-5000
PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
TO
dassitication)
ADDRESS
Monterey, California
9
3c
NAME OF MONITORING ORGANIZATION
Naval Postgraduate School
SOURCE OF FUNDING NUMBERS
DEVELOPMENT OF A BOUNDARY LAYER CONTROL DEVICE FOR TIP
CLEARANCE EXPERIMENTS IN AN AXIAL COMPRESSOR.
S
TARIGAN, MULIUKUR
DATE OF REPORT
4
{Year,
Month, Day)
15
PAGE COUNT
Master's Thesis
3^ = = -£VE\*A3' \G'A'o\ The views expressed in this thesis are those of the author and do not
reflect the official policy or position of the Department of Defense or the United States
Government.
"8
SuBjECT TERMS {Continue on reverse
if
necessary
and
identify by block
number)
Boundary Layer Spires, Axial Compressor, Case-Wall
Boundary Layer Control
i
-3S" = AC"
Cornrue on reverse
if
necessary
and
identify by block
number)
boundA boundary layer control device was designed to change significantly the case-wall
The device was
ary layer thickness entering a large-scale, multistage axial compressor.
intended to double the boundary layer thickness in order to evaluate the influence of the
inlet boundary layer in controlled tip clearance experiments being conducted on the
compressor.
The boundary layer characteristics expected to be produced by the control
Kiel, cobra,
device were predicted empirically and experimental verification was required.
and impact probes were used in the experiments and pressures were recorded manually using
water manometers. The geometry of the boundary layer control device, an annular array of
spires, was derived from shapes developed for simulating the atmospheric boundary layer in
A significantly thicker boundary layer was measured
large rectangular section wind tunnels.
However, the results were interpreted and recommendain the compressor than was intended.
tions were made for geometry changes necessary to achieve the intended control for the tip
clearance investigation.
S-= 3.'
ON AvA ,AS
.\C.ASS- ED
-^N. V:"
OF ABSTRACT
SAME AS RP
21
n
D DTIC
22b TELEPHONE f/nc/ude Area Code)
/E O- 3ESP0NS BlE
1473,
:2c
a
1
ton may be used
until
exhausted
OFFICE
SYMBOL
67Sf
408/646-2593
SHREEVE
DDFORM
ABSTRACT SECURITY CLASSIFICATION
UNCLASSIFIED
USERS
SECURITY CLASSIFICATION OF THIS PAGE
otner editions are obsolete
UNCLASSIFIED
Approved for Public Pele^.se;
Development of
a
list ribut ion
is
unli-riite'i
noundarv T^aver Control Pevice
in an Axial Conpressor
tor Tip Clearance experiments
B.S.,
''luli'.ikur Tariqan
Major, Indonesian Mr Force
Indonesian Air Force Acade^iy,
1966
Subrnitted
in partial fultill^ient of the
requirements for the deqree of
MASTE]R OF SCIENCR
IT.'
AE^ONAfJTIC A L FNCU MFF PI ^IG
NAVAL POSTGPADUATF SCF^OOL
iMarch
1988
boundary
A
layer control device was designed
significantly the case-wall boundary layer
ina
was
a
to change
thickness enter-
large-scale, nultistaae axial compressor.
The device
intended to double the boundary layer thickness in order
to evaluate
the
influence of
the
inlet
boundary
layer in
controlled tin clearance exoer iments being conducted
coniDressor. The
boundary layer characteristics
on
the
expected to
be oroduced by the control device were nredicted empirically
and experimental
verification was
and impact probes were used
in
reguired.
were recorded manually using water manometers.
of
the boundary
layer control
Kiel,
cobra,
the experiments and pressures
The geometry
device, an annular
array of
soires, was derived from shapes developed for simulating the
atmospheric boundary layer in large rectangular section wind
tunnels. A significantly thicker boundary layer was measured
in
the compressor than was
intended.
However,
were interpreted and recommendations were made
changes necessary
tin clearance
to achieve
investigation.
the results
for geometry
the intended control
for the
r
r^'TRODUCTinM
.
I
CO'^^PPES^OR A^tD OPrpATTpM
IT.
^.
rx
ITT
.
B.
IV.
r-;M[->nj^c,.,.^p
p-qpoi
p.oriMpapY
A.
>^
r«ip;N'T'xTT^'^:
^
aH
r^
nA^^
r^..-^ r
,-_
p r-y j
n ^f
LAVFP CO^TPOL nrviCF
r^Fqrc-i
i^:tfnt
i.
Re'7;.'
2.
Oesiqn
ir^nent.s
s<
Aporoach
CONSTRUCTTO^
TEST PROGRAM!
1
2]^
22
26
28
28
B.
STATIC PRESSURE DISTRIBUTION MEASUREMENTS
31
C.
BOUNDARY LAYER flEASUPEflENTS
31
ANALYSIS AND DISCUSSION
3g
A.
CALCULATION OF DISPLACEMENT THICKNESS
36
B.
COMPARISON WITH DESIGN INTENT
39
C.
VI.
]_
9
21
PRELiniMAPY TEST
V.
ji
1.
Poundary Layer Thickness
2.
Profile Shape
3.
Other Considerations
ACHIEVING THE REQUIRED CONTROL
CONCLUSIONS AND RECOMMENDATIONS
39
4-7
47
48
5O
LIST OF REFERENCES
BIBLIOGRAPHY
INITIAL DISTRIBUTION LIST
54
jj ,
r-
-
'
'
:
r.CHOOL
LIST OF TAPLRS
I.
Dimensions of Standard Hal^-ridth Snires
24
II.
Throttle Element ^rranqenents
29
III.
Test Proqraf^ Sui^nary
30
IV.
Pressure Dist r i^^ut ion Data for
Conf ioura t ion 1
33
V.
Pressure Distribution Data for
Configuration 2
34
VI.
Intenrated Velocity Profile Parameters
for Conf iourat ion 1
40
VII.
Integrated Velocity Profile ParaTieters
for Configuration 2
41
LIST OF FIGHRES
1.
Compressor Test Rio
2.
View of the Inlet Belli^outh Without
With (b) a Protective Mesh Screen
3.
(a)
and
Conoressor Bladinq
a.
Flow Path View
7
b.
Stage Tip Section
8
4.
Velocity Oiaqrar^
10
5.
View of the Compressor Test Riq
11
6.
Flectric Motor Drive
12
7.
Throttle Showing Removable Elements
13
8.
Arrangement of the Instrumentation
15
9
Probes
10.
11.
12.
a.
Probe Tip Geometries
16
b.
Views of Probe Tips and Actuators
17
Meriam VJater Micromanometers
a.
Connections to the Probes
19
b.
View of the Instruments
20
Compressor Throttle Section
a.
Element Positions
23
b.
Spire and Plate Elements
23
Spire Element
a.
Circular Arrangement as Built
25
b.
Linear Arrangement as Designed
25
13.
Soire Geometry
a.
As Oesianed
27
t:..
As Built
27
Static Pressure Distribution Ahead of the TGV's
- P-.^ = 7.25 Inches of ^^ater
at =,-
32
15.
Boundary Laver Profiles with Screens
Configuration 1) and Spires (Configuration
35
16.
Deter-ninat ion of the Displacement Thickness
For Configuration 1
14.
17.
2)
42
Determination of the Displacement Thickness
For Configuration
7
43
18.
Comparison of Boundary Layers With and
w'ithout Spires
44
19.
Streamline Location for Uniform Inviscid Flow
45
LIST OF SYMBOLS
p
Cj_^
-
Density
-
Center line
-
Static oressure
-
Static oressure at the case-wall
p^
-
Total pressure
P^
-
Total pressure at the center line
P
-
Radius of the pipe
Local radial displacement
Pg
i
PcV]
r
-
D
-
Diameter
V
-
Velocity
-
Absolute velocity into the rotor blade
-
Absolute velocity out of the rotor blade
~
Reference velocity derived from measurements
-
Velocity at the center line
VI
V2
Voo
(Eq.(2))
V^
-
Local velocity
Wi
-
Relative velocity into the rotor
W2
~
Relative velocity out of the rotor
U
-
Wheel speed
s
-
Spire height
h
-
Compressor annulus height
A
-
Displacement of uniform-flow, inviscid streamline
passing through the spire tip, from the case-wall at
the compressor inlet
V-[
LIST OF SY"POL?
(continueri
)
Boundary layer t^'ijckness
'boundary layer disnlace^ent
~
i -^
: 1
>^
a ra
Value of V/Voo
"'.adial
Pot'^r
thickness
nce
^-t
t*~.e
di snlacer-ent
edne of the boundary layer
in'.;ards
fron the case-wall
inlet relative airflow anqle wit*^ respect to
the axis
Potor outl'^t relative airflow anole
the axis
v/ith
resnect to
"Jotor inlet absolute airflcDw anqle with respect to
the axis
xOtor outlet absolute airflow anqle with respect to
the axis
ACKr'Ov.'LFnC'fi^IT
The
1.
aijt>nr wnuld
nrnfv-.c;snr
R.
i^
li>e to
sion Laboratory,
California and
iJav/al
hi?=
assistance
and
Postoraduate School, "onterey,
staff,
guidance
for their
director of the Turboproou 1-
^^^r^ove,
.
in
Tan Moyle and Thaddeus ^est,
project.
this
benefit
the
fnllov/inM neo^l-:
the
t}-\3n\<
of
VJithout
their
their experience,
T
could not have completed this thesis.
2.
The NTS Photo Division for
their effort
tion on the photoqranhs repaired for
3.
My wife,
and
-ny
Enny Sitepu;
dauphters,
"iy
Mena
sons,
and
rny
and coopera-
thesis.
Suranta and Ferianta;
Sabarina,
sunported ny spirit durinq ny studies.
who
alwavs
P2TP0D^CTT0^'
I.
Axial connressors in aircraft engines are
nearly unifor"^ inlet
boundary layers, but
ranoe of
t'">.at
flow/
nenerally ie-
condition
^/it'-^
are reauired to ooer=ite stablv
inlet flow distortion.
thin
over
a
found in
oractice
-'ulti-staoe core connressors, once develooed,
are tv^en
upry sensitive
anv change
to
the tio-cao result
In view of
in
the
in
rotor blades and the case-wall.
ciency and
It
tin-qap
in
to develop
the need
changes in
compressor effi-
the
loss of stability narqin in
a
betv;een the
In particular,
re'-'uction
a
is
distorted inflov/.
military
aircraft enaines
with improved thrust-to-weiaht ratio and increased tolerance
to distortion,
attention
understand ina the
flow
in.
has been liven in recent
effects of
axial compressors.
tip clearance
A
years to
chanqe
on
the
larqe scale, multistaae axial
conoressor facility, which would allow experimental investigations
to be
made under well-controlled
test conditions,
was installed and prepared for that purpose.
(Re^^.
The first experiment to be conducted in the
one in
which
the tip
clearance will
be
1)
facility
is
increased syste-
m.atically and data will be obtained from instrumentation de-
signed to
flov/
be
field.
limited
resolve
the
important changes
However, the
to the
in
the
internal
results of this experiment would
particular
compressor-face
inlet
flow
conditions which the test installation qenerated.
a
priori,
is
case-wall boundary
importance of the inlet
the
layer profile,
and thickness,
in
ly
The purpose
nap.
Thus,
the boundary layer as
important to chanqe
of
the physical
relation to
gao between the rotor and the case-wall.
clearance
fJnknown,
equal-
is
it
channe the
to
oresent study was
the
to
control the
develop an experimental technique to be used to
boundary layer thickness entering the compressor.
The approach was
removable element
as a
in
a
installed
"spires"
throttle housing in
and design
The presence,
duct.
use an array of
to
the
the throttle,
of
inlet
and the
difficulty of alternate approaches such as providing suction
on the scale
reguired,
that is reported here,
and
a
suggested this solution.
Tn the work
the design of
is
the spires
reviewed
program of measurements to evaluate the effect
spires on the compressor inlet case-wall boundary
reported.
The m.easurements
of
layer,
the
is
showed that the boundary layer
with the spires was thicker than was intended, and
that the
velocity profile was unacceptably distorted.
It was concluded
that compromises which were made
manufacture of the spires, which involved
slender
Also,
cusped shapes with linear cuts,
the installation
within
a
round
qeometry within
data on
of
duct was
a
which the
the
design was
spires as
approximating the
circular array
a
from
tunnel that had
based.
the
were unacceptable.
quite different
larqe wind
in
the linear
provided the
Recommendations were
made for
changes to be
nade to
the soires
goal of doublinq the inlet boundarv layer
changing the shane of the velocity profile.
to
achieve the
thickness without
II
COMPRESSOR AND QPIJRATIOH
.
The compressor
and details of
test facility,
collection method,
data
instrumentation used are '^iscussed
the
in
the
fnllowina sections.
\.
COMPRESSOR
compressor
three-staqe axial
The
with
assembled
desiqned
a
new
to serve
as
a
research
tool for
bladinq, was
variety
a
The facility is shown in Fiqure
experiments.
presently
facility,
synnetrical
desiqn of
of
1.
Two (iifferent sizes of inlet bellmouths were orovided in
measure flow
adjust the pressure differential and
order to
rate to the com.pressor satisfactorily for various staqe con-
fiqurations and operatinq speeds.
larqer bellmouth was used (Ref.
with and without
a
protective
are shown in Fiqure 2.
enters the
throuqh
(Ref.
20-foot lonq
a
1)
3.
mesh screen over
in
the
the
duct containinq
a
inlet,
the airflow
the fiqure,
outside the buildinq
and flows
throttle device
bladinq
the test
inlet quide vanes
first staqe
is
shown
in
the
two
The bladinq is removable and adjustable.
present experiments,
row of
the present work,
Views of the bellmouth,
to the compressor test section.
The compressor
Fiqure
As seen
compressor from
In
2).
in
Fiaure
3
(IGV),
section was bladed
two symmetrical
was removed),
parts of
For the
with one
staqes (the
and one row
of exit
SCO
>=>!
tl=^;
V
J)
u o
^
CJ
u
a
iV
view ne thp Inlet Rellnouth Without (a)
and I'ith (b) 3 Protective riesh Screen
XvXvXv
^ioure
3.
(Continued) Conore!=;sor Tlaciim
(b)
Stane Tin Section
vanes
aui'ie
(EGV).
Each staqe consisted of 30 rotor blades
and 32 stator blades.
The function of the TGV
the radial distribution
of
piaure
is
4,
sv^metrical
the
3,
belt drive
is
radius
(Both the
oroduce
equal to the
an
driven by
ria
a
150
ninimum loss
two-dimensional compressor
to the compressor
is
shown
Figure
in
a
The speed of the
(Fiaure 6).
RP^I
The low speed drive
by
(1610
was used in the present experiment.
The throttle,
located approximately mid-way
inlet bellmouth and the compressor face
spaces
(8
useable)
throttle plates.
7.
The
5.
HP electric motor coupled by
ch?,nqind the belt drive pulleys.
10
At
incidence
compressor can be chanoed nominally from 1600 to 2200
PPM)
in
axial
and 5).
conpressor test
compressor
is
vary with radius.)
designed to
rotor blade
4,
radii.
all
for the correspondinc
cascade (Refs.
The
at
by the
stage design
shown at one
flow angles
TGV is
with resnect to the
incidence
sypnetric
^
velocity diaaran,
velocity comnonent and
each radius,
turn the flow to produce
flow anole renuired
the
rotor in the svr^^retric staqe.
one in which the
to
is
s
The
compressor
throttle
be
the
throttle
chanqed only
is
by
between the
1),
for inserting different
A view of
can
(Fiq.
contained
screens and
shown
in
Figure
stoppinq
the
Symmetric Blading Requires a^=62 & ot2=6l at All Radial Stations.
Figure 4.
Velocity niaqran
10
11
12
13
the connections between
All
the
inlet bell'^outh and the
comnressor were sealed to eli'ninate leaks that niqht effect
the experimental
B.
results.
TNSTRUMF:MTATIO^] AND DATA C0LLF:CT I ^)ri
tlany
The arrangement of
conoressor case.
the present experiments
of
the
is
IGV
8.
pressure distributions
In
Cobra probe were
airflow.
The Kiel probe was
surveys
always
probe was
that the
centerline of the duct.
and the
facing the
installed and held fixed on the
Pressure distributions in the probe
were referenced
to corresponding
values
pressure on the centerline and static pressure at
wall.
9a and
pressure ports on each side of the
connected to the U-tube manometer
such
static
same location (Figure
in the
the latter case,
probe rotated
travers-
a
used similarly to mea-
A cobra probe was
pressure and, by prior calibration,
sure stagnation
b).
the com-
Stagnation pressure
pressor periphery shown in the figure.
impact probe.
inches in
were located 1.0
distributions from tip to hub were m.easured using
ing
in
static
VJali
the selected stations around
at
s
instrumentation used
shown in Figure
survey stations
taps and probe
frorit
along the
oorts for probes are orovided around and
of
total
the case-
The differences between probe total pressure and wall
static pressure
and between
probe
were
pressure
adjusting
twr
'^eriam
probe total pressure
recorded
manually
micromanometers
.
after
and Kiel
carefully
The connections
to
Traverse Probe (Impact Probe)
Center Line Probe (Kiel)
Case-Wall
Static Pressure Tap
Figure
8.
<
Arrangement of the Instrumentation
15
VIEW
A - A
VIEW
B
c=;'^-,
-
,032'*
B
•
A
I
A.
Cobra Probe
VT'
VIEW
C.
Fiqure
(a)
Kiel Probe
9.
Probes
Probe Tip Geometries
16
A - A
^i
Figure
(b)
9.
(Continued)
Probes
Views of Probe Tins and Actuators
and
a
view of the nicromanoneters are shown in Fiaure 10. (a)
and Fiqure 10. (b),
The scales
respectively.
of
the
rieria-ii
manometers were adjusted to zero jefore each experiment.
meter to measure
the
counter and timing
the compressor.
ture
were
resi^ec
t i
vely
total temperature
inlet
using
barometer
a
located insid
gible and no correction
tions
-he
not
and
a
oulse
speed of
building.
required
thermometer
and
Ourina each test
jre was observed to
for change
in
watermanometers was required.
v/ere
digital thermo-
wheel to record the rotational
the change in ambient tempe.
Meriam
a
The local atmospheric pressure and temoera-
measured
,
included
instrumentation
Auxiliary
due to
Air density
the use
velocity derived from the centerline probe.
18
be
water density
of
a
negliin
the
correc-
reference
Kiel Prohe
Static Pressure Tap
Figure 10,
(a)
rieriam V/ater Micromanorneters
Connections to the Probes
19
'^iaure
10.
(Continued)
(b)
Heriam Water iMicromanoneters
View of the Instruments
20
Ill
QOUNDARY LAYE^ CONTROL DEVICE
,
The boun^'^arv
layer control
device
described
is
followinq section and the desi-in intent, actual
in
the
desiqn,
and
construction are discussed.
DESITAT
A.
1
INTE\'T
Requirements and Anoroach
.
boundary
''lor^al
layer conditions at
known to be
5
=
1.1
inches
dary layer thickness
and
5*
=
.13
layer control device.
Either by
could be reduced.
In
the
screens, were
inches.
The boun-
addim
could be chanqed by
wall, or by coolinq the case-wall,
ness
the compressor
from earlier experiments with uniform
inlet,
suckinq air
laver thick-
planned experiment,
required, to increase the boundary laver thickness
factor of two and obtain measurements with
of
the
tip clearance
in
the case-
at
the boundary
boundary
a
it
(5*)
was
by a
different values
the compressor.
Thus,
it
would be
possible to obtain data at two values of tip clearance while
holding
the
ratio
of
boundary to
clearance
tip
[6*/t]
constant.
To increase
elements
reiected.
readily
6*
by a factor of
and graduated
The
for
rouqhness
tests
with
case-wall rouqhness
two,
screens were
elements
both
could
thin boundary
21
considered, but
not be
layers.
removed
Previous
experience with
suqqested that
graded screens
to produce axi-symmet
difficult
(graduated screen
uniformity with
r ic
2
installed
blank throttle elements as shov/n in Fiaure
the
one
in
11.
Pes iqn
.
The system
produce
spires
of
boundary
thickened
particularly
tunnels,
in
has been
layers
experiments
previously to
used
rectangular
in
simulation of the atmospheric boundary
7).
In
Reference
Table
I
produced
were given for
a
used in the
for the
a
to account
configuration and annular
The spire
flow con-
geometry and total number
spires were selected using the information in Reference
for experiments carried out
in
a
wind tunnel.
spire configuration was chosen for the
to avoid possible
in
These data were
but were corrected
present design,
traction (Figure 12).
the data
non-dimensional spire qeometry that
known boundary layer thickness.
circular duct
the
layer (Ref.
based on many experiments,
7,
wind
required
which
careful
of
larqe
a
located so far from the compressor face.
The approach taken was to use "spires"
of
would be
it
A
7
29-element
circular arrangement
resonance due to the wakes from the spires
interacting with the
30
blades of the rotor.
1.5"
1
2
3
A
3.
^le-^ent Positions
b.
Fioure 11.
5
6
7
9
10
Spire and Plate Eler^ents
Conpressor Throttle Section
23
TAT
DI^E^ISIONS OF stan:.
HEIGHT
:
-^.D
I,
HALF--WIDTH SPIRES
SPIRE WIDTH
0.5000
0.0016
0.3750
0.0111
0.3100
0.0167
0.29 3 3
0.0250
0.2733
0.04 17
0.2483
0.0833
0.2100
0.12 50
0.1850
0.1667
0.1650
0.2500
0.1350
0.3333
0.1117
0.4167
0.0917
0.5000
0.0750
0.5833
0.0600
0.6667
0.0450
0.7500
0.0333
0.8333
0.0217
0.9167
0.0117
1.0000
0.0
a.
Circular ^rranaenent as Built
)
Linear Arranqer-ent as Designed
.
FIGUPF 12.
Spire Elenent
25
B.
CONSTRUCTION
qeometry (Fiqure 13(a) could not
The spire
and therefore,
locally,
linear cuts
as
from aluminum
in
Figure 13(h).
alloy,
.25
inches thick,
were arranged uniformly in
attached
tav
three flush,
the
of
the
rim
the screws were
throttle
screen was
ring
rina.
a
aooroxinated
was
the j^hape
shown
he nachine^
The shane
and
the
with
was cut
29-elements
Fach snire was
circular ring.
countersunk Phillips-head screws to
Mo locking
wire was necessary since
prevented from backing out hy
installed in
the throttle
positioned dov/nstrean
of
housing,
the spires
any possibility of dam.aae to the compressor.
the adjacent
and
a
to prevent
(a)
As Desiqned
Piqure 13.
(b)
Spire Geometry
27
As
Built
Tf^ST
IV.
program consisted
test
The
PROGRAM
exploratory flow
of
field neasure^ents
detailed boundary layer surveys to
the snires on
the effect of
the compressor
A.
nunber
a
initial,
of
and then
series of
a
establish quantitatively
the boundary
thickness at
layer
face.
PRELIMI^TARY TEST
conductinq
^^efore
establish an
conducted to
elements
detailed flow
the throttle
in
one which
open throttle
stall and
It was
was
and spire
by
selected
a
The desired through-
removed
was well
boundaries on the
from both
compressor map.
found during this procedure that the position
spire element in
were
similar average
a
that produced
combination of screens only (Ref. 8).
flow condition
tests
arrangement of screen
which produced
through-flow to
mass rate of
surveys,
the throttle housing was
of
important.
the
Less
stable compressor operation was found when the spire element
was the
tests,
most downstream
the
element.
two arrangements
As
a
summarized in
Table
selected, and the test program summarized in Table
carried out.
28
these
result of
II
Illf
were
was
THROTTLE ELEMENT
LOCATIOM
UPSTREAM
co^'^^iGijRArrox
(1)
screen: o^:ly
SC^EE\
C0>:
FIGURATION
(2)
SCRFEM
&
il
SCREEN *3
SCREEN #5
SCREEN ^1
29
|
I
S^IRE
TARLR iir
1
TEST PROGRz\M S(U1MARY
1
TFIPOTTLE ELEMENT
TEST
CONFIGURATION
SURVEYS
1
1
COBRA PROBE
2
1
IMPACT PROBE
&
KIEL PROBE
3
2
IMPACT PROBE
&
KIEL PROBE
KIEL PROBE
&
STATIC PRESSURE DISTRISUTION flEASURETlEMTS
B.
In Test
tion
a
1,
establish
with the throttle elements
1,
survey
was conducted
outer to inner case-wall.
The procedure
confiquraorobe to
cobra
oressure
distribution of static
the radial
was to measure
using the
in
in
fror^
this experiment
"indicated" static pressure trom the side
the
holes of the cobra
orobe as
step-by-step from
was moved
it
the outer wall to the inner.
The relationship of the cobra
probe "indicated" static pressure was established
usinq the
case-wall static pressure and cobra orobe measurements adjaThe survey measurements were individually
cent to the wall.
referenced to
total
pressure from
probe
the Kiel
at
the
center of the axial line A-A (Figure 8).
so obtained
The static pressure
It
is
seen
to decrease linearly
is
from,
shown in Figure 14.
the wall
(Pc
)
to
the
hub surface.
C.
BOUNDARY LAYER MEASUREMENTS
In
Tests
2
and
3,
surveys
were made using
orobe from the outer case wall to where the
became
constant.
almost
(configuration
configuration
in
15,
of
1)
2
and test
(Table II).
Test
3
2
was
with
was with the spire
The results
the impact
impact pressure
screens only
element
in
are shown plotted
Pigure 15 and listed in Table IV and Table V.
In
Figure
the value of the velocity measured on the center line of
the annulus has been used to normalize the profiles.
s^j
Figure 14.
(Inches of VJater)
Static Pressure Distribution Ahead of the Inlet
at P^
-Psw = 7.25 inches of \Jater
Gui(ie Vanes
TABLE IV.
PRESSURE DISTRIBUTION DATA FOR CONFIGURATION
Ri = RADIAL DISPL. INWARDS FROM CASE-WALL,
PRESSURES ARE IN INCHES WATER
V/Vp
^L
-^1
1
V/Voo
0.01
7.100
-.060
7.040
-.0902
.0923
0.02
7.092
-.318
7.058
-.0656
.0671
0.1
3.410
3.6466
7.075
.7018
.7179
0.2
2.3195
4.9920
7.197
.8142
.8328
0.3
1.539
5.5672
7.073
.8673
.8872
0.4
1.078
6.0044
7.038
.9030
.9237
.9264
.9467
.9814
0.5
.755
6.3286
7.048
0.7
.38 8
6.7828
7.043
.Q594
0.9
.185
6.9750
7.060
.9717
.9948
1.1
.010
7.1472
7.125
.9792
1.0016
1.5
-.010
7.3271
7.065
.9798
1.0018
2.0
-.0465
7.2882
7.061
.9980
1.0184
2.5
.0
7.3402
7.6500
.9987
1.0187
3.0
.065
7.3518
7.0835
.9991
1.0188
3.4
.071
7.3967
7.1890
.9915
1.0143
3.6
.070
7.4400
7.110
1.0000
1.0299
3.8
.101
7.4702
7.149
1
.0002
1.0225
4.2
.009
7.5859
7.1590
1.0068
1.0293
4.5
.006
7.6200
7.126
1.0109
1.0340
5.8
.091
7.7524
7.199
1.0015
1.0337
6.0
.012
7.6776
7.117
1.0153
1.0386
•
1
1
PRESSURE DISTRIBUTION DATA FOR CONFIGURATION
R^ = PADIAL DISPL. INV7ARDS FROfl CASE-WALL,
PRESSURES ARE IN INCHES WATER
V/Vc^
^c,
Ri
0.01
7.009
.02
-.2179
--w
7.216
v/v»
-.1710
-.1737
6.690
.4292
7.1170
.2419
.2456
0.1
3.800
3.2621
7.051
.6698
.6802
0.2
2.634
4.4082
7.0850
.7768
.7887
0.3
2.090
5.0133
7.07
.8292
.8421
0.4
1.848
5.4029
7.207
.8529
.8658
0.5
1
.84
5.3526
7.191
.8495
.8627
0.7
1.733
5.5393
7.195
.8614
.8748
0.9
1.497
5.7915
7.189
.8839
.8976
1.1
1.397
5.7872
7.062
.8915
.9053
1.5
1.373
6.1133
7.193
.9078
.9219
2.0
1.000
6.36 3 2
7.141
.9295
.9439
2.5
.915
6.3938
7.123
.9308
.9474
3.0
.484
6.7183
7.136
.9555
.9703
3.4
.032
6.8056
7.157
.9891
.9799
3.6
.156
7.4020
7.178
1.0000
1.0155
3.8
-.117
7.5492
7.264
1.0000
1.0194
4.2
-.227
7.9146
7.221
1.0310
1.0469
4.5
-.911
8.1556
7.098
1.0556
1.0719
5.8
-.411
8.1954
7.143
1
.0550
1.0771
6.0
-.337
8.1086
7.105
1.0520
1.0654
34
BOUNDARY LAYER
.
.
-«^0.1 0.0
.
0.1
M
OJ
l
I
0.9 0.4 0.5 0.6 0.7 0.8 0.9 1.0
1.1
IJ
1.9
V/V(1NF)
Screens
Figure 15. Boundary Layer Profiles With (Conf iqurat ion
(Confinuration 1) and Spires
^
35
2)
ANALYSIS AND DISCUSS lOM
V.
report describes
section of the
This
to
the neasurenents
results of
in
the method
thickness,
boundary layer
calculate the
used
discusses the
cornparison with
the desiqn
expectation, and sunimarizes the results.
A.
DISP LACEP1ENT THICKNESS
CALCrjLA.TK^M OF
The displacenent
the present v/ork was
Pgs.
=
+
1/2
(Pt-
)
9)
configuration
and for conf iqura t ion
1
in
The nethod was as follows:
qives, at any
rad ial stat ion
.
(1)
based on total pressure
Va,f
static pressure at the wall,
and
2
P^i2
The reference velocity,
centerline
.13
inches.
.53
Bernoulli's equation (Ref.
Ptt.
obtained for
thicl-'ness
from earlier experiments waS
the
at
(P^
)
is
^'..^
^Cl
given by
Pt
=
Cr,
Ps
^Cr
+
1/2
Using the results given
static pressure
pVco
-
Pc
=
7.25
(2)
Figure 14, which shows
in
dropped linearly
from the tip to the hub at
P,.
^
a
to -0,8
distance of
inches of water,
36
h =
inches
7.2
that the
of water
inches,
at
—
0.11
=
]
(7-4] =
n.ni53Ri
(3)
^Cl
Usinq
equations (1),
each ooint
(R^)
and
(2)
the
(3),
for configuration
(2)
is
ratio of velocitv at
given,
in
the static pressure distribution for configuration
+
terms of
1,
by
0.0153Ri
(4)
"sw
^tc,
and the centerline velocity by
VCL
=
Using
/
Equation
1
+
(4)
(5)
0.0153Ri
and
Equation (5),
the
ratio
of
velocity at the centerline
velocity at each point to the
the
is
given by
(6)
37
From experiment
-
P»-
and
P^
P^ ^i
W
^Cl
i
The
Location from the case-wall to the hub.
configuration
profiles for
shown
and
(
Figure
in
6
)
boundary layer
conf iaurat ion
and
1
are known at each
P_
which are
2
were obtained using Eguations
15
(5),
(4),
.
The boundary layer displacement thickness was deternined
from boundary
layer profile data before normalising
passane centerline
displacement
(y)
of
Referring
velocity.
to
the
the following
and area A-2 were calculated
the area A-1
sketch,
to
for each
the probe from the case-wall using
-
if
A-1
=
A-2
then y
v/v„
x-y
-J^d
-
At - /^(i -
where
X
total area
is
T7^) dy
^)
^y
(6)
=
^-1
the "free-stream" value of
under the
boundary
38
layer
(7)
V/V„
and A^ is the
profile.
The
areas
described by equations
tion of
y
to
were plotted as
a
funcThe
(7)
results for configuration
data listed
and
determine the dispacement thickness (6*).
1
Table VI.
in
(8)
are shown
Fiqure 16, with the
in
The results
for conf iqurat ion
2
are shown in Fiqure 17, with the data listed in Table VII.
R.
CO^'lPARISO^I
1
.
WITH DESIGN INTENT
Boundary Layer Thickness
The
effect
layer thickness
illustrates the
boundary
of
the
spire element
clearly to increase
is
effect
of
layer profile
on
and
5
6*
spire element
the
and the
boundary
the boundary
layer thickness.
the
layer can
qeometry of
a
combined
However,
chanqinq the shape of the
results
shown in
Fiqure 18,
layer thickness
with the
spire element
the
with
the qoal
by
the
is
doubling of the overall and displacement thick-
nesses without
in
inches.
to 0.53
clearly be controlled
the spire,
selection and position of the screens.
to produce
4.5
thickness was increased from 0.13
The boundary
inches.
selecting
18
both the
on
The overall thickness was increased from 1.1 to
The displacem.ent
Figure
.
the
profile. Clearly,
overall boundary
exceeds
the half-
heiqht of the compressor annulus.
We first examine how far the effect of the
be expected to extend when the
compressor annulus.
the
streamline
from
spire should
flow from the duct enters the
Referring to Figure 19, the position of
the
tip
of
the
spire,
when
it
TABLE VI.
1
1
INTEGRATED VELOCITY PROFILE PARAMETERS EOR
CONFIGURATION 1
|
y
Ri
/
(1-V/V„)dy
0.0
1
x-y-J (1-V/V„)dy
o
o
A^-/ (1-V/V<„)dy
o
.1473
0.0
0.0
0.1
.0461
.0539
.1012
0.2
.0703
0.3
'
.1297
.0770
2
.2138
.0611
7
.3023
.0483
0.5
:
.39 38
.0411
0.7
.1177
.5823
.0296
0.9
.1246
.7754
.0227
1.1
.1295
.8705
.0178
1.5
'"7
...._
""7
0.4
.
2.0
/
/
/
/
2.5
/
/
3.0
/
/
/
/
/
3.4
1
1
1
3.6
.1473
3.4527
0.0
3.8
.1473
3.6527
0.0
4.2
.1473
4.0527
0.0
4.5
.1473
4.3577
0.0
5.8
.1473
5.6527
0.0
6.0
.1473
5.8537
0.0
1
1
TABLE VII.
1
1
1
INTEGRATED VELOCITY PROFILE PARAMETERS FOR
CO^JFIGURATION
Y
Ri
/
y
(l-V/V„)dY
o
0.0
2
Y
x-y-J (1-V/V„)dy
0.0
A^-/ (1-V/V„)dv
o
o
1
.5716
0.0
0.1
.0536
.0470
.5130
0.2
.0901
.1210
.4815
0.3
.1140
.20 27
.4576
0.4
.1343
.3880
.4473
0.5
.1537
.4741
.4179
0.7
.1933
.5456
.3783
0.9
.2262
.7238
.3454
1.1
.2580
.9032
.3136
_____
""1
1.5
2.0
1
/
/
1
/
2.5
/
1
/
/
3.0
1
/
/
3.4
I
/
/
1
1
3.6
/
3.8
/
/
1
/
1
1
1
4.2
/_
/
1
4.5
.5716
4.1786
0.0
5.8
.5716
5.5501
0.0
6.0
.5716
5.7620
0.0
1
41
LEGEND
XY-lNT(l-V/y(INF))pY
0.6
o.e
X
Fiqure 16.
Determination of the Disolacement
Thickness for Configuration 1
42
Fiqure 17.
Determination of the Displacement
Thickness for Conf iqurat ion 2
43
•
o
1
3-
•
!
3-
1
•
3-
•i
P
51
*3
i
LEGEND
lO
•i-
M-
9
3-
i t
O
5* (Config. 2)
p
C[_/
3-
6* (Config.
o
-0^0.1 0.0
0.1
ojt o.a 0.4 o.s o.a 0.7 o.a o.o 1.0
1.1
u
i.a
1.4
1.9
V/V(INF)
Fic]urtj
la.
Comparison of Boundary Layers
With and Without Spires
44
1)
the conpressor
approaches
blading can be
calculated
usin-i
the continuity' equation,
pAV
constant
=
assumino the airflow
(
unizorm and axial at
is
Then the streamline displacement
is
both stations.
compressor face
the
at
(A)
9
niven by
Ptt(r2
-
(P.-s)^)
^
pTT(p2
-
(R
-
A)^)
pTr(R2
_
(r
_
h) 2)
defined
in
Figure 19.
(10)
PTT
r2
where the notation
the
spire,
=
S
8
is
inches,
height at station
2,
eguation (10),
4.56
iment
in
Figure
from the
inches
results given
of spires
in
Ref.
the airflow extends
7,
height.
The difference
tunnel results
layer in
to approximately
comparison,
In
the effect of a
the
of
applied in
may be
spire height from
between the
the
the
in
linear arrangement
the test section
arriving at
due to
data in
the circular
of
six spire heights downstream,
approximately only 0.8
This factor was
the annulus
and
Using these
the effect
case-wall.
on the boundary
large wind tunnel,
inches,
The height
The results from the expp^-
inches.
show that
18
18
inches.
7.2
h =
arrangement on
spire
4.5
A =
R =
the
of a
extended to
the wall.
present spire
present and the
wind
significantly different
blockage presented by the spires to the pipe flow, or may
result from the bluntness of the present shapes.
46
2
.
Profile Shane
conoaring
In
the two
sir-nificant difference
noted.
Pirst,
departure
away
fron the wall,
sinnle power-law profile.
There
dient in the velocity deficit
dary layer.
This
is
the soire shape used
to
enphasize
Fiqure
in
noticeable
there is
sioniFicant
a
seen to be
one
are
from a
2
a
sharp gra-
out to the edpe of
the boun-
is
nrobably the result of the bluntness of
in
the experinients
the necessity
cusped profile shown
13,
sinilarity
orofile shape for conf iqurat ion
the
in
profiles
and one
to achieve
Fiqure
in
The results appear
.
very slender,
the
rather
13(a)
than
use the
blunt approximation shown in Fiqure 13(b).
Second,
profile close
the
affected by the spires.
of
turbulence
in
seems
wall
little
This may be the result of the level
inherent
the pipe-flov;
in
rouqhness, but suggests that
a
to the
and
thicker boundary layer using spires which
is
pipe-
the
nay be difficult
it
to produce
also similar
profile shape both near to and far from the wall.
3
.
Other Considerations
The
degree
of
axi-symmetry
briefly
examined
was
during the experiments by rotating the spire element
throttle
housing.
detected, however, the data
reasonably representative
The degree
of
Fiqure
in
of
unsteadiness
the
in
47
in
non-uniformity
peripheral
Some
18
are believed
average
the flow,
the
was
to be
radial behavior.
which
made
it
difficult
during probe surveys,
stable readings
to obtain
did not measurably affect the accuracy of the nean velocityprof i les
C.
REQUIRED CONTROL
-'XCHievING THE
The velocitv nrofile shown for configuration
18
can be characterized
as
nearly uniform
a
1
in
Fiqure
with
flow
well-defined turbulent boundary layer at the case-wall.
velocity profile shown for conf iaurat ion
ized as
a
a
The
can be character-
2
irregular case-wall boundary layer, or
very thick,
alternately described as
wholly distorted velocity profile
a
across more than half the compressor blade height.
Since
compressor investiga-
the purpose of the overall
tion is to examine the effects of tip gap on
and performance
flow field
tip clearance gap size
blade height,
than the
radial
of
distortions
in
which
investigation.
boundary layer must
small
compared
This
a
range of
very much
always
smaller
make sense to introduce
is
not within the scope of
Any variation in the case-wall
be confined
to the
in
the flow field which extend over much
the span of the blading.
the proposed
is
does not
it
the compressor
characteristics,
to a
blade height
which remains
scale
or
the
investigation
becomes one of investigating inlet flow distortion.
spire
Thus,
if
necessary
to
order
recover
to
elements
use the
the
are to
be used,
original, sharply
asymptotic
outer
cusped
it
is
very
spires in
boundary
layer
profile Shane.
5
of
A
reduction
in
the spire heiqht
inches, nay be necessary to achieve no more
the natural
boundary lavor thickness.
to
(^
or even
than doublino
CONCLUSIONS AND RECOMMENDATIONS
VI.
this study,
In
exoer imental ly
two was intended,
experinent
a
spires
^irst,
approximated
the present
circular pipe, with
turbulence
levels
control
to
the present arranqement was based,
slender cusped
the
the
in
The
present
the
in
on which
are thought
shape of the
present case
thicker profile which could be nachined with
Second,
factor of
a
large rectangular wind tunnels,
in
understood.
spires was
between results
obtained usinq
and results
boundary layers
the design of
nhile an increase
factor close to four was measured.
differences
quantitative
and shown
the case-wall
the thickness of
and disolacenent thicknesses by
the overall
to be
spires was designed
of
increase
layer at the connressor inlet.
boundary
in
rim
a
to
by
a
blunt,
straight cuts.
installation of the spires was
within
much laraer area blockage
and higher
a
present in
than were
wind
the
a
tunnel
anpl icat ion
It was
found
installed in the
element in
order
that the
ring-element of
throttle upstream
to reduce
of at
oscillations
spires
least
in
the
must be
one screen
flow and
vibrations in the compressor to acceptable levels.
The
following
recommendations
are made
achieve the intended boundary layer control:
50
in
order
to
1.
The
orioinallv
specified
neomet rica 1
spire elenents should he nachined,
for
reasons of
exoense.
should be reduced to
2.
The inlet
"^
or
oipe iunctions
The
5
shane of
without conpromise
heiqht
of
the
spire
inches.
should he
snoothed and all
flanged connections should he tiqhtlv sealed.
51
the
REFERENCES
Evaluation of the PerforT^ance and Flow
i-Jaddell, J. L.
S
Thesis, Naval Postgraduate
in an Axial Coripressor
School, Monterey, California, October 1932.
,
^
""I
.
.
Pressure
Distribution
The
in
a
Tobiason, E. A.,
Naval
M.S. Thesis,
Postgraduate School,
Bel l^outh
Monterey, California, June 1973.
,
AT-i.erican Institute of Aeronautics and Astronautics,
Turbine and Rocket
^eroth.ernodynanics
of
Gas
Inc.,
Propulsion Gates, G. C, Editor, 1984.
,
Design of
Staff of Lewis Research Center,
Aerodyna^iic
Axial-Flow
Co^^nressors
NASA SP-36,
reproduced by
National Technical
Information Service,
Washington,
D.C., 1965.
,
Vavra, M. H.,
Aerothermodynanics
machines Uiley, New York, 1974.
Flow
and
in
Turbo-
,
Zucker, R. D.,
Aeronautics,
California.
Aerodynamic Analysis
Department of
Naval
Postgraduate School,
Monterey,
,
Standen, N., M.,
A Spire Array
for Engineering
Thick
Turbulent Shear Layers for Natural wjnd Simulation in
tJind Tunnels
Ottawa, Canada, 1972.
,
TPL Technical Note 82-01,
Waddell, J. L.,
Multistage
Compressor - Flow Losses in Throttle Screens and ^lates,
Feburary 1982.
Zucker, R.
Publishers,
D.
,
Inc.,
Schlichting, H.,
Inc.,
1979.
Fundamental
of
Gas
Dynamics,
Matrix
1977
Poundary-Layer
Theory
,
M.cGraw-Hill
BI3LrOGRAPHY
^.
I
A
,\
Fduca t ion
Con'^onep ts
,
Series, Aerot'iernodvnanics of Aircraft. Fnnine
(jorrion
C.
^
Cat-^-,
"^
i
:^ r
,
l-jr, 3
.-^/-inos u"^ on r'lov/ Research on '^laiin':],
c'-.e
"^low Research on Bla>iim
Brown, '^overi
Co"^^a^y Limited,
Baden, Switzerland, 1969.
'rnc^-^i,-
: i -v :.M
•>»:
L
,
Sc
'J.
H. and La^
S. ^., Principles of ^luid "'echanics
\ddison l/esley Publishing Conoany, 1964
Li,
,
Goldstein, p. J., Fluid 'Mechanics Measurements
Pub 11 shim Cor nor at ion, 1933.
,
,
Hemisphere
Moyle, I., Progress Report - Multistage Axial Conoressor
Progran on Tip Clearance Effects \PS Contractor Reoort
(MPS67-31-01CR)
"'ay 1979 - August 1981.
,
,
Dearock
Peviev^
,
,
R.E., Blaie Tin Gap Effects in Turbo^ach i nes Final Report No MPS67-8 1-0 16
October 198U.
\'PS
\
,
McKligot, Cant. D.M., Uniform inlet Conditions for the MPS
Subsonic Cascade Wind Tunnel Project Report, NPS No.
NPS67-31-019PR, January 1981.
,
I. and Zebner, H., Multistage Compressor Installation of Cast-Enoxy Blades TPL Technical rJote 80-05,
October 1980.
Moyle,
,
Mcyle, I. \'., Multistage Compressor - Initial Measurement
One Stage of Symmetrical Blading TPL Technical Note
80-09, October 1980.
;;ith
,
Neuhoff, F., Calibration and Application of a Combination
Temperature-Pneumatic Probe for Velocity and Rotor Loss
Distribution '^leasurements in a Compressor NPS Contractor
Report No: NPS-67-81-03CR, December 1981.
,
53
IMITI/\L
DISTRIBUTION LIST
No.
of Copies
Defense Technical Information Center
Caneron Station
Alexandria, Virginia 22304-6145
2
Super inten jent
Library, Code 0142
Attn:
Naval Postgraduate School
Monterey, California 93943-5000
2
Chair-^an
1
Department of Aeronautics and Astronautics, Code 67
Naval Postgraduate School
Monterey, California 93943-5000
Professor R. P. Shreeve, Thesis Advisor
Director, Turbooropuls ion Laboratory, Code 67Sf
Department of Aeronautics and Astronautics
r]aval Postgraduate School
Monterey, California 93943-5000
4
Naval Air Systems Command
(Attn:
AIR 931E)
Washington, D.C. 20361
1
Naval Air Propulsion Center
(Attn: Vernon Lubosky)
P.O. Box 7176
Trenton, New Jersey 08628
1
Professor G. J. Walker
Department of Civil and Mechanical Engineering
University of Tasmania
GPO Box 252C
Hobart, Tasmania 7001
Austral ia
1
Dirpers, Mabes TNI-AU
1
Jl Gatot Subroto 72
Jakarta,
INDONESIA
Dirdik, flabes TNI-AU
Jl Gatot Subroto 72
Jakarta, INDONESIA
Komandan Koharmatau
Lanuma Husein Sastanegara
Bandung, INDONESIA
.
1
2
INITIAL DISTRIBUTION LIST (CONTINUED)
No.
11.
Perpustakaan PT Dirgantara
Lanuma Husein Sastaneqara
Banduna, INDONESIA
12.
Muliukur Tarigan
Komplek Kartanegara D-5
Singosari-Malang
INDONESIA
13.
Director S. M. A. Negeri
Kaban jahe
Sumatera Utara
INDONESIA
14.
of Copies
.
1
4
1
Wing Pesbang 30
Lanunia Abd
Saleh
1
.
Malang
INDONESIA
15.
Perpustakaan
Lanuma Abd. Saleh
Malang
INDONESIA
16.
Wing Pesbang 10
Lanuna Husein Sastraneqara
Bandung
INDONESIA
17.
Perpustakaan I.T.B.
Bandung
INDONESIA
18.
Perpustkaan Universitas U.S.U.
Medan, Sumatera Utara
INDONESIA
1
19.
Perpustakaan Universitas Dirgantara
Lanuma Halim P.
Jakarta
INDONESIA
1
20.
Marie Hashimoto
Department of Aeronautics and Astronautics, Code 67
Naval Postgraduate School
Monterey, California 93943-5000
1
1
1
1
55
jThesis
T13835
c.l
Thesis
T13835
c.l
Tarigan
ww^^m
Development of a
boundary layer
control device
for tip clearence
explrj!
-ents in an axial
coT
pressor.
Tarigan
Development of a boundary layer control device
for tip clearence experiments in an axial compressor.