You are on page 1of 14

ELSEVIER

Experimental Investigations on Tandem


Compressor Cascade Performance at
Low Speeds
U. K. Saha • Minimizing the number of axial flow compressor stages for a specific
B. Roy work output, and thereby lowering the engine size and weight, has always
Department of Aerospace Engineering, been the designer's goal. Recently, with the emergence of low aspect ratio
Indian Institute of Technology, blades, there has been a renewed interest in tandem blading for maximizing
Bombay, India the blade loading further besides improving its off-design performance
characteristics.
This paper presents aerodynamic performance evaluations of three as-
sorted cascades of single (CDA 43), tandem (CDA 21-21), and a newly
developed tandem (CDA 32-21) blading. Fundamental investigation has
been attempted in a low speed cascade wind tunnel to find the high
deflection capability of tandem cascades. Although limited performance of
tandem CDA 21-21 has been affected by certain grey areas covering both
geometric and aerodynamic shortfall, tandem CDA 32-21 derived from
CDA 43 and CDA 21-21 has shown high flow deflection characteristics and
hence high loading over a wide range of angle of attack. © Elsevier
Science Inc., 1997

Keywords: aspect ratio, axial compressor, boundary layer, cascade,


controlled diffusion, airfoil, diffusion factor, loss coefficient, tandem
blade

INTRODUCTION ratios beyond the present limits with a minimum number


of blades forces higher aerodynamic loading and even the
For more than a decade, the quest for improved aero controlled diffusion airfoil shows its limitation due to the
engine gas turbine power and fuel economy has stimu- presence of incipient flow separation [8].
lated much research interest in the design of high perfor- This has brought a renewed interest in tandem blading
mance compressor airfoils. This has helped the evolution
[9-12] where boundary layer growth is expected to be
of the modern aircraft gas turbine engine into an attrac-
effectively controlled for the prevention of flow separation
tive power plant in a small package. Axial flow-compres-
sors of compact design are obtained by increasing the and subsequent stall. This has been possible with the
suction capacity and the stage pressure ratio. Aerody- inception of wide chord, low aspect ratio blades [13, 14]
namic optimization with such a design is aimed at increas- for higher Mach number flow regime with the intention of
ing loading to a maximum, with minimum penalty of total raising blade loading and improving off-design perfor-
pressure losses, leading to a minimum number of stages in mance characteristics.
a multistage arrangement and a minimum number of Benefits of low aspect ratio blades, controlled diffusion
blades per stage [1-4]. airfoil, and tandem blading have been appreciated in
Intensive research to obtain compressor blade sections terms of structural lifetime, high loading, and off-design
conforming to prescribed velocity distributions on com- flow control; however, these individual effects are to be
pressor blades has resulted in the development of con- unified together into an efficient design.
trolled diffusion of airfoils [5-7] to meet the requirements An innovative design of a high pressure ratio axial flow
of high efficiency and high blade loading capability, thus compressor, however, requires systematic study, and to
leading to reduced developmental cost. Current stage begin with low speed testing provides qualitative informa-
designs are, therefore, very efficient (90%) with relatively tion with which to constitute a framework for high speed
high pressure ratios ( > 1.6). However, increasing pressure analysis. An understanding of the two-dimensional linear

Address correspondence to Dr. Bhaskar Roy, Department of Aerospace Engineering, Indian Institute of Technology, Bombay 400 076, India.
Experimental Thermal and Fluid Science 1997; 14:263-276
© Elsevier Science Inc., 1997 0894-1777/97/$17.00
655 Avenue of the Americas, New York, NY 10010 PII S0894-1777(96)00125-2
264 U . K . Saha and B. Roy

cascade model may provide qualitative information for


preliminary guidance in the design of three-dimensional
compressor blades. These studies may also help to develop
and validate computational codes [15-18].
The work reported here was aimed at finding high
deflection capability of a developed tandem cascade ( C D A
32-21), compared with that of a single ( C D A 43) and an
existing tandem ( C D A 21-21) cascade. The motivation
behind this investigation was to find a blade profile to
achieve an attached flow behavior on the rear half of the
blade under varying operating conditions (design and
off-design) either by controlled diffusion characteristics of
the airfoil or by setting two blades in tandem, keeping CDA 32-21
similar aspect ratios. Figure 1. Blade profiles under present investigations.
The authors' earlier experimental studies [10, 12] with
single and tandem cascades ( C D A 43 and C D A 21-21)
showed interesting capabilities of the latter within a lim- EXPERIMENTAL METHODS
ited operating range. This motivated us to develop a new
tandem configuration so as to have a wide operating range Test Facility
with good loading capability. The developed tandem cas- The present investigation was carried out in a low speed
cade was configured out of a thick C D A blade profile at cascade wind tunnel (Table 2). Essentially, the flow envi-
the front (for wider operating range), followed by a scaled ronment under consideration is incompressible. This open
down blade profile at the rear, as shown in Fig. 1. The circuit facility (Fig. 3) is driven by a centrifugal blower
gap-nozzle geometry was given a cusp shape to promote located at the upstream end, followed by a diffuser, a
necessary and sufficient flow guidance for efficient mo- header containing a honeycomb and four mesh screens,
mentum transfer on the rear blade suction over a wide and then a 12-to-1 contraction ratio nozzle. The nozzle
range of angle of attack. C D A 32-21 tandem cascade was leads to the test section of a rectangular duct equipped
cut out from an existing C D A blade profile (43 ° camber) with end-wall movable flaps. The cascade along with the
truncated at the middle, followed by an existing, C D A 21 ° circular disks could be rotated in the test section about a
camber blade profile. Detailed blade coordinates of C D A c o m m o n axis to vary the angle of attack. Suction slits were
32-21 are presented in the Appendices. Blade coordinates made on either side of the side walls 0.25 times blade
of C D A 43 and C D A 21 profiles have already been chord upstream for boundary layer removal. Additional
reported [7, 19]. suction was created by employing two centrifugal blowers,
In both cases, tandem cascades were set and tested at each driven by a 0.75-kW motor, on either side of the side
48 ° camber (overall) to find an extra flow deflection of 5 ° walls. End-wall movable flaps were used to improve the
compared with the single cascade with 43 ° camber. In the desired flow uniformity and periodicity.
present investigations, attention was focused only upon
the blade surface and wake flow behavior to evaluate the Measurement Procedure
cascade performance characteristics. A direct cascade per-
formance assessment was made for all three configura- All the measurements in the present study were made
tions on the basis of static pressure coefficient (Cp) distri- with standard pitot static probes and static pressure tap-
bution, diffusion factor, static pressure rise coefficient, pings. While the upstream pitot static probe was used to
loss coefficient, and wake velocity profiles, as well as total record the inlet flow conditions and check flow uniformity,
pressure loss profiles. Detailed cascade blade geometry as the downstream probe was used to record wake total
given in Table 1, and the cascade notations showing pressure and velocity covering two blade passages. Static
various geometric and flow parameters are depicted in pressure tappings made on the blade surfaces were used
Fig. 2. to record static pressures and velocities (with respect to

Table 1. Single and Tandem Cascade Geometry


Blade Single Tandem New Tandem
Variables Notations CDA 43 CDA 21-21 CDA 32-21
Camber angle (degrees) q~, ~p,, 43 ° 48 ° (overall) 48 ° (overall)
Stagger angle (degrees) 7, "Ye 22° 220 (effective) 22° (effective)
Blade chord (mm) c 100 60 (FB) 60 (FB)
c -- 40 (RB) 40 (RB)
Blade pitch (mm) s 50 50 50
Blade span (mm) b 150 150 150
Solidity ( c / s ) tr 2 1.9 1.9
Aspect ratio ( b / c ) AR 1.5 1.56 1.56
No. of blades N 10 2 × 10 = 20 2 × 10 = 20
Experimental Investigations on Tandem Compressor Cascade 265

7 . I Y
z

~ plone

/ /'~t, I
/~IKI / ~ Axiol direclion, x l Figure 2. Cascade nomenclature.

inlet total pressure) at various chordwise positions to find Static Pressure Coefficient
Cp distributions, the diffusion factor, and the static pres-
sure rise coefficient. The mass averaged loss coefficient Single Cascade (CDA 43) Static pressure coefficient
was evaluated by traversing the downstream total pressure (Cp) distributions of the single cascade (CDA 43) at
probe over a length of 150 mm covering the two blade various angles of attack are presented in Fig. 4. At a = 0°,
passages. Wake total velocity over the same length was after an initial acceleration on the suction surface (up to
also recorded by the same probe. 30% of the chord from the leading edge), flow starts
diffusing toward the trailing edge (Fig. 4a). This accelera-
RESULTS AND DISCUSSION tion and deceleration effect diminishes at a = 5°, showing
a flatter Cp distribution (Fig. 4b). At a = 10°, steep
Low speed studies were attempted in the present investi-
gation to find qualitative information on the tested cas- acceleration around the leading edge (LE) is followed by a
cades on the basis of fundamental aerodynamic perfor- diffusion process at 6% of the chord from the LE, and
mance parameters. thereafter Cp remains constant (Fig. 4c). Between a = 0°

WAKE P R O B E r'-CASCADE 20x20MESH SCREEN "-~


TRAVERSE P L A N E UPSTREAM
WALL TAPS
/-.,,..FOR
- FLOW UNIFORMITY ~ - 25x25 HONEY COMB
, h.--\ ~.EC. u~,,~,, ,~ow ...-~,,,, ~-t-._ A S P L ' ' . ~ , o ~ L f l E~

I*--/
,
: SECTION ',
i
IIIIIB
x,..JI ! 1 1 I l l . L - -~
a
', !
i
' " -

. . . . . . . . . . . . I ,,o _1. ,o°o i J tl ,2soi=,wL ,so,, .V=sol.


B

SLITS lSOeAC.
(ALL DIMIENSlON5 ARE IN ram)

Figure 3. Low speed cascade wind tunnel.


266 U . K . Saha and B. Roy

Table 2. Cascade Wind Tunnel Data


Wind Tunnel Drive Suction System Drive Test Section Data
Blower type: radial Blower type: radial Size: rectangular
I)': 3.30 m 3 / s 12:0.056 m3/s Air velocity: 20.0 m / s
Pressure rise: 9.96 kPa Pressure rise: 2.98 kPa Mach No.: 0.06
Motor power: 55 kW Motor power: 0.75 kW Reynolds No. 1.369 × 105

Cp vs. c~ P l o t s

~-3.0
• • • • - Suction Surface • • • • • Suction Surface "• "•" S u c t i o n Surface
• ---- Pressure Surface • ~,•A Pressure Surface -•,•- Pressure Surface
~-2.0 I

-I.0

Cp m - 0 . 0

o~
o 1.0. ~E TE

co 2.0 ......... , ......... , ......... , ......... r ..... ~


20 4O flO 80 100 20 40 60 80 I00 0 20 40 60 BO 100
Percentage of Chord Percentage of C h o r d Percentage of Chord

(a) A n g l e of a t t a c k = 0° (b) Angle of attack = 50 (c) Angle of attack = lO °

~-3.0
v • • • • • Suction Surface • • • • • Suction Surface • =••= Suction Surface
"U "•''" Pressure Surface • A.., Pressure Surface " .4... Pressure Surface
~-2,0

~-I.0

Cp
m -0.0

1.0 LE TE

1
2.o .......
'2'd......i'd......'d'd......"gd......
f0o 0 20 40 60 80 100 o ....... '2'd ...... 'i'o ...... 'g'd ....... gd'"'Too
Percentage of C h o r d Percentage of Chord Percentage of C h o r d

(d) Angle of attack = 15 ° (e) Angle of attack : 180 (f) A n g l e of attack = 20 °

~-3.0
ID
• • • ~ • Suction Surface =••=•Suction Surface • • • • • Suction Surface
AAA.APressure S u r f a c e I &..A pressure Surface
~-2.0

- 1.1)

Cp
-0.0
P

1.0 E TE ~ LE TE
1
o~ 2.0 t.........i.........,...................,......... ......... r. . . . . . . . . v......... w......... , . . . . . ~
........ 2~....... 4~....... d'd.......e~......f0o 0 20 40 60 80 100 0 20 40 60 80 100
Percentage of C h o r d Percentage of Chord Percentage of C h o r d

(g) Angle of a t t a c k = 220 (h) Angle of attack = 25 o (i) A n g l e of a t t a c k = 30 o

Figure 4. Static pressure coefficient (Cp) distributions of single cascade (CDA 43) at varying angles of
attack.
Experimental Investigations on Tandem Compressor Cascade 267

and a = 10°, the cascade is operating at high negative ment. A separation bubble was also created in past experi-
incidences. As a result, sharp acceleration and diffusion ments with C D A [20, 21]. The suction surface pressure
processes were been observed anywhere on the blade gradient starts leveling off with an increase in the angle of
surfaces. At a = 15 ° to 22 °, this initial steep acceleration attack ( a = 25 ° and 30°), which is a precursor to flow
is followed by a sharp diffusion process, which is com- separation (Figs. 4h and 4i). Evidence of the onset of
pleted within 10% of the chord from the leading edge trailing edge separation is clearly observed in this range.
(Figs. 4 d - 4 g ) . This sharp acceleration and deceleration At a = 30 °, the Cp plot still shows the existence of a
evidently created a separation bubble. The separation separation bubble. At this angle of attack, the separation
bubble caused a local flow acceleration (between 10% and point has moved toward the LE and merged with the
35% of the chord), followed by a steady diffusion on the decaying separation bubble. The pressure surface Cp dis-
rest of the blade suction surface. The local separation tribution remains consistent at all angles of attack. This
bubble, while altering the Cp distribution, may have in- may be due to the controlled diffusion characteristics of
duced an early transition, which in turn, effected reattach- the cascade profile.

Cp vs. (x Ptots
~- 3.0
..~ • • • • • Suction Surface • • • • • S u c t i o n Surface I =~==• Suction Surface
- - . . 6 Pressure Surface 4A4~A4 Presaure Surface --*-- Pressure S u r f a c e
~-~.0
LE
~LE
Cp
- 1.0

-0.0
~ TE

~ 1.0. FRONT BLADE REAR BLADE FRONT BLADE DE FRONT BLADE JREAR
BIREAR
BLADE

m . Z.O ......... l ......... i ......... I ......... , .........


20 40 60 80 It 20 40 60 80 100 0 20 40 60 80 100
Percentage of Chord Percentage of Chord P e r c e n t a g e of C h o r d
(a) Angle of attack = 0° (b) Angle of attack = 50 (c) Angle of attack = 10 °

-~- 3.0
..~ • • • • • Suction Surface I • ~• • • Suction Surface _~=~ Suction Surface
O -**-- Pressure Surface • - ~ * , Pressure Surface Pressure Surface
~-2.0

- 1.0

Cp m -0.0

o 1.0 DE

~o 2.0 . . . .

20 40 60 80 100 o........2~.......i'o.......66.......g'o......foo
Percentage of Chord Percentage of Chord Percentage of C h o r d
(d) Angle of attack = 150 (e) Angle of attack --- 1 8 0 (f) Angle of attack = 20 o

"~- 3.0 . • = • = • S u c t i o n S u r f a c e • • • • • • Suction Surface • • • • • Suction Surface


~.'~-2.0 --* Pressure Surface ~...A. Pressure Surface --~-- Pressure Surface

LE ]L~ LE

TE TE E
Cp~
m -0.0

c~
--~'
B 0" I'O FRONT
L BLADE
A [RII.:.,I..i~.,DE
D E ] .F:I,N.I
FRONT~ I,.I~.I.E.
BLADE
R A E PRI ~ : . , B . ~ FRONT BLADE REAR BLADE

Z.O ....... , ......... , .................. , . . . . . . . . . . . . . . . . . . . . . . . .


20 40 60 80 100 0 Z0 40 60 80 100 20 40 60 80 100
P e r c e n t a g e of C h o r d P e r c e n t a g e of C h o r d P e r c e n t a g e of C h o r d
(g) Angle of attack = 220 (h) Angle of attack = 250 (i) Angle of attack = 300

F i g u r e 5. S t a t i c p r e s s u r e coefficient (Cp) distributions of tandem cascade (CDA 21-21) at varying angles of


attack.
268 U . K . Saha and B. Roy

Tandem Cascade ( C D A 21-21) Static pressure coeffi- m i n i m a l static p r e s s u r e d i f f e r e n c e n e a r the trailing e d g e


cient distributions of the tandem cascade ( C D A 21-21) arc ( T E ) of the front blade suction and pressure surfaces, a
shown in Fig. 5. At a = 0 °, steady acceleration is notice- w e a k e r m o m e n t u m transfer by the gap-nozzle flow was
able on the front blade suction surface up to 3 0 % of the effected.
chord (total) from the L E (Fig 5a). The acceleration zone T h e m i n i m a l difference of static p r e s s u r e at low angles
diminishes gradually with an increase in c~ (Figs. 5a-5c). of attack was caused by the i n c o m i n g flow, which is
Within the range of a = 0 ° to I0 °, the front blade pres- incident well o n t o the blade suction surface r a t h e r than
sure surface shows almost a constant Cp distribution; high o n t o the LE. T h e flow s e p a r a t i o n t e n d e n c y toward the T E
deflection over the rear blade suction surface does not of the front b l a d e suction surface o b s e r v e d earlier is
sccm to bc effective in this range owing to inefficient p r e v e n t e d by the g a p - n o z z l e flow at c~ = 15 ° (Fig. 5d),
m o m e n t u m transfer by the gap-nozzle flow formed be- forcing r e a t t a c h i n g on the r e a r blade suction surface. This
tween the blades. At lower angles of attack, owing to f u r t h e r aids the d e c e l e r a t i o n process almost to the rear

Cp vs. G Plo'ts
~-3.0 • • • • • Suction Surface [ i I i I ~ u c ~ o n Surface ,,a.•. Suction Surface
• •* *- Pressure Surface ----- Pressure Surface • - • • - Pressure Surface
o ~HE
~_.-2.0
o
-l.0 ~ T E
P
Cp ~-0.0 1 ~ %'~%• TE

1.0, FRONT BLADE REAR BLAD[ FRONT BLADE REAR BLADE FRONT BLADE REAR~B~DE

w 2.0 ....... , ......... , .................. ,..,

40 80
20 40 80 flO tO0 . . . . . . . . 2 ~ . . . . . . . .fii . . . . . . . 6 d . . . . . . . 6 ~ . . . . . . i o o 0 20 60 I00
Percentage of Chord Percentage of C h o r d Percentage of Chord

(a) Angle of attack = 0° ( b ) A n g l e o f a t t a c k = 50 (C) Angle of attack = lO °

"~-3.0
¢J
. , - , , = Suction Surface i l i l l Suction Surface
*---- Pressure Surface • - - - - Pressure Surface
~.-2.0 -
~J HE
-I.0

Cp
~-0,0
e,
1.o FRONT BLADE REAR BLADE I RONT BLADE REAR BLADE FRONT BLADE REAR BLAD£

r.~ 2.0 ......... l ......... , ......... t ......... , ......... , .................. , .........


...................,
......... ~ ......... , ......... ,

20 40 60 80 I00 0 20 40 80 80 100 20 40 80 60 I00


Percentage of C h o r d Percentage of Chord Percentage of Chord
(d) Angle of a t t a c k = 15 o (e) A n g l e of attack = 18 ° (f) A n g l e of a t t a c k = 200

~-3.0 ~ •••'• SucUon Surface • "'" • Suction Surface t l O , , o Suction Surface


•~ ssure Surface - - - * , Pressure S u r a f c e , 8 * * , Pressure Surface
~12.o
v LE

Cp
~-L.O
~
~-0.0
~ ~ T E
~ E

e~
o l.O. FRONT BLADE REAR BLADE FRONT BLADE REAR B L A D E FRONT BLADE REAR BLADE

oq 2.0 .........,............................
,.........
0 20 40 60 O0 tO0 ........ ~'6....... gd....... 8d....... go...... i'oo 0"....... ~'6....... ~'6....... 8d....... S'0...... i'00
Percentage of Chord Percentage of Chord Percentage of Chord
(g) Angle of a t t a c k = 22 o ( h ) A n g l e of a t t a c k : 250 (i) A n g l e of a t t a c k : 300

Figure 6. Static pressure coefficient (Cp) distributions of new tandem cascade ( C D A 32-21) at varying
angles of attack.
Experimental Investigations on Tandem Compressor Cascade 269

blade TE. Pressure surface Cp distribution at the same 0.8


angle of attack indicates continued flow attachment. Dif- o

fusion characteristics on the front blade suction surface 0.6


r=.
improve significantly at at = 18° and 20 ° (Figs. 5e and 5f);
the rear blade suction surface also indicates comparatively ~0.4
o
higher flow diffusion. From at = 22 ° to 30 °, the diffusion
process is milder on the front blade suction surface (Figs. ~ 0.2
5g-5i), whereas the rear blade suction surface shows its
attached diffusion characteristics. 0.0 .... , .... i .... ~ .... , .... i .... ) .... , ....
-5 0 5 10 15 20 25 30 35
Flow control, together with good diffusion characteris- Angle of a t t a c k (in degrees)
tics, effected by the gap-nozzle m o m e n t u m transfer was
realized in the range of angles of attack from at = 15 ° to (a) Single CDA 43
at = 30 °. The Cp distributions of C D A 21-21 within the
angles of attack from at = 15 ° to 30 ° do not indicate the
presence of a separation bubble, was observed for C D A
43. The absence of a separation bubble is presumably due 0.8
1. (~cX:K9 DFp.er.l ;~r
to the suction created by the gap-nozzle effect. ,~) : ~+ DFm.dm.a
2
~0.6
r..

N e w T a n d e m C a s c a d e ( C D A 32-21) Cpdistributions of the <~ 0 . 4


o
newly developed tandem cascade ( C D A 32-21) are pre-
sented in Fig. 6. At at = 0 °, steeper acceleration is notice-
able up to 40% of the chord (total) on the front blade
~ 0.2

suction surface after which flow decelerates until momen- 0.0


-5 0 5 10 15 20 2,5 30 35
tum exchange takes place by the gap-nozzle at 62% of the Angle of attack (in degrees)
chord (Fig. 6a). A kink believed to be due to local flow
separation and reattachment on the rear blade suction (b) T a n d e m CDA 21-21
surface is observed; this is clearly an indication of the
formation of a separation bubble, as observed in the
earlier investigations [20, 21]. Constant Cp distribution
was predominant on the blade pressure surface at the 0.8
same angle of attack. At at = 5 °, flow acceleration reaches (D(~D(~ DFgenera]
2 I I tt I DFmodifie d
to 30% of the chord, followed by continuous deceleration o 0.6
to the rear blade T E with the interim energy transfer by r.,

the gap-nozzle flow; pressure surface flow decelerates on ¢0.4


o
the front blade, whereas it accelerates on the rear blade
(Fig. 6b). The same effect with a milder acceleration and
deceleration is traced on the front blade at at = 10° and
15 ° (Figs. 6c and 6d). At a t = 18° and 20 °, the flow 0.0
undergoes sharp acceleration around the L E and then .... 8 .... ~'"i'd"'fs"2'6"z'g"a'0"as
decelerates on both suction and pressure surfaces of the
Angle of attack (in degrees)
front blade, indicating a significant amount of diffusion on (C) N e w T a n d e m CDA 32-21
the front blade suction surface while the flow sustains the
diffusion process on the rear blade suction surface (Figs. Figure 7. Variation of diffusion factor with angles of attack
of CDA 43, CDA 21-21, and CDA 32-21.
6e and 6f). With an increase in the angles of attack from
a = 22 ° to 30 °, a separation tendency on the front blade is
initiated, thus limiting the diffusion process on the suction
surface. However, flow behavior on the rear blade suction turn, prevented bubble formation on the front blade suc-
surface shows continuous deceleration, which is indicative tion surface. A strong gap-nozzle flow on the rear blade
of an attached flow behavior. This implies flow reattach- suction surface prevented the formation of the bubble on
ment, showing high overall flow deflection and diffusion the rear blade, and, even if there was a tendency to form a
characteristics of the cascade (Figs. 6g to 6i). At a = 30 °, bubble, it was captured within or carried away by the
the leading edge separation is indicated on the front accelerated gap-nozzle flow.
blade, but, interestingly, this flow has been reattached and
diffused on the rear blade. In the entire range of angle of
attack, mild acceleration was observed near the TIE of the Diffusion F a c t o r
rear blade pressure surface.
This newly developed tandem cascade shows its capabil- Variation of the diffusion factor (DF) with angles of
ity for preventing flow separation under high angles of attack (at) for the tested cascades is shown in Fig. 7. C D A
attack. Also, the formation of a separation bubble was 43 shows a narrow useful range of angel of attack from
been observed except at at = 0 ° on the front blade, which a t = 15 ° to a t = 2 5 ° (Fig. 7a), and this could also be
is actually the front half of the C D A studied under the visualized from the corresponding Cp plots. The highest
single cascade scheme. This indicates that tandem effect amount of diffusion occurs between at = 15 ° and 25 °,
created a pseudosuction at the front blade TE, which, in giving its maximum value of D F = 0.44 at at = 20 °.
270 U . K . Saha and B. Roy

Two definitions of diffusion factor have been used for varying chordwise location of Vmax on the blade suction
the tested t a n d e m cascades: (1) DFg. . . . . 1, based on classi- surface. A t a = 20 ° to 30 °, Pl and Vmax a r e at the same
cal definition, and (2) DFmodified, based on summation of location, and as such both C R and D F show similar
local diffusion factors of the front and the rear blades. loading characteristics.
F o r both definitions, the D F values of C D A 21-21 are
much higher than those of C D A 43 within a range of
a = 15 ° to 20 ° (Fig. 7b) The highest magnitude of diffu- Loss Coefficient
sion factor DFg. . . . . 1 = 0.67 and DFmodified = 0.75 is ob-
Mass averaged loss coefficients were calculated from the
served at a = 20 °. D F values drop sharply after 20 °, which
wake survey by means of traversing a total pressure probe.
is also noticeable from the corresponding Cp plots. High
The survey was parallel to the cascade and very close to
blade loading can be achieved by C D A 21-21 within a the blade TE, where the flow is nonuniform. Data were
range of a = 15 ° to 20 ° . taken at close intervals in the wake region and over coarse
C D A 32-21 shows the highest D F at a = 18 ° and 20 ° intervals outside the wake region (i.e., free-stream passage
(Fig. 7c). DFge,~ral shows a constant magnitude of D F = flow.
0.42 at a = 18 ° and 20°; whereas, in the same range, The variation of loss coefficients with angles of attack
DFmodifie d shows a constant magnitude of D F = 0.50. The for the tested cascades are shown in Fig. 9. Loss coeffi-
D F vs a plot of C D A 32-21 traces a similar variation of cient values are found to be low for single C D A 43
diffusion as that of C D A 43 with angles of attack, without cascade over the entire range. The tandem C D A 21-21
the peaky characteristics of C D A 21-21. cascade, even though it showed higher magnitudes of the
diffusion factor, gives high loss coefficients. This may be
Static P r e s s u r e Rise Coefficient due to higher surface flow velocity, leading to higher skin
friction, and to mixing loss incurred in the process of rear
The static pressure rise coefficient ( C a ) , which is also a blade suction surface flow energization. The new t a n d e m
measure of the flow diffusion or loading capability, was C D A 32-21 showed lower magnitudes of loss coefficients
evaluated for the tested cascades (Fig. 8). C D A 43 shows a than did C D A 43 over a range of a = 5 ° to 15 °. The high
reasonable a m o u n t of loading capability within the range loss coefficients over the rest of the angles of attack are
of angle attack from a = 15 ° to 25 ° (Fig. 8a). C D A 21-21 due to the thicker but attached growth of boundary layer
shows higher values of C R within a limited range (Fig. set by the t a n d e m configuration. This indicates that tan-
8b). C R values for C D A 32-21 (Fig. 8c) trace a similar d e m cascade is capable of controlling the boundary layer
path as those of D F with angles of attack, avoiding the growth, thereby preventing the separation tendency, and
peaky characteristics of C D A 21-21. The observed varia- the expected rise in the loss coefficients also is held within
tion of C R with D F from a = 5 ° to 15 ° was due to the tolerable limits.

4.0

(.} 2.0

~-0.0
-z.o_d.. '6 .... g"" i g " ig'" ~-'g'"i~g" a'o'" a5
Angle of attack (in degrees)
(a) Single CDA 43

4.0 4.0

m 2.0 2.0

-0.0 -o.o

-2.0_d.. -z.0 d'"i'o"i's"z'6"'~'5"3ii"is


'"6 ....
'6 . . . . d ' " i ' d ' " t ' 5 " ' ~ ' 6 " ' ~ , ' ~ " a ' 0 ' "
Angle of attack (in degrees) Angle of attack (in degrees)
(b) T a n d e m CDA 21-21 (c) N e w T a n d e m CDA 32-21
Figure 8. Variation of static pressure rise coefficient with angles of attack of C D A 43, C D A 21-21, and C D A
32-21.
Experimental Investigations on Tandem Compressor Cascade 271

1.5 The total pressure loss curve for C D A 43 exhibited


t;II:IGGDSingle CDA 43
4 - - 4 - T a n d e m CDA 2 1 - 2 1 behavior (Fig. 11) similar to that of velocity profiles in the
: : : : ~ New T a n d e m CDA 3 2 - 2 1 /
.. s.o entire range of angles of attack ( a = 0 ° to 30°). Suction
surface at a = 30 ° indicates higher growth of the bound-
@
ary layer, which is also evident from the corresponding
~ 0.5 wake velocity profile.
0
The wake velocity profiles for C D A 21-21 are symmetri-
,,=1

0.0 cal about the wake center from a = 0 ° to a = 10°, which


-5 0 5 10 15 20 25 30 35 distort slowly as the angle of attack increases (Fig. 12).
Angle of a t t a c k (in d e g r e e s )
However, the flow seemed to be highly unstable from
Figure 9. Variation of loss coefficient with angles of attack of c t = 2 2 ° to a = 3 0 °.
CDA 43, CDA 21-21, and CDA 32-21. The loss distribution for C D A 21-21 (Fig. 13) is symmet-
rical about the wake center at lower values of angle of
W a k e Profile Characteristics attack ( a = 0 ° to 10°). The wake profile becomes flatter
and distorts highly as the angle of attack increases. It is
Normalized wake velocity and total pressure loss profiles probable that the two adjacent wakes are coalescing to-
are plotted for the tested cascades in Figs. 10-15. The gether in the higher range of angles of attack ( a = 22 ° to
wake-measuring plane was selected in such a way that the 30°).
center of the horizontal axis (shown in the plots) corre- The wake velocity profiles for C D A 32-21 are more
sponds to the center of the blade pitch of the middle two asymmetrical about the wake center (Fig. 14) as compared
blades. Measurements were taken covering the two cen- with the velocity profiles of C D A 43. Beyond a = 20 °, the
tral blades in either case. The wake velocity measured at a width of the wake increases, indicating a thicker growth of
distance of 45 m m aft of the cascade line is normalized by boundary layer on the suction surface. The increase in
the reference inlet free-stream velocity. Similarly, the wake width could be ascribed to the additional 5 ° camber
wake total pressure measured at a distance of 15 m m aft of the newly developed cascade, where the boundary layer
of the cascade line was normalized by the reference total grew but remained attached, giving efficient diffusion.
pressure. The loss profiles for C D A 32-21 (Fig. 15) show a thicker
Wake velocity profiles at all angles of attack for C D A wake profile compared with that of C D A 43, indicating a
43 are asymmetrical about the wake center, indicating the thicker growth of suction surface boundary layer and
differential growth of boundary layers developed on the hence high losses of the cascade, which are reflected in
suction and pressure surfaces (Fig. 10). The defect in the the mass averaged loss coefficient.
wake velocity profile is large at high angles of attack
( a = 30°).

1.8
1.8
I~_^AAA ~ = 0 0 t ~ k : t ~ k ~ = ~ 0 O O O O O C ( = 1 0 0
c¢=0 ° ***** a=5 ° o o o o o ¢¢=10 °
~" 1.3
1.3
~' 0 . 8
,~. 0.8
I
;> 0.3
0.3 (2,
-0.2
-0.2 . . . . , . . . . | . . . . , . . . . i . . . . i . . . . , . . . . i . . . .

,. 1.8
t.8 A&AAA 0f=15 ° t:Ct~:$ CC=[8 ° OOOOOCC=20°
&~,~z~& c¢=15 ° * * * * * or=IS ° o o o o o ¢t=20 ° K
0" 1.3
t.3
Z
A

oN 0 . 8
~,, 0.8 a,
I
;> 0.3
0.3 D.,
-0.2 . . . . i . . . . i . . . . i . . . . i . . . . i . . . . i . . . . J . . . .
-0.2 . . . . i . . . . i . . . . , . . . . i . . . . , . . . . i . . . . i . . . .

t.8 [.8
auA4& = = 2 2 ° ~-~-,~-t c~=~.5 ° ooooo a=30 ° a&&&& a_-22 ° *~,~,* a=25 ° ooooo a=30 °

1.3 ~ 1.3
A

> 0.8 a,0~ 0 . 8


I
0.3 0.3
o.,
-0.2 . . . . i . . . . i . . . . i . . . . , . . . . i . . . . ; . . . . i . . . .

-8 -6 -4 -2 0 2 4 6 8
Trailing Edge Blade to Blade Direction, cm Trailing Edge Blade to Blade Direction, c m
Figure 10. Wake velocity profiles of CDA 43 at varying Figure IL W a k e total pressure loss profiles of C D A 43 at
angles of attack. varying angles of attack.
272 U . K . Saha and B. Roy

1.8 1.8
n ~ ¢(=0 ° ***** C1=5 ° o o o o 0 0 = 1 0 0 ,,~,at~ c<=O° **~** o~=5 ° n..tUL~_nc~=lO °

"d 1.3 1.3


a,

;> 0.8 0.8


N
;>.
0.:3 0.3

-0.2 -0.2
1.8 1.8
~zzzz ~x=15 ° **~, ~=18 ° ooooo c~=20 ° z z z z ~ c~=15 ° * * * * * (x=18 ° o o o o o o=200

1.3 1,3
~J
~e

0.8 ~.~ 0.8


t~
~> 0.3 z> 0.3

-0.2 -0.2
1.8 1.8
~zzz~ ~=:~2 ° * ~ k * * ~=25 ° o o o o o c~=30 ° ~t~,~t, ~x=22 ° ***~, c~=25 ° ooooo <~=30 °
1.3 1.3

> 0.8 ~ 0.8


e,l

~ 0.3 ;> 0.3

-0.2
-0.z_g,. '-'d'' ' - ' i ' ~ z ' '6 . . . . ~ . . . . i . . . . d . . . . a
Trailing Edge Blade to Blade Direction, c m T r a i l i n g Edge Blade to Blade Direction, c m

Figure 12. W a k e velocity profiles of C D A 21-21 at varying Figure 14, Wake velocity profiles of CDA 32-21 at varying
angles of attack. angles of attack.

1.8 1.8
~ 4 A 4 ~ 0~----.0° :,k*:,k:,k:,k 0 < = 5 ° o n o o o (x=].O ° ~e,z~z~ (x=O ° ***** 0~=5 ° 0 o 0 0 0 ~=lO °

~ 1.3 O" 1.3

~:~ 0.8 0.8


i I
0.3 0.3
0.,
-0.2 -0.2 . . . . i . . . . i . . . . i . . . . J . . . . i . . . . , . . . . i . . . .

1.8
~z~ c(=15 ° ***** o(=18 ° ooooo c~=20 °
O' 1.3 C~ 1.3

0.8 g 0.8
t I
0.3 N 0.3

-0.2 -0.2
1.8
~4z4~ ~=22 ° **~** (x=25 ° ooooo ct=30 ° ~,,~ 0<=22 o , ~ , ct=25 ° ovoov ~=30 °
O~ L,3

r~ 0.8 ~' 0.8


I I
0.3 0.3
P.
-0.2 . . . . i . . . . J . . . . , . . . . i . . . . i . . . . , . . . . , . . . .
- 0 . 2 8 -. ,'~- 6i ; " , , ,~ 4i . . . . - ~] , , , , 0i . . . . 2~ . . . . 4i . . . . 6~ ' ' ' -8 -6 -4 -2 0 2 4 6 8
Trailing Edge Blade to Blade Direction, c m Trailing Edge Blade to Blade Direction, c m

Figure 13. W a k e total pressure loss profiles of C D A 21-21 at Figure 15. W a k e total pressure loss profilcs of C D A 32-21 at
varying angles of attack. varying angles of attack.
Experimental Investigations on Tandem Compressor Cascade 273

U N C E R T A I N T Y ANALYSIS aggravated further by an extra camber of 5 °. Furthermore,


skin friction drag may be high because of a higher magni-
Extreme care was exercised to ensure the highest possible tude of surface velocity, especially on the rear blade. It
reliability and accuracy of the acquired test data. Each could be said that, after a judicious trade-off between
data point on the blade surfaces represents an average of higher blade loading and higher losses, CDA 21-21 has
three repetitive pressure and velocity readings. The time the potential compared with CDA 43 over a limited range
between successive readings was 5 s. In surveying the of angles of attack, and further optimization with the
wake flow manually, each time the probe was moved to a cascade geometry is required for its improved perfor-
new position, we allowed a settling time of 10-15 s to mance over a wider range. However, a compressor meant
record the data. The accuracy of the micromanometer was only for design-point operation may stand to gain by the
0.1 mm H 2 0 for the pressure reading and 0.1 m / s for the use of the CDA 21-21 tandem blade.
velocity reading. The potential of CDA 32-21 was observed in maintain-
The uncertainties relative to cascade parameters are as ing a sustained flow attachment within a reasonable range
follows: stagger angle (~/) = _+0.5°, angle of attack ( a ) = of angle of attack from a = 5 ° to a = 30 °. This indicates
+0.5 °, overall camber (~Po)= +1.0°, blade pitch ( s ) = that its flow deflection capability is superior to that of
-+ 0.5 mm, blade profile tolerance = -+ 0.25 mm. CDA 43. The reason for its efficacy could be attributed to
a guided flow energization by the cusp-shaped gap-nozzle
formed between the blades and to higher magnitude of
PRACTICAL SIGNIFICANCE/USEFULNESS front blade leading edge thickness to total chord ratio. An
The results of the present study would be extremely useful additional camber of 5 ° was added for higher flow deflec-
from a practical point of view because it provides signifi- tion in the newly developed tandem CDA 32-21 cascade,
cant guidance for designing practical machines meant for with acceptable loss coefficients over this range of angles
higher loading limits. The present data could be used as of attack.
guidance for designing machines for low speed applica-
tions, as in the case of air handling equipment, ventilation
fans, cooling fans, and so forth. FUTURE RESEARCH
For high speed applications, further optimization of the
Future research needs to address the cusp shape of the
blade shape is necessary to take into account the forma-
front blade trailing edge and the consequent gap geometry
tion of shocks and compressibility effects. Such high speed
more accurately. The flow conditions near the gap are to
data may be useful for designing tandem stator and rotor
be accurately assessed through boundary layer studies,
blades for aircraft gas turbines, for vehicular gas turbines
and this knowledge could be pursued for developing a
(e.g., cars, trucks, hovercrafts, ships, etc.), and for indus-
suitable computational code. Such an investigation would
trial power generating units. Potential can also be realized
not only lead to accurate analysis of the flow in and
in tandem vaned diffusers of centrifugal compressors and
around the tandem cascade configurations, but also pro-
in turbopump feed systems in liquid propellant rocket
vide guidelines for future experimentation.
engines.
The present work has been addressed more to the
control of diffusion, which occurs on the rear half of the
blade surface. The front half, with high acceleration, sepa-
CONCLUSIONS
ration bubble, and onset of sharp diffusion, would require
In the present investigations, an attempt has been made close investigation, especially at high speed where com-
to achieve an extra flow deflection of 5 °, using a tandem pressibility, turbulence, and shock interactions would be
blade configuration, in comparison with a single blade significant.
profile, with minimum loss penalty. CDA 43 has shown Recently, research has been undertaken to find the
good diffusion capability over a range of t~ = 15° to 25 °, operational feasibility of the developed tandem cascade
whereas CDA 21-21 has shown higher diffusion capability for the off-design flow condition. Knowledge of off-design
within a range of a = 15° to 20 °. Thin airfoil characteris- performance of a high pressure ratio axial flow compres-
tics of CDA 21-21 configured out of two scaled down sor is extremely important because it affects operating
CDA profiles were observed. A thin airfoil is character- costs and component life. In case of aero engines, idling,
ized by a narrow range of operation. When used in a take-off, and cruise will always result in different flow
tandem configuration, both the front and the rear blades situations within the compressor stages. Similarly, other
have a low leading edge radius to chord (total) ratio. modes of propulsion, such as marine and industrial gas
Hence, flow guidance over both front and rear blade turbines, where axial compressors are employed often
leading edges is limited to a narrow incidence range. This need to operate away from the design point for significant
was observed in the present experimental findings. lengths of time. Thus to maintain the performance of the
CDA 21-21 has a higher total pressure loss coefficient machine at off-design condition, flexibility in the opera-
than that of CDA 43. This may be caused by the mixing tion of the component is required. One way to improve
loss incurred as a result of the suction surface flow ener- this situation is to operate the cascade in a variable
gization. The other reason could be the lower magnitude camber mode. In our next experimental venture, the de-
of thickness to total chord ratio, yielding poor flow guid- veloped tandem cascade will be tested at various camber
ance, especially near the leading edges. Wake flow is settings to assess its off-design flow control capability.
274 U . K . S a h a a n d B. Roy

APPENDIX A APPENDIX B

B l a d e C o - o r d i n a t e s o f C D A 43 ( t r u n c a t e d ) B l a d e C o - o r d i n a t e s o f C D A 21 ( s o l i d )

xss Xss xps Y~ xs~ x~ x~ r~s


- 0.02 0.53 0.34 - 0.36 - 0.42 0.69 - 0.39 0.51
- 0.06 1.00 0.82 - 0.65 -0.41 0.90 - 0.35 0.35
- 0.06 1.13 1.36 - 0.83 -0.39 1.13 - 0.28 0.23
O. 10 1.90 2.00 - 0.89 -0.31 1.42 - 0.20 0.12
0.28 2.74 2.71 - 0.90 - 0.12 1.89 0.00 0.00
0.59 3.56 3.49 - 0.86 0.07 2.24 O. 11 - 0.04
1.12 4.26 4.32 - 0.74 0.33 2.65 0.22 - 0.05
2.16 5.20 5.24 - 0.53 0.65 3.13 0.33 - 0.05
3.46 6.19 6.30 - 0.34 1.05 3.65 0.45 - 0.04
5.17 3.37 7.36 - 0.12 1.51 4.21 0.57 0.01
7.11 8.62 8.51 0.14 2.05 4.83 0.69 0.02
9.28 9.90 9.70 0.42 2.63 5.47 0.83 0.07
11.58 11.19 10.98 0.72 3.26 6.13 0.95 0.13
13.86 12.39 12.29 1.06 3.91 6.77 1.09 0.19
15.96 13.43 13.60 1.45 4.55 7.39 1.23 0.25
17.80 14.29 15.14 1.79 5.17 7.97 1.37 0.33
19.41 14.99 16.67 2.18 5.77 8.48 1.51 0.42
20.79 15.57 18.26 2.60 6.31 8.95 1.67 0.51
22.03 16.06 19.92 3.03 6.81 9.36 1.83 0.61
23.15 16.48 21.64 3.48 7.27 9.73 1.99 0.71
24.18 16.85 23.44 3.93 7.69 10.05 2.15 0.81
25.16 17.17 25.32 4.40 8.07 10.33 2.31 0.93
26.07 17.46 27.28 4.87 8.41 10.58 2.49 1.05
26.95 17.73 29.27 5.34 8.73 10.81 2.67 1.17
27.81 17.97 31.42 5.81 9.03 11.01 2.85 1.31
28.63 18.18 33.63 6.27 9.29 11.19 3.04 1.44
29.43 18.38 35.96 7.05 9.79 11.51 3.23 1.58
30.21 18.56 38.36 8.20 10.21 11.77 3.65 1.88
30.98 18.72 40.00 9.04 11.03 12.23 4.09 2.21
31.72 18.87 41.00 9.34 12.03 12.73 4.56 2.56
32.43 19.00 43.00 10.09 13.09 13.21 5.07 2.95
33.13 19.11 45.00 11.18 13.63 13.45 5.35 3.16
33.81 19.21 47.00 12.08 14.77 13.90 5.94 3.61
34.46 19.29 49.00 13.02 15.35 14.12 6.26 3.86
35.07 19.36 52.00 14.10 16.59 14.55 6.60 4.12
35.63 19.42 55.00 14.65 17.25 14.78 7.34 4.69
36.16 19.45 57.00 14.95 18.62 15.23 8.19 5.35
36.67 19.48 59.00 15.12 19.35 15.47 9.18 6.09
37.16 19.50 61.00 15.23 20.10 15.71 9.74 6.51
38.91 19.50 20.89 15.97 11.04 7.45
40.22 19.52 21.71 16.23 11.79 7.99
41.56 19.47 22.55 16.52 12.64 8.58
43.01 19.38 23.45 16.82 13.59 9.22
44.45 19.25 24.38 17.13 14.67 9.93
45.94 19.09 25.36 17.47 15.92 10.70
47.44 18.89 26.39 17.83 17.35 11.55
48.97 18.66 27.48 18.21 20.99 13.53
52.11 18.08 28.63 18.63 23.27 14.65
55.39 17.36 29.83 19.06 25.90 15.85
58.77 16.52 31.07 19.51 28.67 17.07
61.00 15.58 32.31 19.97 31.59 18.21
33.43 20.41 33.77 19.05
34.21 20.72 34.79 19.47
Experimental Investigations on Tandem Compressor Cascade 275

or solidity ( c / s ) , dimensionless
q~ camber angle, degrees
1 / ..... s.o,io° s°..oe % overall camber angle (for tandem blade), degrees
.............. ;'--'7 77"7 ..... Subscripts
-1 9 19 29 39 49 59 69 79 89 99
X-axis 1,2 cascade inlet and outlet planes
Controlled Diffusion Airfoil (32 g C a m b e r )
loc, ref local and reference conditions
max maximum on the blade suction surface
x, y, $ axial, tangential, and spanwise directions

REFERENCES
~t0 1. Wennerstrom, A. J., Highly Loaded Air Flow Compressors: His-
Suction Surface
/ / -~--~-e=Pressu,'e Surface tory and Current Developments. ASME J. Turbomachinery 112,
- 0 q~lr,. , . . . . . . . . . . . . . . i 567-578, 1990.
-I 10 20 30 40 2. Cumpsty, N. A., Compressor Aerodynamics. Longman, Bombay,
X-axis
India, 1989.
Controlled Diffusion Airfoil (210 Camber}
3. Starken, H., Performance of Controlled Diffusion Blades, Axial
Flow Compressors, VKI Lecture Series 1992-02, 1992.
4. Behlke, R. F., The Development of a Second Generation of
Controlled Diffusion Airfoils for Multistage Compressors. ASME
J. Turbomachinery 108, 32-41, 1986.
NOMENCLATURE 5. Stephens, H. E., and Hobbs, D. E., Design and Performance
AR aspect ratio ( b / c ) , dimensionless Evaluation of Supercritical Airfoils for Axial Flow Compressors.
United Technologies Corporation Report FR 11455, 1979.
b blade span, m m
6. Hobbs, D. E., and Weingold, H. D., Development of Controlled
Cp static pressure coefficient (Cp = P l o c - P1/Q1) Diffusion Airfoil for Multistage Compressor Application. ASME
CR static pressure rise coefficient (C R = P2 - J. Eng. Gas Turbines Power 106, 271-278, 1984.
P l / Q I ) , dimensionless 7. Bledsoe, M. R., The Method of Complex Characteristics for
CDA controlled diffusion airfoil Design of Transonic Blade Sections. Paper No. D O E / E R /
c blade chord, m m 03077-273/UC-32, Elsevier Science, Inc., New York, 1986.
8. Sturm, W., Scheugenpflug, H., and Fottner, L., Performance
DF diffusion factor ( D F = Vm~ - V 2 / V 1)
Improvements of Compressor Cascades by Controlling the Pro-
FB front blade file and Sidewall Boundary Layers. ASMEJ. Turbomachinery 114,
LE leading edge 477-493, 1992.
N n u m b e r of blades 9. Wu, C.-H., Zhuang, B., and Guo, B., Experimental Investigation
static pressure, kPa of Tandem Blade Cascades with Double Circular Arc Profiles.
P
ASME Paper No. 85-GT-94, 1985.
/'0 total pressure, kPa
10. Roy, B., Marathe, B. V., and Gangal, D. B., Comparative Cas-
PS pressure surface cade Studies of Some High Diffusion Compressor Bladings. Proc.
Q dynamic pressure, kPa Ninth Int. Symp. on Air Breathing Engines, Athens, Greece, pp.
RB rear blade 1281-1289, 1989.
SPRC static pressure rise coefficient, dimensionless 11. Sachmann, J., and Fottner, L., Highly Loaded Tandem Compres-
sor Cascades with Variable Camber and Stagger. ASME Paper
SS suction surface
No. 93-GT-235, 1993.
$ blade pitch, m m 12. Roy, B., Walvekar, A. K., Saha, U. K., and Marathe, B. V., Low
TE trailing edge Speed Cascade Studies of Highly Cambered Single and Tandem
V flow velocity, m / s Compressor Blading. Int. J. Turbo Jet Engines 12, 129-137, 1995.
volume flow rate, Cu m / s 13. Gostelow, J. P., Cascade Aerodynamics. Pergamon, Oxford, 1984.
14. Wennerstrom, A. J., Low Aspect Ratio Axial Flow Compressors:
Greek Symbols Why and What It Means. ASMEJ. Turbomachinery 111, 357-365,
angle of attack, degrees 1989.
flow angle, degrees 15. Wisler, D. C., Loss Reduction in Axial Flow Compressors
/3 Through Low-Speed Model Testing. ASME J. Gas Turbine Power
8 deviation angle, degrees 107, 354-363, 1985.
t u r n i n g / d e f l e c t i o n angle (~ =/31 -/32), degrees 16. Hirsch, C., Advanced Methods for Cascade Testing. AGARD-
T stagger angle, degrees AG-328, 1993.
% effective stagger angle (for t a n d e m blade), 17. Yocum, A. M., and O'Brien, W. F., Separated Flow in a Low
degrees Speed Two Dimensional Cascade I: Flow Visualization and Time
1" incidence angle, degrees Mean Velocity Measurements. ASME J. Turbomachinery 115,
409-420, 1993.
K blade angle, degrees 18. Yocum, A. M., and O'Brien, W. F., Separated Flow in a Low
ae0 total pressure loss, kPa Speed Two Dimensional Cascade II: Cascade Performance.
to loss coefficient (to = A P o / Q 1 ) , dimensionless ASME J. Turbomachinery 115, 421-434, 1993.
276 U . K . Saha and B. Roy

19. Marathe, B. V., Cascade Studies of Highly Loaded Compressor 21. Elazar, Y., and Shreeve, R. P., Viscous Flow in a Controlled
Blades. M. Tech. Thesis, Aero. Eng. Dept. Indian Inst. Tech., Diffusion Compressor Cascade with Increasing Incidence. ASME
Bombay, 1988. J. Turbomachinery 112, 256-265, 1990.
20. Sanger, N. L., and Shreeve, R. P. Comparison of Calculated and
Experimental Cascade Performance for Controlled Diffusion
Compressor Stator Blading. ASMEJ. Turbomachinery 108, 42-50,
1986. Received April 27, 1995; revised August 30, 1996

You might also like