A slip factor calculation in centrifugal impellers based on linear cascade data

Dr. Abraham Frenk Becker Turbo System Engineering (2005). (Presented by Dr. E. Shalman)

Abstract Accurate modeling of the flow slip against direction of rotation is essential for correct prediction of the centrifugal impeller performance. The process is characterized by a slip factor. Most correlations available for calculation of the slip factor use parameters characterizing basic impeller geometry (review of Wiesner (1967) and Backstrom (2006)). Approach presented below is based on reduction of radial cascade to equivalent linear cascade. The reduction allows to calculate characteristics of radial blade row using well established experimental data obtained for linear cascades, diffusers and axial blade rows.

Rotors with splitters or with high loading must be divided into few radial blade rows. In the case of multiple rotor blade rows the slip angle is calculated for each blade row. The slip angle of the rotor is a sum of the slip angles obtained for each row. Suggested reduction of radial blade rows allows also calculating of other parameters essential for impeller design. Suggested method allows also determine additional causes influencing slip factor. The slip factor depends not only on parameters characterizing basic impeller geometry, but on difference of inlet flow angle from stall flow angle. In the rotors with the same basic geometry parameters slip factor depends on the length of the blade. Slip factor increases with blade length.

Axial compressor

Centrifugal compressor

Slip flow in Centrifugal compressor

Cu2 Ca2 C2
2k

Busemann (1928) called this displacement flow; other authors refer to its rotating cells as relative eddies.

v s =1 R2
Stodola (1927)

(definition)

v s = R2 sin 2 k Z sin 2 k =1 Z

Axial compressor
Xf t

Centrifugal compressor
Ca Cu Z-number of blades

2k

Howell (1945)
= 0.23(2 x f )2 + 0.002( 2 k + ) t

Slip factor or slip angle = 2 - 2k.

Deviation angle = 2 - 2k.

ca 2 = 1 (tan ( 2 k + ) tan ( 2 k )) u2
= 1
sin 2 k Z 0.7
(Wiesner)

 = 0.23(2 x f )2 + 0.002( 2 k + ) t

General blade row

2

cos(1 ) w2 P = 1 = 1 K D 2 1 w1 2 w1 cos( 2 )
W1 Ca1 1,eqv W2 Ca2 2,eqv

Equivalent linear cascade P = Peqv

2 =2,eqv

General Inviscid
w2 P = 1 2 1 w1 2 w 1
Peqv
2

cos(1,eqv ) = K D cos(1 ) eqv = 1,eqv 2

2

In Linear cascade Ca1= Ca2

cos(1,eqv ) w2 = 1 = 1 2 cos( 1 w1 2 2, eqv ) w1

eqv i + = eqv 2 = 2k + 1 = 1k + i

 eqv = 0.23(2 x f ) + 0.002( 2 k + ) eqv teqv

2

Calculation of equivalent pitch (solidity)

2 2t 2

teqv
b

t 2 2t 2 = = t1 + t 2 1 b b t1 + t 2

2 k , eqv = 2 k + 0.23(2 x f )2 + 0.002( 2 k + ) eqv teqv 1

) (

ca 2 (tan ( 2k ,eqv + eqv ) tan ( 2k ,eqv )) = 1 u2

cos(1,eqv ) = K D cos(1 )

Calculation of the coefficient KD

1. Influence of the channel height h.
Stratford (1959) obtained the height (h) of the diffuser with given length (b) that has maximal static pressure rise coefficient. The velocity ratio w2/w1 depends on the ratio b/h. The experimental data for linear cascade were obtained for h=2b. The experimental data of Stafford may be approximated by equation

Experimental data for diffusers

/ w1 2b = (w2 / w1 )h = 2b h
2

(w

0.1

cos(1,eqv ) = K D cos(1 )

10

Calculation of the coefficient KD

2. Influence of the ratio Ca1/Ca2.
W1 1,eq
v

W2

2,eq
v

Cascade with constant height and solidity:

Ca1

Ca2

w2 ca 2 cos(1 ) = w1 ca1 cos( 2 )

If the solidity is changed, then velocity ratio in the equivalent cascade must be multiplied by the pitch ratio

cos(1 ) w2 = KT cos( 2 ) w1

ca 2 t1 2b KT = ca1 t 2 h

0.1

cos(1,eqv ) = K D cos(1 ) Calculation of the coefficient KD

For axial compressors

11

W1 Ca1

1,eq
v

W2 Ca2

2,eq
v

w2 cos(1,eqv ) cos(1 ) = = KT w1 cos( 2,eqv ) cos( 2 )

ca 2 t1 2b where KT = Hence, K =K T D ca1 t 2 h
0.1

In General case

KD = K

1 2 KT T

where

ca 2 t1 2b KT = ca1 t 2 h

0.1

(effect of boundary layer separation and blockage)

12

Impellers with high blade loading and splitters

The rotor with splitters is divided into sequence of rotors. Each rotor has the same number of blades at inlet and exit. For high loaded rotors the loading of each equivalent linear cascade must be less then maximal allowed.

Impellers with high blade loading and splitters

Rotor 2

13

Calculate slip angle for each rotor. The slip angle of the rotor is a sum of the slip angles obtained for each row

Rotor 1

High blade loading

cos(1,eqv ) = K D cos(1 )
KD = K
1 2 KT T

14

ca 2 t1 2b KT = ca1 t 2 h

0.1

Stall conditions

15

Maximal pressure gradient

ca 2 t1 2b KT = cos 1,eqv = K D cos 1 KD = K ca1 t 2 h Analysis of experimental data of Howell (1945), Emery (1957), Bunimovich (1967)allows obtaining following equation for coefficient of static pressure rise at stall conditions (maximal pressure rise)

( )

1 2 KT T

0.1

w2, st Pst = 1 2 w 1 w1 2 1, st
b

cos(1eqv, st ) = 1.124 = 1 1+ m t cos( ) 2 , eqv st st

2 2

mst = (wSuctionSide w1 )db

0

(w w
1 0

Pr essureSide

)db

Parameter mst characterizes type of the cascade. E.g. For cascades studied by Howell mst =1. t = t / b is relative pitch of the linear cascade

16

Compressor Stages used to test the model

1. Stage of turbojet Olympus 2. Stage tested by Stechkin 3. Stage designed by Beker Engineering and tested at Concept. 4. Stage designed by Beker Engineering , not tested 5. Stage C1 (tandem blades) (data from USSR) 6. Monig et al. Trans ASME. Journal of turbomachinery 1993,v115, p.565-571. 7. Stage C9 (data from USSR) 8. Musgrave D., Plehn N.J. Mixed flow stage design and test results with a pressure ratio of 3:1. ASME Gas turbine conference presentation 87-GT-20.

17

Results
Table 1. Rotor parameters. Rotor
No.

Z
number of blades

Rotor pressure ratio

G
Kg/s

2k
(outlet blade angle)

r2 r1
1.83 2.30 1.89 2.1 2.12 2.0 2.18 1.78

t1,mean
(inlet relative pitch)

t2,mean
(outlet relative pitch)

1
(inlet flow angle )

ca2
m/s

u2
m/s (tip velocity)

1 2 3 4 5 6 7 8

6/12 12 11/22 7/14 27/27 15/30 18/36 12/24

4.35 3.95 5.23 3.87 4.54 3.86 6.79 6.36

0.41 3.0 0.48 1.0 2.7 2.0 2.75 2.89

16.99 33.58 12.0 48.28 9.0 8.48 6.25 19.9

1.0 0.50 1.27 1.50 0.25 0.74 0.56 1.0

0.695 0.397 0.641 0.346 0.316 0.175 0.612

57.8 32.7 57.48 56.33 46.65 50.47 46.25 60.05

180 170 189.3 182.9 185.3 160 161 165

472.3 483 500 550 455 425 520 510.6

Results
Table 2. Slip factor and exit angle. Rotor
No.

18

2
measured

2
Present work

measured

Present work

Wiesner

Stodola

Stechkin

Eck

1 2 3 4 5 6 7 8

31.8 27.0 20.17 23.32 17.9 35.9

32.0 43.8 27.0 56.3 20.6 23.0 17.8 35.0

0.881 0.887 0.915 0.899 0.933 0.893

0.878 0.896 0.888 0.875 0.901 0.896 0.934 0.890

0.832 0.840 0.888 0.874 0.902 0.908 0.919 0.895

0.860 0.72 0.848 0.762 0.876 0.890 0.910 0.887

0.718 0.823 0.880 0.834 0.910 0.914 0.931 0.861

0.784 0.840 0.865 0.870 0.903 0.896 0.926 0.877

19

Measured and calculated slip factor for 6 impellers

20

Calculated slip factor for 8 impellers

21

Conclusions
Suggested method allowing reduction of arbitrary rotor blade passage to equivalent linear cascade Method allows establishing equivalence between axial and centrifugal impellers Method allows calculating slip factor based on well established linear cascade correlations for deviation angle