Int. J. Multiphase Haw Vol. 6, pp. 363-367 Pergamon/Elsevier, 19~0.

Printed in Great Britain

BRIEF COMMUNICATIONS
TWO-PHASE FLOW IN BENDS

D. CHISHOLM NationalEngineeringLaboratory,East Kilbride,Scotland

(Received 29 December 1978; in revised form 15 October 1979) Abstract--An equation is developed for use in predicting the two-phase multiplier for pressure drop in bends; the equation simplifies the use of an existing method. The method is also compared for the first time with data at high density ratios (ptJpc = 560). 1. I N T R O D U C T I O N

In this note an existing method for predicting pressure drop in bends (Chisholm 1971) is developed so that the two-phase multiplier can be evaluated from a simple equation, avoiding the use of a graphical procedure. In addition the method is compared for the first time with data at higher density ratios ( p t ] p a = 560).
2. P R E S S U R E DROP A T T R I B U T A B L E T O A B E N D

The pressure drop attributable to a bend can be defined as

Apb = Pa -- Pc + Dpaza + Dpczc ,

[11

where the subscripts a and c refer to properties at points a and c distant za and zc upstream and downstream of the bend; Pa and Pc are static pressures at points a and c; and equilibrium flow exists at a and c and the static pressure gradients at these points are Dpa and/)Pc. The pressure drop attributable to the bend in single-phase flow is frequently estimated using a resistance or pressure drop coefficient k defined by the equation
6 2

Apb = kgz-.
~p

[2]

3. AN E L E M E N T A R Y MODEL

Romie, as quoted by Hoopes (1957), developed an equation which in terms of the momentum flux can be expressed
MF MFLo = x ~ p.b_ .+ (1 - x ) a p~ 1- a

2 [3]

where
MFLo = G2/pL .

[4]

This can be approximated (Chisholm 1971), except where the density ratio approaches unity, as

MFLo =

1+ ( p ~ -

[5]

While [5] is not a precise transformation of [3] when the phase density ratio approaches unity, nevertheless [5] holds at the critical point.
363

364

BRIEF O M NC TO C M U I AI N

For flow in a pipe of uniform cross-section, a change of momentum flux due to a change in velocity ratio K can therefore be expressed A(MF) 1)x ( 1

MFIo - ( ~ -

-x)A(1) "

161

This equation of course assumes an incompressible and non-evaporating flow. Let us assume that the bend separates the two phases with a resultant increase in the velocity ratio within the bend, then downstream of the bend there will be a further pressure loss as the momentum flux increases to the equilibrium value. Assume also that in single-phase flow the downstream effects are small, and the single-phase loss occurs within the bend, then the two-phase pressure drop, where the plane of the bend is in the horizontal plane, can be expressed as

APh = APhLo4~o + AfMF),

I71

where the change of momentum flux is obtained from [6] with A(I/K) corresponding to the change in velocity ratio discussed above Assume the two-phase multiplier within the bend is given with sufficient accuracy by homogeneous theory

From [2]
G2
APbLo : kLog'--

zpl,

191

From [41 and [6]-[9]

/~PbLO = 1 + (PL-- ] Apb

[1o1

where

kLo 4. THE C O E F F I C I E N T B

It has already been shown (Chisholm 1%9) that the assumption that nO/K) is a constant gives values of B following the general trend of experimental results for 90 bends, but overpredicts B at higher values of the pipe radius-to-diameter ratio (R]D). Further examination has now suggested the following empirical equation 112]

The magnitudes obtained from this equation are consistent with the model. From [11] and [12] B=I+ -

2.2

[131

kLo (-

L)l

BRIEFCOMMUNICATION

365

Table 1. Comparison of predicted and experimental B-coefficients for 90 bends Test series R

1 0 1.25 1.8 1.9 1,2 Tee

2 1.0 0.310 3.4 3.4 1,2 Disturbance upstream

3 1.5 0.174 4.5 4.6 1,2

4 1.5 0.282 3.4 3.2 1,2 Disturbance upstream

5 2.36 0,250 Figure 1 3.0 3

6 5.0

7 5.02

kLo
B Experiment Equation [12] Author Comments

0.234 0.300 2.4 Figure 1 2.3 2.0 1,2 3 k smaller in test 6 as higher Re

Authors: l--Chisholm (1971); 2--Fitzsimmons (196,1); 3--Sekoda et al. (1969)

Table 1 compares previously reported values of B for 90 bends in the horizontal plane with predicted values using this equation; the agreement is to within 6 per cent. Equation [13] successfully correlates the data (test series 2 and 4) with an upstream disturbance (a change of section) which had not previously been satisfactorily correlated. 5. COMPARISON AT HIGH DENSITY RATIOS Empirical values of B were previously obtained from data of Fitzsimmons (1964) with steam-water mixtures at pressures in excess of 55 bar (PL/pZ <~27.0). The data of Sekoda et al. (1969) were obtained with air-water flow in 90 bends at 1.5 bar (PL/PC--560). These data therefore allow the procedure to be checked at higher density ratios. The form of data presentation used by Sekoda et al. necessitates the use of the equation

Apb = 1 + c[APb~ ] '12~ aPbz

'

[14]

which can be transformed (Chisholm 1973). to

Apb = 1 + (F 2 - 1 ) { B x 2 - n / 2 ( 1 - x ) 2 n/2 + X 2 - n } , ApbL0

[ 15]

where

r2= ap~co = ko__~o = & (go ~" aaL

ApbLO kLo PG
and approximately C = BF.

PG \I"~L/ '

[161

[171

In Fitzsimmon's tests, the resistance coefficient k was independent of Reynolds number (Re-106), whereas in Sekoda's tests k was slightly a function of Re(n = 0.08). Where k is independent of Re, [15] reduces to [10]. For an air-water mixture at 1.5 bar with n = 0.08
r = (560)1/2/(56)o-4 - 2 0 . [18]

BRIEFCOMMUNICATION

"L~'~_ _~,,.~ ~
, x

k~

D
o

= 18mm !

,./D = 2 3 6

,q,/D=5.02

x~,

t;?/O= 2 3.6. C= 6,0

8=20

I0

40

60

0 I00

(APbL I A Pt~) l/2

Figure 1. The root of the two-phase multiplier for 90o bends to a base of the Lockhart-Martinelli parameter.

Using [13] and [17] gives therefore for Sekoda's tests

R/D= 2.36 C = 6 0 R/D=5.02 C = 4 0 RID = ct C = 20.

For simplicity the variation of n with R/D is ignored here. Figure 1 shows that [14] with these values of the coefficient C give good agreement with experiment. This tends to confirm that [13] can be used at high density ratios, though consideration of the model does not necessarily suggest this. While Sekoda's analysis has necessitated consideration of the dependence of k on Re, this is a refinement not yet justified in practice; assume in practice n = 0 and k~o = kLo. 6. BENDS OTHER THAN 900 BENDS IN THE HORIZONTAL PLANE Little evidence is available of the effect of the plane of the bend. Data of Peshkin (1961) for flow in rectangular channels with a horizontal inlet and vertical outlet showed little difference between downward and upward flow at outlet. The present method is recommended meanwhile with all geometries. For bends other than 90, pending experimental confirmation, our recommendation is to use [13] to evaluate the coefficient B. With a 180 bend for example, as k will be larger for the same RID thant the 90bend, both B and the two-phase multiplier will be lower than for the 90 bend. This is consistent with trends observed with tests on 90 and 180 bends using gas-solid mixtures (Uematsu 1964).
7. C O N C L U S I O N S

It has been demonstrated that, at least for the available data, the two-phase multiplier for a 90 bend can be evaluated using [10] with the coefficient B evaluated from B = 1~ 2.2 [13]

Acknowledgements--This not is published by permission oof the Director, National Engineering

Laboratory, Department of Industry. It is British Crown copyright reserved.

BRIEF COMMUNICATION

367

REFERENCES

CHISHOLM,D. 1969 Prediction of pressure losses at changes of section, bends and throttling
devices during two-phase flow. In Designing, For Two-phase Flow. Report of a meeting at NEL, 17 January 1968. Part IV. NEL Report No. 388. East Kilbride, Glasgow: National Engineering Laboratory. CHISHOLM,D. 1971 Prediction of pressure drop at pipe fittings during two-phase flow. 13th Int. Cong. of Refrigeration, 1971, Washington, Vol. 2, pp. 781-789. International Institute of Refrigeration, Paris. CHISHOLM,D. 1973 Pressure gradients due to friction during the flow of evaporating two-phase mixtures in smooth tubes and channels. Int. J. Heat Mass Trans[er 16 347/355. FITZSIMMONS,D. E. 1964 Two-phase pressure drop in piping components. HW-80970, Rev. 1. General Electric Hartford Laboratories, Richland, OH. HooPES, J. W. 1957 Flow of steam-water mixtures in a heated annulus and through orifices. AIChE J. 3, 268-275. PESHKIN, M. A. 1961 About the hydraulic resistance of pipe bends to the flow of a gas-liquid mixture (in Russian). Tepolenergetika 8 79--80. SEKODA,K., SATO,Y. KARIYA, 1969 Horizontal two-phase air-water flow characteristics in a S. disturbed region due to 90 bend. Trans. Japan Soc. Mech. Engrs 35(279), 2227-2233. UEMATSU, 1964Pneumatic conveyance of granular solids through a pipe. In The 1st Australasian T. Con[. on Hydraulics and Fluid Mechanics, Perth, Australia (Edited by Sylvester, R.), pp. 69-80. Pergamon Press, Oxford.