You are on page 1of 10

Nuclear Engineering and Design 39 (1976) 257-266 North-Holland Publishing Company

A STUDY OF THE CRITICAL FLOW MODELS USED IN REACTOR BLOWDOWN ANALYSIS K.H. ARDRON and R.A. FURNESS Central Electricity Generating Board, Berkeley Nuclear Laboratories, Berkeley, Gloucestershire, UK Received 11 February 1976 Two-phase critical flow models widely used in safety calculations are compared with extensive published experimental measurements of the choked flowrates of steam-water mixtures. Comments are made on the applicability of the models in their supposed regions of validity and a mathematical upper limit to discharge flowrate is derived. The data fall below this limit and generally above a homogeneous isentropic flow model in thermal equilibrium. Suitable measurements of critical flowrates in pipes of diamater in the range of direct relevance to reactor blowdown calculations are still unavailable.

1. Introduction The ability to calculate the discharge rates of single and two-phase fluids from pipes, nozzles and orifices is particularly important in the safety analysis of water-cooled nuclear reactors. The discharge flowrate controls heat transfer in the core and the rate of system depressurization, thus influencing both the design of emergency core cooling systems and the containment. Theoretical models to predict single-component two-phase discharge flowrates have been advanced which assume homogeneous flow with the phases in thermodynamic equilibrium [ 1], separated flow in thermodynamic equilibrium [ 2 - 4 ] , and nonequilibrium between phases [ 5 - 7 ] . These have been reviewed in detail by Furness [8] who concluded that present theories fail to satisfactorily account for the observed effects of pipe length, diameter and entrance geometry on the critical flowrate. Since there is no completely general theory of two-phase critical flow each model is only applicable in a limited range of conditions. The object of this paper is to compare the various theories and the extent to which they agree with each other and with the published data in their supposed regions of validity.

through an outlet, the flowrate (this term is defined in the appendix) is always less than a certain maximum or 'critical' value. After this critical condition has been reached further reductions in receiver pressure leave the flowrate unaltered, serving only to form a steep pressure gradient at some location in the outlet passage. The outlet section in which the pressure gradient becomes infinite is referred to as the 'exit plane', and flow properties in the exit plane are said to have attained 'critical' values, being virtually independent of receiver pressure. The critical flowrate of a one-component twophase mixture is controlled by the degree of interphase heat, mass, and momentum transfer occurring as the fluid travels between the stagnation state * and the exit plane. Different assumptions about these processes have led to the formulation of the different flow models described below. For convenience the models will be placed in two categories: (1) equilibrium theories which assume that thermodynamic equilibrium exists throughout the expansion, and (2) non-equilibrium theories which allow for some departures from an equilibrium path.

2. Theories of critical two-phase flow


When a fluid expands from a compressed state to arbitrary ambient (or 'receiver') conditions in passage 257 * The stagnation state is defined, as is usual, as the thermodynamic state of a fluid particle were it to be brought isentropically to rest.

258

K.H. Ardron, R.A. Furtu~ss / Critical/low models used in blowdown analysis

2.1. Equilibrium theories


Several models have been based on the assumption of complete inter-phase thermal equilibrium, which is expected to be valid in flows where fluid transit time is long compared with the timescale of the evaporation process. This condition will be satisfied in the limit of very long, or very large bore pipes [9]. The theory of Moody [3] has been widely used in reactor depressurization analysis. The theory is based on the assumption of isentropic annular flow with no liquid entrainment and a critical phase velocity ratio (or slip ratio) given by: u(g)
u(L) -K =

~ F o u s k ~ S l i p

[2J Theoretical
Ratio
=

"~

Xe

"----~.~ip

Ratio

ell

[p(L)ll/3
~,-~/

(1)
Xe 0 0 2
~,~ ~ Linni;,~ ~0~ Klingebil A K lingebi=l K lingcbil Q Kf/ngebiel [IO] Ill] ~lll [If~ [Ill O = 4 7ram D . 1 2 7ram D=12.Tmm D=I2 7ram O=12 7ram

which is shown by Moody [3] to maximize the isentropic flowrate with respect to K. Direct measurements by Linning [10] and Klingebiel [11] shown in fig. 1, suggest that critical slip ratios of steam-water mixtures have a strong quality dependence and are much smailer than the predictions of eq. (1). Similar results can be inferred from steam-water jet momentum measurements discussed by Moody [12]. The observed tendency a r K to approach unity as pressure increases is also supported
4"0E_.xperimen t a l ~' Manely [I] Nozzle. D =ll mm 3-5- Fauske [2] L / D = 4 0 8 , D - 3 m m

.2

~Z3 \

u >

~ =o.o5

-...o Xe=O'31
~o I!0 2~0
Exit

static pressure (bars)

3tO

4~0

X: O.70 5!0

Po =8-5bors
Po - 6 " C h a r s

Fig. 1. Comparison of experimental and theoretical slip ratios for s t e a m - w a t e r mixtures.

A Faletti [IS] Annulus, D.14mm

Po .7"3 bars, Model Predictions /-'-" Upper bound flow (Frictionless} --Homogeneous Equilibrium Theory (Moneely [ I |)(Frict ionless) -- -- Moody [3] (Frictionless) .......... Storkmon t al [SJ (Frictionless) w , - - Homogeneous Frozen Theory (Henry t Fauske |7])(Frictionless) .... Henry e Fousk. [7](Frictionless)

Friedrich [32] L/D=[,S . D=4mm Po=6.Obars

3"0.
0
X

2s

"~v,~ ~

2"0. ~ - ~

~ -'~ ~
"-~.
_ _ _

%
1"5.

~~
_ _ _

o
o - o o l . . . . . .

Reservo.ir

b!o,

stagnation quality X o

~!,

,!o

Fig. 2. Comparison of critical flow models with data for Po = 6.9 bar (100 psia).

K.H. Ardron, R.A. Furness / Critical flow models used in blowdown analysis

259

N.
4.0'

~
"~ ~ ~, \ ~ ~'x ~X~ ~ . ~ ~ . ~ ~ .'.'/ --

Model Predictions
Upper bound flow (Frictionless) Homogeneous Equilibrium Theory (Monely [ I ])(Frictionles ) - - - - Moody [3] ( F r i c t i o n l e ) .......... Starkmon et al [5](Frictionles) - - . - - Homogeneous Frozen Theory (Henry Fouske [7])(Frictionless) .... Henry Fausk [7](Frictionles)

35

V "'"" "'~ "" ~ . .

~r

3.02.5-

"\

x
~

. ~ ..........................
Exaerimenta, ~

~ , ~

~.. 2"0-

1"5-

" ~ Q ~~: - . :

"\~
~.~=~.~ ~,

,7

~l.O0

Moneety Ill Nozzle D = 6 mm P0--3~5 bar Fr=edrich [32] L / D = I ' 5 . D=4mm P o = 3 ~ S b o r s

0"5-

0
0.001

. . . . . . 6.'o, . . . . . . Reservoir stagnation quality

6!,
Xo_

. . . . .

i! O

Fig. 3. Comparison of critical flow models with data for P0 = 34.5 bar (500 psia).

Experimental.
~ 8"0, ~ . ~ ' ~ 7.0
,It

5tarkman etal [5] Nozzle D =6ram , Po =65-5 bar L/D:O.3 . D:12"7mm . D=12.7mm 4~L/D=40 LID= , D=12"7mm.

. ~.~.

& L/D=3-5

53 .

~
~ ~ "~

~ ~,~ " ~

0 LID=4S,

L/ " 0 = 6 . 5 . D=I2 7ram L / D = 12 . D=12.7mm 21 , D=12.7mm

O=12-7mm

L/ + , "D = 13 X LID = 9

eL/De 2 . D:

D=12.7mm / P e 620.8bars |.., and .gmm |5uthcrlond . D= 54 mm t [21] . D= 76 mm )

~",~ ~ . .

"~L/D=

0 6.o.
x u 5'0. m n .
3.C

~%=(~riedrich

[32] L / D = I 5

D=4mm , Po=bO.Sbar

A ..... . . ^ ----._... _ 7 V ~ & . _ . ~ v . . . .~ ~ . ~&~"- ~ ~ N. ~ ~0 ~ ~ ....~. ~ ~ ~ ~*~

o2.o
U.

.. ~ == ~ '" ~ . ....... ........


_ v

=a

. ~

~.o:---'~...._-- ~ , ~ ......- ~ . . .....

C ~ -~ ~ i . : ~ . --'~.'-~::

.'".;.... ~ _

~%

\.

~----..o~e.~.F ..... Theo. I [Henry , Fauk[7=Frictionl) ~ou. ]~,,~..,on,e, ~ \\ ~ - - ~ ..... ~

I / / . " Upper bound flow (Frictionle) I Homogeneous Equilibrium Theory I (Moneely [ I ] ) (Frictionle) I - - - - M o o d y [3] (Frietionles) I ......... 5torkmon I t al [S] (Frictionle)

I.O.

0-001

.... b!ol . . . . . . b!l Reservoir stagnation quality Xo

.......

i.~o

Fig. 4. Comparison of critical flow models with data for Po = 62 bar (900 psia).

260

K. tl. Ardron, R.A. Furness / Critical flow models used in blowdown analysis

by measurements on air-water mixtures by Vogrin [13] although some of Vogrin's results were not in agreement with the results of later experiments by Fauske [14]. In view of the above, the agreementshown in fig. 2 between the Moody [3] theory and the small bore pipe data of Fauske [14] and Faletti [15] is surprising and has led several authors (e.g. Simon [9], Porter [20]) to suggest that appreciable slip only occurs in critical discharge from either very long pipes or from pipes of very small bore; this inference is supported by the tendency of measured critical tlowrates in nozzles to fall below the predictions of Moody's [3] theory, as is seen in figs. 2-4. This over-estimate can only be explained by a departure from the Moody slip value.

In an earlier isentropic equilibrium theory, Fauske [19] used a slip value o f K = (p(L)/p(g))l/2 which minimizes the momentum flux at the outlet. The use of this higher slip Value gives critical flowrates very close to those predicted by Moody [3]. A homogeneous * equilibrium theory [ 1] has been widely used to predict critical flowrates from long pipes. Inspection of fig. 5 shows good agreement between predictions of this model and measured saturated water critical flowrates in long tubes D = 6 ram, LID = 40. This theory provides a lower bound to all the data reported in this paper.
* In h o m o g e n e o u s flow t h e o r i e s the m i x t u r e is t r e a t e d as a single-phase p s e u d o - f l u i d w i t h average p r o p e r t i e s , prec l u d i n g relative m o t i o n b e t w e e n the phases, i.e. K = l.

Experimental 0 L/D. 0 : L / D ' I 2 t Fauske [19] , D-6mm L/D=3 . . L / D=40 L/D:3'S ~ L'/O =18 Sozziand Sutherfand 12.0- ~[~ L/D=6. s ]IKL/D=29} [21[ , D-12.Tmm r 0 Friedrich [32] I-/D=I.5 , 0 = 6ram Uchida Nariai [331 LtD-I2 . D I 4 mm I1"O- O Zaloudek [34[ L/D = 6 , D = 12 7 mm ./ IO-O

0 0 0 0 x ~ 6.0 ~5"0
~ 4'0

7.e

O L.

U.

0 I

3"0

~7"'/J
i

,,///
~1)%

~ ~

~[['""

---

2,0

' /
~

< ~" .W'~....-'"

Model Predictions Upper tx)und flow (Frictionlem) ~ Homos(nt~ul Equilibrium Theory [Maneely [l~Frictlonlesl )

I'O' 0

P5

30 45 Stagnation

I-.--Henry Fouske [7](Frictionless) [....-..Edwards |17~. L/D-12 I----Henry [18] /D 112 / (Frlctlonless) 60 75 90 105 120 pressure ( b a r s )

Fig. 5. Discharge o f s a t u r a t e d w a t e r t h r o u g h orifices, nozzles and pipes.

K.H. Ardron, R.A. Furness / Criticalflow models used in blowdown analysis 2.2. Non-equilibrium theories

261

A number of theories have explicitly allowed for the possibility of thermal non-equilibrium between phases due to finite evaporation rates. This is expected to be particularly important in flow through short pipes, nozzles, and orifices, where the fluid transit time is short. In the simple isentropic homogeneous frozen composition theory (discussed, for example, in ref. [7]), non-equilibrium is treated by imposing a condition of zero quality change between stagnation and the throat. The vapour is assumed to expand isentropically and independently of the liquid fraction. The tendency of this limiting theory to over-estimate the nozzle flowrates shown in figs. 2-4 indicates that partial inter-phase thermal equilibrium is established even when the flow path is short. Starkman et al. [5] have used an approximate version of the homogeneous frozen composition theory in which the enthalpy of the liquid fraction is neglected in the energy balance. For steam-water flows this approximation breaks down for qualities less than 0.1 when the liquid enthalpy dissipation required to maintain constant quality becomes large. However, as discussed in section 3, this simple method provides a reasonable correlation of short pipe and nozzle data for stagnation qualities above 1-2%. Henry and Fauske [7] have refined the no-slip frozen composition picture by allowing for the existence of partial inter-phase equilibrium in the exit region, whilst maintaining a constant flow quality between stagnation and the throat. This amounts to the adoption of a reduced sonic velocity in the exit region. Inspection of the figures shows the model to be in reasonable agreement with short pipe and nozzle data (LID < 5) for qualities greater than 0.001. Below this limit, however, agreement is less satisfactory and the model underestimates the flow of saturated water through an orifice by 50% (see fig. 5). Edwards [ 17] has developed a numerical theory which allows non-equilibrium effects in the 'pipe' and 'exit' regions to be treated by use of a consistent bubble growth model. A pipe wall friction term is included. This theory has the advantage of providing a relation between discharge flow rate and pipe geometry and is applicable to any outlet configuration. The theory, however, does not model the flow field

at the pipe inlet and consequently cannot satisfactorily describe the effect of pipe diameter on flowrate, as discussed by Simon [9]. A simpler phenomenological model was used by Henry [18] to describe the critical flow of saturated water through long pipes (LID > 12). Non-equilibrium effects are described, in this theory by empirically determined parameters which characterize (i) the pipe length traversed before boiling initiation, (ii) the subsequent rate of vapour generation, and (iii) the non-equifibrium throat quality. Application of the method with a single choice of fitting parameters gives good agreement with the saturated water flowrates measured by Fauske [19] as is discussed in subsection 3.2 below, but Porter [20] has pointed out that more experiments will be necessary before this set of parameters can be confidently applied to the discharge from large bore pipes. It is of interest and importance that all the data and model predictions lie below the upper bound shown in the figures. To determine this bound, evaporation is forbidden and the flowrate is maximized with respect to slip by adopting the slip ratio of eq. (1). Friction and entrance losses are ignored. Details of this calculation; which is similar to an analysis performed numerically by Porter [16], are given in the appendix.

3. Detailed comparison between theories and experiment


The principal data trends and overall applicability of the models described in section 2 are summarized below. Most data considered were obtained in steady state tests when outlet diameters were necessarily small, D <~ 15 mm. Of all the reported transient tests known to the present authors only recent work by Sozzi and Sutherland [21] has involved determination of stagnation properties of the discharging stream, thus yielding data suitable for comparison with theory (see section 4).
3.1. Critical discharge o f saturated steam-water mixtures (X o > O)

Published data are plotted in figs 2-4. There is a shortage of suitable large bore pipe data (D > 20 mm) at all pressures.

262

K.H. Ardron, R.A. Furness / Critical/low models used in blowdown analysis

The short pipe and nozzle data, L/D ~< 5, are in reasonable agreement (-+ 20%) with the Henry and Fauske [7] theory down to very low qualities X -0.001; the simple non-equilibrium theory of Starkman et al. [5] gives a satisfactory representation (-+10%) of this data for X 0 2> 0.01. Moody's [3] model gives good agreement (-5%) only with the long narrow bore pipe data (D = 3 mm) at low pressures, P0 = 6.9 bar. The fact that this model predicts flowrates higher than those observed for stagnation qualities above 0.01 has led to its adoption for US water-reactor licensing calculations [22]. At pressures of reactor interest, P0 ~ 60 bar, data for pipes more than 20 dia. long lie within 20% of, although generally above, predictions of the homogeneous equilibrium theory.
3.2. Critical discharge o f saturated and subcooled water X 0 = 0

Measured critical flowrates of saturated water for a range of stagnation pressures 3.5 bar < P0 < 120 bar are plotted in fig. 5. Flowrates of initially subcooled water observed by Sozzi and Sutherland [21] follow similar trends. Short pipe and nozzle data (i,/1) < 5) fall up to 50% above the predictions of the Henry and Fauske [7] non-equilibrium theory, although this model is widely used with short pipe and nozzle configurations. Inclusion of irreversible flow contraction losses, or friction losses, makes the disagreement worse. With the adoption of an empirical liquid flow contraction coefficient CO = 0.65 [23], orifice flows closely follow predictions of Bernoulli's equation, indicating non-evaporating hydraulic flow (which is the upper bound flow described in subsection 2.2). The flow of highly subcooled water will, of course, always be hydraulic. Long pipe data (LID > 20) lie within 20% of, but generally above, the homogeneous equilibrium line, even without the inclusion of frictional losses. Henry's [18] model (discussed above) describes these results to within 5%.

4. Application of discharge models to large scale blowdown analysis


A particular critical flow model will relate the

two-phase choked flowrate to the stagnation enthalpy and pressure of the fluid close to the inlet of tile exit pipe or nozzle. The correct choice of discharge model is not, therefore, sufficient to determine the blowdown characteristics of a large vessel; details of the time-varying fluid quality in the neighbourhood of the outlet must also be known. Several workers have obtained an approximation to the inlet stagnation quality by assuming that the voids within the vessel are distributed according to tile simplified phaseseparation model incorporated in the RELAP blowdown codes [24] and have thus been able to compare observed blowdown pressure histories with those predicted by use of the critical flow theories. On this basis it has been usual to characterize observed blowdown flowrates by a gross parameter, the Moody multiplier CM, which can be defined as the time-averaged ratio of the experimental instantaneous blowdown flowrate to that given by the use of the Moody [3] theory with the adopted assumption about voidage distribution * Reported values of Moody multipliers for a large number of published blowdown experiments are listed in table 1. The fact that CM has been always found to be less than unity is probably due to the effects of (a) rapid formation of high vapour concentrations close to the inlet to the exit pipe, and (b) a tendency of the Moody theory to over-estimate discharge flowrates for stagnation qualities greater than 1% (see section 3). An attempt to calculate the voidage distribution in greater detail has been made by Henry [25] in the analysis of vessel blowdown data reported by Hutcherson et al. [26]. Henry attempted to determine the true inlet stagnation quality by considering the growth of surface voids within the vessel (ignoring bulk nucleation). His method, however, remains to be confirmed by direct quality measurements. Because the inlet stagnation quality has apparently been measured onlyin tests using outlet diameters up to 75 mm it has not been possible by direct comparison to determine which of the critical flow models considered in this note best describes discharge flows from the large breaks typical in a reactor blowdown calculation (D ~> 200 ram). The margin of uncertain* CM so defined has no physical basis and is distinct from the flow contraction coefficient CD applicable to singlephase flow through an orifice.

K.H. Ardron, R.A. Furness / Critical flow models used in blowdown analysis

263

Table 1. Moody flowrate multipliers calculated in blowdown of large pressure vessels Reference [281 Initial conditions Po = 39 bar (saturated) Po = 39 bar (saturated) P0 = 67 bar (saturated) Break location bottom bottom Orifice dia./Pipe dia.= % area 4 8.16 Moody multiplier 0.85 0.73

[28]

bottom bottom bottom bottom

2.04 4.0 8.16 12.8 4 8.16 12.8 50 7.5 10 12.5 15 25 2

0.80 0.73 0.63 0.6 0.73 0.63 0.60 0.60 0.8 0.8 0.7 0.7 0.65 0.8

[28]

Po = 98 bar (saturated)

bottom bottom bottom bottom bottom bottom bottom bottom bottom bottom

[29] [30]

Po = 52 bar (saturated) Po = 52 bar (saturated)

[31]

Po = 155 bar (subcooled) Po : 39 bar (saturated) P0 = 67 bar (saturated) Po = 155 bar (subcooled)

[28]

top top top top top

0.51 4 0.51 2.04 10

0.75 0.6 0.8 0.65 0.6

[281

[28]

ty is greatest in the prediction of critical flowrates of low quality and subcooled fluid (X 0 < 0.01). This has particular importance in analysis of reactor blowdown since in several postulated break locations a reactor 'vessel' is expected to deliver low quality fluid (X 0 < 0.01) until the coolant mass is almost entirely depleted [27].

5. Conclusions
A comparison with all available measurements of the critical flowrates of s t e a m - w a t e r mixtures (in

selected regions of stagnation pressure) has shown that the critical flow models in common use agree with observations only in a restricted range of thermodynamic and geometric conditions. The worst discrepancies appear in the prediction of low quality flows in pipes of length less than 12 dia. where neither the theories of Moody [3] or Henry and Fauske [7] predict the approach to Bernoulli flow which can be observed in practice. Uncertainties here can have an important influence on the predicted form of depressurization transient in a water-reactor loss-of-coolant accident. All reported data have been found to lie in a region

264

K.H. Ardron, R,A. Furness / Critical flow models used in blowdown ana(vsis
integral form leads to a relationship between flow variables in the exit region and the stagnation conditions. For steady uniform isentropic flow this can be written 2"~eLg .(g)2e + l ( l - Xe)U(L)2 =

b o u n d e d f r o m b e l o w be p r e d i c t i o n s o f the h o m o geneous e q u i l i b r i u m m o d e l a n d from above b y a n o n - e v a p o r a t i n g flow limit derived in this paper. Wide b o r e pipe data d i r e c t l y relevant to r e a c t o r b l o w d o w n safety c a l c u l a t i o n s are lacking over the e n t i r e range o f s t a g n a t i o n qualities. T h e r e is an absence o f a general critical flow t h e o r y w h i c h satisfactorily describes o b s e r v e d effects o f o u t l e t g e o m e t r y on discharge flowrates, and can thus be c o n f i d e n t l y applied in r e a c t o r b l o w d o w n calculations.

= Xo(h g)
- (h~g) h~L))(Xe - Xo). (A4a)

With the use of eqs. (A1)-(A3) eq. (A4a) can be written in an alternative form derived by Moody [3]:

ho-he= 2
Acknowledgements This p a p e r is p u b l i s h e d b y p e r m i s s i o n o f the Central Electricity G e n e r a t i n g Board.

G2 [-Ke(1 ~- X e ) + X e 7 2

L p~L)

V +l- l p~g)j LXe K~X ej'

(A4b)

where h is the enthalpy per unit mass of the flowing fluid h = Xh (g) + (1 X)h (L) .

A2. M a x i m u m ]low conditions


A p p e n d i x : D e r i v a t i o n o f m a x i m u m critical f l o w r a t e To obtain the quantities X, K, p(L), and p(g) in the exit region, and hence determine G from eq. (A4b), details of the thermodynamic path followed by a fluid particle during transit between stagnation and the throat are required. However, to determine an upper limit to the critical discharge it is only necessary to assign values to these variables (treated as independent) which maximize the outlet flow. Now consideration of eq. (A4a) shows that since the enthalpy of evaporation h (g) ' h (L) is positive the kinetic energy flux, and hence the flowrate, increases with decreasing outlet quality. For our bounding model we therefore preclude flashing of the liquid phase in the flow passage, and adopt a simple constant quality condition X e = X o. (Condensation, which is allowed strictly in an isentropic expansion of the vapour phase, is forbidden merely for simplicity; for a steam-water blowdown the 'error' so introduced in the bound is only significant at the lower mass flows appropriate to stagnation qualities greater than 30%.) With the constant quality condition inter-phase heat transfer, which is irreversible, is precluded by the isentropic assumption implicit in eq. (A4). Therefore the vapour expansion follows an isentropic law: p(g) = p~g) r/1/'r , (A5)

In this appendix we present the model of critical discharge which gives the upper-bound mass flow rates in figs. 2-5. The analysis is independent of the flow regime, and is appropriate to uniform steady isentropic two-phase flow through a general outlet configuration such as a duct, nozzle or orifice.

A1. Conservation equations


In two-phase flow the flowrates of the separate phases w, and the 'flowrate' G, are defined by w(g ) = p(g)u(g)a, w (L) = p(L)u(L)(1 G = w (g) + w (L) , c~), (A1)

where the void fraction c~ is defined as the fraction of the total flow area occupied by the gas phase. The flow quality and the slip ratio, respectively, are defined by

X = w(g)/G,

K = u(g)/u (L) ,

(A2)

where r/= P/Po is the pressure ratio. The liquid phase is treated as incompressible so that p(L) = const. Since X and p(L) are fixed, the necessary conditions that the outlet flow is a maximum are

from which, using eqs. (A1), there follows a general relation between void fraction and flow quality: p(L) . Application of the conservation of energy equation in (A3)

(aG/aK)r~ = 0

and

(aG/arl) K = 0

(A6)

Eqs. (A6) are independent simultaneous equations in the critical outlet parameters r/* and K* which maximize the flow when inserted in eq. (A4b). Moody [3] has shown that the first of eqs. (A6) is satis-

K.H. Ardron, R.A. Furness / Critical flow models used in blowdown analysis
fied when the slip value is given by eq. (1); hence, using eq. (A5): K* = (p(t)/p(g)*)l/3 = (p(L)/p~g))l/3 n . - 1 / 3 3 ' . (A7) G h K L P S u w X 3' n p = total mass velocity (=w (g) + w(L)), abbreviated to 'flowrate' = enthalpy = slip ratio (=u(g)/u (L)) = pipe length; length of convergent section of nozzle = pressure = entropy = velocity = mass velocity (= Ou) = quality or dryness fraction = void fraction = specific heat ratio of vapour phase = dimensionless pressure (= P/Po) = density.

265

If the second of conditions (A6) is applied to eqs. (A4b) and (A5) a second relation between r~* and K* is obtained: bo (1 - r/*)-- + 1 (1 - n,l_(1/3") ) 3" 3 " - 1

-1~*1-(1/3')(1
where

+b*)(1 + K'b*)= 0 ,

(AS)

[ 1 - X~ p (g) b= ~ X ]p(L) '


and in forming eq. (AS) we have used the thermodynamic identities valid in constant quality isentropic flow:

Superscripts
( )* = critical condition ()(g) = vapour property ()(L) liquid property

Subscripts
( )o ( )e ()~ = stagnation or reservoir property = exit plane property = ambient conditions.

(ah)

XI-X
+ p(L) '

a-P S = 0 - ~

ho - h = XoC~g)2{(1 - n ) ~

+~ 1

(1 - r71 - (1/3'))) References

A3. Maximum critical flowrate


Eliminating K* between eqs. (A7) and (A8) gives a transcendental equation in r/*, the critical pressure ratio. Solving for r/* the maximized flowrate Gma x is obtained from equation (A4b) f which, when eqs. (A.6) are satisfied, can be rearranged in the form: 2 _ long) c(0g)12r~.l+(1/3") 1__ 1 +b* Gmax Xo (1 + K'b*): 2 (A9)

"

Nomenclature b C CD CM D = dimensionless parameter = sonic velocity = contraction coefficient = Moody multiplier = orifice diameter; pipe internal diameter; nozzle throat diameter

~" Note that when 7" falls below the back pressure ratio 7,0, hydrodynamic flow is established across the exit plane. In this condition the mass flowrate must be obtained from eq. (A4b) with the exit pressure set equal to the ambient pressure. This means that in the saturated/subeooled liquid limit, Xo = 0, the bounding calculation predicts Bernoulli flow.

[ 1 ] D.J. Maneely, A study of the expansion process of low-quality steam through a de Laval nozzle, University of California, Livermore Report UCRL 6230 (1962). [2] H.K. Fauske, Contribution to the theory of two-phase, one-component critical flow, Argonne National Laboratory, Report ANL-6633 (1962). [3] F.J. Moody, Maximum flow rate of a single component two-phase mixture, Trans. ASME, J. Heat Transfer 87c (1965) 134. [4] S. Levy, Prediction of two-phase critical flow rate, Trans. ASME, J. Heat Transfer 87c (1965) 53. [5] S. Starkman, V.E. Schrock, K.F. Neusen and D.J. Maneely, Expansion of a very low quality two-phase fluid through a convergent-divergent nozzle, Trans. ASME, J. Basic Eng. 86D (1964) 247-256. [6] R.E. Henry, A study of one- and two-component twophase critical flows at low qualities, Argonne National Laboratory, Report ANL-7430 (1968). [7] R.E. Henry and H.K. Fauske, The two-phase critical flow of one-component mixtures in nozzles, orifices and short tubes, ASME paper 70-WA/HT-5 (1971). [8] R.A. Furness, A review of theoretical and experimental studies on the critical discharge flow rates of fluids, CEGB Report RD/B/N2997 (to be published) (1976). [9] U. Simon, Blowdown flow rates of initially saturated water, Topical Meeting on Water-Reactor Safety, Salt Lake City, Conf-73034 (1973) 172-195. [10] D. Linning, The adiabatic flow of evaporating fluids

266

K.H. Ardron, R.A. Furness / Critical flow models used in blowdown analysis
in pipes of uniform bore, Proc. Inst. Mech. Eng. (London) 18 (2) (1952) 64. W.J. Klingebiel, Critical flow slip ratios of steam-water mixtures, Ph.D. thesis, University of Washington (1964). F.J. Moody, Prediction of blowdown thrust and jet forces, ASME paper 69-HT-31 (1969). J.A. Vogrin, An experimental investigation of twophase two-component flow in a horizontal convergingdiverging nozzle, Argonne National Laboratory, Report ANL-6754 (1963). H.K. Fauske, Two-phase two- and one-component critical flow, Proc. of Symposium on Two-phase Flow, Vol. 3 SG101, University of Exeter (1965). D.W. Faletti, Two-phase critical flow of steam-water mixtures, Ph.D. thesis, University of Washington (1959). W.H.L. Porter, A method for analysing critical flow of steam water mixtures, European Two-phase Flow Meeting, Paper WP/a, Haifa (1975). A.R. Edwards, Conduction controlled flashing of a fluid and the prediction of critical flow rates in a onedimensional system, UKAEA Report AHSB(S)R147 (1968). R.E. Henry, The two-phase critical discharge of initially saturated or subcooled water, Nucl. Sci. Eng. 41 (1970) 336-342. H.K. Fauske, The discharge of saturated water through tubes, Chem. Eng. Progr. Syrup. Ser. 61 (1965) 210. W.H.L. Porter, Unpublished work (1973). G.L. Sozzi and W.A. Sutherland, Critical flow of saturated and subcooled flow at high pressures, General Electric Company, Report NEDO-13418 (1975). USAEC Docket RM-50-1, Public rulemaking hearing on acceptance criteria for ECCS for light-water-cooled nuclear power plant, Concluding statement of the regulatory staff (1973), W.J. Clarke, Flow Measurement, Pergamon Press, Oxford (1965). [24] W. Rettig, G. Jayne, K. Moore, C. Slater and M. Uptmor, Relap 3: A computer program for reactor blowdown analysis, Idaho Nuclear Corp., Report IN-1321 (1970). [25] R.E. Henry, Private communication (1975). [26] M.N. Hutchinson, R.E. Henry and D.E. Wallersheim, Experimental measurements of large pipe transient blowdown, Trans. 1st. Conf. Eur. Nucl. Soc., Paris, Trans. Amer. Nucl. Soc. 20 (1975) 488 490. [27] R.T. Alleman, A.J. McElfresh, A.S. Neuls, W.C. Townsend, N.P. Wilburn and M.E. Witherspoon, tfigh-enthalpywater blowdown tests from a simple vessel through a side outlet, Battelle Memorial Institute, Richland, Washington, Report BNWL-1470 (1971). [28] H. Shimamune, M. Shiba, H. Adachi, K. Namatame and M. Sabajame, Current status of the ROSA program, Proc. CREST Specialists Meeting on ECC for LWRs, Munich (1972). [29] A.R. Edwards and C. Jones, An analysis of phase llA blowdown tests - The discharge of high enthalpy water from a simple volume into a containment volume, UKAEA Report SRD/R/27 (1974). [30] A. Azzalin, An experimental investigation on blowdown in pressure tube reactor conditions, CISE Report CISE-R-342 (1973). [31 ] C. Slater, Comparison of predictions from the reactor primary system decompression code (RELAP3) with decompression data from the semiscale blowdown and emergency core cooling (ECC) project, Idaho Nuclear Corp. Report IN-1444 (1970). [32] H. Friedrich, Flow through single-stage nozzles with different thermodynamic states, Energie 12 (1960) 3. [33] H. Uchida and H. Nariai, Discharge of saturated water through pipes and orifices, Proc. of Third International Heat Trans. Conf. Vol. 5 (1966) 1. [34] F.R. Zaloudek, The critical flow of hot water through short tubes, General Electric Company, Report HW77594, Hanford Works (1963).

[ 11 ] [12] [13]

[ 14 ]

[15] [16]

[ 17]

[18]

[19] [201 [21]

[22]

[23]

You might also like