You are on page 1of 16

4 Two-Phase Flow Regimes – I

4.1 Introductory Remarks


Gas–liquid two-phase mixtures can form a variety of morphological flow configu-
rations. The two-phase flow regimes (flow patterns) represent the most frequently
observed morphological configurations.
Flow regimes are extremely important. To get an appreciation for this, one can
consider the flow regimes in single-phase flow, where laminar, transition, and tur-
bulent are the main flow regimes. When the flow regime changes from laminar to
turbulent, for example, it is as if the personality of the fluid completely changes as
well, and the phenomena governing the transport processes in the fluid all change.
The situation in two-phase flow is somewhat similar, only in this case there is a
multitude of flow regimes. The flow regime is the most important attribute of any
two-phase flow problem. The behavior of a gas–liquid mixture – including many of
the constitutive relations that are needed for the solution of two-phase conserva-
tion equations – depends strongly on the flow regimes. Methods for predicting the
ranges of occurrence of the major two-phase flow regimes are thus useful, and often
required, for the modeling and analysis of two-phase flow systems.
Flow regimes are among the most intriguing and difficult aspects of two-phase
flow and have been investigated over many decades. Current methods for predicting
the flow regimes are far from perfect. The difficulty and challenge arise out of the
extremely varied morphological configurations that a gas–liquid mixture can acquire,
and these are affected by numerous parameters. Some of the physical factors that
lead to morphological variations include the following:

a) the density difference between the phases; as a result the two phases respond
differently to forces such as gravity and centrifugal force;
b) the deformability of the gas–liquid interphase that often results in incessant coa-
lescence and breakup processes; and
c) surface tension forces, which tends to maintain one phase dispersal.

Flow regimes and their ranges of occurrence are thus sensitive to fluid properties,
system configuration/and orientation, size scale of the system, occurrence of phase
change, etc. Nevertheless, for the most widely used configurations and/or relatively
well defined conditions (e.g., steady-state and adiabatic air–water and steam–water
flow in uniform-cross-section long vertical pipes, or large vertical rod bundles with
uniform inlet conditions) reasonably accurate predictive methods exit. The literature
also contains data and correlations for a vast number of specific system configura-
tions, fluid types, etc. Although experiments are often needed when a new system
121

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
122 Two-Phase Flow Regimes – I

Camera

Tube
Figure 4.1. A simple flow-regime-observation experi-
mental system for vertical pipes.

Air, QG Mixer

Liquid, QL

configuration and/or fluid type is of interest, even in these cases the existing methods
can be used for preliminary analysis and design calculations.
In this chapter the major flow regimes and the empirical predictive methods
for adiabatic two-phase flow in straight channels and rod bundles will be discussed.
The discussion of mechanistic models for regime transitions will be postponed to
Chapter 7, so that the necessary background for understanding these mechanistic
models is acquired in Chapters 5 and 6.
Also, in this chapter only conventional flow passages (i.e., flow passages with
DH ≥ 3 mm) and rod bundles will be considered. There are important differences
between commonly used channels and mini- or microchannels with respect to
the gas–liquid two-phase flow hydrodynamics. Two-phase flow regimes and condi-
tions leading to regime transitions in mini- and microchannels will be discussed in
Chapter 10.

4.2 Two-Phase Flow Regimes in Adiabatic Pipe Flow

4.2.1 Vertical, Cocurrent, Upward Flow


Consider the simple experiment depicted in Fig. 4.1, where steady-state flow in a long
tube with low or moderate liquid flow rate is considered. Experiment proceeds with
constant liquid volumetric flow rate QL , whereas the gas volumetric flow rate QG is
started from a very low value and is gradually increased.
The major flow regimes that will be observed are depicted in Fig. 4.2. We will
postpone for the moment discussion of the finely dispersed bubbly regime and focus
on the others.
In bubbly flow [Fig. 4.2(a)] distorted-spherical and discrete bubbles move in a
continuous liquid phase. The bubbles have little interaction at very low gas flows,
but they increase in number density as QG is increased. At higher QG rates, bubbles
interact, leading to their coalescence and breakup.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.2 Two-Phase Flow Regimes in Adiabatic Pipe Flow 123

(a) Bubbly (b) Dispersed (c) Slug (d) Churn (e) Annular/
Bubbly Dispersed
QG

Figure 4.2. Major flow regimes in vertical upward pipe flow.

Bubbly flow ends when discrete bubbles coalesce and produce very large bub-
bles. The slug flow regime [Fig. 4.2(c)] then develops; it is dominated by bullet-shaped
bubbles (Taylor bubbles) that have approximately hemispherical caps and are sepa-
rated from one another by liquid slugs. The liquid slug often contains small bubbles.
A Taylor bubble approximately occupies the entire cross section and is separated
from the wall by a thin liquid film. Taylor bubbles coalesce and grow in length until
a relative equilibrium liquid slug length (Ls /D ∼ 16) in common vertical channels
(Taitel et al., 1980) is reached.
At higher gas flow rates, the disruption of the large Taylor bubbles leads to churn
(froth) flow [Fig. 4.2(d)], where chaotic motion of the irregular-shaped gas pockets
takes place, with literally no discernible interfacial shape. Both phases may appear
to be contiguous, and incessant churning and oscillatory backflow are observed. An
oscillatory, time-varying regime where large waves moving forth in the flow direction
are superimposed on an otherwise wavy annular-dispersed flow pattern involving a
thick liquid film on the wall is also referred to as churn flow. Churn flow also occurs
at the entrance of a vertical channel, before slug flow develops. This is a different
interpretation of churn flow and represents the irregular region near the entrance of
a long channel where eventually a slug flow pattern will develop.
Annular-dispersed (annular-mist) flow [Fig. 4.2(e)] replaces churn flow at higher
gas flow rates. A thin liquid film, often wavy, sticks to the wall while a gas-occupied
core, often with entrained droplets, is observed. In common pipe scales, the droplets
are typically 10–100 μm in diameter (Jepsen et al., 1989). The annular-dispersed
flow regime is usually characterized by continuous impingement of droplets onto the
liquid film and simultaneously an incessant process of entrainment of liquid droplets
from the liquid film surface. Figure 4.3 depicts the cross section of a tube in the
annular-dispersed regime (Srivastoa, 1973).
The inverted-annular regime and dispersed-droplet regime, depicted schemati-
cally in Figs. 4.4, should also be mentioned here. These regimes are not observed
in adiabatic gas–liquid flows. They do occur in boiling channels, however. In the
inverted-annular flow regime a vapor film separates a predominantly liquid flow

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
124 Two-Phase Flow Regimes – I

Figure 4.3. Cross-sectional view of annular-


dispersed flow. (From Levy, 1999, based on
Srivastoa, 1973.)

from the wall. The liquid flow may contain entrained bubbles. This flow regime takes
place in channels subject to high wall heat fluxes and leads to an undesirable con-
dition called the departure from nucleate boiling. In the dispersed-droplet regime
an often superheated vapor containing entrained droplets flows in an otherwise dry
channel. This regime can occur in boiling channels when massive evaporation has
already caused the depletion of most of the liquid.
Flow regimes associated with very high liquid flow rates are now discussed. In
these circumstances, in all flow regimes except annular (i.e., all flow regimes where the
two phases are not separated), because of the very large liquid and mixture velocities
the slip velocity between the two phases is often small in comparison with the average
velocity of either phase, and the effect of gravity is relatively small. Furthermore, as
long as the void fraction is small enough to allow the existence of a continuous
liquid phase, the highly turbulent liquid flow does not allow the existence of large
gas chunks and shatters the gas into small bubbles. Bubbly flow is thus replaced
by a finely dispersed bubbly flow regime, where the bubbles are quite small and
nearly spherical [Fig. 4.2(b)]. No froth (churn) flow may take place; furthermore, the
transition from slug to annular-mist flow may only involve churn flow characterized
with the oscillatory flow caused by the intermittent passing of large waves through a
wavy annular-like base flow pattern.

Vapor Liquid Figure 4.4. Inverted-annular and dispersed-


droplet regimes.

Inverted-annular Dispersed-droplet

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.2 Two-Phase Flow Regimes in Adiabatic Pipe Flow 125

FLOW DIRECTION
O P Q R

10

ANNULAR
MIST
(Water
Dispersed)
Superficial Water Velocity, jL, ft/sec.

H I J K L M N

1.0

H s
E d) ) OT ase
BL rse
d FR h Ph ed)
B e UG rse ot ers
BU Disp SL ispe (B isp
r D D E D
A i B C ir F G
(A (A

0.1

0.1 1.0 10 100


Superficial Gas Velocity, jG, ft/sec.

Figure 4.5. Flow regimes for air–water flow in a 2.6-cm-diameter vertical tube. (From Govier
and Aziz, 1972.)

It must be emphasized that the flow regimes shown in Fig. 4.2 are the major and
easily distinguishable flow patterns. In an experiment similar to the one described
here, transition from one major flow regime to another is never sudden, and each pair
of major flow regimes are separated from one another by a relatively wide transition
zone. Figure 4.5, borrowed from Govier and Aziz (1972), displays schematics of

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
126 Two-Phase Flow Regimes – I

Gas, QG

Liquid, QL
Mixer

Camera

Figure 4.6. A simple flow-regime-observation experimental system for horizontal pipes.

flow regimes and their range of phase superficial velocities for air–water flow in a
2.6-cm-diameter vertical tube.

4.2.2 Cocurrent Horizontal Flow


Let us now consider the simple experiment displayed in Fig. 4.6, where we establish
a fixed liquid volumetric flow rate QL . We then start with a small gas volumetric flow
rate QG , and increase QG while visually characterizing the flow regimes.
First, consider flow regimes at low liquid flow rates. For “low liquid flow rate”
conditions assume QL is low enough so that during drainage of liquid from the pipe
when QG = 0, as shown in Fig. 4.7, the liquid occupies less than half of the pipe’s
cross-sectional height (i.e., hL < D/2). The major flow regimes are shown in Fig. 4.8.
The stratified-smooth flow regime occurs at very low gas flow rates and is character-
ized by a smooth gas–liquid interphase. With increasing gas flow rate, the stratified-
wavy flow regime is obtained, where hydrodynamic interactions at the gas–liquid
interphase result in the formation of large-amplitude waves.
The slug flow regime occurs with further increasing gas flow rate. In comparison
with the stratified-wavy regime, it appears as if the “waves” generated at the surface
of the liquid grow large enough to bridge the entire channel cross section. The slug
flow regime in horizontal channels is thus different from the slug flow defined for
vertical channels. The gas phase is thus no longer contiguous. The liquid can contain
entrained small droplets, and the gas phase may contain entrained liquid droplets.
The annular-dispersed (annular-mist) flow regime is established at higher gas
flow rates. The flow regime resembles the annular-dispersed regime in vertical tubes,
except that here gravity causes the liquid film to be thicker near the bottom.
The flow regimes at high liquid flow rates are now described. Referring to Fig. 4.7,
we are now considering cases where, in the absence of a gas flow, liquid drainage out of
the tube would result in hL > D/2. The major flow regimes are depicted in Fig. 4.9.

D
hL

Figure 4.7. Drainage of liquid out of a horizontal pipe.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.2 Two-Phase Flow Regimes in Adiabatic Pipe Flow 127

Figure 4.8. Major flow regimes in a horizontal pipe with low liquid flow rates.

The following flow regimes are observed as the gas flow rate is increased. In the
bubbly flow regime, discrete bubbles tend to collect at the top of the pipe owing
to the buoyancy effect. The finely dispersed bubbly flow regime is similar to the
finely dispersed bubbly flow pattern in vertical flow channels. It occurs only at very
high liquid flow rates. It is characterized by small spherical bubbles, approximately
uniformly distributed in the channel. The plug or elongated bubbles flow regime is the
equivalent of the slug flow regime in vertical channels. Finally, the annular-dispersed
(annular-mist) flow regime is obtained at very high gas flow rates.
It is once again emphasized that the flow patterns in Fig. 4.9 only display the
major flow regimes that are easily discernable visually and with simple photographic
techniques and are commonly addressed in flow regime maps and transition models.
Many subtle variations within some of the flow patterns can be recognized by using
more sophisticated techniques (Spedding and Spence, 1993). Figure 4.10, borrowed
from Govier and Aziz (1972), displays schematics of flow regimes and their range of
phase superficial velocities for air–water flow in a 2.6-cm-diameter tube.

(a) Bubbly

(b) Dispersed Bubbly

(c) Plug / Elongated Bubble

(d) Annular /Dispersed

Figure 4.9. Major flow regimes in a horizontal pipe with high liquid flow rates.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
128 Two-Phase Flow Regimes – I

20
S T DISPERSED BUBBLE
10

O
N Q R
Superficial Water Velocity, jL, ft/sec

LE ED
2.0
B AT
BU NG L SLUG M
B
LO

FLOW
1.0
E

DIRECTION
ANNULAR
G I MIST

F
J

H K
0.2
A B
C D
0.1
D E
E
FI
TI th D
A oo ce) FI
E
R m fa WAVE
ST (S ter TI ly
A ipp ce)
In R fa
ST (R ter
.02 In
0.1 1.0 10 100
Superficial Gas Velocity, jG, ft/sec

Figure 4.10. Flow regimes for air–water flow in a 2.6-cm-diameter horizontal tube. (From
Govier and Aziz, 1972.)

With regards to the two-phase flow regimes, the following points should be borne
in mind:

1. Flow regimes and conditions leading to regime transitions are geometry depen-
dent and are sensitive to liquid properties. The most important properties are
surface tension, liquid viscosity, and liquid/gas density ratio. Important geomet-
ric attributes include orientation with respect to the gravitational vector, the size
and shape of the flow channel, the aspect ratio (length to diameter) of the channel,
and any feature that may cause flow disturbances.
2. The basic flow regimes such as bubbly, stratified, churn, and annular-dispersed
occur in virtually all system configurations, such as slots, tubes, and rod bundles.
Details of the flow regimes of course vary according to channel geometry.
3. The apparently well-defined flow regimes described here do not represent a com-
plete picture of all possible flow configurations. In fact, by focusing on the flow
regime intricate details, it is possible to define a multitude of subtle flow regimes
(e.g., see Spedding and Spence, 1993). However, flow regime maps based on
the basic regimes presented here have achieved wide acceptance over time. The
regime change boundaries are generally difficult to define because of the occur-
rence of extensive “transitional” regimes.
4. Bubbly, plug/slug, churn, and annular flow also occur in minichannels (i.e., chan-
nels with 100 μm ≤ D ≤ 1 mm)
5. In adiabatic, horizontal flow, often for simplicity the regimes are divided into four
zones:
r stratified (smooth and wavy),
r intermittent (plug, slug, and all subtle flow patterns between them),

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.3 Flow Regime Maps for Pipe Flow 129

105

104

103
Annular Wispy Annular
ρGj2G (kg/m.s2)

102
Churn

10 Bubbly

Bubbly-Slug
1.0 Slug

0.1

10 102 103 104 105 106


ρL j2L (kg/m.s2)

Figure 4.11. The flow regime map of Hewitt and Roberts (1969) for upward, cocurrent vertical
flow.

r annular-Dispersed, and
r bubbly.
6. Flow regimes in boiling and condensing flows are significantly different than those
in adiabatic channels. They will be discussed later.

4.3 Flow Regime Maps for Pipe Flow


Flow regime maps are the most widely used predictive tools for two-phase flow
regimes. They are often empirical two-dimensional maps with coordinates repre-
senting easily quantifiable parameters. The coordinate parameters in the majority of
widely used maps are either the phasic superficial velocities (Mandhane coordinates,
after Mandhane et al., 1974) or include the phasic superficial velocities as well as some
other properties. Most of the widely used regime maps are based on data for vertical
or horizontal tubes with small and moderate diameters (typically 1 ≤ D ≤ 10 cm)
and for liquids with properties not too different from water. They also primarily rep-
resent “developed” conditions, with minimal channel end effects. Experimental data
and regime maps for a wide variety of scales, geometric configurations, orientations,
and properties, can also be found in the open literature.
The flow regime map of Hewitt and Roberts (1969) is displayed in Fig. 4.11. This
flow regime map is for cocurrent, vertical upward flow in pipes. The coordinates are
defined as

(Gx)2
ρG jG2 = , (4.1)
ρG
[G(1 − x)]2
ρL jL2 = . (4.2)
ρL

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
130 Two-Phase Flow Regimes – I

50 Dispersed
Froth
20 Annular
Wavy
Gx / λ (kg/m2.s)

10
5.0

2.0 Slug

1.0
0.5 Stratified Bubbly
Plug
0.2
0.1
10 20 50 100 1,000 5,000 20,000
G(1 − x)ψ (kg/m2.s)

Figure 4.12. The flow regime map of Baker (1954) for cocurrent flow in horizontal pipes.

The flow regime map of Baker (1954), shown in Fig. 4.12, deals with cocurrent
horizontal flow in pipes. The data base of this flow regime map is primarily air–water
mixture. The property parameters are defined as
1 
ρG ρL 2
λ= , (4.3)
ρa ρW
 1
 σ   μ   ρ 2 3
W L W
ψ= . (4.4)
σ μW ρL

and are meant to account for deviations from air and water properties. In these
expressions the subscript a stands for air, W for water, G for the gas of interest, and
L for the liquid of interest. In Eq. (4.4), σW represents the air–water surface tension
and σ is the surface tension of the gas–liquid pair of interest.
The flow regime map of Mandhane et al. (1974), displayed in Fig. 4.13, is probably
the most widely accepted map for cocurrent flow in horizontal pipes. The range of
its data base is as follows:
Pipe diameter 12.7–165.1 mm
Liquid density 705–1,009 kg/m3
Gas density 0.80–50.5 kg/m3
Liquid viscosity 3×10−4 –9×10−2 kg/m·s
Gas viscosity 10−5 –2.2×10−5 kg/m·s
Surface tension 0.024–0.103 N/m
Liquid superficial velocity 0.9×10−3 –7.31 m/s
Gas superficial velocity 0.04–171 m/s

4.4 Two-Phase Flow Regimes in Rod Bundles


The thermal-hydraulics of rod bundles is important because the cores of virtually all
existing power-generating nuclear reactors consist of rod bundles. Two-phase flow
occurs in the core of boiling water reactors (BWRs) during normal operations and
in pressurized water reactors (PWRs) during many accident scenarios.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.4 Two-Phase Flow Regimes in Rod Bundles 131

10
Bubbly

1.0
Plug Slug
jL (m/s)

Annular
10−1
Wavy
Stratified

10−2

10−3
10−2 10−1 1.0 10 102
jG (m/s)

Figure 4.13. The flow regime map of Mandhane et al. (1974) for cocurrent flow in horizontal
pipes.

Adiabatic experimental studies (i.e., experiments without phase change) using a


20-rod bundle (16 complete and 4 half-rods) with near-prototypical bundle height,
rod diameter, and pitch have indicated that the flow patterns include bubbly, slug,
churn, annular, and possibly dispersed-bubbly (Venkateswararao et al., 1982). In
bubbly flow, the bubbles are typically small enough to move within a subchannel
defined by four rods in bundles with rectangular pitch and three rods in bundles
with triangular pitch. The slug flow regime can have at least three configurations
(Venkateswararao et al., 1982): Taylor bubbles moving within subchannels (cell-type
slug flow); large-cap bubbles occupying more than a subchannel; and Taylor-like
bubbles occupying the test sections entire flow area in a 20-rod bundle (shroud-type
Taylor bubbles). The churn flow regime is characterized by irregular and alternating
motion of liquid and can result from the instability of “cell-type” slug flow. Figure 4.14

5.00
DISPERSED BUBBLE C
B
B C
A
1.00 D
0.50
BUBBLE CHURN
jL (m/s)

Annular
SLUG

0.10

0.05 EXPERIMENT E−1


THEORY E
A D
E−2
A D
0.01
0.01 0.05 0.10 0.50 1.0 5.0 10 50
jG (m/s)

Figure 4.14. The rod bundle flow regime data of Venkateswararao et al. (1982).

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
132 Two-Phase Flow Regimes – I

0.0 αBS αCD αSA αAM 1.0


Inverted
IAN/ Inverted Mist

)
Post-dryout annular

PO
(IAN) ISL slug (ISL) (MST)

)(M
IAN/ SLG/ ISL-

PO
Transition BBY/IAN ISL- SLG/ ANM/MST

/M
ISL

PR
SLG ANM

) (M
Bubbly Slug SLG/ Annular

PR
Pre-CHF ANM mist (ANM)
(BBY) (SLG)

(M
Unstratified
UTB
G
ρ Transition
0.5UTB TG – TI

Vertically stratified (VST)

0.0 αBS αDE αSA αAM 1.0

Figure 4.15. Schematic of RELAP5-3D vertical flow regime map (RELAP5-3D Code Devel-
opment Team, 2005). Hatchings indicate transitions.

displays the experimental flow regime map of Venkateswararao et al. (1982). These
authors showed that their data could be predicted by the flow regime transition
models of Taitel et al. (1980) (designated as theory in the figure), to be described in
Chapter 7, with modifications to account for rod bundle geometric configuration.
Flow regime maps and models that are used in reactor thermal-hydraulic com-
puter codes usually assume that the basic flow regimes include bubbly, slug/churn, and
annular, and they often include relatively large regime transition regions as well. For
thermal-hydraulic codes, the following points should be noted. First, hydrodynamic
parameters that are not easily measurable can be readily used in the development
of regime models because these parameters are calculated and therefore “known”
by the code. Second, what is really important for reactor codes is the correct pre-
diction of regime-dependent parameters such as interfacial friction, heat transfer
rates, etc.
The two-phase flow regime models of a well-known thermal-hydraulic code are
now briefly discussed as examples. These models utilize the void fraction and volu-
metric fluxes, based on the argument that in transient and multidimensional situations
they are the appropriate parameters that determine the two-phase flow morphology
(Mishima and Ishii, 1984).
The RELAP5–3D code (RELAP5–3D Code Development Team, 2005) uses sep-
arate flow regime maps for vertical and horizontal flow configurations. The verti-
cal flow regime map is used when 60◦ ≤ |θ | ≤ 90◦ , the horizontal flow regime map
is applied when 0◦ ≤ |θ | ≤ 30◦ , and interpolation is applied when 30◦ < |θ | < 60◦ ,
where θ is the angle of inclination with respect to the horizontal plane. (Note that
regime maps are primarily used for the calculation of parameters such as interfacial
area concentration, interfacial heat transfer coefficients, etc. Interpolation is used for
the calculation of these parameters.) The flow regime map for vertical flow is shown
in Fig. 4.15. Distinction is made between precritical heat flux (pre-CHF) and post-
CHF (post-dryout) regimes. Flow boiling and critical heat flux and postcritical heat
flux (post-CHF) will be discussed in Chapters 13 and 14, respectively. The post-CHF
regimes occur when, because of boiling, sustained or macroscopic physical contact

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
4.4 Two-Phase Flow Regimes in Rod Bundles 133

0.0 αBS αDE αSA αAM 1.0


Annular
Bubbly Slug SLG/ Mist
mist
(BBY) (SLG) ANM (MPR)
UG − UL UG − UL = Ucr; (ANM)
and G = G*2 SLG/
BBY- SLG- ANM- MPR-
G ANM-
UG − UL = 0.5Ucr; HST HST HST HST
HST
G = G*1
Horizontally stratified (HST)

Figure 4.16. Schematic of RELAP5-3D horizontal flow regime map (RELAP5-3D Code
Development Team, 2005). Hatchings indicate transitions.

between the surface and the liquid is interrupted. Post-CHF regimes are assumed
when TG − TI > 1 K. The parameters in Fig. 4.15 are defined as follows:

UTB = 0.35 g Dρ/ρL (Taylor bubble rise velocity in vertical tubes) (4.5)
⎧ ∗
⎪ αBS for G ≤ 2,000 kg/m2 ·s, (4.6)


∗ ∗ G − 2,000
αBS = αBS + (0.5 − αBS ) for 2,000 < G < 3,000 kg/m2 ·s, (4.7)

⎪ 1,000

0.5 for G ≥ 3,000 kg/m2 ·s, (4.8)

αBS = max{0.25 min[1, (0.045D∗ )8 , 10−3 ]}, (4.9)

D∗ = D/ σ/gρ, (4.10)
αCD = αBS + 0.2, (4.11)
  
f
αSA = max αAMmin
, min αcrit , αcrit
e
, αBS
max
, (4.12)


f min{[ g Dρ/ρG /UG ], 1} for upward flow, (4.13)


αcrit =
0.75 for downward or countercurrent flow, (4.14)
 
3.2  
2 1/4
αcrit
e
= min σ gρ/ρG ,1 , (4.15)
UG

0.5 for pipes, (4.16)
αAM
min
=
0.8 for bundles, (4.17)
αBS
max
= 0.9, (4.18)
αDE = max (αBS , αSA − 0.05 ) , (4.19)
αAM = 0.9999. (4.20)

For a vertically stratified flow regime to occur at a point in the computational


domain (i.e., in a control volume), the void fraction above that point (i.e. in that
control volume) should be greater than 0.7, and there must be at least a void fraction
difference of 0.2 across the control volume.
The RELAP5–3D horizontal flow regime map is displayed in Fig. 4.16. The param-
eters in the flow regime map are defined as follows:
 
1 ρgα A
Ucrit = (1 − cos θ  ) , (4.21)
2 ρg D sin θ 

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
134 Two-Phase Flow Regimes – I

θ′
Gas
Figure 4.17. Definition of the angle θ  .

LIQUID



⎨0.25 for G ≤ G∗1 , (4.22)
αBS = 0.25 + 0.00025 (G − G∗1 ) for G∗1 < G < G∗2 , (4.23)


0.5 for G ≥ G∗2 , (4.24)
where G∗1 = 2,000 kg/m2 ·s, G∗2 = 3,000 kg/m2 ·s, αDE = 0.75, αSA = 0.8, and θ  is the
angle defined in Fig. 4.17 when stratified flow is assumed.
These flow regime transition models are sometimes modified and improved for
various conditions (see, e.g., Hari and Hassan, 2002).

4.5 Comments on Empirical Flow Regime Maps


Empirical flow regime maps have been in use for decades. They suffer from several
shortcomings, however. Some of their major shortcomings are as follows:
1. These flow regime maps generally address “developed” flow conditions and are
not very accurate for short flow passages.
2. Empirical flow regime maps often attempt to specify parameter ranges for various
flow regimes using a common set of coordinates. Since mechanisms that cause
various regime transitions are different, a common set of coordinates may not be
appropriate for the entire flow regime map.
3. Most flow regime maps are based on data obtained with water, or liquids whose
properties are not significantly different than water, in channels with diameters
in the 1- to 10-cm range. The maps may not be useful for significantly different
channel sizes or fluid properties.
4. Closure relations are necessary for the solution of conservation equations (e.g.,
interfacial area concentration, interfacial forces and transfer process rates, etc.)
and these closure relations depend on flow regimes. A flow regime change thus
implies switching from one set of correlations and models to another. This can
introduce discontinuities and can cause numerical difficulties. This difficulty is
mitigated to some extent by defining flow regime transition zones.
Two-phase flow regimes will be further discussed in Chapter 7, after the two-phase
model conservation equations are discussed in the next chapter.

PROBLEMS
4.1 Saturated liquid R-134a is flowing in a vertical heated tube that has a diameter
of 1 cm. The pressure is 16.8 bar, which remains approximately constant along the
tube. A heat flux of 100 kW/m2 is imposed on the tube.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
Problems 135

a) Assuming that friction and changes in kinetic and potential energy are negligible,
prove that the first law of thermodynamics leads to
d xeq 4qw
G = ,
dz Dhfg
where z is the axial coordinate.
b) Assuming the flow regime maps based on adiabatic flow apply, using the flow
regime map of Hewitt and Roberts (1969) determine the sequence of two-phase
flow regimes and the axial coordinate where each regime is established for the
following mass fluxes: G = 200, 500, 1,500 kg/m2 ·s.

4.2 Repeat Problem 4.1, this time assuming that the tube is horizontal, and use the
flow regime maps of Baker (1954) as well as Mandhane et al. (1974). Compare and
discuss the predictions of the two flow regime maps.

4.3 A horizontal pipeline that is 15 cm in diameter is at 20◦ C and carries a


mixture of kerosene (ρL = 804 kg/m3 ; μL = 1.92 × 10−3 kg/m·s) and methane gas
(M = 16 kg/kmol; μG = 1.34 × 10−5 kg/m·s). Because of pressure drop considera-
tions, it is important that the flow regime remains stratified or wavy, but not intermit-
tent. The pressure along the pipeline varies in the 1- to 10-bar range. Using the flow
regime map of Mandhane et al. (1974), determine the allowable range of methane
mass flux for the following kerosene mass fluxes: GL = 10, 35, 75 kg/m2 ·s. Discuss
the validity of the flow regime map of Mandhane et al. for the described system.

4.4 The fuel rods in a PWR are 1.1 cm in diameter and 3.66 m long. The rods are
arranged in a square lattice, as shown in Fig. P4.4, with a pitch-to-diameter ratio
of 1.33. For a period of time during a particular core uncovery incident, the core
remains at 40-bar pressure, while saturated liquid water enters the bottom of the
core. The heat flux along one of the channels is assumed to be uniform and equal
to 6.0 × 103 W/m2 . The flow is assumed to be one dimensional and the equilibrium
quality at the exit of the channel is 0.12.

Pitch
z
Flow
Channel L Figure P4.4. Figure for Problem 4.4.

Fuel Rods

a) Assuming that quality varies along the channel according to AG dxeq /dz =
pheat qw /hfg (where A is the flow area and pheat is the heated perimeter), calculate
the coolant mass flux.
b) Using the flow regime map of Hewitt and Roberts (1969), determine the sequence
of flow regimes and the approximate axial location of regime transitions.

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007
136 Two-Phase Flow Regimes – I

4.5 In an experiment with a test section that includes a vertical pipe with D = 5.25 cm
inner diameter, cocurrent upward two-phase flow regimes are to be studied. Liquid
superficial velocities are set at jf = 0.2, 1.0, and 2.5 m/s, jg is varied from 0.1 to 10 m/s,
and the flow regimes and their transition conditions are recorded. Using the flow
regime map of Hewitt and Roberts (1969), and the flow regime map of the RELAP5-
3D code, find the flow regimes and the conditions when they are established for
saturated steam–water mixtures at 1- and 5-bar pressures. Compare the predictions
of the two methods, and comment on the results. For void fraction calculation, when
needed, use the following correlation for the slip ratio:
  
ρf
Sr = 1 − x 1 − .
ρg

Downloaded from https://www.cambridge.org/core. Universidad Nacional de Mexico (UNAM), on 23 Dec 2018 at 07:49:00, subject to the Cambridge Core terms of
Cambridge Books
use, available at https://www.cambridge.org/core/terms. Online © Cambridge University Press, 2010
https://doi.org/10.1017/CBO9780511619410.007

You might also like