ANSYS - Multiphase Flow

2010 ANSYS, Inc. All rights reserved. 1 ANSYS, Inc. Proprietary 2010 ANSYS, Inc. All rights reserved.
ed. 1 ANSYS, Inc. Proprietary

Devel opment and
Val i dat i on of Eul er i an
Mul t i phase Fl ow
Model s i n ANSYS CFD
Devel opment and
Devel opment and
Mul t i phase Fl ow
Mul t i phase Fl ow
Thomas Frank
ANSYS Germany
Thomas.Frank@ansys.com
Thomas Frank
Thomas Frank
ANSYS Germany
ANSYS Germany
Thomas.Frank@ansys.com Thomas.Frank@ansys.com
2010 ANSYS, Inc. All rights reserved. 2 ANSYS, Inc. Proprietary
Out l i ne

Introduction

MPF model validation for
adiabatic air-water flows

Polydisperse MPF model
validation

MUSIG model

Validation of the RPI wall
boiling model

The FRIGG testcase for
boiling flow in rod bundles

RPI & MUSIG model

Summary & Outlook
ANSYS as Par t of t he Ger man CFD
Net w or k i n Nuc l ear Reac t or Saf et y
GRS
FZ Dresden-
Rossendorf /
TOPFLOW
ANSYS
Germany
FZ Karlsruhe
AREVA
TU Mnchen:
TD, Nucl. Energy
Univ. Applied Sciences
Zittau-Grlitz: IPM
Univ. Stuttgart:
Nuclear Energy
Becker Technologies
ThAI test facility
International Guest
Visitors
Met hodol ogy of CFD Model
Devel opment & Val i dat i on
Experiment Experiment
phys. phys.- -math. math.
Model Model
Validation Validation
3d CFD Model 3d CFD Model
Complex Complex
Geometry Geometry
Complex Flow Complex Flow
Conditions Conditions
Combination with Combination with
other Models other Models
Eul er i an MPF Model i ng
- The Par t i c l e Model
Mass weighted averaged conservation equations

Mass, momentum, energy transport equations for each phase
( ) ( )
1
N
k k k k k kl
l
l k
r r
t

=
=
c
+V = I
c

U
( ) ( )
( )
k
k k k k k k k k k k k
r r r P r
t

c
+V = V V H + +
c
U U U F I

turbulence models for each phase
(e.g. k-c

/ k-e

SST model, 0-eq. disp. phase turb. model, Sato model)

heat transfer equations for each phase with interfacial transfer

closure

interfacial forces need empirical closure

high void fraction effects, bubble induced turbulence, etc.

secondary drag lift turbulent wall virtual mass
mom. transfer dispersion lubrication
L WL TD M k D V I
= + + + + + F F F F I F F
Li f t f or c e, Wal l l ubr i c at i on f or c e &
t ur bul ent di sper si on
Lift force:

due to asymmetric wake and deformed asymmetric particle shape
Tomiyama C
L

correlation
Wall lubrication force:

surface tension prevents bubbles from approaching solid walls
Antal, Tomiyama & Frank W.L.F. models
Turbulent dispersion force:

turbulent dispersion = action of turb. eddies via interphase drag
( )
L G L L L G L
r C = V F U U U (Re ,Re , )
L L P
C C Eo
V
=
( )
2
WL G L rel rel W W W wall
C r = F U U n n n
P
(Eo, y/d )
wall W
C C =
3
4
tF D P F
TD F F P P
P rF
P F
C r r
U U r
d
r r
v
o
| |
V V
=
|
\ .
F
FAD model by
Burns et al. (ICMF04)
Bubbl y Fl ow Model Val i dat i on
FZR MT-Loop and TOPFLOW Dat abase

TOPFLOW
MT-Loop
CFX Model Val i dat i on
MT-Loop & TOPFLOW Test Mat r i x

M01 experimental test series on MT-Loop

evaluation based on air volume fraction profiles at L/D=59,2
(z=3.03m) from the sparger system

-

numerically investigated test case conditions
superficial air velocity J_G [m/s]
s
u
p
e
r
f
i
c
i
a
l

w
a
t
e
r

v
e
l
o
c
i
t
y

J
_
L

[
m
/
s
]
finely disperse bubbly flow
bubbly flow with near wall
void fraction maximum
bubbly flow in the transition
regime
bubbly flow with void
fraction maximum at pipe
center
bubbly flow with void
fraction maximum at pipe
center, bimodal
slug flow
Val i dat i on: Bubbl y Fl ow s
Tur bul ent Di sper si on For c e
0,00
1,00
2,00
3,00
0 5 10 15 20 25
Radius, mm
N
o
r
m
.

A
i
r

V
o
l
u
m
e

F
r
a
c
t
i
o
n
k-eps + RPI TD (0.5)
k-eps + FAD TD
SST + RPI TD (0.5)
SST + FAD TD
Data
Model improvement
Monodi sper sed Bubbl y Fl ow
MT-Loop Test Case FZR-019
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0.00 5.00 10.00 15.00 20.00 25.00
Radius [ mm]
n
o
r
m
a
l
i
z
e
d

v
o
l
u
m
e

f
r
a
c
t
i
o
n

[
-
]
Experiment FZR-019
Antal W.L.F., Grid 2
Tomiyama W.L.F., Grid 2
Frank W.L.F., Grid 2
FZD-019:
J
L
=1.017 m/s
J
G
=0.004 m/s
d
P
=4.8 mm
Grace drag
Tomiyama lift
T./A./F. Wall L. Force
FAD Turb. Disp.
SST turb. model
Sato model
At=0.002s
2210 Iterations
Monodi sper sed Bubbl y Fl ow
MT-Loop Test Case FZR-052
FZD-052:
J
L
=1.017 m/s
J
G
=0.0151 m/s
d
P
=4.4 mm
Grace drag
Tomiyama lift
T./A./F. Wall L. Force
FAD Turb. Disp.
SST turb. model
Sato model
At=0.002s
2400 Iterations
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
0.00 5.00 10.00 15.00 20.00 25.00
Radius [mm]
n
o
r
m
a
l
i
z
e
d

v
o
l
u
m
e

f
r
a
c
t
i
o
n

[
-
]
Experiment FZR-052
Antal W.L.F., Grid 2
Tomiyama W.L.F., Grid 2
Frank W.L.F., Grid 2
movable diaphragm
wire-mesh sensor
movable diaphragm
TOPFLOW Test Fac i l i t y @FZD
gas injection
TOPFLOW-074 Test Case
Condi t i ons f r om Test Mat r i x

Selection of test case conditions:

TOPFLOW-074 test case was subject of validation in the past

Superficial velocities: J
G

=0.0368 m/s

J
L

=1.017 m/s

Wire-mesh sensor measurements at locations:

z=10, 15, 20, 40, 80, 160, 250, 520mm
3d Bubbl y Fl ow Ar ound Obst ac l e
Wat er Vel oc i t y Compar i son

Comparison

CFD Experiment

Absolute water
velocity distribution in
symmetry plane

Import of exp. data
into CFX-Post

Pre-interpolation of
exp. data to
Az=0.01m
CFD
Exp.
Ai r Voi d Fr ac t i on Compar i son

Comparison

CFD Experiment

Air void fraction
distribution in
symmetry plane
CFD
Exp.
1) z=10mm
2) z=15mm
3) z=20mm
4) z=40mm
5) z=80mm
6) z=160mm
7) z=250mm
8) z=520mm
1)
2)
3)
4)
6)
7)
8)
5)
Ai r Voi d Fr ac t i on Compar i son
Cr oss-Sec t i onal Ai r Voi d Fr ac t i on
5) z=80mm
TOPFLOW Experiment CFX Simulation
ANSYS CFX

Ex per i ment al Dat a
Quant i t at i ve Compar i son

Quantitative data comparison @ cross sections

z=10, 15, 20, 40, 80, 160, 250, 520mm:
absolute water velocity
air volume fraction
y=0mm
x=-35mm x=+35mm
z=520mm
z=-520mm
obstacle
ANSYS CFX

z=-80mm
y=0mm
0.0
0.5
1.0
1.5
2.0
2.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
N
o
r
m
.

A
i
r

V
o
l
u
m
e

F
r
a
c
t
i
o
n

[
-
]
Experiment (z=-80mm)
CFX Simulation (z=-80mm)
0.0
0.5
1.0
1.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
A
b
s
o
l
u
t
e

W
a
t
e
r

V
e
l
o
c
i
t
y

[
m
/
s
]
ANSYS CFX

z=-20mm
y=0mm
0.0
0.5
1.0
1.5
2.0
2.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
N
o
r
m
.

A
i
r

V
o
l
u
m
e

F
r
a
c
t
i
o
n

[
-
]
0.0
0.5
1.0
1.5
2.0
-100 -75 -50 -25 0 25 50 75 100
x [mm]
A
b
s
o
l
u
t
e

W
a
t
e
r

V
e
l
o
c
i
t
y

[
m
/
s
]
ANSYS CFX

z=80mm
y=0mm
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
N
o
r
m
.

A
i
r

V
o
l
u
m
e

F
r
a
c
t
i
o
n

[
-
]
Experiment (z=80mm)
CFX Simulation (z=80mm)
0.0
0.5
1.0
1.5
2.0
2.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
A
b
s
o
l
u
t
e

W
a
t
e
r

V
e
l
o
c
i
t
y

[
m
/
s
]
Experiment (z=80mm)
ANSYS CFX

z=250mm
y=0mm
0.0
0.5
1.0
1.5
2.0
2.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
N
o
r
m
.

A
i
r

V
o
l
u
m
e

F
r
a
c
t
i
o
n

[
-
]
Experiment (z=250mm)
0.0
0.5
1.0
1.5
2.0
2.5
-100 -75 -50 -25 0 25 50 75 100
x [mm]
A
b
s
o
l
u
t
e

W
a
t
e
r

V
e
l
o
c
i
t
y

[
m
/
s
]
Experiment (z=250mm)
Pol ydi sper sed Bubbl y Fl ow Caused by
Br eak up & Coal esc enc e
Transition from disperse bubbly
flow to slug flow:
Balance between:

coalescence

of bubbles

turbulent bubble

breakup
bubble size distribution;

polydisperse bubbly flow
counter-current radial motion of
small and large bubbles;

more than one velocity field
new population balance model
(inhomogeneous MUSIG)
I nhomogeneous MUSI G Model

momentum equations are solved for N gas phases

(vel. groups)

size fraction equations for M
i

bubble size classes in each vel. group

bubble coalescence and break-up over all EM
i

MUSIG groups
d
P
d
P,krit
N(d
P

)
coalescence/

evaporation
d
1
d
3
d
4
d
5
d
6
d
7
d
8
d
2
v
1
v
2
v
3
Velocity group mass transfer
Break up
Coalescence
breakup/

conden-

sation

Basic population balance equations

Size fraction equations
( )
B B C C
dn(m, r,t)
n(m, r,t) U(m, r,t)n(m,
n(m, r,t) m(
r,t)
dt
r,t)
m t t r
B D B D
c c c
c c
c
= + + =
c c
= +
MUSI G Model Ex t ensi on

B B C C
j
i d i i d i i B D B D
j
i
( r f ) ( r U f ) S S S S
t x
S
c c
+ = + +
c c
i i
i 1 i
i i 1 i 1 i
m m
m m m m
+
I I

i i
i i 1
i i 1 i 1 i
m m
m m m m
+
+
I I

S
i
=
for evaporation
for condensation
Bubble number
density
New terms
Breakup/Coalescence terms
i
i
i
i
i
i

Courtesy of FZD
L
~

8
m
D=195mm
Val i dat i on of 3x 7 I nhomogeneous
MUSI G Model on TOPFLOW-074

good agreement at levels A, L through R

too fast spreading of the bubble plume from inlet due to too
intensive turbulent dispersion
0.0
5.0
10.0
15.0
20.0
25.0
0.0 25.0 50.0 75.0 100.0
x [mm]
A
i
r

v
o
l
u
m
e

f
r
a
c
t
i
o
n

[
-
]
Exp. FZR-074, level A
Exp. FZR-074, level C
Exp. FZR-074, level F
Exp. FZR-074, level I
Exp. FZR-074, level L
Exp. FZR-074, level O
Exp. FZR-074, level R
CFX 074-A, Inlet level (z=0.0m)
CFX 074-A, level C
CFX 074-A, level F
CFX 074-A, level I
CFX 074-A, level L
CFX 074-A, level O
CFX 074-A, level R
Condensat i on Test Case

P=2 [MPa]

J
w

=1.0 [m/s]

J
s

=0.54 [m/s]

T
s

=214.4 [C]

T
w

=210.5 [C] T
w

=3.9 [K]

D
inj

= 1 [mm]

Detailed experimental data:
Bubble size distribution
Radial steam volume fraction distribution
Dirk Lucas, Horst-Michael Prasser: Steam bubble condensation in sub-cooled water
in case of co-current vertical pipe flow,
Nuclear Engineering and Design, Volume 237, Issue 5, March 2007,

Pages 497-508
Physi c al Model Set up

Standard MUSIG & Extended MUSIG
25 bubble size classes
3 velocity groups:
03 [mm],36 [mm], 630 [mm]
Arranged in accordance with critical Tomiyama bubble
diameter for bubble size dependent lift force
Break up model: Luo & Svendsen (F
B

=0.025)
Coalescence model: Prince & Blanch (F
C

=0.05)
TOPFLOW Condensat i on Test c ase
Inlet BC Inlet Position WLF TD Force Heat Transfer
Config 1 D
inj
= 4mm
Source point @
Wall
- CTD=1.0 Nu=2+0.6Re
p
0.5
Pr
0.3
Config 2 D
inj

= 4 mm
Source point @
75 mm
F
WLF
CTD=1.5 Nu=2+0.15Re
p
0.8
Pr
0.5
Config 3 D
inj

= 1 mm
Source point @
75 mm
F
WLF
CTD=1.5 Nu=2+0.15Re
p
0.8
Pr
0.5
Resul t s: Vapor Vol ume Fr ac t i on
Config 2 Config 3 Config 1
Resul t s: Ver t i c al Aver aged St eam
Di st r i but i on
Resul t s: Radi al St eam Di st r i but i on
Resul t s: Radi al St eam Di st r i but i on
Resul t s: Bubbl e Si ze Di st r i but i on
Resul t s: Bubbl e Si ze Di st r i but i on
CFD Si mul at i on f or Fuel
Assembl i es i n Nuc l ear Reac t or s
Material Properties Material Properties
Multiphase Flow Multiphase Flow
Modeling Modeling
Wall Boiling & Wall Boiling &
Bulk Condensation Bulk Condensation
Conjugate Heat Conjugate Heat
Transfer (CHT) Transfer (CHT)
FSI: Stresses & FSI: Stresses &
Deformations Deformations
Validation against Validation against

Experiments Experiments
Turbulence Turbulence
Mul t i phase Fl ow Regi mes f or
Boi l i ng Wat er Fl ow
subcooled
flow
bubbly
flow
slug flow
ONB OSB
T
sat
T
x
wall temperature
mean fluid
temperature
subcooled
boiling
nucleate boiling
(saturated boiling)
annular
flow
spray
flow
Fl ow s w i t h Subc ool ed Boi l i ng (DNB)
RPI -Wal l Boi l i ng Model
E Q F Wall
q q q q

+ + =
convective heat flux
1
( )
F F W L
q A h T T =
evaporation heat flux

E G L
q m (h h ) =

quenching heat flux
2
( )
Q Q W L
q A h T T =
u
t
*
m
*
m
y
1
A
2
A
2
A
Quenching heat
flux
*
m
Convective heat
flux
Mechanistic wall heat partioning model:
RPI -Wal l Boi l i ng Model
Submodel s f or Model Cl osur e
Submodels for closure of RPI wall boiling model:
Nucleation site density:

Lemmert & Chawla , User Defined
Bubble departure diameter:

Tolubinski & Kostanchuk, Unal, Fritz, User Defined
Bubble detachment frequency:

Terminal rise velocity over Departure Diameter, User Defined
Bubble waiting time:

Proportional to Detachment Period, User Defined
Quenching heat transfer:

Del Valle & Kenning, User Defined
Turbulent Wall Function for liquid convective heat transfer coefficient

Correlation for bulk flow mean bubble diameter required:

e.g. Kurul & Podowski

correlation via CCL

Supported combination of wall boiling & CHT in the solid
GGI & 1:1 solid-fluid interfaces
RPI Wal l Boi l i ng Model i n t he
ANSYS CFX-Pr e 12.0 GUI
I nvest i gat ed Boi l i ng Test c ases

Bartolomei et al. (1967,1982)

Lee et al. (ICONE-16, 2008)

OECD NEA PSBT subchannel
benchmark (1987-1995, 2009)

Bartolomei with recondensation

(1980)
R = 7.7
mm
Z
=

2

m
q
=
0
.
5
7
M
W
/
m
2
G
in

=900 kg/(s m
2
)
2010 ANSYS, Inc. All rights reserved. 45 ANSYS, Inc. Proprietary 2010 ANSYS, Inc. All rights reserved. 45 ANSYS, Inc. Proprietary
FRI GG-6a Test Case
(Angl ar t & Nyl und,
1967, 1996 & 1997)
FRI GG
FRI GG
-
-
6a Test Case
6a Test Case
1967, 1996 & 1997)
1967, 1996 & 1997)
FRI GG-6a Test Case
Desc r i pt i on
1
3
.
8
m
m
2
1
.
6
m
m
3
5
.
5
m
m

Geometry (FT-6a)
Six electrically heated
rods placed in a vertical
adiabatic pipe

Flow Parameters
Upward directed
subcooled water flow
Mass flux @Inlet
G
in
= 1163 kg m
-2
s
-1
Pressure @Inlet
p
in
= 5 MPa
Adiabatic
Wall
Heated
Rod
Rod wall heat flux
q
Rod
= 0.522 MW m
-2
Liquid subcooling @Inlet
T
sub
= 4.5 K
L
H

=4.2 m
FRI GG-6a Test Case

Determination of
experimental data by
gamma ray attenuation
method:
Measurements of area
averaged gas volume
fraction in different
cross-sectional zones
along the test section
Defintion of Zones:

Zone1 (r < 14.6 mm)

Zone2 (14.6 mm < r < 28.6 mm)

Zone3 (r > 28.6 mm)
Zone1
Zone2
Zone3
r
FRI GG-6a Test Case
Mesh Ref i nement Hi er ar c hy
Mesh01 Mesh02 Mesh03
No. Elements
699 x 150
(104 850)
2796 x 300
(838 800)
11184 x 600
(6 710 800)
No. Nodes 116 421 884 639 6 892 869
Max y
+
180 94 51
Min Angle
[deg]
51.9 50.4 49.64
Min
Determinant
0.84 0.91 0.98
Numerical
Effort
~ 90 minutes
@ 6 CPUs
~ 17 hours
@ 16 CPUs
~ 6 days
@ 40 CPUs
FRI GG-6a Test Case
Basel i ne Set up: SST

St eam VF
Outlet Outlet
Two cross-sectional distributions of steam
volume fraction (Mesh03,SST)
Plot of steam volume fraction (Mesh03, SST)
FRI GG-6a Test Case
Basel i ne Set up: SST

T
L
Outlet Outlet
Two cross-sectional distributions of liquid
temperature (Mesh03,SST)
Plot of liquid temperature (Mesh03,SST)
FRI GG-6a Test Case
Mesh Compar i son
Tur bul enc e Model i ng i n Rod
Bundl es

So far good comparison, but
Wall friction in rod bundles leads to secondary flows
Anisotropic turbulence
SST BSL RSM
Does not influence so much cross-sectional averaged
flow properties
Secondary flows affect steam & temperature distributions
on heated wall surfaces

Can be relevant for safety!
FRI GG-6a Test Case
Tur bul enc e Model Compar i son
FRI GG-6a Test Case
Outlet
Steam volume fraction
SST model
Outlet
BSL RSM model
FRI GG-6a Test Case
Contour plot of steam volume fraction (Outlet) Liquid velocity in cross section (Outlet)

SST model NO secondary flows
FRI GG-6a Test Case
Contour plot of steam volume fraction (Outlet) Liquid velocity in cross section (Outlet)

BSL RSM model secondary flows
New R& D Consor t i um
R&D Initiative:
Modeling, Simulation &
Experiments for Boiling
Processes in Fuel
Assemblies of PWR
Coupl i ng of RPI Wal l Boi l i ng Model
w i t h Homog./I nhomg. MUSI G
Bubble departure
Wall Boiling Model
(RPI)
Bubble breakup
& coalescence
(MUSIG Model)
Condensation
(MUSIG model
extension)
C
o
u
p
l
i
n
g
Future model extensions:
Bubble departure diameter
computed from equilibrium of
forces
Include further phenomena
CFX 12.1
CFX 12.1
Cust.
Solv.
v3
Cust.
Solv.
v1
Cust.
Solv.
v2
DEBORA Test c ase: RPI & MUSI G
0 0.002 0.004 0.006 0.008 0.01
R [m]
0.0
0.2
0.4
0.6
[
-
]
Exp
CFX-12
Gas1
Gas2
0 0.002 0.004 0.006 0.008 0.01
R [m]
1.5
2.0
2.5
3.0
3.5
[
m
/
s
]
U
GAS
Exp
U
GAS1
CFX-12
U
GAS2
CFX-12
U
LIQUID
CFX-12
0 0.002 0.004 0.006 0.008 0.01
R [m]
350
355
360
365
370
[
K
]
Exp
CFX-12
T
SAT
0 0.002 0.004 0.006 0.008 0.01
R [m]
0.0000
0.0002
0.0004
0.0006
[
m
]
Exp
CFX-12

Inhomog.

MUSIG

Phase

change

Breakup &

Coales-

cence

RPI
Volume
Fraction
Velocity
Temperature
Mean Sauter
diam.
dashed lines d
B

=f(T
sat

-T
L

) ; solid lines d
B

as mean Sauter diam. from MUSIG group
By courtesy of
E. Krepper, FZD
Modeling, Simulation & Experiments for
Boiling Processes in Fuel Assemblies of PWR

Ultrafast electron beam X-ray CT of fuel rod bundle in
titanium pipe on TOPFLOW @ FZD:
Images by courtesy of U. Hampel, FZD
Modeling, Simulation & Experiments for
Boiling Processes in Fuel Assemblies of PWR
Wall boiling
simulation in
3x3 rod
bundle with
spacer grid:
Wall
superheat
T
W

-T
Sat
Summar y & Out l ook

Overview on ANSYS CFD multiphase flow model
development and validation

Continuous effort in model improvement, R&D

Emphasis in validation on BPG, comparison to data,
geometry & grid independent modeling

High interoperability of physical models

Outlook:
Ongoing & customer driven CFD model development
Research cooperation with Industry & Academia
More & more complex MPF phenomena
Coupling of wall boiling model to inhomogeneous MUSIG
Extension of the wall heat partitioning in wall boiling model
Thank You!

ANSYS - Multiphase Flow

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ANSYS - Multiphase Flow

Uploaded by

Copyright:

Available Formats

2010 ANSYS, Inc. All rights reserved. 1 ANSYS, Inc. Proprietary 2010 ANSYS, Inc. All rights reserved.

ed. 1 ANSYS, Inc. Proprietary

MUSI G Model Ex t ensi on

2010 ANSYS, Inc. All rights reserved. 27 ANSYS, Inc. Proprietary

evaporation heat flux

You might also like