Transient Two-phase Flow in a Slide-gate Nozzle and Mold...
Transcript of Transient Two-phase Flow in a Slide-gate Nozzle and Mold...
CCC Annual ReportUIUC, August 20, 2014
Transient Two-phase Flow in
a Slide-gate Nozzle and Mold with
Double-ruler EMBr
POSTECH: Seong-Mook Cho and Seon-Hyo KimUIUC: Brian G. Thomas
Department of Mechanical Science & Engineering
University of Illinois at Urbana-Champaign
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 2/34
Slide-gate and Double-ruler EMBr in Continuous Slab Casting
Middle plate
SEN
OR
IR
Towardsright NF
Towardsleft NF
<Schematic of continuous steel slab casting>
<Slide-gate middle plate on SEN>
Slide-gate produces asymmetric open area for flowing molten steel
Double-ruler EMBr locates magnet rulers at two regions: just above the port and below nozzle port
Upper ruler
Lower ruler
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 3/34
Research Scope Previous works
- Quantified transient two-phase (molten steel-argon) flow in slide-gatenozzle and mold without EMBr, using LES coupled with DPM model andnail board tests
- Investigated effects of double-ruler EMBr on single-phase (molten steel)flow in slide-gate nozzle and mold using standard model and nailboard tests
Objectives to- Gain insight of effects of double-ruler EMBr on transient two-phase flow in
slide-gate nozzle and mold- Develop and validate LES coupled with DPM and MHD to predict two-
phase flow considering EMBr effect- Compare steady-state standard model and LES to predict transient
flow variations in the mold Methodologies
- Plant Experiments: nail board tests, eddy-current sensor measurements- Computational Models: standard model, LES coupled with DPM and
MHD model
εk −
εk −
εk −
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 4/34
Process ConditionsCaster Dimensions
Nozzle bore diameter (inner/outer) 90 mm (at UTN top) to 80 mm (at bottom well) /160 mm (at UTN top) to 140 mm (at SEN bottom)
Nozzle bottom well depth 19 mm
Nozzle port area 80 mm (width) 85 mm (height)Nozzle port angle
*2008: 52 to 35 down degree step angle at the top, 45 down degree angle at the bottom
*2010: 35 down degree angle at both top and bottom
Mold thickness 250 mmMold width 1300 mm
Domain length 4648 mm (mold region: 3000 mm (below mold top))
Process ConditionsSteel flow rate 552.5 LPM (3.9 tonne/min)Casting speed 1.70 m/min (28.3 mm/sec)
Argon gas flow rate & volume fraction9.2 SLPM (1 atm, 273 K); 33.0 LPM (1.87 atm, 1827 K) & 5.6 % (hot)
Submerged depth of nozzle 164 mm
Meniscus level below mold top 103 mm
EMBr current (both coils) EMBr off: 0 A EMBr on: 300 A
×
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 5/34
Plant Measurements: Nail Board Tests (2013 CCC Report)
<Time-averaged surface velocity> <Surface velocity fluctuation>
~50% fluctuation of mean velocity
<Time-averaged surface level> <Surface level fluctuation>
~8 mm
<Slag motion without EMBr>
<Slag motion with EMBr>
k: coefficient of slag motion
k=slope
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 6/34
Plant Measurements: Eddy Current Sensor Measurement (2013 CCC Report)
AVG level: ~103 mm
<Surface level by time>
<Power spectrum of level variation>
With EMBr, surface level fluctuation at quarter point (located midway between SEN and NF) is reduced by ~33%: 0.6 mm (Without EMBr) and 0.4 mm (With EMBr)
~0.03 Hz (~35 sec)
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 7/34
Governing Equations
Molten steel flow field: Large Eddy Simulation (LES)
Argon gas bubble motion: Discrete Phase Model (DPM)
Electromagnetic force induced by EMBr: Magneto-Hydro-Dynamics (MHD) model
( ) mass shell,ii
Sρux
=∂∂
( ) ( ) ( ) iL,imom,Ar,i mom,shell,i
j
j
it
ji
*
jij
i FSSx
u
x
uμμ
xx
puρu
xρu
t+++
∂∂
+∂∂+
∂∂+
∂∂−=
∂∂+
∂∂
iradient,pressure_giss,virtual_maibuoyancy,idrag,iAr, FFFF
dt
du+++=
( ) ( )iAr,i2ArAr
Didrag, uu
dρ
ReμC
4
3F −⋅=
μ
uuρdRe ArAr −
=
( )
( )
( )
( )We8 3
8
ReWe
2065.1
3
We
Re100 Re
6.3
100Re0.49 Re
20.68
0.49Re Re
16C
2.6
0.385
0.643
D
<=
<=
<=
<<=
<=
argonsteel
2
ArAr
σ
uuρdWe
−
−=
iAr
Aribuoyancy, g
ρ
ρ-ρF = ( )iAr,i
Ariss,virtual_ma uu
dt
d
ρ
ρ
2
1F −=
i
ii
Ariradient,pressure_g x
uu
ρ
ρF
∂∂=
( ) ΔtmFFFFS Ari gradient,_ pressurei ss,virtual_mai buoyancy,i drag,i Ar,mom, +++−=
( )bBjF 0L
+×=
( )bBμ
1j 0
+×∇= ( ) ( )( ) ( ) 00
2 B uu bBbμσ
1b u
t
b
∇⋅−∇⋅++∇=∇⋅+∂∂
V
AρuS casting
massshell, −=Mass conservation: Mass/momentum sink terms to account for solidification of the molten steel:
Momentum conservation:
icasting
imom, shell, uV
AρuS −=
Bubble motion equation:
Force equations:
Drag coefficient:
Lorentz force:
Induced current density: Induced magnetic field:
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 8/34
Domain, Mesh, and Boundary Conditions
<Front view> <Side view>
Total cells in the domain: ~1.8 million
Boundary conditions
Inlet0.00938 m/sec, 5.6 % volume fraction of Argon
Outlet0 Pa (gauge pressure), particle escape
SurfaceStationary wall with a no slip shear condition,particle escape
Interfacebetween the molten steel /the steel shell
Stationary wall with a no slip shear condition, molten steelmass/momentum sink,particle reflect
(sec)t 2.94(mm) S =
Inlet
Outlet
Surface
Symmetric plane
0 500 1000 1500 2000 2500 30000
5
10
15
20
25
30
35
Measurement
Sh
ell t
hic
kne
ss (
mm
)
Distance from meniscus (mm)
Steel shell thickness profile:
Half domain (assuming symmetry between NFs, containing the electrically-conducting steel shell region as a solid zone)
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho • 9/34
Double-ruler EMBr Field: External Magnetic Field
Weaker
Magnetic field applied by the double-ruler EMBr has high peaks in two regions: one centered just above the port and the other below the nozzle port
Inner oval showing region of strong field over 0.13T
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •10/34
Double-ruler EMBr Field: Current Density & Electromagnetic Force
<Induced current density> <Electromagnetic force>
High current density and electromagnetic force are generated in two regions: near the nozzle port and near the NF 600mm below the mold top
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •11/34
Comparison of Time-averaged Mold Flow Patternbetween Standard Model and LES (Single Flow)
Time-averaged ~23 s Time-averaged ~32.3 s
εk −
εk−Standard
LES
EMBr off EMBr on
EMBr off EMBr on
Standard model and LES both predict double-role flow pattern and high surface velocity in the mold without EMBr
Both two models show EMBrproduces less flow up the NF (600 mm below mold top), resulting in slower surface velocity and stronger downward flow along the NF
In EMBr on case, steady-state standard model shows better agreement with LES than EMBr off case
The limited accuracy of this steady-state model is perhaps due to complex transient flow variation (anisotropic variations)
εk −
εk −
Inner oval showing region of strong field over 0.13T
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •12/34
Comparison of Mold Flow Fluctuations between Standard Model and LES (Single Flow)
( ) ( ) ( )2'2'2' wvu == ( )2'v ( )2'w( )2'u
εk −RMS velocity
(m/sec)
EMBr off
EMBr on
εk−Standard
LES
LES
εk−Standard
Two models both predict EMBrinduces smaller flow variation towards the top surface LES predicts EMBr is more effective to reduce the variations along mold thickness; same trend with R. Singh et al. result[*]
Steady-standard model well-
predicts isotropic variations however, shows limited accuracy for anisotropic variations
Time-averaged
~23 s
Time-averaged ~32.3 s
εk −
LES is much more appropriate model to predict more complex flow variations induced by two-phase flow
*R. Singh, B. G. Thomas, and S. P. Vanka, Metall. Mater. Trans. B., 2014, vol. 45B, pp. 1098-1115
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •13/34
Transient Swirl Flow in Slide-gate Nozzle (Two-phase Flow)
Time-averaged ~55.2 s
EMBr off
EMBr on
IR OR
IR OR
Time-averaged ~62.8 s
EMBroff
EMBron
Clockwise23.4 s(43%)
14.4 s(23%)
Counter clockwise
29.4 s(54%)
46.2 s(74%)
Intermediate1.2 s(3%)
1.8 s(3%)
With EMBr, period of the counter-clockwise swirl flow becomes longer: EMBr makes the flow (from asymmetric open area in the middle plate to nozzle bottom) go down by longer path with imposing electromagnetic force
*19.92s *63.30s
**13.35s **42.21s
*After argon gas injection
**After EMBrapplication
Clockwise Counter clockwise
ClockwiseCounter clockwise
Period of flow directions
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •14/34
Swirl Flow Videos
EMBr off EMBr on
Real flow time ~55.2 s Real flow time ~62.8 s
Swirl in nozzle[*] (Water model experiment)
*Hershey, D. E., B. G. Thomas, and F. M. Najjar, "Turbulent Flow through Bifurcated Nozzles," International Journal for Numerical Methods in Fluids, 17:1, 23-47, 1993.
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •15/34
Nozzle Port Velocity Histories (Two-phase Flow at P1-Nozzle Port)
Intermediate period
10 15 20 25 30 35 40 45 50 55 60 65 700.0
0.5
1.0
1.5
2.0
2.5
3.0
Vel
ocit
y m
agn
itu
de
(m/s
ec)
Time after EMBr application (sec)
P1_nozzle port (EMBr_on) P1_nozzle port (EMBr_on)_0.3s_time-averaging
15 20 25 30 35 40 45 50 55 60 65 700.0
0.5
1.0
1.5
2.0
2.5
3.0V
eloc
ity
mag
nitu
de (
m/s
ec)
Time after gas injection (sec)
P1_nozzle port (EMBr_off) P1_nozzle port (EMBr_off)_0.3s_time-averaging
<EMBr off> <EMBr on>
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •16/34
Jet Velocity Histories (Two-phase Flow at P2-Jet Flow)
15 20 25 30 35 40 45 50 55 60 65 700.0
0.5
1.0
1.5
Vel
ocit
y m
agn
itu
de
(m/s
ec)
Time after argon gas injection (sec)
P2_jet flow (EMBr off) P2_jet flow (EMBr off)_0.3s_time-averaging
10 15 20 25 30 35 40 45 50 55 60 65 700.0
0.5
1.0
1.5
Vel
ocit
y m
agni
tude
(m
/sec
)
Time after EMBr application (sec)
P2_jet flow (EMBr_on) P2_jet flow (EMBr_on)_0.3s_time-averaging
Intermediate period
AVG_lag time between P1 and P2 with and without EMBr: ~0.29s
<EMBr off> <EMBr on>
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •17/34
Transient Flow Pattern in the Mold(Two-phase Flow)
Velocity magnitude (m/sec)
Time-averaged ~55.2 s
EMBr off
EMBr on
Time-averaged ~62.8 s
*19.92s *63.30s
**13.35s **42.21s
With and without EMBr, double-role pattern is induced in two-phase flow in the mold
Without EMBr, up-and-down wobbling of the jet flow induces variations of velocity magnitude and direction at the surface, and changes the jet flow impingement point on the NF
EMBr makes the slightly thinner jet flow and reduces the wobbling, resulting in more stable surface flow
*After argon gas injection
**After EMBrapplication
Inner oval showing region of strong field over 0.13T
Avg_lagtime:~3.4s
Avg_lagtime:~4.1s
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •18/34
Mold Flow Videos
Real flow time ~55.2 s Real flow time ~62.8 s
EMBr off EMBr on
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •19/34
Velocity Fluctuations & Turbulent Kinetic Energy in the Mold (Two-phase Flow)
Turbulent kinetic energy (m2 /sec2)
++2
'2
'2
' wvu2
1RMS velocity (m/sec)
'u 'v 'w
Time-averaged ~55.2 s
EMBr off
EMBr onTime-averaged ~62.8 s
++2
'2
'2
' wvu2
1'w'v'u
Two-phaseflow shows more variations in nozzle port and upper-role region than single-phase flow does
EMBr is moreeffective to reduce the variations produced by the jet flow wobbling in upper-role region; especially, casting direction and mold thickness direction
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •20/34
Calculation of Jet Flow Angle
u
v
w
Flow direction
( )ii-1
i vutanα /−=Instantaneous vertical angle:
( )ii-1
i vwtanβ /−=Instantaneous horizontal angle:
along casting direction
along mold width
along mold thickness
Averaged vertical angle:
Averaged horizontal angle:
Vertical angle fluctuation (standard deviation):
Horizontal angle fluctuation (standard deviation):
=
=n
1iiavg α
n
1α
=
=n
1iiavg β
n
1β
( )=
−=n
1iavgiα αα
n
1σ
( )=
−=n
1iavgiβ ββ
n
1σ
<Instantaneous angle histories of two-phase flow without EMBr (at P1)>
15 20 25 30 35 40 45 50 55 60 65 70-90-80-70-60-50-40-30-20-10
0102030405060708090
Ver
tica
l an
gle
(deg
ree)
Time after argon injection (sec)
Point-1 without EMBr
avgα
avgβ
ασ
βσ
45.4
26.1
21.1
5.1
iα
15 20 25 30 35 40 45 50 55 60 65 70-90-80-70-60-50-40-30-20-10
0102030405060708090
Hor
izon
tal a
ngl
e (d
egre
e)
Time after argon gas injection (sec)
Point-1 without EMBr
iβ
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •21/34
Jet Flow Angle in the Mold (Effect of EMBr on mean and variation)
IR
OR
NFSEN
<Top view: horizontal jet angle>
<Front view: vertical jet angle>
NF
Center line
Surface
IR
OR
NFSEN
NF
SEN
Surface
“EMBr off” “EMBr on” EMBr reduces jet wobbling along both casting direction (vertical angle) and mold
thickness direction (horizontal angle)
21.145.4 ±
22.026.6 ±
13.245.4 ±
17.424.9 ±
26.15.1 ± 22.91.0- ± 18.65.2 ± 21.02.8- ±
<Front view: vertical jet angle>
SEN
avg stdev
Center line
P1 P2
Reduction in angle variation by
EMBr
P1-nozzle
P2-jet
Vertical angle
fluctuation37.4 % 20.9 %
Horizontal angle
fluctuation28.7 % 8.3 %
P1
P2
<Top view: horizontal jet angle>
Horizontal angle fluctuation is slightly larger than vertical angle fluctuation
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •22/34
Transient Surface Flow Pattern (Two-phase Flow)
Time-averaged ~55.2 s Time-averaged ~62.8 s
*19.92s
*63.30s
**13.35s
**42.21s
*After argon gas injection
**After EMBrapplication
<EMBr off> <EMBr on>
Decreased jet wobbling by EMBr reduces the surface flow variations along both mold width and thickness; the horizontal velocity fluctuation and the asymmetric flow between IR and OR are suppressed
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •23/34
Surface Flow Videos
Real flow time ~55.2 s Real flow time ~62.8 s
EMBr off EMBr on
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •24/34
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
Tu
rbu
lent
kin
etic
ene
rgy
(m2/
sec2
)
Distance from mold center (mm)
Without EMBr With EMBr
Velocity Fluctuations & Turbulent Kinetic Energy at the Surface (Two-phase Flow)
<RMS u velocity> <RMS v velocity>
<RMS w velocity> <Turbulent kinetic energy>
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
RM
S u
vel
ocit
y (m
/sec
)
Distance from mold center (mm)
Without EMBr With EMBr
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
RM
S v
vel
ocit
y (m
/sec
)
Distance from mold center (mm)
Without EMBr With EMBr
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
RM
S w
vel
ocit
y (m
/sec
)
Distance from mold center (mm)
Without EMBr With EMBr
offEMBr
'uoffEMBr
'v
offEMBr
'w
onEMBr
'u
onEMBr
'v
onEMBr
'w onEMBr TKE
offEMBr TKE
33.7 %
10.8 %
39.9 %
45.6 %
'u
'v
'w
++2
'2
'2
' wvu2
1
Reduction in Velocity fluctuations & TKE by EMBr
=0.059 m/s
=0.039 m/s
=0.065 m/s
=0.058 m/s
=0.074 m/s
=0.045 m/s
=0.0068 m2/s2
=0.0037 m2/s2
EMBr mainly reduces surface velocity fluctuations along mold thickness (between IR and OR) and casting direction; the swirl flow variations produced by slide-gate can be controlled by the EMBr
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •25/34
Surface Velocity Histories (Two-phase Flow at W/4 at the surface)
15 20 25 30 35 40 45 50 55 60 65 700.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55V
eloc
ity
mag
nit
ud
e (m
/sec
)
Time after argon gas injection (sec)
Without EMBr
10 15 20 25 30 35 40 45 50 55 60 65 700.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
Vel
ocit
y m
agn
itu
de
(m/s
ec)
Time after EMBr application (sec)
With EMBr
Variations of surface velocity magnitude is related to the swirl flow direction in the nozzle bottom with slide-gate
Clockwise swirl induces high momentum jet flow, resulting in high surface velocity after lag time (without EMBr: ~3.4s(avg), with EMBr: ~4.1s(avg)). On the other hand, counter clockwise swirl produces lower surface velocity
Intermediate period
Avg_lagtime: ~3.4s Avg_lag time: ~4.1s
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •26/34
Frequency Analysis at the Surface(W/4 at the surface)
0.01 0.1 1 10 1001E-201E-191E-181E-171E-161E-151E-141E-131E-121E-111E-101E-91E-81E-71E-61E-51E-41E-30.01
Pow
er s
pec
tru
m (
m2/
s2)
Frequency (Hz)
Without EMBr With EMBr
~0.054 Hz (~18.5sec)
~0.095 Hz (~10.5 sec)
1E-3 0.01 0.10.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018EMBr off: EMBr on:
Pow
er s
pec
tru
m(m
m2 )
Frequency(Hz)<Power spectrum of surface level><Power spectrum of surface velocity>
Prediction
Eddy-current measurements
The characteristic frequencies (EMBr off: ~0.054 Hz, EMBr on: ~0.095 Hz) of surface velocity seem to produce similar peaks as the measured surface level fluctuations
The strong maximum peak (~0.03 Hz) at the surface, both with and without EMBr, might be produced by low frequency sloshing between right and left narrow face; the half model of LES fails to capture low peak ( 0.03 Hz) OR this peak would be shown by longer flow time (over 70 s)
~0.057 Hz (~17.5 sec)
~0.096 Hz (~10.4 sec)
~0.03 Hz (~35 sec)
≤
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •27/34
Model Validation with Nail Board Tests:Surface Velocity
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45S
urf
ace
velo
city
mag
nit
ud
e (m
/sec
)
Distance from mold center (mm)
Predicted MeasuredWithout EMBr: With EMBr:
0 50 100 150 200 250 300 350 400 450 500 550 600 6500.000.010.020.030.040.050.060.070.080.090.100.110.120.130.140.15
Su
rfac
e ve
loci
ty
mag
nit
ud
e fl
uct
uat
ion
(m
/sec
)
Distance from mold center (mm)
Predicted MeasuredWithout EMBr: With EMBr:
<Surface velocity magnitude> <Surface velocity fluctuation>
The predicted surface velocity magnitude and its fluctuation profiles show remarkable agreement with measurements for both EMBr off and on case
With EMBr, surface flow is slightly slower (by ~17 %) with smaller fluctuations (by ~43 %)
This finding suggests that use of the double-ruler EMBr may help to reduce defects caused by surface flow instability.
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •28/34
Model Validation with Nail Board Tests:Surface Level
Region 1(near SEN: 0~235mm)
Region 2(midway: 235~485mm)
Region 3 (near NF: 485~650 mm)
Without EMBr 0.85 0.74 0.82
With EMBr 0.97 0.63 0.65
<Coefficient (k) of slag motion at the surface regions with and without EMBr>
( )gk)ρ(1ρ
PPH
slagsteel
avgii −−
−=
0 50 100 150 200 250 300 350 400 450 500 550 600 650-8-7-6-5-4-3-2-10123456789
10
Su
rfac
e le
vel h
eigh
t (m
m)
Distance from mold center (mm)
15 sec (prediction) 18 sec (prediction) 21 sec (prediction) 24 sec (prediction) 27 sec (prediction) 30 sec (prediction) Nail board (measurement)
0 50 100 150 200 250 300 350 400 450 500 550 600 650-8-7-6-5-4-3-2-10123456789
10
Su
rfac
e le
vel h
eigh
t (m
m)
Distance from mold center (mm)
11.7 sec (prediction) 14.7 sec (prediction) 17.7 sec (prediction) 20.7 sec (prediction) 23.7 sec (prediction) 26.7 sec (prediction) Nail board (measurement)
<EMBr off> <EMBr on>
New Semi-empirical equation to relate pressure to surface level:
The improved pressure-based surface level prediction agrees with measured surface level profiles with and without EMBr
Predictions are less accurate near SEN without EMBr; perhaps due to low frequency and high amplitude wave motion near SEN
The model might benefit from true free-surface analysis, instead of simple surface-pressure method
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •29/34
Bubbles Distribution in the Mold (1/3 Scale Water Model Experiment)
<35.0 LPM (Water)_0.2 SLPM (Air)> <35.0 LPM (Water)_0.8 SLPM (Argon)>
<35.0 LPM (Water)_1.6 SLPM (Argon)>
With higher gas flow rate, bubbles in the mold are larger
With small gas flow rate, bubbles are carried throughout larger mold region; (because they are small), causing more chance to touch the NF
With high gas flow rate, most bubbles float to the surface near the SEN (because the flow cannot carry bigger bubbles as easily), so less are found near the NF
window
Volume fraction:
0.6 %
Volume fraction:
2.4 %
Volume fraction:
4.6 %
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •30/34
Effect of EMBr on Argon Bubble Distribution in the Mold
<EMBr off> <EMBr on>
- Transient gas distribution changes (agrees with measured) are induced by jet flow wobbling.
- Most bubbles found in upper recirculation region; this trend agrees with the water model measurement
- Predicted bubble spreadingacross top region (differing from measurement) might be due to incorrect assumption of constant bubble size (1mm) vs. (1-5mm with 2.5mm mean)
- With EMBr, more argon bubbles float up to the surface near the SEN wall. In the region 600~1200 mm from the mold top, many bubbles have longer residence time near NF.
Volume fraction: 5.6 %
5.6 % volume fraction
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •31/34
Argon Gas Distribution Videos
Real flow time ~55.2 s Real flow time ~62.8 s
EMBr off EMBr on
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •32/34
Summary: Modeling
LES coupled with DPM and MHD model is used to predict transient molten steel-argonflow in slide-gate nozzle and mold with and without EMBr, and the model is validated withnail board tests and eddy-current sensor measurements showing remarkable agreements. Steady-standard model well-predicts isotropic variations, however, shows limited
accuracy for anisotropic variations, especially produced by swirl flow near nozzle port Predicted surface velocity magnitude and its fluctuation profiles show remarkable
agreement with the measurements for both EMBr off and on case The improved pressure-based surface level prediction (new semi-empirical equation to
relate pressure to surface level) agrees with measured surface level profiles with andwithout EMBr Half model of LES predicts the characteristic frequencies of surface velocity fluctuations,
which seem to produce similar peaks as the measured surface level fluctuations. Themeasured strong peak (~0.03 Hz) by an eddy-current sensor at the surface, both with andwithout EMBr, might be produced by low frequency sloshing between right and left narrowface; the model fails to capture the low peak (0.03 Hz) OR this peak would be shown bylonger flow time (over 70 s) LES coupled with DPM shows reasonable match with water model measurements of
argon bubble distribution For better prediction of bubble distribution in the mold, bubble size distribution is needed to
be implemented to transient two-phase flow model
εk −
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •33/34
Summary: Swirl Flow and Its Effect Slide-gate induces swirl flow in nozzle bottom region with clockwise, counter-clockwise, and
intermediate directions. EMBr makes the flow (from asymmetric open area in middle plateto nozzle bottom) go down by longer path with imposing electromagnetic force, resulting inlonger period of the counter-clockwise flow Swirl flow induces jet wobbling showing high jet flow angle fluctuations (vertical and
horizontal angle) in the mold. Both angle fluctuations with EMBr are decreased. EMBr decreases jet wobbling, reducing flow variations produced by slide-gate, including
variations in thickness direction in both jet & top surface, and along NF (in casting direction) Clockwise swirl induces high momentum jet flow, resulting in high surface velocity after
lag time for the jet flow to move toward to surface. On the other hand, counter clockwiseswirl produces lower surface velocity Transient gas distribution changes are induced by jet flow wobbling. With higher gas flow rate, bubbles in the mold are larger and most bubbles float to the
surface near the SEN (because the flow cannot carry bigger bubbles as easily), so less are found near the NF With small gas flow rate, bubbles are carried throughout larger mold region; (because they
are small), causing more chance to touch the NF Most bubbles found in upper recirculation region with and without EMBr With EMBr, more argon bubbles float up to the surface near the SEN wall. In the region
600~1200 mm from the mold top, many bubbles have longer residence time near NF. Thisphenomena may induces more chances for small bubbles to be entrapped by solidifyingsteel shell (13.5~19.1 mm below slab surface).
Pohang University of Science and Technology • Department of Materials Science and Engineering • Seong-Mook Cho •34/34
Acknowledgments
• Continuous Casting Consortium Members(ABB, ArcelorMittal, Baosteel, MagnesitaRefractories, Nippon Steel and Sumitomo Metal Corp., Nucor Steel, Postech/ Posco, Severstal, SSAB, Tata Steel, ANSYS/ Fluent)
• POSCO (Grant No. 4.0004977.01)• Yong-Jin Kim, POSCO for help with the
plant measurements