Continuous Casting Defects from Transient Flow & Inclusion...

1

ANNUAL REPORT 2005Meeting date: June 1, 2005

Quan Yuan (Ph. D. Student) &Brian G. Thomas

Department of Mechanical & Industrial EngineeringUniversity of Illinois at Urbana-Champaign

Continuous Casting Defects from Transient Flow & Inclusion Entrapment

University of Illinois at Urbana-Champaign • Metals Processing Simulation Lab • BG Thomas 2

Acknowledgements

National Science Foundation– Grant NSF-GOALI DMI 01-15486

Continuous Casting Consortium at UIUC (Nucor, Postech, LWB Refractories, Accumold, Algoma, Corus, Labein, Mittal ISG Riverdale)National Center for Supercomputing Applications (NCSA) at UIUCFluent, Inc., CFX, UIFLOW (P. Vanka)Other Graduate students, especially B. Zhao

2


Defects related to Mold Flow

Inclusion defects from:- nozzle clogging- air entrainment- entering nozzle from upstream- mold slag entrainment

Surface defects from:- Top surface level profile(Poor flux infiltration)

- Level fluctuations- Surface hook formation (meniscus freezing)

Shell thinning & breakouts

Water

Spray

Molten Steel Pool

Solidifying Steel Shell

Flux Rim

Submerged Entry Nozzle

Support Roll

Roll Contact

Ferrostatic Pressure

Bulging

Roll

Nozzle

Nozzle

copper mold

Liquid Flux

Air Gap

Flux Powder

jet

nozzle portargon

bubbles

Inclusion particles and bubbles

Resolidified Flux

Contact Resistances

Oscillation Mark

entrainment

CL


Model of Flow in 3-port SEN:water model and steel caster

Transient Inlet Jets

Constant Pressure at Bottom

2400

mm

934mm

Free-Slip Bounary Condition

zy

x

Shell WithdrawalSpeed = 25.4mm/s

80mm

132mm984mm

Fix Fluid Velocity

Pressure BoundaryCondition at Bottom

984mm

1200

mm

132mm

No-Slip Boundary

Free-Slip Bounary

Casting Speed = 25.4mm/s

zyx

SEN

Stopper Rod

Tundish Region

Steel caster domain(open bottom)Steel caster domain(open bottom)

Water-model domainWater-model domain

Distance below meniscus (mm)

Shel

lthi

ckne

ss(m

m)

0 500 1000 1500 2000 2500 30000

5

10

15

20

25

30

NF (CON1D Prediction, Y. Meng)WF (COND1D Prediction, Y. Meng)NF (Measurement. B.G. Thomas et al., 1998)WF (Measurement. B.G. Thomas et al., 1998)

Shell thickness obtained from CON1D prediction (Y. Meng), compared with measurements (B.G. Thomas et al, 1998)

Shell thickness obtained from CON1D prediction (Y. Meng), compared with measurements (B.G. Thomas et al, 1998)

3


Casting Conditions Simulated

57.Superheat (°C)

100, 250 and 400Particle Diameter (µm)

2700Particle Density (kg/m3)7020Fluid Density (kg/m3)

7.98 × 10-7Fluid Dynamic Viscosity (m2/s)

1.524Casting Speed (m/min)127SEN Submergence Depth (mm)

32Bottom Nozzle Port Diameter (mm)

75 × 32 (inner bore)Nozzle Port Height × Thickness (mm × mm)

2400Domain Length (mm)

13279.48

Domain Thickness (mm) top (mold thickness)bottom

984934.04

Domain Width (mm) top (mold width)bottom

1200Mold Length (mm)


Transient Flow near Stopper Rodx (m)

z(m

)

-0.05 0 0.05

-0.3

-0.25

Stopper Rod

1.0m/s:

Clogging: ~8% of particles touch wallsClogging: ~8% of particles touch wallsAsymmetric flow due to turbulenceAsymmetric flow due to turbulence

4


Transient Flow near Nozzle Portx (m)

z(m

)-0.1 -0.05 0 0.05 0.1

0

0.05

0.1

0.15

0.2

10.80.60.40.20

(Scale: 1.0m/s)

(Vx2+Vz

2)1/2 (m/s)

x (m)

z(m

)

-0.1 -0.05 0 0.05 0.10

0.05

0.1

0.15

0.2

10.80.60.40.20

(Scale: 1.0m/s)

(Vx2+Vz

2)1/2 (m/s)

Deep jet angle on both sidesDeep jet angle on both sides Asymmetrical jetsAsymmetrical jets


Transient Flow Pattern

LES in water model

Measured dye injection

LES in steel caster

5


Numerical Validation: Time-averaged Velocities from Different Grid Resolutions0.1 0.2 0.3 0.4

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.20.5m/s:

0.1 0.2 0.3 0.4

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.20.5m/s:

0.1 0.2 0.3 0.4

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.20.5m/s:

20,000 cells20,000 cells 400,000 cells400,000 cells 800,000 cells800,000 cells


Effect of Grid Resolution

Number of computational cells (× 106)

Mea

ner

roro

ftim

e-av

erag

edve

loci

ties

(m/s

)

Rel

ativ

em

ean

erro

r(%

)

0 0.2 0.4 0.6 0.80

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0.11

0.12

0.13

0.14

0.15

0

10

20

30

40

50

60

70

80

Number of Computational Cells (× 106)

Mea

ner

roro

frm

sve

loci

ties

(m/s

)

Rel

ativ

eer

roro

frm

sve

loci

ties

(%)

0 0.2 0.4 0.6 0.80

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0

5

10

15

20

25

30

35

40

45

Errors of time-averaged (left) and rms (right) velocitiesErrors of time-averaged (left) and rms (right) velocities

Error calculated by:Error calculated by:( ) ( ) ( )2 2 2exact exact exact

x, i x,i y, i y,i z, i z,i1

V V V V V VN

iErrorN

=

− + − + − =∑

17% error

(0.03m/s)

17% error

(0.03m/s)

6


Top Surface Level Profiles

Distance from center, x (m)Ste

elsu

rface

:liq

uid

leve

lpro

file

(m

0 0.1 0.2 0.3 0.4

-12-10

-8-6-4-202468

10

Predicted Level, RightPredicted Level, LeftMeasurement

SEN

Instant t = 20.3 in simulation

Water modelWater model

Steel CasterSteel Caster

( )( )

mean

steel flux

p ph

gρ ρ−

=−

Liquid level is calculated from predicted pressure:Liquid level is calculated from predicted pressure:

Distance from center, x (mm)

Wat

ersu

rface

:men

iscu

sle

vel(

mm

)

0 0.1 0.2 0.3 0.4-4

-2

0

2

4

measurement, instant 1measurement, instant 2measurement, instant 3LES(t=100s, left side)

SEN


Velocity Fluctuations on Top Surface

Horizontal (x) velocity towards SEN at a point 20mm below top surface, mid-way between SEN and narrow face.

Time (s)

xve

loci

tyto

war

dsth

eSE

N(m

/s)

0 10 20 30 40 50 60 70

0

0.1

0.2

0.3

PIVLES, left pointLES, right point

Frequency (Hz)

Four

ierc

oeffi

cien

tam

plitu

de(m

/s)

0 0.5 1 1.5 210-5

10-4

10-3

10-2

10-1

PIVLES, right side

A spectral analysis of the above signals.

(PIV data from S. Sivaramakrishnan, 2000.)

PIV measurementLES calculation (right side)PIV measurementLES calculation (right side)

Yuan et al, Met Trans B, 2004Yuan et al, Met Trans B, 2004

7


Instantaneous temperature field

Mean temperature profile - validationDistance below meniscus (mm)

T(°

C)

-20 -10 0 10 20 30 40 50 60 70 80

1510

1520

1530

1540

1550Simulation without SGS modelSimulation with SGS modelProbe insertionProbe withdrawn Measurement

position

SEN

50 mm

NF

Profile location

0 0.1 0.2 0.3 0.4 0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1832.01829.61827.31824.91822.51820.11817.81815.41813.01810.61808.31805.91803.51801.11798.81796.41794.01791.61789.31786.91784.51782.11779.81777.41775.0

Temperature field

T (K)

z

x

measured

computed

Unsteady Heat Transfer in Thin Slab Caster

Thomas, B.G., B. Zhao, et al; 2004 NSF Design, Service, and Manufacture and Industrial Innovation Grantees and Research Conf. Proceedings, (Dallas, TX, Jan. 5-8, 2004), pp. T/BGT/1-41.


Instantaneous temperature field

0 0.1 0.2 0.3 0.4 0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1832.01829.61827.31824.91822.51820.11817.81815.41813.01810.61808.31805.91803.51801.11798.81796.41794.01791.61789.31786.91784.51782.11779.81777.41775.0

Temperature field

T (K)

z

x

Unsteady Heat Transfer in Thin Slab Caster

Thomas, B.G., B. Zhao, et al; 2004 NSF Design, Service, and Manufacture and Industrial Innovation Grantees and Research Conf. Proceedings, (Dallas, TX, Jan. 5-8, 2004), pp. T/BGT/1-41.

X

Z

Y

449KW

12%26%

35%

4%

Superheat

Removal

Fractions

Superheat

Removal

Fractions

8


Inclusions: Types and Sources

10µm

Dendritic alumina Alumina cluster

Refs:1). Cramb et al, 2001 Steelmaking Conference Proceedings, ISS2). Flemings, Metall. Trans., 1972. 3). L. Zhang & BG Thomas, unpublished report to IMF4). L. Zhang & BG Thomas. 2002 Steelmaking Conference, ISS.

400µm

Inclusion Sources

1) Deoxidation products;2) Reoxidation products (oxidized

by slag or by air); 3) Slag entrapment; 4) Exogenous inclusions from

other sources, such as, broken refractory brickwork & ceramic lining particles;

5) Chemical reactions, such as dissolution of refractory walls

Slag inclusions (globules)

10µm

Bubble with inclusions


Lagrangian trajectories of 15000 particles(plastic beads in water chosen to match 300 micron alumina inclusions in steel)

Time (s)

Parti

cles

rem

ove d

tom

enis

cus/

parti

cle s

ente

r ing

(%)

P ar ti

c le

rem

o val

rate

tom

eni s

cus

(%s- 1

)

0 50 100 150 200 250 300

10

20

30

40

50

60

70

80

90

100

1

2

3

4

5

6

7

8

9

10

(a)

(b)

(c)

(d)

(e)

(f)Removal rate

Removal fraction to top surface

measured

Particle Transport Model: validation with water model data

calculated

Measure particles trapped by screen

9


pushing entrapment engulfment

Particle Capture by Dendritic Interface

particle pushing and capture by engulfment and entrapment.

χ

η

Forces acting on a spherical particle close to dendriticfront of mushy-zone:

Dendrite shape computed by N. Provatas, N. Goldenfeld, and J. A. Dantzig, 1998.

Buoyancy: FBDrag force: FDχ and FDηLift force: FLLubrication force: FLubVan der Waals force: FIConcentration Gradient force: FGradReaction force: FN(Press gradient, stress gradientadded mass and Basset forcesneglected.)

Solidification direction


Particle Capture Criterion

( ), 2 cos 0L D Lub Grad IF F F F Fχ θ− − − − <

( ) ( ) ( ), , ,cos sin sin 2D B L D Lub Grad IF F F F F F Fη η χθ θ θ+ + − > − −

( ) ( ) ( ), , ,cos sin sin 2D B L D Lub Grad IF F F F F F Fη η χθ θ θ− + − > − −

Particle size larger than PDAS (dp≥PDAS)?no Capture

yes

In solidification direction, repulsive force smaller than attractive force?noDrift back to flow

Particle contacts a boundary representing mushy-zone front?yes

noContinue

yes

Can cross-flow and buoyancy drive particle into motion though rotation?

, if FDη and FBη in same direction

, if FDη and FBη in opposite directionand FDη≥FBη

or

or, if FDη and FBη in opposite direction

and FDη<FBη

yes no

( ) ( ) ( ), , ,cos sin sin 2B D L D Lub Grad IF F F F F F Fη η χθ θ θ− + − > − −

10


Cross-Flow Velocity Effect on Capture of Slag Droplets

Critical downward cross-flow velocity (relative to shell) to capture slag droplets in solidifying steel dendritic interface

χ

η

Particle diameter, dp (µm)

Crit

ical

dow

nwar

dflo

wve

loci

tyre

lativ

eto

shel

l(m

/s)

0 100 200 300 400 500 600 700 800-0.12

-0.1

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

0.12

PDAS = 75µmPDAS = 100µmPDAS = 150µmPDAS = 200µm

Particle drifts downward

Particle drifts upward

Particle is captured

rd = 2.13µm, vsli=500µm/s


Validation of Particle Capture Criterion –Slag Droplets in Quiescent Steel

Particle radius (m)

Forc

em

agni

tude

sat

h 0cr(N

)

10-6 10-5 10-410-10

10-9

10-8

FiFLubFGradh0

cr calculated based on Kaptay 2001

Comparison of predicted and measured critical solidification speeds for pushing of slag particles in quiescent steel.

Radius of Particle (µm)

Inte

rface

adva

ncin

gsp

eed,

V Sol

(µm

/s)

0 10 20 30 400

5

10

15

20

25

Engulfed (Experiment, H. Shibata et al., 1998)Pushed (Experiment, H. Shibata et al., 1998)Predicted critical velocity, h0

cr from Kaptay, 2001Predicted critical velocity (no interfacial gradient force)

Comparison of the Van der Waals, lubrication and concentration gradient force.

11


Validation of Particle Capture Criterion -CrossFlow

Radius of spherical PMMA particle (m)

Flow

spee

dne

eded

topu

shpa

rticl

ein

tom

otio

n(m

/s)

0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.00070

0.0025

0.005

0.0075

0.01

0.0125

0.015

PMMA, R=4.2µm/sPMMA, R=4.2µm/s, experimentPMMA, R=68.8µm/sPMMA, R=68.8µm/s, experiment

Experimental data from Q. Han, 1994.

experimental setup (Q. Han 1994).

Comparison between predicted and measured critical flow speeds to push PMMA particle into motion.

water

PMMA particles settled on ice-water interface before introducing flow

100µm PMMA (Q. Han, 1994)


Particle Motion in Steel Caster

Animation-small

particles

Animation-small

particles

Animation-100 µmAnimation-100 µm Animation-mold slag particle entrainmentAnimation-mold slag particle entrainment

12


What Causes Asymmetry?

Time (s)

Dow

nwar

dsve

loci

ty,w

(m/s

)

20 25 30 35 40 45

-0.2-0.1

00.10.20.30.40.50.60.7

LeftRight

Time (s)

Dow

nwar

dsve

loci

ty,w

(m/s

)

20 25 30 35 40 45

-0.2-0.1

00.10.20.30.40.50.60.7

LeftRight

Particle Injection Period

Time (s)

Dow

nwar

dsve

loci

ty,w

(m/s

)

20 25 30 35 40 450

0.1

0.2

0.3

0.4

LeftRight

Particle Injection Period

zy

x


Inclusion Removal to top surface(originally entered from nozzle ports)

Time after first particle entered mold (s)0 20 40 60 80 100 120 140

0

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm

9s duration of particle injection

Time after first particle entered mold (s)

Frac

tion

ofpa

rticl

esre

mov

edby

the

top

surfa

ce(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100

12%12%

42%42%

70%70%

Particle density (silica): 2700 kg/m3

Particle density (silica): 2700 kg/m3

Very large particles can be removed with a straight-walled mold(owing to better buoyancy and more difficult to entrap)

Very large particles can be removed with a straight-walled mold(owing to better buoyancy and more difficult to entrap)

13


Removal and Entrapment History for Large Particles from Nozzle Ports

Time after first particle entered mold (s)0 20 40 60 80 100 120 140

0

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm



Frac

tion

ofpa

rticl

esre

mov

edby

the

top

surfa

ce(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100



Frac

tion

ofpa

rticl

esca

ptur

edin

the

uppe

r2.4

m(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm

11.29%24.30%39.01%Capture Fraction69.89%42.50%12.58%Removal Fraction400µm250µm100µmParticle Size

Fractions of particles removed to top surface and captured by upper 2.4m shell.


InclusionSize

SEN walls

Mold slag (top surface)

40µm 7% 8% Simulation

200µm 7% 42% Measurement

(tundish to slab) All 22%

Inclusion Removal in the Mold: Simulation and Measurements

Measurements: Zhang et al, AISTech 2004, Nashville, TN

SEN clogging accounts for some inclusion removal

SEN clogging accounts for some inclusion removal

14


-5

-4

-3

-2

-1

0

1

2(m)

0

1

2(m)

Inclusion Distribution in Solid Slab-5

-4

-3

-2

-1

0

1

2(m)

Casting Speed: 25.4mm/s

Meniscus Location whenInclusions Start EnteringLiquid-Pool

Last Inclusion Capturedin Solid Steel Strand

Narrowface viewNarrowface view

Zoom inZoom in

Wideface viewWideface view

9s sudden burst of smallinclusions into mold region9s sudden burst of smallinclusions into mold region

Inclusion size:

10 and 40 µm

Inclusion density:

2700 and 5000 Kg/m3

Inclusion size:

10 and 40 µm

Inclusion density:

2700 and 5000 Kg/m3


Predicted Total Oxygen ( Continuous Injection of Small Particles)

Predicted oxygen concentration in final steel slab

(10ppm oxygen from continuous injection of particles from nozzle ports).

( )op

p

(48/102) MC

x∆y∆z (1 / )Mpρ ρ ρ=

∆ + −

Oxygen concentration is computed from computed positions of entrapped small particles (≤40µm) by:

where:CN

i=1

3d

6p pπ ρ

∑Mp=

15


Mold Slag Entrainment

x

y

0 0.1 0.2 0.3 0.4-0.06

-0.04

-0.02

0

0.02

0.04

0.060.2 m/s

z = 0.59849 m

SEN

Vortexing around nozzle - erodes nozzle walls and entrains mold slagVortexing around nozzle - erodes nozzle walls and entrains mold slag

Instantaneous

Velocity vectors

(transient)

Instantaneous

Velocity vectors

(transient)

Time average

streamlines

Time average

streamlinesx

y

0 0.1 0.2 0.3 0.4

-0.05

0

0.05

Stream lines in the plane

SEN

Horizontal sections 38mm below top surfaceHorizontal sections 38mm below top surface


Transport of Large Particles Entrained from Top Surface

t = 1.80sdp = 100µm

t = 2.88sdp = 100µm

t = 5.58sdp = 100µm

t =7.38sdp = 100µm

4000 particles (100µm) introduced from volumes near top surface.

16


Removal of Large Particles (originally entrained from top surface)

1.8s of particle injection


Frac

tion

ofpa

rticl

esre

mov

edby

the

top

surfa

ce(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm


Removal of Large Particles Entrained from Top Surface



Frac

tion

ofpa

rticl

esre

mov

edby

the

top

surfa

ce(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm



Frac

tion

ofpa

rticl

esca

ptur

edin

the

uppe

r2.4

m(%

)

0 20 40 60 80 100 120 1400

10

20

30

40

50

60

70

80

90

100

100µm250µm400µm

0.40%4.03%24.93%Capture Fraction99.05%92.58%44.60%Removal Fraction400µm250µm100µmParticle Size

Fractions of particles removed to top surface and captured by upper 2.4m shell.

17


ConclusionsLES and Lagrangian particle transport can predict turbulent liquid-particle flow in continuous steel casting molds with reasonable accuracy.A simple particle capture criterion based on force balance appears to agree with prior experimental results.Particle entrapment into a solidifying dendritic interface depends on whether many forces can balance including: drag from transverse flow and surface energy gradient forces from sulfur concentration gradients Removal fraction to top surface slag layer of slag droplets from nozzle ports is 70% of 400µm but < 12% for ≤100µm; Re-entrainment of slag particles at the top surface depends on particle size. >92% of 250µm particles return to the slag but >50% of 100µm incs are eventually entrapped in the steel


Implications

Difficult to remove inclusions in the moldOptimize upstream processes to remove inclusions before they get to the moldOptimize flow in the mold:– Avoid skewed surface profile, level fluctuations,

slag entrainment, and other problems– Avoid meniscus freezing and hooks

Computational modeling is a powerful tool to predict transient flow, level fluctuations, surface defects, and inclusion behavior

Continuous Casting Defects from Transient Flow & Inclusion...

Documents

Transcript of Continuous Casting Defects from Transient Flow & Inclusion...