hook formation near slab corners and prediction of …ccc.illinois.edu/s/2006_presentations/04_gogi...

ANNUAL REPORT 2006Meeting date: June 15, 2006

Go-Gi Lee(Visiting Scholar from POSTECH, South Korea)

Brian G. Thomas

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

Hook formation near slab corners and prediction of argon gas bubble size

in the slab casting mold

University of Illinois at Urbana-Champaign • Metals Processing Simulation Lab • Go-Gi Lee 2

Acknowledgements

• Continuous Casting Consortium at UIUC(Accumold, Algoma steel, Bosteel, Corus, Mittal Riverdale, LaBein, LWB Refractories, Nucor, Postech, Steel Dynamics)

• POSCO

• Korea Research Foundation

• Professor Brian G. Thomas

• Professor Seon-Hyo Kim

• Dr. Ho-Jung Shin, POSTECH


1. Investigation of 3-D Frozen Meniscus Shape


3-D Frozen Meniscus Region- corners of slab

< Narrow face results >

H.-J. Shin, CCC Annual Report, 2005

• Micrographs of the hook come from 2-D vertical section view of certain location • Near the corners of the slab: deepest hook depth at corners; complex 3-D hook shape

3-D shape of frozen meniscus at corners is not clear

Curved hook at 20mm distance from the corner

-110 -90 -70 -50 -30 -10 10 30 50 70 90 1100.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

OutsideCenter of narrow face

Inside

Mea

n ho

ok d

epth

(m

m)

Perimeter of narrow face (mm)

2002 POSCO, measured 2003 POSCO, measured 2004 POSCO, measured

0 2mm

Scale

Bubble

J. Sengupta et al., Acta Materilia, 2006


Metallurgical Method

20mm

Corner

CL

115mm

Sample for microscopy analysis

13mm Wide face

Casting direction

Sample size: 20mm(width) * 13mm(depth) * 30mm(height)

OM

Narrow face

• Ten micrographs (2-D vertical section view) were obtained from YZ planeaccording to X distance(~6mm) from corners

< Location of sample > < Definition of axes >

Narrow face

Wide faceX

Y

ZCORNER

OM

Observing plane (YZ plane)

Casting Direction


Casting Conditions

0.2410.100121745.0

Positive strip time (sec)

Negative strip time (sec)

Modification ratio for non-sinusoidal mode (%)

Frequency(cpm)

Stroke(mm)

- Slab thickness: 230 mm; Slab width: 1300 mm; Casting speed: 1.45 m/min; Pour temperature: 1571˚C

~ 0.040.050.01

~ 0.020.010.01~ 0.01~ 0.015≤ 0.0050.08< 0.005

AlTiCuNiCrSPSiMnC

3.21118011701149

Viscosity at 1300 oC(Poise)

Melting Temperature(oC)

Softening Temperature(oC)

Solidif. Temperature (oC)Properties

0.3506.720.113.430.030.030.340.185.970.8439.836.31.10

Li2OB2O3FK2ONa2OP2O5MnO2Fe2O3TiO2Al2O3MgOCaOSiO2BasicityChemicalcomposition

(wt. %)

Steel composition of ultra-low carbon steel (wt. %)

Mold powder composition and properties

Mold oscillation conditions


Actual Measurement of Micrograph

Micrograph of YZ plane at X = 1.7mm

Blue color: Upper line of frozen meniscusRed color: Lower line of frozen meniscus Gray color: Profile of oscillation markGreen color: Tip of frozen meniscus

1.8

1.7

1.6 02

46

810

12

-25

-20

-15

-10

-5

0

3rd OM

2nd OM

Ca

stin

g di

r ect

i on

Wide face (mm)

Narrow face (mm)

1st OMScale

0 2mm

1st

OM

2nd

OM

3rd

OM

Z=0

Slab surfaceY=0

DT

DT : Distance from slab surface to tip of frozen meniscus


Analysis of Micrographs

6 5 4 3 2 1 0 02

46

810

12-25

-20

-15

-10

-5

0

5

3rd OM

2nd OM

Ca

stin

g d

ire

ctio

n (

mm

)

Wide fa

ce (m

m)

Narrow face (mm)

Corner

1st OM

2.5mm3.5mm

5.0mm

< Micrographs according to the X direction >

Blue color: Upper line of frozen meniscusRed color: Lower line of frozen meniscus Gray color: Profile of oscillation markGreen color: Tip of frozen meniscus

X = 2.5mm 3.5mm 5.0mm

Scale

0 2mm

1st

OM

2nd

OM

3rd

OM


Investigate of Frozen Meniscus Shape

Hook characteristics:Continuous defect along the OM; 3-D structure

0 2 4 6 8 10 12 14 16 18 20 220

2

4

6

8

10

12

14

16

18

20

22

Wid

e F

ace

(mm

)

Narrow Face (mm)

DT at 1st OM

DT at 2nd OM

DT atf 3rd OM

DT: Distance from slab surface to

the tip of the hook structure

Corner

Max. hook depth point

65

43

21

0 0 2 4 6 8 10 12-10

-8

-6

-4

-2

0

2

4

Ca

stin

g D

ire

ctio

n (

mm

)

Wide Face (mm)

Narrow Face (mm)

Upper lines of forzen meniscus Lower lines of forzen meniscus Tip of frozen meniscus

Corner

Front view of frozen meniscus at 1st OM Tops view of frozen meniscus


Conclusions

• Frozen meniscus shape near the corner of slabs is investigated with metallurgical analysis:- Micrographs show the 3-D frozen meniscus

shape

- Hook is continuous defect at the oscillation mark running completely around the strand perimeter- Deepest hook depth appears near corners


2. Investigation of Argon Bubble Sizein Water Model using Porous Nozzle Refractory


Water Model Apparatus (1/3 scale)

Water flowmeter

Surface

Ф 25*11 exit

Tundish

500mm

Slide gate

Inner diameter of SEN: 25mm

1200mm

Submergence depth

Gas flowmeter

Water flowmeter

Water bath

Refractory

Pump77mm

Gas injection

High speed camera

Size of MgO refractory brick: 14mm(W)*44mm(L)*17mm(D)


Schematic of Upper Tundish Nozzlein Water Model

UTN

Slidegate

SEN

Bottom of the tundish bath

140mm

45mm

340mm

10mm

10mm

Ar injection

Acrylic cap

Ar injection

High speed

camera

Refractory

Outlet port

80mm

22mm

Nozzle wall

25mm

Top view

Side view


Conditions of the Experiment

< Properties of MgO refractory bricks >

•Fixed test conditions - Submergence depth: 60mm - SEN outport angle: 35°downward - Height of tundish bath: 330mm

• Factors - Total gas flow rate: 0.1, 0.2, 0.3 l /min - Mean water velocity: 0.96, 1.10, 1.25 m/s

Brick 1 Brick 2Fired B.D. (g/cc) 2.9 2.9Porosity (%) 16.2 17.6Permeability (nPm) 7.52 16.32Modulus of rupture (psi) 1085 1119

< Test conditions in water model >

Note: MgO uses to simulate the dolomite refractories


Similarity Analysis between Steel Caster & Water Model

2

2 2

3

( )

( )

lliq

gas

l

l

l g

g g

l g

l

li

lq

UInertial forcedueto liquid momentum

Buoyancy force Dg

d VInertial forcedueto gas momentum

Buoyancy force D g

DUInertial forcedueto liquid momentum

Viscous forceof liq

Fr

F

uid

r

Re

Re

ρρ ρ

ρρ ρρμ

= =−

= =−

= =

2 2

2 2

g g

g

l l

g g

gas

liq

gas

dVInertial forcedueto gas momentum

Viscous forceof gas

D UInertial forcedueto liquid momentum

Surfacetension force D

d VInertial forcedueto gas momentum

Surfacetension force

We

eD

W

ρμ

ργ

ργ

= =

= =

= =

Ref.) H. Bai: Ph. D. Thesis. UIUC, Urbana, IL, 2000

,

: , : ,

: , : , : ,

: , : , : ,

: /

l g l

g l g

where

D averagebubblediameter at a pore d gas injection porediameter

liquid density gas density liquid viscosity

gas viscosity U liquid velocity V gas velocity

liquid gas surfacetensioncoefficient

ρ ρ μμγ , :g gravitational acceleration

**) Note: Gas density was calculated for the temperature in liquid and the pressure due to the height of liquid in tundish (see appendix 2)

Dimensionless groups

Physical properties used in calculation

ParametersWater-air

systemSteel-hotAr system

Liquid density (ρl , kg/m3) 1000 7020

Gas density (ρg , kg/m3) 1.331* 0.446*

Liquid viscosity (μl , kg/m·s) 0.001 0.0056

Gas viscosity (μg , kg/m·s) 1.70E-05 7.42E-05

Surface tension coefficient (γl/g , N/m) 0.073 1.192


Liquid Flow Rate between Two Systems

( )( )

( )( )

( )

3

3

,

,

2,

,

,

,

,

,

,

W

S

W nozzle

S nozzle

W nozzle

Here

Q is the water flowrate m s

Q is the steel flowrate m s

U is the water velocity in water mo del nozzle m s

U is the steel velocity in actual nozzle m s

A is the cross sectional area of water model nozzle m

A

−

( )( )

2, ,S nozzle

C

is thecross se ctional area of actual nozzle m

V is thecasting speed m min

−

Table. Relationship between the liquid flow rate and casting speed

We choose the same liquid velocity between two systems, because the refractory geometry scale is 1:1

, ,

, ,

W nozzle S nozzle

S S nozzle S nozzle

U U

Q U A

=

∴ = ⋅

Note) Bore diameter of SEN is 25mm in water model and 75mm in steel caster, respectively

Water flow rate (l /min) Qw 28.2 32.5 36.8Liquid velocity in nozzle (m/sec) Uw,nozzle(=Us,nozzle) 0.96 1.10 1.25

Steel thoughput (m3/min) Qs 0.25 0.29 0.33

Casting speed (m/min, 230mm*1500mm slab) Vc 0.74 0.85 0.96

SC

QV

Cross setional area of the strand

⎛ ⎞∴ = ⎜ ⎟−⎝ ⎠

, , ,,

WW W nozzle W nozzle W nozzle

W nozzle

QQ U A U

A= ⋅ ∴ =

From mass balance


Gas Flow Rate Ratiocaused Surface Area of Refractory

, ,

air in air inair out

T W U W W

Q QQ

N N A− −

− = =⋅

QAr-hot

QAr-cold

, ,

Ar cold Ar coldAr hot

T S U S S

Q QQ

N N A

β β− −−

⋅ ⋅= =⋅

Qair-out

Qair-in

d

WaterFlow

Liquid Steel Flow

UU

< Schematic of water model experiment > < Schematic of steel caster >

Ar hot W Ar cold

air out S air in

Q A Q

Q A Qβ− −

− −

=:

: *

*) 1

W

S

Ar gas expansincoefficient

A Surfacearea ratioof two refractoryA

seetheappendix

β

:

:

:

T

U

N Total active sites oninner wall surfaceof refractory

N Active sites per unit areaof refractory

A Inner wall areaof refractory

d


Summary of Dimensionless Number Analysis

0 1 2 3 4 5 6 7 8 9 100.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

High liquid velocity (U>1.4m/s) Fr

liq Re

liq We

liq

Frgas

Regas

Wegas

Qa

ir-in (

l/min

)

QAr-cold

(l/min)

At high liquid velocity (U>1.4m/sec):Dair≈DAr at the same gas flow rate

Qair-in Qair-in

QAr-cold QAr-cold

(This work) (Bai's calculation)Frliq 0.07 0.285Reliq 0.087 0.356Weliq 0.03 0.123Frgas 0.0152 0.067Regas 0.005 0.026Wegas 0.01 0.044

Dimensionlessnumber

Similarity requirement for gas flow rate

Matching all of the dimensionless groups at the same time is impossible


Active Sites in Stagnant Flow

LWB RefractoriesLow permeability(7.52nPm) Qg = 0.3 l/min

22 active sites

Chosun RefractoriesPorosity = 20% Qg = 0.018 l/min

40 active sites0 1mm

Scale

0 1mm

Scale

0.0 0.1 0.2 0.3 0.4 0.50

5

10

15

20

25

30

35

40

45

50

55

60

Permeability (LWB Ref.) 7.52 nPm 16.32nPm

Porosity (Chosun Ref.) 20%

Act

ive

site

s in

sta

gnan

t flo

w (

#)

Total gas flow rate (l/min)

• Number of active site and bubble size through the porous refractory in stagnant flow are mainly dependent of the texture of porous refractory

• Generally speaking, active site increases as gas flow rate increases


Comparison of Mean Active Sites between Flowing and Stagnant Flow

• Active sites in downward flow are about 2 times those in stagnant flow

• Gas flow rate per site of high permeability is slightly higher than that of low permeability in downward-flowing flow at the same total gas flow rate

0.0 0.1 0.2 0.3 0.4 0.50

5

10

15

20

25

30

35

40

45

50

55

60

Downward-flowing flow Low permeability High permeability

Stagnant flow Low permeability High permeability

Mea

n ac

tive

site

s (#

)

Total gas flow rate (l/min)

Test conditions- Low permeability (7.52nPm)- Mean liquid velocity: 1.25 m/s- Total gas flow rate = 0.1 l/min- Camera speed: 4000 frames/s

t=0 t=0.25ms t=0.5ms

< Example of experiment photograph series >


Calculation of Total Active Sites in the Steel Caster Nozzle Wall

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.400

1

2

3

4

5

6

7

8

9

10

Downward-flowing flow Stagnant flow Low permeability Low permeability High permeability High permeability

NU (

#/cm

2 )

QS (ml/s)

*) see appendix 1

0 1 2 3 4 5 6 7 8 9 10 11 120

5

10

15

20

25

NU (

#/c

m2)

QT (l/min, cold)

Low permeability High permeability measurement measurement estimated value estimated value

Downward-flowing flow

Nu (#/cm2) = 2.48 lnQT (l/min) + 9.8

Nu (#/cm2) = 2.15 lnQT (l/min) + 8.7

( ) ( )( )

TS

T

Q Total gas flowrateQ Mean gas flowrate per site

N Total active sites=

• Low permeability is much more active sites

Water Model Steel CasterQT (Total gas flow rate-cold, l/min) 0.2 7

NU (Mean active sites per unit area, #/cm2) 5.5 13.732

A (Inner wall area of refractory*, cm2) 6.2 353.4NT (Total active sites, #) 34 4853


Comparison of Estimated Mean Gas Flow Rate per Site between Two Systems

0.00 0.05 0.10 0.15 0.20 0.250.00

0.05

0.10

0.15

0.20

0.25

Me

asu

red

mea

n g

as

flow

ra

tep

er

site

in w

ate

r m

od

el (

ml/s

, ho

t)

Estimated mean gas flow rate per site in steel caster (ml/s, hot)

Similarity of mean gas flow rate per site between two systemsIt makes the result of water model is meaningful

( )* 2 2

, /

( ) (#/ )T

S U

Q cold l min

A cm N cm

β ×=

×

Estimated mean gas flow rate per sitein steel caster (ml/s, hot)

*) AS = 353.4cm2, see appendix 1

NU (Mean active sites per unit area on inner wall of refractory)= 2.3154 lnQT(cold, l/min) + 9.2263

QT (Total gas flow rate-cold, l/min) 0.1 0.2 0.3 0.4 5 7 9 11β (Gas volume expansion coefficient**)NT (Total active sites on inner wall area, #) 24 34 40 44 4578 4853 5058 5223Estimated mean gas flow rate per site (hot, ml/s) 0.069 0.098 0.125 0.151 0.073 0.096 0.119 0.140

1 4

Steel CasterWater Model


Bubble Size Distributions:measured in water model nozzle

1.0 1.5 2.0 2.5 3.0 3.5 4.00

10

20

30

40

50

60

70

80

90

100

Bub

ble

volu

me

per

cen

tag

e (

%)

The range of bubble diameter (mm)

Low permeability High permeability 0.96 m/s 0.96 m/s 1.1 m/s 1.1 m/s 1.25 m/s 1.25 m/s

Mean liquid velocity

Total gas flow rate = 0.1 l/min

1.0 1.5 2.0 2.5 3.0 3.5 4.00

10

20

30

40

50

60

70

80

90

100

Low permeability High permeability 0.96 m/s 0.96 m/s 1.1 m/s 1.1 m/s 1.25 m/s 1.25 m/s

Bu

bbl

e vo

lum

e p

erc

en

tag

e (

%)

The range of bubble diameter (mm)

Mean liquid velocity

Total gas flow rate = 0.3 l/min

0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.20.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.20.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Low permeability (7.52nPm)

Sta

nda

rd d

evia

tion

of b

ubb

le d

iam

eter

(m

m)

Mean bubble diameter (mm)

Correlation coefficient = 0.83047Correlation coefficient = 0.72677

High permeability (16.32nPm)

• Bubble size variations increase with increasing gas flow rate and increasing bubble diameter


Effect of Liquid Velocity and Gas Flow Rate on Bubble Diameter (water model)

• The mean bubble size increases with increasing gas flow rate and decreasing liquid velocity

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.40.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

Low permeability High permeability

Mea

n bu

bble

dia

met

er (

mm

)

Gas injection volume fraction (%)

*100( )

Total gas flowrateGas injection volume fraction

Total gas flowrate Liquid flowrate=

+


Comparison of Measured and Predicted Gas Bubble Size

0.0 0.1 0.2 0.3 0.4 0.50.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Mea

n bu

bble

dia

met

er (

mm

)

Mean gas flow rate per site (ml/s)

d=0.1mm, predicted in water model d=0.4mm, predicted in water model Low permeability, measured High permeability, measured Liquid steel argon prediction

Mean liquid velocity = 0.96 m/s

d: Gas injection hole diameter

0.0 0.1 0.2 0.3 0.4 0.50.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Mea

n b

ubb

le d

iam

eter

(m

m)

Mean gas flow rate per site (ml/s)

d=0.4mm, predicted in water model d=0.1mm, predicted in water model Low permeability, measured High permeability, measured Liquid steel argon prediction

Mean liquid velocity = 1.25 m/s

d: Gas injection hole diameter

Ref.) Bai & Thomas, Met. Trans., 2001

Bubble size prediction model:

Total gas flowrateMean gas flowrate per site

Mean active sites=


Comparison of Estimated Ar Bubble Size between Steel Caster and Water Model

using the Bai’s PredictionRef.) Bai & Thomas, Met. Trans., 2001

0.0 0.5 1.0 1.5 2.0 2.50

1

2

3

4

5

6

7

8

Mea

n bu

bble

dia

met

er (

mm

)

Mean liquid velocity (m/s)

Air flow into water Ar flow into steel 0.069 ml/s 0.073~0.14 ml/s 0.151 ml/s

Gas injection flow rate per site (hot)

Gas injection hole diameter = 0.1mm


Conclusions

• Investigation of Ar bubble size in water model nozzle using porous refractory- Micro texture of the porous refractory greatly affects injected gas behavior in stagnant flow- Active sites in downward-flowing flow are about two times compare with stagnant measurement- Although matching different dimensionless groups at the same time is impossible, the similarity of mean gas flow rate per site between water model and steel caster is good- Bubble size variations increase with increasing gas flow rate and increasing bubble diameter- Mean bubble size increases with increasing gas flow rate and decreasing liquid velocity- The match of mean bubble size between the previous model prediction by Bai and the experimental data is reasonably good


Appendix 1. Surface Area Ratio

Effective length of UTN considered distribution of Ar gas velocity

0.00

0.05

0.10

0.15

0.20

0.25

0.000.010.020.030.04

Velocity (m/s)

Dis

tan

ce fr

om

bo

ttom

to to

p o

f th

e n

ozz

le (

m). Slab P

LeffLUTN

75mm

150mm

Distribution of Ar gas velocity

Cylinder-like refractory in actual plant

Surface area (AS)= 353.4cm2

Rectangular refractory fixed in water model

Surface area (AW)= 6.2cm2

vs 44mm

14mm

17mm

B. G. Thomas & Z. Hashisho, CCC Annual Report, 2005


Appendix 2. Calculation of Gas Density

Table. Properties used in calculation

Standard pressure Pstd 101325 Pa

The Density of the liquid (kg/m3)

Density of water ρW 1000 kg/m3

Density of steel ρS 7020 kg/m3

The Density of the gas (kg/m3, STP)

Density of Ar gas ρAr 1.783 kg/m3

Density of air gas ρAir 1.29 kg/m3

The height of steel in tundish Hs 1.0 m

The height of water in tundish Hw 0.33 m

( )(1) '

( )

(2) '

.(1) '& (2) '

g

g

g

is density of gas

weight m

volume V

Ideal gas eqution is PV nRT

Combining Eqs gives

mP

nRT

ρ

ρ

ρ

=

=

=

( )( )

( ) ( )

273

273

3

273

273

273

1.032

1.032* 1.331

Air K std std W W

Air std K stdstd

Air AirK std

T P g HP

P T P

kg m

ρ ρρ

ρ ρ

⎛ ⎞⎛ ⎞ ⎛ ⎞+ ⋅ ⋅ ⎛ ⎞= =⎜ ⎟⎜ ⎟ ⎜ ⎟ ⎜ ⎟⎝ ⎠⎝ ⎠⎝ ⎠ ⎝ ⎠

=∴ = =

( )( )

( ) ( )

1833

1833

3

1833

273

1833

1 11.679* 0.25

6.714 4

0.25* 0.446

Ar K std std S S

Ar std K stdstd

Ar ArK std

T P g HP

P T P

kg m

ρ ρρ

ρ ρ

⎛ ⎞⎛ ⎞ ⎛ ⎞+ ⋅ ⋅ ⎛ ⎞∴ = =⎜ ⎟⎜ ⎟ ⎜ ⎟ ⎜ ⎟⎝ ⎠⎝ ⎠⎝ ⎠ ⎝ ⎠

⎛ ⎞= ≈ =⎜ ⎟⎝ ⎠

∴ = =

( )( )

3

3

0.446

1.331

Ar hot

air out

kg m

kg m

ρ

ρ−

−

∴ =

∴ =

∴Ar volume expansion coefficient relative to STP = 4

1) Steel-Ar system

2) Water-air system

hook formation near slab corners and prediction of …ccc.illinois.edu/s/2006_presentations/04_gogi...

Documents