Investigation of Oscillation Marks and Hook Formation in ULC Steels using Metallurgical Analysis and Models Ho Jung Shin Department of Mechanical & Industrial Engineering University of Illinois at Urbana-Champaign May 10, 2004 University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 1
Acknowledgement Professor Brian G. Thomas Professor Seon-Hyo Kim Dr. Ya Meng Mr. Go-Gi Lee, POSTECH Continuous Casting Consortium POSCO University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 2
Introduction Initial solidification during continuous casting of steel slabs greatly affects surface quality of slabs Molten steel flow Superheat Casting speed Heat flux to gap Mold level fluctuation Mold oscillation Initial solidification Mold powder properties and consumption rate Other process parameters Most operation parameters have strong interdependencies This makes prediction and interpretation of initial solidification phenomena difficult University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 3
Importance of Initial Solidification Initial solidification behavior affects the formation of oscillation marks (OM) and subsurface hooks Initial solidification features related to surface quality problems - Deep oscillation mark : Transverse crack formation - Deep hook : Easy Entrapment of mold flux and inclusion-laden gas bubbles < Transverse Crack at valley of OM> < Bubble captured by hook > University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 4
Modeling analysis Objectives Plant production Metallurgical analysis Gathering plant operation data during production of slabs Gathering samples after production of slabs Model validation Prediction of the phenomena in the mold for different conditions Analysis of the OM profiles Depth, width, pitch Etching analysis of hook characteristics Shape, size, primary dendrite arm spacing Interpretation & Understanding of Initial solidification behavior Investigate effect of casting speed, EMS and superheat Suggest solutions to achieve small hook on slabs University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 5
Test Conditions (2003) - Steel grade : Ultra low carbon steel ( C 0.005% ) - Slab thickness : 230 mm - Nozzle submergence depth : ~ 160 mm - Oscillation asymmetry : 59 % 160 mm 104 mm 35 mm 80 mm Test 1 Test 2 Test 3 Test 4 Test 5 Casting speed (m/min) 1.75 1.42 1.22 1.47 1.47 Pour Temperature ( ) 1560 1564 1564 1571 1571 Electromagnetic current (A) 313 234 0 277 0 Slab width (mm) 1300 1570 Working mold length (mm) 775 796 790 782 774 Oscillation stroke (mm) 7.5 6.83 6.43 6.93 6.94 Oscillation frequency (mm) 187 155 135 159 160 University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 6
Mold Powder Properties Melting Characteristics Physical Properties Density (g/ ml ) 2.7 Crystal Temp. ( ) Softening, Melting, Flowing point ( ) 1,145 1400 1.28 Crystal Ratio (%) 6.80 Viscosity (poise) 1300 2.62 1200 1100 5.85 ~ 1000 SF-Tension (d/cm) 48.72 Solid Temp ( ) 1,101 Shape Granule µ = ' TS T µ s T T fsol fsol n 50 40 Mold flux n : 1.15 n : 1.75 µ s ' T S : viscosity at 1300 : 1300 n : Exponent for temperature dependency of viscosity Viscosity (poise) 30 20 10?? 0 {( T s ' - T fsol ) / ( T - T fsol )} n Solid Temperautre = 1101 ( o C) 0 n = 1.15 was chosen for this modeling 1050 1100 1150 1200 1250 1300 1350 1400 1450 1500 Temperature ( o C) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 7
Mold Calibration (Heat Flux Profile) Differences of cooling water temperature and mean heat flux Measured ΔT ( ) Measured mean heat flux ( kw/m 2 ) Calculated ΔT ( ) Calculated mean heat flux ( kw/m 2 ) Test 1 6.75 1665.9 6.75 1709.4 Test 2 5.78 1428.5 5.78 1422.8 Test 3 5.72 1413.1 5.71 1417.2 Test 4 7.08 1429.8 7.08 1450.6 Test 5 7.49 1515.9 7.46 1544.8 Distance of thermocouple location below meniscus Test 1 Test 2 Test 3 Test 4 Test 5 Mean temperature from hot thermocouple Test 1 Test 2 Test 3 Test 4 Test 5 1 st (mm) 173.8 160.1 149.3 162.6 159.7 1 st ( ) 173.8 160.1 149.3 162.6 159.7 2 nd (mm) 126.2 128.9 113.8 133.2 129.0 2 nd ( ) 126.2 128.9 113.8 133.2 129.0 3 rd (mm) 115.7 109.7 102.6 113.3 112.0 3 rd ( ) 115.7 109.7 102.6 113.3 112.0 University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 8
Calibration with measured temperatures of thermocouples Mold 40 mm Hot- Thermocouple Contact resistance 21 mm Measured Temperature (T 1 ) Cooling water Mold 22 mm Predicted Temperature by CON 1D (T 2 ) Cooling water channel Mold 22 mm (T 2 ) (T 1 ) Cooling water Mold? mm (T 1 ) Predicted Temperature by CON 1D (T 2 ) < Fig 1. Thermocouple location> < Fig 2. Predicted temperature > by CON 1D < Fig 3. Temperature gradient between > water channel and thermocouple < Fig 4. Concept of Offset> University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 9
Matching model with mold temperature 210 200 200 200 190 Predicted Thermocouple Temperature ( 7mm offset) Measuered Thermocouple Temperature 180 Predicted Thermocouple Temperature ( 7mm offset) Measuered Thermocouple Temperature 180 Predicted Thermocouple Temperature ( 7mm offset) Measuered Thermocouple Temperature 180 Temperature ( o C) 170 160 150 140 130 120 Temperature ( o C) 160 140 120 Temperature ( o C) 160 140 120 110 100 100 100 90 0 100 200 300 400 Distance below meniscus (mm) < Test 1 > 80 0 100 200 300 400 Distance below meniscus (mm) < Test 2 > 80 0 100 200 300 400 Distance below meniscus (mm) < Test 3 > 200 180 Predicted Thermocouple Temperature ( 7mm offset) Measuered Thermocouple Temperature 200 180 Predicted Thermocouple Temperature ( 7mm offset) Measuered Thermocouple Temperature Temperature ( o C) 160 140 120 Temperature ( o C) 160 140 120 100 100 80 0 100 200 300 400 Distance below meniscus (mm) < Test 4 > 80 0 100 200 300 400 Distance below meniscus (mm) < Test 5 > University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 10
Heat Flux Comparison Heat Flux (W/m 2 ) 8 7 6 5 4 3 2 Test 1 (Vc = 1.75 m/min, Hf = 1.71 MW/m 2 ) Test 2 (Vc = 1.42 m/min, Hf = 1.43 MW/m 2 ) Test 3 (Vc = 1.22 m/min, Hf = 1.42 MW/m 2 ) Test 4 (Vc = 1.47 m/min, Hf = 1.45 MW/m 2 ) Test 5 (Vc = 1.47 m/min, Hf = 1.54 MW/m 2 ) 1 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 11
Mold Temperature Comparison Temperature of hot face mold ( o C) 350 300 250 200 150 Test 1 (Vc = 1.75 m/min, Hf = 1.71 MW/m 2 ) Test 2 (Vc = 1.42 m/min, Hf = 1.43 MW/m 2 ) Test 3 (Vc = 1.22 m/min, Hf = 1.42 MW/m 2 ) Test 4 (Vc = 1.47 m/min, Hf = 1.45 MW/m 2 ) Test 5 (Vc = 1.47 m/min, Hf = 1.54 MW/m 2 ) 100 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 12
Mold Temperature Comparison Temperature of cold face mold ( o C) 150 140 130 120 110 100 90 80 70 60 Test 1 (Vc = 1.75 m/min, Hf = 1.71 MW/m 2 ) Test 2 (Vc = 1.42 m/min, Hf = 1.43 MW/m 2 ) Test 3 (Vc = 1.22 m/min, Hf = 1.42 MW/m 2 ) Test 4 (Vc = 1.47 m/min, Hf = 1.45 MW/m 2 ) Test 5 (Vc = 1.47 m/min, Hf = 1.54 MW/m 2 ) 50 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 13
Shell Temperature Comparison Temperature of shell ( o C) 1550 1500 1450 1400 1350 1300 1250 1200 Test 1 (Vc = 1.75 m/min, Hf = 1.71 MW/m 2 ) Test 2 (Vc = 1.42 m/min, Hf = 1.43 MW/m 2 ) Test 3 (Vc = 1.22 m/min, Hf = 1.42 MW/m 2 ) Test 4 (Vc = 1.47 m/min, Hf = 1.45 MW/m 2 ) Test 5 (Vc = 1.47 m/min, Hf = 1.54 MW/m 2 ) 1150 1100 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 14
Mold Powder Layer Thickness in Gap 1.8 1.6 1.4 Thickness (mm) 1.2 1.0 0.8 0.6 0.4 0.2 Test 1 (Solid layer), (Liquid layer) Test 2 (Solid layer), (Liquid layer) Test 3 (Solid layer), (Liquid layer) Test 4 (Solid layer), (Liquid layer) Test 5 (Solid layer), (Liquid layer) 0.0 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 15
Shell Thickness Comparison 20 18 16 Shell Thickness (mm) 14 12 10 8 6 Test 1 (Vc = 1.75 m/min, Hf = 1.71 MW/m 2 ) Test 2 (Vc = 1.42 m/min, Hf = 1.43 MW/m 2 ) 4 Test 3 (Vc = 1.22 m/min, Hf = 1.42 MW/m 2 ) Test 4 (Vc = 1.47 m/min, Hf = 1.45 MW/m 2 ) 2 Test 5 (Vc = 1.47 m/min, Hf = 1.54 MW/m 2 ) 0 0 100 200 300 400 500 600 700 800 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 16
Measurement of OM Profile X axis Method Y axis Starting point -115 0 Center of narrow slab face 115 Casting Direction Profile of oscillation mark (mm) Y = 0 mm, Y = 10 mm, Y = 20 mm, Y = 30 mm, Y = 40 mm, Y = 50 mm X axis : 0 ~ 80 mm Y axis : 0 ~ 200 mm (10 mm Interval) 0 20 40 60 80 X Axis (mm) < Example of Oscillation Mark Profiles > University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 17
Average Results of OM Profiles OM Pitch (mm) Calculated Measured OM Depth (mm) OM Width (mm) Oscillation stroke (mm) Oscillation frequency (cycle / minute) Test 1 9.35 9.44 0.251 6.12 7.50 187 Test 2 9.16 9.10 0.328 6.76 6.84 155 Test 3 9.00 8.91 0.280 6.66 6.42 135 Test 4 9.20 9.36 0.314 6.29 6.95 160 Test 5 9.19 9.09 0.272 6.57 6.92 159 Oscillation mark depth (mm) 1.0 0.8 0.6 0.4 0.2 Averaged OM depth of of Test 1 Averaged OM depth of of Test 2 Averaged OM depth of of Test 3 Averaged OM depth of of Test 4 Averaged OM depth of of Test 5 0.0 1.0 1.2 1.4 1.6 1.8 2.0 Casting speed (m/min) Stoke Set Point = A S + B S V C Frequency Set Point = A F + B F V C A F, B F, A S, B S : Coefficients, F: Frequency, S: Stroke, V C : Casting speed Same tendency with stroke and frequency of mold oscillation University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 18
16 14 Test 1 (Vc : 1.75 m / min, cating width : 1300 mm) Mean pitch of oscillation mark Calculated oscillation mark pitch (= Vc / Frequency ) Results of Mean Pitch of OM around Strand Perimeter 16 16 14 14 Test Test 2 (Vc (Vc: : 1.42 1.42 m // min, min, cating cating width width :: 1300 1300 mm) mm) Mean Mean pitch pitch of of oscillation oscillation mark mark Calculated Calculated oscillation oscillation mark mark pitch pitch (= (= Vc Vc // Frequency Frequency )) 16 16 14 14 Test Test 3 (Vc (Vc: : 1.21 1.21 m // min, min, cating cating width width :: 1300 1300 mm) mm) Mean Mean pitch pitch of of oscillation oscillation mark mark Calculated Calculated oscillation oscillation mark mark pitch pitch (= (= Vc Vc // Frequency Frequency )) Mean Pitch of OM (mm) 12 10 8 Mean Pitch of of OM (mm) 12 12 10 10 8 Mean Pitch of of OM (mm) 12 12 10 10 8 6-100 -50 0 50 100 Distance from center of narrow slab face (mm) < Test 1 - EMS 313 A > 6-100 -100-50 -50 0 50 50 100 100 Distance from center of of narrow slab face (mm) 6-100 -100-50 -50 0 50 50 100 100 Distance from center of of narrow slab face (mm) < Test 2 - EMS 234 A > < Test 3 - EMS 0 A > 16 14 Test 4 (Vc : 1.47 m / min, cating width : 1570 mm) Mean pitch of oscillation mark Calculated oscillation mark pitch (= Vc / Frequency ) 16 14 Test 5 (Vc : 1.46 m / min, cating width : 1570 mm) Mean pitch of oscillation mark Calculated oscillation mark pitch (= Vc / Frequency ) Mean Pitch of OM (mm) 12 10 8 Mean Pitch of OM (mm) 12 10 8 6 6-100 -50 0 50 100-100 -50 0 50 100 Distance from center of narrow slab face (mm) Distance from center of narrow slab face (mm) < Test 4 - EMS 277 A > < Test 5 - EMS 0 A > Higher EMS causes greater pitch variations University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 19
Relation between Level Fluctuation and OM Pitch Measured Level fluctuation with Eddy Current Sensor (mm) 3 Test 1 Test 2 Test 3 Test 4 Test 5 2 1 0-1 -2-3 0 5 10 15 20 25 30 35 40 45 Time (sec) Standard deviation of OM pitch difference (mm) 3.5 3.0 2.5 2.0 1.5 1.0 Test 1 ( Vc= 1.75 m/min) Test 2 ( Vc= 1.42 m/min) 0.5 Test 3 ( Vc= 1.22 m/min) Test 4 ( Vc= 1.47 m/min) Test 5 ( Vc= 1.47 m/min) 0.0 0.4 0.5 0.6 0.7 0.8 Standard deviation of level fluctuation (mm) - Standard deviation of level fluctuation : average value during 45 seconds while the samples were produced - OM pitch difference = ( Measured OM Pitch ) ( Calculated OM Pitch ) * Calculated OM Pitch = casting speed / Oscillating frequency University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 20
Relation OM Pitch and Depth OM Depth (mm) 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 Test 1 Test 2 Test 3 Test 4 Test 5 Mean depth of OM (mm) 0.5 0.4 0.3 0.2 Test 1 (( Vc= 1.75 m/min) Test 2 (( Vc= 1.42 m/min) Test 3 (( Vc= 1.22 m/min) Test 4 (( Vc= 1.47 m/min) Test 5 (( Vc= 1.47 m/min) 0.1 0.0-8 -6-4 -2 0 2 4 6 8 10 OM pitch difference (mm) 0.1 0.5 1.0 1.5 2.0 2.5 3.0 Standard deviation of OM pitch difference (mm) OM pitch difference = ( Measured OM Pitch ) ( Calculated OM Pitch ) * Calculated OM Pitch = casting speed / Oscillating frequency University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 21
Definition of Hook Characteristics Hook angle Effective Hook Depth (EHD) Hook length Hook Shell thickness Oscillation mark depth University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 22
Measured Hook Characteristics Effective hook depth (mm) Hook Length (mm) Hook Angle (degree) Hook Shell thickness (mm) OM Depth (mm) Test 1 1.11 1.92 27.0 0.462 0.282 Test 2 1.22 1.76 30.9 0.520 0.300 Test 3 1.29 1.60 34.6 0.723 0.268 Test 4 1.21 1.83 26.1 0.557 0.347 Test 5 1.06 1.65 28.0 0.463 0.236 Effective hook depth (mm) 3.5 3.0 2.5 2.0 1.5 1.0 Average EHD of of Test 1 Average EHD of of Test 2 Average EHD of of Test 3 Average EHD of of Test 4 Average EHD of of Test 5 0.5 0.0 1.0 1.2 1.4 1.6 1.8 2.0 Casting speed ( m/min ) Hook length (mm) 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 Average Hook Length of of Test 1 Average Hook Length of of Test 2 Average Hook Length of of Test 3 Average Hook Length of of Test 4 Average Hook Length of of Test 5 1.0 0.5 0.0 1.0 1.2 1.4 1.6 1.8 2.0 Casting speed ( m/min ) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 23
CON 1D Predictions at Meniscus Region Shell Temperature ( o C ) 1550 1500 1450 1400 Test 1 ( Vc = 1.75 m/min ) Test 2 ( Vc = 1.42 m/min ) Test 3 ( Vc = 1.22 m/min ) Test 4 ( Vc = 1.47 m/min ) Test 5 ( Vc = 1.47 m/min ) Shell Thickness (mm) 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 Test 1 (( Vc Vc = 1.75 m/min )) Test 2 (( Vc Vc = 1.42 m/min )) Test 3 (( Vc Vc = 1.22 m/min )) Test 4 (( Vc Vc = 1.47 m/min )) Test 5 (( Vc Vc = 1.47 m/min )) 0.2 1350 0 5 10 15 Distance below meniscus (mm) 0.0 0 2 4 6 8 10 12 14 Distance below meniscus (mm) - Ignoring local effect of oscillation mark University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 24
Comparison hook shell thickness with modeling prediction 3 CON1D: Solidus and liquidus locations in mold Starting point of solidification 1 mm Location (mm) 2.5 2 1.5 1 Starting point of solidification test1 Solidus test1 Liquidus 1 mm 0.5 1 mm 1 mm 0 0 5 10 15 20 Distance below meniscus (mm) Time (sec) 0.0 0.1 0.2 0.3 1.0 CON 1D Predition Measured data 0.8 1 mm Shell thickness (mm) 0.6 0.4 0.2 1 mm < Test 1 > 0.0 0 2 4 6 8 10 Distance below meniscus (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 25
Typical Hook Shape Matching to Model Prediction Time (sec) 0.0 0.1 0.2 0.3 0.4 1.2 CON 1D Prediction Measured data T p? 0.5 mm Melting? Hook thickness (mm) 1.0 0.8 0.6 0.4 0.2 Melted? T p = 0.312 sec 0.0 0 2 4 6 8 10 Distance below meniscus (mm) 1 mm < Test 3 > University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 26
Comparison of Shell Thickness Measured shell Thickness just above OM (mm) 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Average Test 1 Test 2 Test 3 Test 4 Test 5 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 Predicted shell thickness at T p (mm) - Scatter due to local variations in undisturbed hook solidification time or level fluctuation University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 27
Relation between EHD and shell thickness 2.5 Effective Hook Depth (mm) 2.0 1.5 1.0 0.5 Test 1 ( Vc = 1.75 m/min ) Test 2 ( Vc = 1.42 m/min ) Test 3 ( Vc = 1.22 m/min ) Test 4 ( Vc = 1.47 m/min ) Test 5 ( Vc = 1.47 m/min ) 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Measured Hook Shell Thickness (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 28
Relation between Hook and Superheat 6.5 Length of solid shell left 6.0 Test 1 Test 2 Test 3 Test 4 Test 5 Superheat = Tundish Temperature Liquidus Temperature Length of solid shell during T p (mm) 6.57 6.52 6.32 6.50 6.51 Length of left solid shell (mm) 1.92 1.76 1.60 1.83 1.65 Remelted length of solid shell (mm) 4.65 4.76 4.72 4.67 4.85 Average 3.5 Test 1 Test 2 Test 3 3.0 Test 4 Test 5 2.5 25 30 35 40 45 Tundish temperature ( ) 1560 Liquidus temperature ( ) - CON 1D 1533.9 1533.9 1533.9 1533.9 1533.9 University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 29 Melted length of solid shell (mm) 5.5 5.0 4.5 4.0 1564 1564 1571 1567 Tundish Temperature - Liquidus Temperature ( o C) (Superheat) Occurrence hook rate (%) 97.8 91.8 93.5 67.7 100
Occurrence rate of Hook with Superheat 100 90 Hook rate Segregation rate 100 90 Occurence rate of hook (%) 80 70 60 50 40 30 20 10 80 70 60 50 40 30 20 10 Occurence rate of segregation (%) 0 0 24 28 32 36 40 (T Tundish - T Liquidus ) ( o C) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 30
Relation between Melted Shell and Superheat 7.0 6.5 Melted length of solid shell (mm) 6.0 5.5 5.0 4.5 4.0 3.5 3.0 Test 1 Test 2 Test 3 Test 4 Test 5 2.5 25 30 35 40 45 Tundish Temperature - Liquidus Temperature ( o C) (Superheat) - Assumption : no hook occurrence case and segregation hook type are all melted - EMS off : little effect to melt shell - EMS on : longer shell melted with increasing superheat University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 31
Conclusions Modeling analysis of shell formation focusing on initial hook formation was conducted with CON1D and parameters of real plant operation Measurement of OM depth, pitch and hook shape were made from samples of continuous casting slabs for different casting speeds, EMS power and superheat OM pitch is directly related to mold level fluctuations and deviations indicate the stability of level during production Increasing OM pitch deviation is correlated with deeper oscillation mark depth Hook shell thickness can be predicted using CON1D and the hook shell thickness at top of OM correlates with positive strip time Effective hook depths increase in proportion to hook shell thickness Hook length correlates with size and shape : longer hooks are deeper EHD Increasing superheat decreases hook length, likely due to melting. Increasing EMS correlates with more OM pitch deviation, level fluctuation and deeper hooks Increasing casting speed effect is complex because it increases both level fluctuation and superheat. University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Ho-Jung Shin 32