UIUC, August 12, Clogging Effects on Asymmetric Flow and Vortex Formation. Seong-Mook Cho, Seon-Hyo Kim

Size: px

Start display at page:

Download "UIUC, August 12, Clogging Effects on Asymmetric Flow and Vortex Formation. Seong-Mook Cho, Seon-Hyo Kim"

Delphia Greer
6 years ago
Views:

ANNUAL REPORT UIUC, August, Clogging Effects on Asymmetric Flow and Vortex Formation Seong-Mook Cho, Seon-Hyo Kim

University of Science and Technology Materials Science and Engineering Seong-Mook Cho Research Scope - Nozzle

slab caster Level fluctuation near s: disturbs the formation of the initial solidifying shell Clog Back flow -

model) surface velocity vortex formation frequency level fluctuation - Methodology: water model experiments to

1 ANNUAL REPORT UIUC, August, Clogging Effects on Asymmetric Flow and Vortex Formation Seong-Mook Cho, Seon-Hyo Kim Department of Materials Science and Engineering Pohang University of Science and Technology Rajneesh Chaudhary, Brian G. Thomas Department of Mechanical Science and Engineering University of Illinois at Urbana-Champaign Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho Research Scope - Nozzle clogging influences asymmetric flow causing problems: Vortex formation: can entrap mold flux and cause defects in slab caster Level fluctuation near s: disturbs the formation of the initial solidifying shell Clog Back flow - Objective: studying effect of asymmetric flow caused by nozzle clogging on: mold flow pattern (computational model) surface velocity vortex formation frequency level fluctuation - Methodology: water model experiments to quantify surface Nozzle clogging at flow, vortex formation and level fluctuation recirculation region of port computational model to explain the flow pattern. Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

2 Previous Works Flow pattern Paticle motion Temperature distribution Top surface level profile Lifeng Zhang, Yufeng whang and Xiangjun zuo MTB, 8, Vol. 39B, p53~55 Present research: more detailed study of transient surface flow phenomena with nozzle clogging Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3 Schematic of /3 Scale Water Model Stopper-rod Tundish Weir Dam 3mm Nozzle wall thickness:.5mm Bore diameter of SEN: 5mm Left Submergence depth: 6mm Right Free surface Port angle: 35 deg downward Water flow meter 75mm 5mm mm Ф 5* exit Pump Pump Water bath Water flow meter Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

3 Nozzles to Study Clogging Effects No-clog Small-clog Severe-clog Right Dimension of nozzle port Noclog Smallclog

Ratio of area between left and right 67.67 33.33 Ratio of area between two ports and nozzle bore.5.

Materials Science and Engineering Seong-Mook Cho 5 Processing Conditions for Experimental and Computational Works /3 scale

77 Ton/min) Mold width 5 mm 5 mm Mold thickness 75 mm 5 mm Computational ti mold domain width thickness length 5

67 kg/m-s (steel) μ fluid Distance from closed stopper-rod to opened one (mm) Nozzle (well-bottom type) port angle Nozzle

3 3 Nozzles to Study Clogging Effects No-clog Small-clog Severe-clog Right Dimension of nozzle port Noclog Smallclog Severeclog Width (mm) Height (mm) Left Width (mm) Height (mm) Ratio of area between left and right Ratio of area between two ports and nozzle bore Port angle (degree) Left port Right port No-clog Small-clog Severe-clog All cases 3.3mm 3.3mm 6.7m m 8mm 9.mm.8 3.5mm 5.mm 6.7m mm Clog in left nozzle port - Small-clog (33% clog) - Severe-clog (67% clog) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 5 Processing Conditions for Experimental and Computational Works /3 scale water model Real caster Casting speed.97 m/min.59 m/min Water flow rate 3. LPM 537 LPM (3.77 Ton/min) Mold width 5 mm 5 mm Mold thickness 75 mm 5 mm Computational ti mold domain width thickness length 5 mm(no-clog), 5mm (Clog) 37.5 mm mm SEN depth 6 mm 8 mm ρ fluid 998. kg/m 3 (water) 7 kg/m 3 (steel).3 kg/m-s (water).67 kg/m-s (steel) μ fluid Distance from closed stopper-rod to opened one (mm) Nozzle (well-bottom type) port angle Nozzle ports No-clog: 3. Small-clog: 3., Severe-clog: degree No-clog: Symmetric ports Small-clog:.67_asymmetric left port, Severe-clog:.33_asymmetric left port Nozzle bore diameter (inner/outer) 5 mm/6 mm 75mm/38mm Shell no Yes Gas injection no Yes Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 6

Sensors for Measurements Ultrasonic displacement sensors Electromagnetic current sensor Specifications Electromagnetic current sensor Ultrasonic displacement sensor

Sensors in water model> - Electromagnetic current sensor for measuring surface flow velocity - Ultrasonic displacement sensors for measuring level fluctuations

flow Steady-state, 3-D and incompressible Navier-Stokes equations for momentum conservation Standard k ε turbulence model for simulating time-averaged turbulent fluid

domain, clog: half domain) Hexa-cells used in the computational domain to model turbulent nozzle and mold flow (no-clog:.7 million, small-clog:.

4 Sensors for Measurements Ultrasonic displacement sensors Electromagnetic current sensor Specifications Electromagnetic current sensor Ultrasonic displacement sensor Response time Hz Hz Collecting data Hz Hz frequency Collecting data time sec sec Sensor head dimension mm 3mm 8mm mm Measuring direction X,Y vector components < Sensors in water model> - Electromagnetic current sensor for measuring surface flow velocity - Ultrasonic displacement sensors for measuring level fluctuations Vertical direction to sensor head Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 7 Computational Model - Single-phase flow Steady-state, 3-D and incompressible Navier-Stokes equations for momentum conservation Standard k ε turbulence model for simulating time-averaged turbulent fluid flow Using Fluent - Domain and convergence /3 scale water model geometry Combined computational domain of nozzle and mold assuming symmetrical flow (no-clog: quarter domain, clog: half domain) Hexa-cells used in the computational domain to model turbulent nozzle and mold flow (no-clog:.7 million, small-clog:.5 million, severe-clog:.3 million) Convergence in almost all cases until scaled residuals were reduced to stable Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 8

5 Geometry, Mesh and Boundary Conditions o No-clog Small-clog Severe-clog.35m.75m Velocity inlet:.35m.95m/sec input to cylinder surfaces.3m.5m.5m.m Pressure outlet: Pa Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 9 Asymmetric Flow in the SEN with Nozzle Clogging No-clog Small-clog Severe-clog High momentum flow High momentum flow No back flow, Spread jet No Back flow 5% 8% 5% 5% 75% Flow rate ratio - Nozzle port clogging causes asymmetric flow in a nozzle Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

6 Equations to Calculate Jet Characteristics Weighted average velocity of each component Vertical jet angle (degree) average jet speed Weighted average turbulent kinetic energy Weighted average turbulent kinetic energy dissipation rate Back flow zone fraction Hua Bai and Brian G. Thomas, MTB,, Vol. 3B, No. Fjet (Jet Force) = Ujet (Averaged Jet Speed) Q jet (Flow Rate) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho Jet Characteristics No-clog Small-clog Severe-clog Weighted average nozzle port velocity in x- direction (Outward) (m/sec) Weighted average nozzle port velocity in y- direction (Horizontal) (m/sec) Weighted average nozzle port velocity in z- direction (Downward) (m/sec) Left Right Left Right Left Right Vertical jet angle (degree) Horizontal jet angle (degree) Average jet speed(m/sec) Flow rate (kg/sec) Averaged jet force (N) Maximum velocity magnitude (m/sec) Weighted average turbulent kinetic energy(m /s ) Weighted average turbulent kinetic energy dissipation rate (m /s 3 ) Back-flow zone (%) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

Jet Velocity Profiles Distan nce from nozzle

57 (.86).3 (.57) Small-clog.8 (.76). (.96).

increases - Faster jet at non-clogged port

Severe-clog < m/sec > Spread jet Faster jet

clogging causes asymmetric flow in a mold

clogged nozzle, surface flow from right side

flow from left - In severe-clog case, it is

7 Jet Velocity Profiles Distan nce from nozzle port bottom(m m)..3.. At port outlet plane No-clog_L, R Small-clog_L Small-clog_R Severe-clog_L Severe-clog_R Jet force (Flow rate): N (kg/sec) Left Right Total No-clog.57 (.86).57 (.86).3 (.57) Small-clog.8 (.76). (.96).383 (.57) Severe-clog.9 (.5).36 (.7).37 (.57) Flow rate at clogged port decreases, but flow rate at non-clogged port increases - Faster jet at non-clogged port than clogged port because of increasing flow rate - Increased flow rate and velocity magnitude induce higher jet force Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3 Asymmetric Flow in the Mold with Nozzle Clogging g No-clog Small-clog Severe-clog < m/sec > Spread jet Faster jet high speed of downward flow - Nozzle clogging causes asymmetric flow in a mold (Unbalanced double roll pattern) - With clogged nozzle, surface flow from right side cross the surface and suppress the uprising flow from left - In severe-clog case, it is possible for inclusions to penetrate deeply down into a mold Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

8 Velocity near Narrow Faces No-clog Small-clog Severe-clog istance from surface(m) Di < mm from s > Impingement point of jet flow No-clog_L, R Small-clog_L Small-clog_R Severe-clog_L Severe-clog_R Jet impingement point on s is ~.3m below free surface - Mold flow near s is faster on the same side as non-clogged port - More clogging causes more asymmetry between left and right side - There exists stagnant flow region in the mold with clogged nozzle Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 5 Measuring Surface Velocity mm < Front view > L < Measuring positions of surface flow velocity> L L L L < Top view > - Measure velocity profiles near surface with current meter sensors during sec - Compare velocity profiles on ¼, ½ points from s, gap between SEN and mold inner wall - Calculate average velocity magnitude, flow direction angle Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 6

9 Quantifying Averaged Surface Velocity Vector Left Outside +y axis Flow direction angle θ Right / / Inside / / +x axis Determining the flow direction angle from averaged velocity components Quantifying vector-averaged surface flow with flow direction angle and velocity magnitude Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 7 Averaged Surface Flow Pattern (Experimental Results) Measured at mm from free surface and averaged for sec OR.m/sec No-clog / / IR / / OR Small-clog IR OR Severe-clog.6. IR Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 8

10 Velocity magnitude(m m/sec) V agnitude(m/sec) Velocity ma Speed & Direction Variation with No-Clog Nozzle Left_/ degree) ow direction angle(d Flo Velocity magnitude(m m/sec) V Right_/ Left_/.5 Right_/ on angle(degree) Flow directio gnitude(m/sec) Velocity mag Surface flow becomes slower and more chaotic toward SEN degree) low direction angle(d Fl on angle(degree) Flow directio Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 9 elocity magnitude(m m/sec) Ve agnitude(m/sec) Velocity ma Speed & Direction Variation with Small-Clog Nozzle Left_/ degree) low direction angle( Fl elocity magnitude(m m/sec) Ve Right_/ Right /.5 Left_/.5 _ on angle(degree) Flow directio - Surface flow from right is faster and more consistent than left - Surface flow at the left (slower) side is more chaotic than right side Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho agnitude(m/sec) Velocity ma (degree) low direction angle( F tion angle(degree) Flow direct

11 Velocity magnitude(m m/sec) V agnitude(m/sec) Velocity ma Left_/ Speed & Direction Variation with Severe-Clog Nozzle (degree) Flow direction angle( F Velocity magnitude(m m/sec) V Right_/ Left_/ Right _ / tion angle(degree) Flow direct There is same trend as more clogged nozzle, but more severe - Flow direction is more consistent on faster right side Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho degree) ow direction angle(d Flo ion angle(degree) Flow directi elocity magnitude(m m/sec) Ve de(m/sec) Velocity magnitud Speed & Direction Variation in the Gap between SEN and Mold < No-clog > < Severe-clogged > degree) ow direction angle(d Flo gle(degree) Flow direction ang locity magnitude(m m/sec) Vel < Small-clog > More asymmetrical flow between left and right side makes faster flow in the gap between SEN and mold - Flow direction is more consistent with higher velocity Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho degree) Flow direction angle(d

Velocity Magnitude and Turbulent Kinetic Energy (TKE) m/sec m /sec Velocity magnitude TKE No-clog Velocity magnitude TKE Small-clog Smaller vortex area, Severe asymmetry Vl Velocity magnitude itd TKE

12 Velocity Magnitude and Turbulent Kinetic Energy (TKE) m/sec m /sec Velocity magnitude TKE No-clog Velocity magnitude TKE Small-clog Smaller vortex area, Severe asymmetry Vl Velocity magnitude itd TKE Severe-clog Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3 Surface Velocity Results (Prediction and Measurement) Outside Left Right Velocity magn nitude(m/sec) Predicted Measured / No-clog Small-clog Severe-clog SEN,,, Distance from center of mold(m) / Inside / / mm from surface - Prediction well-matches with measurement - Nozzle clogging induces the asymmetric surface flow in a mold - Surface flow at the same side as non-clogged port is faster than clogged port - More asymmetry of flow between left and right side induces faster flow in the gap between SEN and mold - Surface flow at the ¼ region is faster than ½ region in measurement (slope of flow speed decreasing is not so steep as prediction) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho

13 Comparison of Predicted Averaged Surface Velocity with Measured One Velocity magnitude (m/sec) No-clog Left Gap between SEN and Right / / mold inner wall / / Predicted Measured Predicted Small-clog Measured Predicted Severe-clog Measured Model over-predicts experiments in the gap between SEN and mold inner wall; Standard k ε model can t accommodate flow speed with direction variation (Real turbulent flow decreases the averaged surface flow) - In severe-clog case, model predicts that surface flow from right side crosses the surface and suppresses the uprising flow from left - Computational model predicts the region of the fastest surface flows at ½ L region Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 5 Comparison of Predicted Turbulent Kinetic Energy (TKE) with Measured One Turbulent kinetic energy ( m /sec ) Left Gap between SEN and Right / / mold inner wall / / No-clog Small-clog Severe-clog Predicted Measured Predicted Measured Predicted Measured Measured TKE is greater than predicted one; anisotropy of real turbulence; directional variation of surface flow affects TKE together with speed variation - Vortexing flow pattern causes higher turbulent kinetic energy at left region near SEN in measurement Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 6

Counting Vortex Formation Left Outside SEN 3 Right

sesame seeds (tracer particles) at surface of

Counting vortex formation Divide the four region near

( Criterion of vortex: over rotations) Left region VS

: + VS 3 + Pohang University of Science and Technology

Vortex Formation with Clogged Nozzle < No-clog > <

SEN in the cases of no and small-clog cases.

severe-clog case - Nozzle clogging causes asymmetric

region of SEN with the < Sever-clog > nozzles with

14 Counting Vortex Formation Left Outside SEN 3 Right Inside -Visualization of vortex formation: Scatter sesame seeds (tracer particles) at surface of water-model mold Record videos and take pictures - Counting vortex formation Divide the four region near SEN Count the number of each region in a time interval ( Criterion of vortex: over rotations) Left region VS Right region: + VS + 3 Outside region VS Inside region : + VS 3 + Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 7 Vortex Formation with Clogged Nozzle < No-clog > < Small-clog > - Most vortices are formed at regions near SEN in the cases of no and small-clog cases. - All vortices are formed at left region near SEN with severe-clog case - Nozzle clogging causes asymmetric vortex formation - More vortices form at the left region of SEN with the < Sever-clog > nozzles with clogged left port Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 8

15 Vortex Formation Frequency y(#/min) Vortex frequnecy x frquency(#/min) Vortex 5 Left region 3 Right region No-clog Small-clog Severe-clog Nozzles 5 Inside region 3 Outside region Vortex frequency(#/min) Left Right Total No-clog 3.9 (8%).3 (5%) 8. Small-clog 5.5 (63%) 3. (37%) 8.7 Severe-clog 3.3 (%) 3.33 < Left region VS Right region> Vortex frequency(#/min) No-clog Small-clog Severe-clog Outside. (5%).3 (9%) 6 (5%) Inside 3.8 (6%). (5%) 7.3 (55%) Total No-clog Small-clog Big-clog < Inside region VS Outside region> Nozzles - Vortices are caused by asymmetric flow between right and left region - Vortex frequency increases with clogging due to greater diff. of right/left velocity Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 9 No-clog Small-clog Severe-clog Left side Vortex Formation Frequency (Visualization) at minutes minutes Left side minutes Left side Outside.6.8 SEN.3.5 < #/min > Inside Outside 3.. SEN.3. < #/min > Inside Outside 6 SEN 7.3 < #/min > Inside Right side Right side Right side Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3

Measuring Level Fluctuation 5mm 5mm < Measuring positions of level fluctuation > - Measure level fluctuation on the surface near s with ultrasonic displacement sensors - Compare fluctuation profiles

University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3 Histories of Level Fluctuation Level fluctuation(m mm) L Level fluctuatio on(mm) -5-5 Left side - Right side -

16 Measuring Level Fluctuation 5mm 5mm < Measuring positions of level fluctuation > - Measure level fluctuation on the surface near s with ultrasonic displacement sensors - Compare fluctuation profiles at 5mm position from both s - Calculate average level and standard deviation of level - Transfer level fluctuation profiles to power spectrum by FFT (Fast Fourier Transform) analysis Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3 Histories of Level Fluctuation Level fluctuation(m mm) L Level fluctuatio on(mm) -5-5 Left side - Right side < No-clog > Left side - Right side < Severe-clog > evel fluctuation(m mm) Le < Small-clog > Left side Right side Distance from nozzle port top(mm) Left Right No-clog Small-clog Severe-clog Avg 6 6 Stdev Avg 6 6 Stdev.7.9 Avg Stdev.7.9 Slightly higher averaged level with nozzle clogging g Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3

17 Characteristics of Level Fluctuation vel fluctuation(mm m) Lev Level fluctuatio on(mm) Left side Right side evel fluctuation(m mm) Le Left side Right side < No-clog > < Small-clog > Left side - Right side < Severe-clog > 3 - Smooth level fluctuation with nonclogged nozzle - Abruptly severe level fluctuation with clogged nozzle (more severe with more clogged nozzle) Severe level fluctuation disturbs uniform initial solidifying shell Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 33 Power Spectrum of Level Fluctuation.....5Hz ) er spectrum(mm^ Powe (mm ) Power spectrum( E-3 E- E-5 E-6 E-7 E-8 E-9.. E-3 E- E-5 E-6 E-7 E-8 E-9 No_clog, L < No-clog > E-7 < Small-clog > E-8 No-clog, R E-3.. Severe-clog, L Frequency(Hz) <Se Severe-clog ereclog > Severe-clog, R E-3...5Hz ) Pow wer spectrum(mm E-3 E- E-5 E-6 E-9 Small-clog, L Small-clog, R E-3.. Frequency(Hz) - Level fluctuation with non-clogged nozzle shows more turbulent trend - With clogged nozzle, there is higher power at the right (non-clogged port side) - Asymmetric flow by nozzle clogging creates periodic asymmetric fluctuation (high power at.5 Hz) Frequency(Hz) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 3

18 More frequent vortex formation Level fluctuation with higher power Asymmetric flow Higher jet force Summary - Nozzle clogging effects has been studied with 3 cases nozzles (No, Small and Severe-clog) - Asymmetric flow Asymmetric jet from clogged nozzle cause asymmetric flow in a mold The surface flow at the side with non-clogged port is faster than clogged port More asymmetry of flow between left and right side induces faster flow at the gap between SEN and mold - Vortex formation Vortices are caused by the asymmetric flow between right and left side More vortices form at the left region of SEN with the nozzle having clogged left port Level fluctuation Phenomena in the mold with clogged nozzle - Level fluctuation The surface level at the same side as non-clogged port is higher with higher power - Comparison of results between experimental and computational works Standard d k ε model shows the well matched trend with measurement Real vortexing flow pattern makes the difference of turbulent kinetic energy between measurement and prediction (need of studying transient modeling) Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 35 Acknowledgements Continuous Casting Consortium Members (ABB, Arcelor-Mittal, Baosteel, Corus, LWB Refractories, Nucor Steel, Nippon Steel, Postech, Posco, ANSYS-Fluent) POSCO: Ho-Jung Shin, Woong-Ryul Choi, Sung- Kwhang Kim, Joong-Wook Cho, Shin-Eon Kang POSTECH: Hyoung-Jun Lee, Soo-Young Seo UIUC: Chuanbo Ji Pohang University of Science and Technology Materials Science and Engineering Seong-Mook Cho 36

Thermal Stress Cracking of Sliding Gate Plates

ANNUAL REPORT 2011 UIUC, August 18, 2011 Thermal Stress Cracking of Sliding Gate Plates Hyoung-Jun Lee, Seong-Mook Cho, Seon-Hyo Kim Department of Materials Science and Engineering, Pohang University of