Water Modeling of Optimizing Tundish Flow Field

Available online at www.sciencedirect.comScienceDirectJOURNAL OF IRON AND STEEL RBSEARCH, INTERNATIONAL. 2007, 14(3): 13-19Water Modeling of Optimizing Tundish Flow FieldLIU Jin-gang'，YAN Hui-cheng' ，LIU Liu2，W ANG Xin-hua'(1. Shougang Research Institute of Technology, Beiing 100041, China; 2. Department of Metallurgical Technology,Central Iron and Steel Research Institute, Beijing 100081, China; 3. Metallurgical and Ecological Engineering School,University of Science and Technology Beiing, Beijing 100083, China)Abstract: In the water modeling experiments, three cases were considered, ie，a bare tundish, a tundish equipped witha turbulence inhibitor, and a rectangular tundish equipped with weirs (dams) and a turbulence inhibitor. Compa-ring the RTD curves, inclusion separation, and the result of the streamline experiment, it can be found that thetundish equipped with weirs (dams) and a turbulence inhibitor has a great effect on the flow field and the inclusionseparation when compared with the sole use or no use of the turbulent inhibitor or weirs (dams). In addition, theenlargement of the distance between the weir and dam will result in a better effect when the tundish equipped withweirs (dam) and a turbulence inhibitor was used.Key words: tundish; flow field; turbulence inhibitor; weir; dam; water modeling experimentSymbol ListCpak一Dimensionless peak concentration;,一 Theoretical residence time, L,=V/Q, s;Fr- - -Froude number;-Flow velocity, (m. s~1);g- - Acceleration of gravity, (m. s~i);V一-Tundish volume, m* ;-Character dimension, m;-Dead volume fraction, V。=1-0.v;Q一-Volumetric flow ratio;Mixing volume fraction, Vn=1- -V,- -Vs;Re- -Reynolds number;V,- Plug volume fraction, V, =the;R:/d- Ratio of plug volume fraction to dead volumeVm- Active volume fraction, Vmm =V, +Vm;fraction, Ru-号;-Model ratio parameter;Rpm/d- - -Ratio of active volume to dead volume fraction,一Ratio of concentration attenuation to thickening,Rmwu=-;4.- - Average residence time,.= CO ,0一-Non-dead volume fraction, 0.=二;2C(I)m一Theoretical residence time twice, S;Ppak- - -Ratio of troak to,;t.-- -Time needed for 90% concentration attenuation, s;0min一Ratio of tain tol,;Time to obtain the maximum concentration, s;Subscripttmim- Minimum breakthrough time, s;p一Prototype parameter;-Time to obtain peak concentration, s;m- -Model parameter.The flow field in the tundish has a great influ- hancement of the metallurgical function of theence on inclusion floatation, molten steel tempera-tund中国煤化工the tundish turbu-ture, and component homogenization. Many metal- lencefor the function oflurgical researchers[1-4] at home and abroad haveHHC N M HSisho, butt at presbuilt a lot of tundish flow field models for the en-ent, there is no definite conclusion as to whether itBlography;LIU Jin-gang(1976-), Male, Doctor;E-mail: liujg9916@163. com; Rerised Date: June 10, 2005.●14●Journal of Iron and Steel Research, InternationalVol. 14has an effect on high quality steel. In addition, inTable 2 Physical parameter in prototype and model tundishEurope, some researchers think that the weir andFirst setSecond setParametersof valuesdam are unnecessary. In the research, a reasonableActual casting speed/(m●min 1 )1.21.4combination setting in tundish was given by a fur-Actual tundish flowrate/(m3 ●h-1)1:15ther study on the function of turbulence inhibitorSimulated flowrate/(m2 .h-1)1.31.and weir (and dam).Nozzle Fr number5.6X10-7 7. 5X10-91 Experimental Principle1.1 Similarity principle1.2 Experimental theory of residence time distribu-In water model experiment, dynamic similaritytion (RTD) curvesand geometrical similarity between the model andExcept the ideal plug flow and mixing flow,the prototype were required. For the dynamic simi-there are shortcut flow and dead volumef- -8] whenlarity，Re number and Fr number in the modelthe fluid flows through a container, and they allshould be equivalent to those in the prototype re-have unfavorable influence on the mass and heatspectively. Since the flow of molten steel within thetransfer of the fluid. The curve of no shortcut flowtundish is turbulent, Re number can certainly meetand dead zone is an ideal RTD curve, and simultane-the requirement. Thus, make sure that Fr numberously, a minimum concentration and a minimum ra-in the model is equal to that in the prototype, thattio of concentration attenuation time to thickeningis,time are obtained.UThe characteristics of the ideal tundish flowFr,=:= Frm(1)control structure include the appropriate prolonga-The geometrical similarity ratio of the model totion of tmia andt, the increase of V,, Ro/a, Vpm, andthe prototype is 1 : 2.5, which is the model ratioRpm/d, and the appropriate decrease of Cpek and t.parameter.2 Experimental Equipment and Scheme(2)Lp 2.52. 1 Experimental scheme designThen volumetric flow ratio of the model to theFour types of tundish are designed, and theyprototype isare the bare tundish, the tundish only with a turbu-Qm=λ/2Q, .lence inhibitor, the tundish only with dams, and theSince λ is known, the experimental water volu-tundish with both a turbulence inhibitor and weirsmetric flowrate can be obtained from the actual cast-(dams).ing speed.After a comparison of the usage effect of someAccording to the similarity theory, the geomet-turbulence inhibitors, three kinds of turbulence in-rical parameters in prototype tundish and modelhibitors with the same dimensions were researched,tundish are shown in Table 1, and the physical pa-and they are the turbulence inhibitor with a big-toprameters are given in Table 2.and small-bottom cone shape, the turbulence inhibi-In this research， the actual casting speed istor with a small-top and big-bottom cone shape, andtaken as1.2 m. min-'(i.e. the simulated flowratethe turbulence inhibitor with an inner helicalis1.3 m3●h-1).groove. The turbulence inhibitor with an inner heli-cal groove is shown in Fig. 1. In addition, a cylin-Table 1 Geometrical parameter in prototype and model tundishder-shaped turbulence inhibitor of 200 mm inPrototypeRectangularheight, 240 mm in outer diameter, and 5 mm in walltundishmodel tundishthickness with a hole of 20 mm in diameter in theTundish dimension/ mm1 00X1 950400X780Tundish liquid level/ mm1 00000lowe中国煤化Iudied.Tundish inlet inner diameter/ mm528and experiences, theTundish inlet length/ mm1 350675MYH.CNMHGat200mm,andheTundish outlet dimension/ mm832height of the weir (dam) within the tundish is 230 mm.Mold dimension/ mm150X1 200Moreover, the top of weir is above the liquid level, andNo.3Water Modeling of Optimizing Tundish Flow Field●15●the dam is fixed on the bottom of the tundish. TheAfter the preliminary experimental condition isstructures and the combination modes of the tundishstable, a certain quantity of saturated NaCl solutionare shown in Fig. 2 and Table 3, and the experimen-was quickly poured into the inpouring ware, and thetal scheme is given in Table 4.conductivity recorder simultaneously was turned onto record the proceeding curve of the changing con-2.2 Experimental methodsductance ratio of the nozzle. The concentration in(1) Steamline and RTDthe RTD curve is the dimensionless concentration ofUnder the irradiation of the slice lamp-house,the trace medium.the varied flow fields during the experiment were re-(2) Inclusion separationcorded with a vidicon, while the instantaneous flowPolystyrene particles of 1 mm in diameter are usedfield was shot with the digital camera.Tundish24.2Helical groovedepth: 10 mmφ160, InpouringIngot area- Ridgearea-师---十哥| BarousdSubmerged nozleinbibitor3中100390Unit mmFig. 1 Schematic of turbulence inhibitor withinner helical grooveFig.2 Top view of structure of tundishTable 3 Experimental conditin of dffent combinations of weir and dammmLocationSpace betweenHeight ofHeight of Bottom of weirCombination modelof weirof damweir and dam bottom of weir.top of dam_ below top of damDamWeir and dam45622023010Bottom of weir above top of dam4506(240-10Widening space between weir and dam3609Narrowing space between weir and dam420Table 4 Tundish experimental schemeNo. Experimental schemeExperimental conditionNotesNo flowcontoling2 Turbulence inhibitor Big-top and smallbottom cone shapeWith normal depthRise up 110 mmSmall-top and big bottom cone shapeCylinder shape7 Weir and damTurbulence inhibitor with an inner helical8 CombinationDam and turbulence inhibitor中国煤花icorie 10 is efieWeir, dam, and turbulence inhibitorBottom{ weir above top of dam, turbulence inhi.MYHCNMHGWidening space between weir and dam, and turbulence inhibitor12Narrowing space between weir and dam, and turbulence inhibitor13Weir and dam, and turbulence inhibitor raised 110 mm●16.Journal of Iron and Steel Research, InternationalVol.14as inclusions. When the condition reaches the exper- the number of inclusion that entered the mould. Theimental requirement, the mixture of 30 mL water more the inclusion number, the lower is the purifica-and 500 polystyrene particles are added to the in-tion of molten steel in the mould, The inclusion ratiopouring ware once for all and record the time imme- is the ratio of the inclusion number to the total num-diately. Based on the tundish RTD experiment andber of the input polystyrene particles.the experience, the molten steel will be completely3 Experimental Results and Discussionuniform in 10 min, and hence, the experimental timeisset at 10 min. After 10 min, the number of poly-3.1 Bare tundishstyrene particles captured is counted in the filter (theThe flow field distribution and RTD curve offilter is located under the submerge nozzle). The the bare tundish are shown in Fig. 3.number of the captured polystyrene particles will beAs shown in Fig. 3 (a), without flow controlling,100「(b)50 t20 t200400Time/sFig. 3 Flow field (a) and RTD curve (b) for experimental scheme No. 1the ladle shroud is disturbed and is in fluctuation.100Disturbance of liquid level is intense. Fig.3 (b)A- No.2B-No.3shows that minimum tmin and te are 12 s and 340 s,C-No.4D- No.5respectively; Lpak and Cpeak are 50 s and 62 on aver-E- No.6age. The results indicate that the shortcut flow andthe dead volume in the tundish lead to the absence ofits function in temperature and composition, and20failure of inclusion floatation within a limited periodof time.06003.2 Tundish with turbulence inhibitorFig.5 RTD curves of experimental scheme No.2 to No.6Fig.4 and Fig. 5 shows that the involving tur-bulence inhibitor can improve the flow field of thehibitor enhances the trend of disturbance of the fluidinpouring area, but the high cylinder turbulence in-surface.Fig.5 shows that the turbulence inhibitormakes the maximum of tmim equal 22 s and Cpeak equal93.5, while other values of Cpeak are above 70 and itsattenuation becomes slower in a short time. Thebetter values of tpak and t。are 106 and 730, respec-tive中国煤化工”, but they are muchlowYHa result, using tur-bulCNMH G shortcut flow andbigger dead volume in the tundish, and the tundishloses its metallurgical character in temperature,composition and inclusion separation, but the turbu-万数据Tlow fed ot eperimntt scheme Na3lence inhibitor with an inner helical groove can beNo. 3Water Modeling of Optimizing Tundish Flow Fieldused to control the flow field in the inpouring area.100 s, and 680 s, respectively， which are longerthan that in most of other arrangements; Cpeak is 46,3.3 Tundish with weir and dam solelywhich is smaller comparatively. Thus, the shortcutAs shown in Fig. 6 (a), when the weir and damflow and the dead volume are smaller. Such tundishare used solely, the flow field of ingot area is regularhas better effect on the composition and temperatureand stable, whereas disturbance in inpouring areahomogenization of the molten steel; what' s more,becomes less intense.the prolonged concentration attenuating time is aFig.6 (b) shows that tain，, tpak, andt. are 56 s,great help to inclusion floatation.(a)50厂b)30 t0|100300500700Time/sFig6 Flow field (a) and RTD curve (b) for experimental scheme No. 73.4 Combination of turbulence inhibitor and weir (dam)To reduce the disturbance of inpouring area in30厂吊tundish and confine the flow field of inpouring areaA-No.8to a small range, the turbulence inhibitor and theB-No.9weir (dam) are used together.C-No10D-NoLIFig. 7 shows that this approach is able to con-R-NoE-No.12trol the flow field and decreases disturbance of the_F-No.13fluid surface efectively. The disturbance of inpour-ing area surface is limited to a small range and the :200400600fluid surface of ingot area is stable. Of all the re-sults, condition No.11 is the best in which the hori-Fig.8 RTD curves for experimental scheme No.8 to No. 13zontal distance is increased, the velocity of flow be-tween the weir and dam is decreased， inclusionfrom those values when the turbulence inhibitor andfloats easily，flowing is stable, and fluid surface isthe weir (dam) are used together. Cpoak is 46, whichnot open to the air.is at average level in combination. So the turbulenceFig. 7 and Fig. 8 shows that lam, tpak, and t。areinhibitor and the weir (dam) used together has a16 s, 42 s, and 543 s, respectively, which greatly dierbetter effect:For condition No. 11, tmn, toak，, and t。are 66 s,120 s, and 714 s, respectively; Cpeak is 56，which aresimilar to other experiments. In addition, there isno shortcut flow and a smaller dead volume in thetundish.中国煤化工RTD curves, themaxin:AHCNMHG: combination is lessthan; ssuuil more slowly, whiletmin，tmx, and tpek doubled, compared with the soleuse of the turbulence inhibitor. Thus， the combi-nation also plays a positive role in inclusion floata-芳另数据" feld for experimental scheme No.11tion.●18●Journal of Iron and Steel Research, InternationalVol. 143.5 Discussion on RTD curves and inclusion separationand Rmw/d are 8.40, 0. 40, and 0. 66，respectively;Data of RTD curves and inclusion separation ofthe inclusion ratio is 2.5%. The effects are im-the above experiments are shown in Table 5. Data ofproved by 367%，286%，178%， and 461% in con-RTD curves and inclusion separation in tundishestrast to the sole use of the turbulence inhibitor.with various structures are shown in Table 6,Compared with the sole use of the weir and dam, τTable 5 and Table 6 shows that: (1) Withoutis improved considerably, the Rp/d is decreased andflow controlling, the minimum values of parametersthe inclusion ratio is increased, Thus, applying theτ, Rp/a, and Rpm/d are 7. 63, 0. 08, and 0. 27, respec-turbulence inhibitor in the tundish with the weir andtively; the inclusion ratio is 6%, which is the lar-dam has a favorable impact on the flow field as agest among these results. Thus, the bare tundish iswhole. (5) When the weir (and dam) with a wid-not a proper structure. (2) When turbulence inhibi-ened space between the weir and dam together withtor is solely used, parameters r, Rp/d, and Rpm/d arethe turbulence inhibitor is used, the inclusion ratio7.63, 0. 15, and 0. 37, respectively, and the inclu-is the smallest, and other parameters, such as tpnksion ratio is 5.3，compared with the non-flow-con-and Cxak， are in good condition. Moreover , the fluidtrolling, and R/da and Repa/d are increased. The resultsurface is stable, the flowing velocity of water is de-indicates that the turbulence inhibitor can improvecreased, and the floatation time of inclusion is pro-the flow field, but since the increases are not great,longed, which benefits the purification of the moltenthe improvement is not satisfactory. (3) As the damsteel. Thus, it is the best tundish structure.is used solely, parameters t, Roiad, and Rm/d are4 Conclusions13.2, 0. 42, and 0. 84, respectively, and the inclu-sion ratio is 1. 3%. In comparison with the sole use(1) The flow field is not satisfactory for theof the turbulence inhibitor, the effects are improvedbare tundish or the sole use of turbulence inhibitor.by 233%, 300%， 227%, and 408%, and Rps and(2) The turbulence inhibitor with an inner hel-Rpa/d are perfected enormously. Thus, using theical groove can improve the flow field in the inpour-weir and dam solely can improve the flow field in the ing area, whereas the sole use of the dam cannot im-tundish, (4) When the turbulence inhibitor and theprove the whole flow field. Thus, the turbulence in-weir and dam are used together, parameters t, Rp/d,hibitor with an inner helical groove and the weir (andTable 5 Experimental resultsInclusionConditions trin tmse lomk tu te Crosk14。OnwV。Rp/d Rpm/dparticle number ratio/ %No.l7.63 346 74 0.21 0. 090.110.27306No.2 22 6(50 691 440 45 10.00 346 106 0.31 0.12 0.17 0.4416.2No.320 2828 691 56376 66.88 346 350.1 0.070.08 0. 115.No.4 20 22 31 691 420 93.5 35.36 346 77 0.22 0.07 0.09 0. 2804No.514 2828 69514 72.5 34.71 346 98 0.28 0.060.08 0. 39No. 624106 691 730 93 7.09 3461340.390.18 0.29 0.64255.0No. 756 96 100 691 68046 13.18 346 158 0.46 0.230.42 0.856.31.3No. 854319.27346 103 0.3 0.080.12 0.4322No.9 22 25 131 691 480 35 3.20 346 135 0.39 0.22 0.36 0. 642.No.1068 8C28 691 62045 8.2046 129 0.37 0. 280.45 0. 593.0Na.11 66 72 120 691 616 56 9.19 346 141 0.41 0.27 0.46 0. 696. 671.No.1252 8674 6971444 4.82 346 141 0.41 0.340.58 0. 69No.13 25 88 126 691 720 46 5.88_ 346_ 161 0.47 0. 22.0.41 0. 899.0中国煤化工Table 6 Experimental results of RTD and inclusion sep.MYHCN MH CecturesStructure of tundishteain lpek te CpeskVp KpVd Rpoud Inclusion ratio/%Bare tundish12 5040 627.66.With turbulence inhibitor19 51 533 76 30.8 90 0.26 0.10 0.14 0.375.3With weir and dam56 100 680 46 13.2 158 0.460.23 0.42 0.84Wt wwle inhibitor and weir (dem)43 120 616 45 8.4 135 0.39 0.24 0.40 0.66_2.5No. 3Water Modeling of Optimizing Tundish Flow Field●19●dam) should be used together.on Mixing in Six Strand Bllet Caster Tundish [J]. Ironmakingand Stemaking, 2002, 29(1); 36-46.(3) Flow-controlling has a great impact on in-clusion separation, whereas the combination of the[2] SchadeJ, Smith M P, Palmer s E. Doubling Tundish Volumeat AK Steel's Middletown Works: Structural Criteria, Designbare tundish and turbulence inhibitor has a bad im-Considerations and Operating Results[J]. I and SM, 1996, 23pact on inclusion separation. The weir (and dam) al-(10); 93- 103.so works effectively, and improves inclusion separa-[3] Godiwalla K M, Sinha S K, Sivaramakrishnan C S. StatisticalAnalyses of Residence Time Distribution of Fluid Elements intion by 461% - 408% compared with using the bareContinuous Casting Water Model Tundish With One Side Entrytundish and the tundish with a turbulence inhibitor.[A]. ISS AIME, eds. Steelmaking Coference Proceedings(4) Compared with the sole use of the turbu-[C]. Chicago: ISS AIME, 1994. 703-712.lence inhibitor, Rpm/d and inclusion separation are im-[4] Hiraki S, Kanazawa T, Kumakura s, et al. Influence ofTundish Operation on the Quality of Hot Coils During Highproved by 200%- 300% when the turbulence inhibi-Speed Continuous Casting [J]. I and SM, 1999, 26(3); 47-tor and the weir (and dam) are used together, so the52.effect similar to the sole use of the weir (and dam)5] Mazumdar Dipak, Guthrie Roderick 1 L. The Physical andcan be obtained. Furthermore, a better control ofMathematical Modeling of Continuous Casting Tundish Sys-tems [J]. ISU International, 199, 39(6); 524-547.the flow field in the inpouring area is obtained.[6] Sahai Y. Fluid Dynamics of Continuous Casting TundishTherefore, it is a better tundish structure.Physical Modeling [A]. ISS AIME, eds. Steelmaking Confer-(5) The combination of the weir and dam withence Proceedings [C]. Washington; ISS AIME, 1986. 677-the widened distance between them and the turbu-68[7] John Knoepke, Joel Masterivch. W ater Modeling Inland Steel'slence inhibitor is the best tundish structure in pres-No. 3 Combination Caster Tundish [A]. ISS AIME, eds.ent study.Steelmaking Conference Proceedings [C]. Washington: ISSReferences:AIME, 1986. 77-788.[8] Chiang L K. Water Modeling of IPSCO's Slab Caster Tundish[1] JhaPK, Dash s K, Kumar s. Effect of Outlet Positions,[A]. ISS-AIME, eds. Steelmaking Conference ProceedingsHeight of Advanced Pouring Box, and Shroud Immersion Depth[C]. Toronto: ISS AIME, 1992. 437-450.中国煤化工MYHCNMHG