Lab experiments on the innovative rapid thick strip casting process

International Journal of Minerals, Metallurgy and MaterialsVolume 19, Number 5, May 2012, Page 391DOI: 10.1007/s12613-012-0569-3Lab experiments on the innovative rapid thick strip casting processRichard Nagy and Dieter SenkDepatment of Frous Matallurgy IEHK, RWTH Aachen University, Aachen 52056, Germany(ecceived: 10 January 2012; revised: 16 February 2012; accepted: 20 February 2012)Abstract: Rapid thick strip casting (RTSC) by Anton Huleck, Inventmetall' ', is an innovative concept for the production of hot strips with afinal as-cast thickness of about 25 mm before rolling. The innovation of the mechanism consists in a vertical mould performing a caterpillarmotion. This moving mould has an unconventional parallelogram shaped cross section. The conventional rectangular shape is formed in theshaping machine, which is placed straight below the mould. Further elements of the technology are state-of-the art. For the investigation ofthis new casting system theoretical calculations were complemented with practical experiments. The investigation focused mainly on two keyaspects: the characteristics of the mould and the shaping process. For the practical analysis a static mould with three pairs of elements inlaboratory scale was developed and commissioned by the Dept. of Ferrous Metallurgy @ RWTH Aachen University. The shapingexperiments were carried out in model scale with two different materials and in variable boundary conditions. The results of theseexperiments delivered important mechanical as well as thermal informations about the casting system.Keywords: continuous casting; rapid solidification; strip metal; moulds; near-net-shape castingof 20 to 30 mm. The final hot strip thickness after direct1. Introductioninline-rolling depends on the number of sequential rollingIf a company wants to keep its market position, it needsstand and steps.to develop innovative technologies for higher quality orincreased productivity. Against this background near-net-RTSC Rapid Thick Surip Casting Technologyshape casting (NNSC) processes have been developed and| SCHEMATIC LAYOUTFor Near Net Shape Cast Stripsoptimized for the production of hot strip steel. They shouldto produce CoilsLiquid steelcomplement the conventional hot rolling technologies.CatepillarThrough developments in this field and with the directtrack mouldcoupling of a continuous caster with a hot strip mill it is国9望possible to produce flat products more economically. Theprocedure specific high productivity and lower investmentShapingmachine-与costs make such technologies highly profitable... 9The rapid thick strip casting (RTSC) technology repre-Calibrationsents a development in this field. The particular innovationNarnet sripof the system is a caterpillar mould, which encloses anhiekoummunconventional, parallelogram-shaped cross- section. StraightPatented inbelow the moving mould, a shaping machine is placed withhina.4县.the secondary cooling zone. Here, the conventional rectan-上县LoopHot Rlling Millgular shape is formed by inline-shaping with liquid core.Subsequently, the solidification ends in a so-called calibrat-Fig.中国国煤化S tchnoloe and theing machine (Fig.1). The final as-cast thickness is in a rangeas-ca-mnit: mm).Corresponding author: Dieter Senk E-mail: dieter senk@ichk .rwth-aachen.deTYRCNMHG◎University of Science and Technology Beijing and Springer-Verlag Berlin Heidelberg 2012空Springer392Int J. Miner. Metall. Mater, VoL19, No.5, May 2012The RTSC technology has been invented in Austria by30 mm thick (Fig. 2). The entire mould height thereforeAJ. Hulek at begin of the 1990s [1-2] as a process of highamounted to 240 mm. The casting thickness was 70 mm andproductivity slab casting. The first practical and theoreticalso the facility capacity accounted for 45 kg of steel melt.studies were carried out at the University of Leoben/AustriaThe angle of the plates amounted to 25° in the bend region.in 2000 [3-4]. The technology was presented intemationallyHence the original mould shape proportion was reproducedat the Korean-German Symposium in 2005 [5]. Subse-realisticly. In the back of the elements cooling fins werequently, an extensive theoretical investigation occurred atmilled with a depth of 18 mm. This was decisive from thethe RWTH Aachen University in 2008 [6~7]. Finally, apoint of cooling efficiency. The cooling process was achie-static test facility in model scale with three pairs of mouldved by fll-cone spraying nozzles produced by Lechlermodules was developed and commissioned in 2010.GmbH.2. Simulation experiments on the 'verticalThe two mould halves were fixed on two separated steelcaterpillar mould'structures. They assured the stability of the whole construc-tion. The appropriate contact pressure of the two mould2.1. Test facilitieshalves was provided by horizontal threaded rods. The deadThe static test facilities consisted of two halves, each withweight of the upper elements at real operation was simulatedthree mould modules. These modules were made by coldadditionally with vertical adjusting screws. The force wasforming of a copper-silver alloy (CuAg0.1P) by Saar-Me-adjusted with a dynamometric tool and it was important fortallwerke GmbH. They were 500 mm long, 80 mm high and contolling the mould tightness during the casting process.(aStrain gaugesS0,Angular sideSolder postsLong sideCooling finsFig. 2. A mould module and preparation for the strain measurement: (旧) dimensions and shape of a mould module; (b) straingauges on the back side of the modules.After the mould has been prepared, the system was set onrate was up to 10Hz.a refractory basement with a centric bottom tapping,In order to follow the deformation strain gauges were putblocked by a ceramic stopper rod. Below the tap hole a boxon the mould at the representative spots. A foil rosette con-with sand was placed in order to pick. up the excess melt. Onsists of three strain gauges, which are arranged exactly atthe box a rocker system was arranged as a help tool for459 to each other; with one rosette a two dimensional de-releasing the stopper rod.formation state could be evaluated. To pursue the distortions2.2. Instrumentation of the mouldof a certain spot in the space, two rosettes are necessary,To get the most possible information about the tempe-which are mounted at the right angle surfaces. Fig. 2 showsrature and deformation behaviour of the modules the mouldsuch a prepared mould module with bonded foil strainwas exploited. In doing that the symmetry of the mould wasgauges at the outside of the mould and in a cooling fin.used to reduce the number of measurement points. For theFinally, as the last link in the chain“Spider" amplifierstemperature profile NiCrNi thermocouples were positionedfrom the company HBM transferred the data of both the ther-in the middle and the lower modules in two different depths,mocouples and the strain gauge rosettes to a computer system.5 and 25 mm below the hot surface. They were placed at the中国煤化工characteristic points of the mould and additionally at the2.positions of the strain gauges. The temperature conditions in.FYHCNMHGxperiments were carriedthe melt were measured at four points as well. The sampling out. Four times a steel grade C50 and twice a peritectic steelR. Nagy et al, Lab experiments on the innovative rapid thick strip casting process393grade were prepared. In order to get the most information1250°C before casting. The mould was heated up by a gasthe boundary conditions like cooling efficiency or superheatburmer from below, through the bottom tap hole.were varied. The test matrix is shown in Table 1.Table 1. Test matrixThe experiments were carried out according to theParameterTrial 1 Trial2 Trial3 Trial4 Trial 5 Trial 6following scheme.Steel gradeC50C50 C50 C50 C10 C10(1) Preparation of the equipment. The mould had beenTig/°C1497 1497 1497 1497 1529 1529assembled, lowered to the refractory bottom and at least theTau/°C16971713 1713 1717 1745 1743thermocouples of the melt were installed into their holders.OT (superheat)/K200216 216 220 216 214The clamping screws for the vertical forces were set and theCooling intensity 14.01.2 1.5 2.5 2.3 2.3stopper system was installed. Finally, the entire facility was(m~s5")covered by refractory material. Fig. 3 shows the completeContact time in the15 20 18- 22mould/sfacility before casting.Zirconium sand (at(2) Preheating of the aggregates. Before the start of thethe stopper tip)experiment all necessary units had been brought up to theFacing (in the rangeright temperature. The centre runner, the stopper and ladleofthe stopper)were preheated. They reached a temperature of aboutNote: “” means no data;“x"" means applies here.a)(b)Mould! MouldI CoolingSpray nozzles jSand boxFig. 3. Basic construction with the mould and the complete faility: (a) basic construction with three pairs of mould modules; (b)facility before casting without refractory cover material.(3) Melting and aloying. In an open 100 kg capacityhad reached the defined residence time in the mould, theinduction furmace 88 kg of iron were melted and the steelstopper was released and the residual melt was tapped intocomposition prepared. The peritectic steel melting operationthe box. The mould cooling was switched off at less thanwas carried out under inert Ar atmosphere. After reaching300°C in the strand shell.the tapping temperature, the steel was aloyed. The targetanalysis of the two steel grades is listed in Table 2.Table 2. Compositions of the two steel gradeswt%Steel grade CMn .SPA0.5 0.80 0.300 0.020 0.0200.060I0.1 0.45 0.015 0.007 0.015 0.044(4) Casting, cooling and tapping. In the first step the meltwas tapped from the furmace into a ladle. The temperaturedrop in the ladle including the tapping process wascalculated in average with 35 K/min. Subsequently, the melt中国煤化工was cast at a predefined velocity through a center runnerfrom the ladle (Fig. 4). The mould cooling was startedYHCNMHGsimultaneously or just right before pouring. Afler the meltFig. 4. Casting process.394Int J. Miner. Metall. Mater, Vol.19, No.5, May 20122.4. ResultsWith the help of the temperature measurement in twoFig. 5 shows the solidified strand shell in the mould fordifferent depths at the same spot important physical valuesexample in trial #3. In Fig. 5(b) the measured shell thicknesscould be calculated: surface temperatures, thermal gradientsand heat fluxes. Additionally, in knowledge of the boundaryat the top can be seen. Fig. 6 presents the background of theconditions, like contact time in the mould and with the helpcalculations. Figs. 7 and 8 show exemplary two results ofof the strand shell's thickness development, it was alsotemperature and deformation measurements in trial #4. Thepossible to estimate the heat transfer cofficients (Eq. (1))time axis is scaled for 60 s.and average solidification cofficients (Eq. (2)):a)(bMouldLong side17.63.4，3.03.4 Angular sideThermocouples10.5- Shell3.3Fig.5. Strand shell at the end of the experiment and after the verification (example trial #3). Shell deformation after cooling bystrong contraction of the Cu-modules: (a) strand shell after tapping in the mould; (b) verification of the strand shell at the top (unit:mm).(a(b)料r 30mmYCold sideHot side| Hot sideStrand shell LAir gapFig. 6. Sketch for the thermal measurements and calculations: (a) calculation of the surface temperature; (b) model to the heattransfer cofficients (Xr, x, dca, dca2: distances of the thermocouples from the cooled surface; a: heat transfer eoffcient.350Midway shew side, mm70006000300Midway long side, 5 mm5000-. Direcion of breadth,。250，25mm4000Direction of Uhickness. roroothickes200300045 downwardshes150l 18另1000e 10050-100036.6.6.-2000058059060061062063057090600610Time /sFig, 8. Measured strains at the midway of the long side,Fig. 7. Measured temperature profile in the middle module inmiddle module in trial #4 (C50; V=3.0 L/min; =18 s).trial #4 (C50; V= -3.0 L/min; =18 s). .ated solidus temperature [8], T(1) the measured tem-K()=7Tsol-Ti(Q)_ s()_ decol(1)pera中国煤化_-flux， s(1) the thicknessof tconductivity of steel,9()入λco入cYHCNMH G copper, and dl thewhere x<(t) is the heat transfer cofficient Tsol the calcu-distance of the thermocouple 1 from the hot surface.R. Nagy et al, Lab experiments on the innovative rapid thick strip casting process395s=Kq05(2)(3) Metal sheet experiments with the parallelogramwhere K is the solidification coefficient (mm/min°), s thecross- section. This analysis can be further divided into coldmeasured shell thickness after the experiment, and 1 theand hot tests. In the following section these experiments aredescribed in more detail.residence time of the melt in the mould (contact time).4. Metal sheet model experiments3. Simulation of the part 'shaping machine'To examine the effect of a strain hardening materialThe shaping machine with several rolls and with thebehaviour during the shaping process experiments withsecondary cooling zone is placed directly below the movingmaterial grade S255 were carried out. Thereby, it had to bemould. The initial parallelogram strand cross-section iscleared how far the concentrated forming effect of thereduced for a conventional rectangular shape with abending points extends into the adjoining cross-section areas.thickness of about 25 mm at this place. The step isIf the deformations appear even through strong local plasticquality-related in the casting process, because the criticaleffects, the local strains could arise strongly. On the otherstrain rates may not be exceeded. Otherwise hot crackshand, if the loads can be distributed on a wider area by strainwould occur at the solidification front. For this reasonhardening, the local strains stay small. Hence, the criticalparticular attention was paid to the behaviour of the cornerstrain rates probably will not be exceeded.and bend region of the cross -section.For these analyses hot and cold tests were carried out.For the investigation several experiments were arranged.The models after the shaping process were investigatedThe tests were divided into the following groups:either with a photogrammetric measurement system [11] or :(1) Experiments with plasticine modelling material. Thewith microstructure analysis.results of this test series have already been published in the4.1. The modelproceedings of the Aachener Stahlkolloquium Conference in2009 [9].As the simulation ofa 8 to 10 mm thick strand shell at theend of the mould flled with liquid or mushy steel metal(2) Test series for the investigation of the shaping processsheets with a thickness of 1.5 mm were bent for the paralle-with liquid core. In these experiments a mould in a modellogram cross-section shape. The model scale amounted toscale was used, at which the long side of the mould wasm= -5.0. Fig. 9 shows the cross-section dimensions. In themovable. Therefore, with a help of a hydraulic cylinder softexperiments two different model lengths were used, indeedreduction and liquid core reduction could be simulated [10].1=500 mm and 1-80 mm. With these two different modelIn this work two different steel grades were investigated.types effects from the length could be determined.(a)1500._100Fig. 9. Metal sheet model: (a) dimensions of the model (unit: mm); (b) model without sand flling.In order to control the resistance to deformation the core 4.2. The rolling millof the cross-sections was fllel in three different ways:The test models were rolled on a model mill at themoulding and tamped sand with different grain sizes andInstitute of Metal Forming (IBF) @ RWTH Aachen Univer-humidity, and modeling plasticine.sity. The mill operates with a pinion gear and can be rebuiltThe surfaces of the models were marked with a gridfrom the duo (hot rlling) to a quarto (cold rolling) machine.patterm squarely using an electrochemical process for theThe roll adjustment is done with the help of an electricdeformation measurement with the Vialux system [9]. Themotd中国煤化工10 kW Leonard drivegrid constant was 1 mm. The bending areas of the CrosS-openYHCNMHGmaximum rlling forcesections were prepared from the inside of the sheets as well.n rllingsg speed is 1200398Int J. Miner. Metall, Mater, VoL.I9, No.5, May 2012local concentration of the strains, the critical strain ratesdes Warmehaushals einer Kupferplatte in Kontakt mit einercould be exeeded theorcically. Because of that the shapingStahlschmelze, Institut ftr Eisenbottenkunde an deprocess should be adjusted to tbe grade of steel in any case.Montanuniversitat Leoben, 2000.4] O. Harrer, Temperaturrechmuang fir die Kofillenplatten zumThe mould radi have to be optimized, too.RTSC Verfahren, Intiut fur Verformungskunde undThe static lab experiments with the variation of severalHuttenmaschinen an der Montanuniversitat Lcoben, 2001.parameters offered important facts influencing the proccss[5] A.J. Hulek, Rapid thick strip casting: a new concept for stipand a plant design. For this reason, the next logical stepcasting, [in] Proceedings of the Koream-German New Steelshould be to design and test a dypamic RTSC-system pilotTechnology Symposium, Disseldor, 2005.plant with moving vertical mould and shaping machine at a[6] D. Senk and R. Nagy, Theoretische Untersuchung des RTSCcapacity of several tons of liquid steel.Verfahrens, Dept. of Ferous Metallurgy, RWTH AachenUniversity, 2007.Acknowledgment[7] R. Nagy, A.J. Hulek, and D. Senk, Rapid thick stuip casting:continuous casting with moving moulds, [i问] Proceedings ofThis work would not have been possible without thethe 6th European Conference on Contimious Casting, Ric-Saar-Metallwerke GmbH[, Lechler GmbH, Dept. of Metalcionc, 2008, p.21.Forming @ RWTH Aachen University and the company[8] M. Suzuki, R. Yamaguchi, K. Murakami, and M. Nakada,Inventmetall with Mr. Anton J. Hulek. They supported thisInclusion particle growth during solidifcation of stainlesswork gratefully.steel, ISU Int, 41(2001), No.3, p.247.[9] R. Nagy and D. Senk, RTSC (Rapid Thick Strip Casting), einReferencesVerfahren mit mitlaufender Kokille, [in] Proceedings of the24th ASK Aachener Sahlolloquium, Aachen, 2009, p.245.[1] A. Hulek, Method for the Vertical Contimous Casting ofa[10] H. Cremers, Eingfuss der Sofl Reduction anf die Mitensei-Steel Snrip, United States Patent, Patent No.US 6945311 B2,genung [Diploma Thesis], RWTH Aachen University, Aachen,2009.[2] A. Hulek, Plant for the Contimuous Casting of Steel, UnitedIEHK, 2000States Patent, Patent No.US 4953615, 1990.[1] Homepage von VIALUX Messtechnik + Bildverarbeitung[3] C. Bemhard, Bericht iber Versuchsserie zur UntersuchungGmbH, htp:/ww.vialux .de/HTML/autogr .htm (2011)中国煤化工MHCNMHG