Development of Fluoride-free Mold Powders for Peritectic

1. Introduction

Mold powders play an important role in ensuring the sur-face quality of the products and the process stability of thecontinuous casting process, especially longitudinal crackingand sticker breakout. In addition, the molten mold powder,in form of a top-covering slag may contribute to refining ofthe melt from inclusions. Mold powders usually contain thefollowing components: CaO, SiO2, Na2O and CaF2. Amongthe chemical compositions of mold powder, the roles of flu-orine are to control viscosity, break temperature and crys-tallization fraction developed in a slag film, which are di-rectly related to both lubrication and heat transfer throughthe slag that infiltrates between the mold wall and the solid-ified steel shell.1) However, the fluorine emissions lead toerosion of plant, acidification of the cooling water and are apotential health and safety hazard.2) Thus the replacementof fluorine with a more benign constituents in the moldpowder is a research area of interest.3)

The melting and viscosity characteristics of the F-freemold powder were investigated in laboratory scale, whichindicated that the melting temperature and the viscosity at1 300°C of F-free mold powders with B2O3 added in placeof F in traditional powders can be controlled in the required

range.4) Industrial trials of the fluoride-free mold powderwere carried out at some steel works with the focus on bil-let and bloom casters.5) For slab casters, especially thosecasting peritectic grade steels, there is currently no substi-tute for fluorine in the mold slag. This is due to that, whilemold slag viscosity is the pertinent design parameter for developing mold fluxes for casting billets, for the case slab casting, the heat transfer through the slag also needs to be considered in order to control surface defects. This is because the longitudinal cracking in peritectic carbon(0.09–0.16 mass% C) steels results from the 4% mismatchin the thermal shrinkage coefficients for the d and g phaseswhich results in stresses which can only be relieved bycracking. The stresses can be minimized by keeping theshell as thin and as uniform as possible. This is achieved byreducing the heat transfer by maintaining a thick layer ofslag film with a significant crystalline fraction inside themold/strand gap. Owing to the inevitably need of controlledcuspidine crystallization in the slag to control the mold heatflux, the issue of F-free slab mold powders by the substitu-tion of B2O3 is not acceptable.1) Therefore, the solution tothe outstanding problem of the F-free slab mold powdersrequires the tailoring of a chemistry that simultaneously

Development of Fluoride-free Mold Powders for Peritectic SteelSlab Casting

Guanghua WEN,1,2) Seetharaman SRIDHAR,2) Ping TANG,1) Xin QI1) and Yongqing LIU1)

1) Department of Metallurgy, College of Materials Science and Engineering, Chongqing University, 174 Shazhengjie,Chongqing 400044, P. R. China. E-mail: [email protected]) Department of Materials Science and Engineering, Carnegie Mellon University, 5000 Forbes, Pittsburgh, PA 15213, U.S.A.E-mail: [email protected], [email protected]

(Received on April 2, 2007; accepted on May 21, 2007 )

In this paper, titanium-bearing blast furnace slags (CaO–SiO2–TiO2) produced at Panzhihua Iron and SteelCompany (P. R. China) is used as the base material to develop fluoride-free (F-free) mold powders to im-prove the heat transfer between the mold and the strand. Effects of the binary basicity (CaO/SiO2), TiO2,Na2O, Li2O, MgO, MnO and B2O3 on the melting temperature, viscosity and heat flux of F-free mold pow-ders are investigated. The laboratory results indicate that 1) the melting temperature and the viscosity of theF-free powder decrease, as expected, with increasing the content of Li2O, B2O3 and Na2O respectively, butthe lowest viscosity is achieved with 6.0 mass% TiO2; 2) the heat flux of the F-free slag film with1.0–6.0 mass% TiO2 is close to that of a conventional mold slag film with 2.0–10.0 mass% F; 3) the effect ofbasicity of the F-free powder on the heat flux is the same as the powder bearing fluoride; 4) the heat fluxchanges significantly with more than 8.0 mass% Na2O and about 4.0 mass% MnO, whereas the effects ofLi2O and B2O3 in the F-free powder on heat flux are not significant. The suitable range of main componentsof the F-free powder with TiO2 is proposed for casting peritectic-grade-steel slabs. The industrial trials ofperitectic steel casting, using the proposed F-free flux, reveals a good surface quality of the slab, and well-controlled heat transfer at the continuous casting mold by the F-free powder with the precipitated crys-talline phase being perovskite (CaTiO3) instead of cuspidine in the conventional mold slags that contain fluo-ride.

KEY WORDS: titanium-bearing blast furnace slag; slabs; continuous casting; F-free mold powder; heattransfer; peritectic steel.

ISIJ International, Vol. 47 (2007), No. 8, pp. 1117–1125

1117 © 2007 ISIJ

lowers viscosity and crystallizes in a manner that results ina desirable control of heat flow through the mold/strandgap.

In the current study, the hot metal bearing vanadium andthe blast furnace slag bearing titanium (CaO–SiO2–TiO2)from the Vanadic Titanomagnetite are obtained as a resultof the blast furnace smelting process at Panzhihua Iron andSteel Company (PanSteel) (P. R. China). The content ofTiO2 in the slag is about 23–25 mass%. The current annualamount of the blast furnace slag bearing titanium is over 5Mt at PanSteel. This blast furnace slag bearing titaniumconsists of bearing–titanium minerals with high crystalliza-tion tendency, and can therefore, unlike common blast furnace slags, not be extensively recycled through use incement manufacturing. The titanium in the slag is distrib-uted as different mineral compounds, such as perovskite(CaO·TiO2), titanaugite (CaO·TiO2·Al2O3) and titaniumdiopside (Ca2(Mg3,Ti) · (Al2,Ti)2· (SiO4)2O12). Content ofTiO2 in perovskite crystal is 60 mass% of total TiO2 amountin the slag.6) Recently, crystallization of CaO–SiO2–TiO2

synthetic slag with the basicity (CaO/SiO2) of about 0.8 asa candidate for F-free mold flux was conducted by H.Nakada et al.7) The result indicates that the CaOSiO2TO2

crystallizes rapidly in the slag film, similar to cuspidine incommercial mold fluxes, but the thickness of the crystallinelayer was found to be smaller than that of the crystallinelayer resulting from cuspidine precipitation. The prescenceof Ti in the glassy could result in a change in opacity andtherefore radiative properties of the slag layer but this hasnot been quantified. According to the equilibrium phase di-agram of CaO–SiO2–TiO2 system,8) the main crystallinephase related to TiO2 should be perovskite since the basic-ity of commercial mold powder is approximately 0.8 to 1.4.If the blast furnace slag bearing titanium could be used as abase chemistry for manufacturing mold powders, the cuspi-dine in the fluxes with fluoride could potentially be re-placed by the perovskite in the slag. It may thus be feasibleto control the mold heat flux for casting of different steelgrades. This would require that the F-free powder with TiO2

must exhibit similar (i) melting temperature and (ii) viscos-ity, in addition to (iii) crystallization and resulting thermalconductivity, as those in the F-bearing powder. Therefore,before applying the blast furnace slag bearing titanium tothe F-free mold powder, the research on these thermophysi-cal properties is needed. This paper investigates the changeof viscosity, melting temperature, and heat flux of the abovementioned slags.

2. Experimental

The melting temperature of mold powders was deter-mined using a high temperature microscope. The specimenis heated and monitored for signs of melting. The test con-sists of heating an agglomerated sample pressed into acylinder (3 mm in diameter and 3 mm in height) at a con-trolled rate, and then monitoring the changes in sample di-mension. Shapes corresponding to “softening”, “hemi-sphere” and “fluidity” are specified, their height is originalheight of 75%, 50% and 25%, respectively (see Fig. 1) andthe temperatures at which the samples achieve these shapesare recorded by a computer. The hemisphere temperature isusually defined as the melting temperature of mold fluxes.9)

The high temperature viscosity of liquid mold fluxes wasmeasured with a rotating viscometer. This instrument meas-ures the torque of spindle rotated at fixed speed in a cru-cible filled with the liquid of 250 g. The crucible was heatedfrom room temperature to 1 300°C in the MoSi2 electricfurnace, and then maintained isothermally at 1 300°C for10 min. The viscosity at 1 300°C is determined by the aver-aged value of 20 measurements which were continuouslymeasured. A calibration measurement was carried out atroom temperature by using standard oil of known viscosity.

It is well known that the heat transfer through the slagfilm in the mold primarily depends on two parameters ofthe mold fluxes, namely break temperature as it governs thethickness of the slag film layer and crystallization tendency,but it is very difficult to accurately quantify them.10) Conse-quently, an experimental apparatus for simulating coppermold was designed to directly measure the heat flux of theslag film, which is schematically shown in Fig. 2. A quan-tity of 1 000 g slag was melted in a graphite crucible whichwas heated in an induction furnace. The water cooled de-tector made of copper was immersed liquid slag at 1 400°C,and a solid slag deposition formed on the copper wall. Sub-sequently, the copper detector was lifted up and the at-tached solid slag was removed after an immersion time of120 s. The liquid slag temperature was measured by a py-rometer. The temperatures of inlet and outlet cooling waterwere recorded simultaneously and exported to a computer.The heat flux through the slag film was calculated based onthe temperature difference of water between outlet and inletagainst immersion time. The cooling water flow rate was0.30 m3·h�1.

A correlation of water temperature difference against im-mersion time can be obtained, as shown in Fig. 3. It can beseen in this figure that there are three significant stages ofwater temperature variation. When the detector is immersedin the liquid slag at the 1 400°C, a sudden increase in tem-

ISIJ International, Vol. 47 (2007), No. 8

1118© 2007 ISIJ

Fig. 1. Images of typical stages during the melting process.

perature difference of DT1 at time t1, which mainly repre-sents the heat transfer of liquid slag between the coppermold and liquid slag during initial stage of solid slag depo-sition. Then, as the solid slag film increases in thickness,the temperature difference decreases from DT1 to DT2 dur-ing the time period from t1 to t2. This is attributed to thecombined effects of formation of the solid slag film, recrys-tallization of the glassy slag layer and formation of the gapbetween the copper mold and the solid slag film. Finally,after the immersion time exceeded t2, the solid slag filmslowly grows, causing the temperature to drop further, al-beit at a slow rate. The result of many trials showed that therange of slag film thickness was 1–5 mm, which is close tothe one of the slag films removed from the operating con-tinuous casting mold11) where the immersion time is closeto the experimental time t2, hence the heat flux of solid slagfilm in mold is defined as the heat flux at this moment.

There were 35 specimens of the F-free powders. Theirchemical compositions were chosen considering seven pa-rameters, namely basicity (CaO/SiO2) 0.6–1.2, and contentsof TiO2 1–9 mass%, Na2O 2–10 mass%, Li2O 0.5–2.5 mass%,MgO 3–8 mass%, MnO 2–6 mass% and B2O3 2–10 mass%.Each parameter included five different levels. TiO2 of thespecimens came from the blast furnace slag bearing tita-nium, the other compositions were achieved by adding pureoxides (CaO, SiO2, MgO, MnO and B2O3), Na2CO3 wasadded as a source of Na2O and Li2CO3 as a source of Li2O.

For comparison of the heat flux developed in the slagfilm with the F-free powders, 10 specimens of the powderswith fluoride were prepared, the basicity 0.8–1.6 and2.0–10.0 mass% F were mainly considered, and the constant

amount of 8.0 mass% Na2O, 3.0 mass% Al2O3, 2.0 mass%MgO was used in all the cases to simulate the conditionssimilar to industrial ones.

All the calculated chemical compositions of studied slagsincluding F-free and F containing are listed in Table 1 andTable 2, respectively.

3. Results and Discussion

3.1. Melting Temperature and Viscosity

Figure 4 shows the effects of the melting temperature and viscosity of the F-free powders as functions of binarybasicity, TiO2, Na2O, Li2O, MgO, MnO and B2O3, respec-tively. The viscosity was in all cases measured at 1 300°C.It can be seen that Li2O is the strongest constituent for low-ering the melting temperature for compositions of Li2O�2.5 mass% in the slag, and B2O3 has a strong influencetoo. Increasing the contents of TiO2, MgO and CaO/SiO2

ratio in the slag increases the melting temperature of the F-free slag bearing titanium. The melting temperature de-creases with an increase of Na2O and MnO but does notchange significantly.

The effect of chemical composition on the viscosity of aliquid slag is relatively well understood when consideringthat slag structure depends on the relative amounts of con-stituents that act as network formers vs. those that act asnetwork breaker. As indicators of the amount of networkbreakers present, the compositions of Li2O, Na2O and theCaO/SiO2 ratio in the F-free powder bearing titanium lower,as expected, the viscosity in different degrees. Li2O lowersthe viscosity more if Li2O �2.0 mass% in the slag. To alesser degree for MgO contents larger than 4.0 mass%,there is a tendency to lower the viscosity. The influence ofMnO on the viscosity is not significant. For B2O3 contentbetween 2.0 and 10.0 mass%, the effect of B2O3 is to lowerthe viscosity, which indicates that B2O3 is not only a net-work former, but also an additive to reduce the viscosity.

The influence of TiO2 on the viscosity seems somewhatcomplex. The viscosity drops with increasing TiO2 to reacha minimum at 6.0 mass% TiO2, and subsequently it in-creases with increasing TiO2 content. This is because TiO2

has both acidity and alkalescence. If the TiO2 content islarger than 6.0 mass%, Ti ions in the slag acts as a networkformer and thereby increases the viscosity, while for con-tents lower than 6.0 mass% it acts as a network breaker anddecreases the viscosity. This result is not completely consis-


1119 © 2007 ISIJ

Fig. 2. Schematic diagram of the experimental apparatus (left) and solid slag deposition (right) for copper detector.

Fig. 3. Schematic diagram of the temperature difference withimmersion time.

tent with reported results on the effect of TiO2 on the vis-cosity of F-containing mold fluxes at 1 300°C,12) which in-dicates that the minimum in viscosity occurs at 10.0 mass%TiO2.

In the F-free powder with TiO2, the sequence of the mainfactors lowering the melting temperature is Li2O�B2O3�Na2O, and the sequence for the viscosity is Li2O�B2O3�Na2O. Therefore, Li2O and B2O3 additions in the F-freepowder can replace CaF2 without compromising the lower-ing of both the melting temperature and the viscosity.

3.2. Effect of TiO2 on the Heat Flux

It is well known that the heat flux decreases with increas-

ing fluorine content due to enhanced crystallinity of theslag film. It is therefore important to determine what effectTiO2 replacement of F has on the heat flux through the slagfilm. Figure 5 shows how the heat flux (measured accord-ing to the experimental setup described in Fig. 2) varieswith TiO2 content. The heat flux varies from 0.44 to0.34 MW·m�2 as the increase of TiO2 content from 1.0 to6.0 mass%, which is similar to the effect of F from 2.0 to10.0 mass% in the conventional powder on the heat fluxfrom 0.48 to 0.37 MW·m�2. When TiO2 content is largerthan 6.0 mass%, there is a tendency to increase the heatflux. This change in behavior at TiO2 contents above6.0 mass% on the heat flux is consistent with the change


1120© 2007 ISIJ

Table 1. Chemical compositions of studied F-free slag specimens (mass%).

Table 2. Chemical compositions of studied F-containing slag specimens (mass%).

observed on the viscosity (see Fig. 4(b)). If the changes dueto that TiO2 become a network former, above 6.0 mass%this would be expected to increase the phonon contributionto thermal conductivity in the melt which is consistent withthe experimental observations. On the other hand, increasedpolymerization and viscosity should not promote crystal-lization, which would increase heat transfer. The effect oncrystallization needs to be studied independently before thenature of TiO2 on heat flux can be further elucidated.

The change of the heat flux for the F-free powder withTiO2 and the conventional powder is 20.4% and 22.9%, re-

spectively. Thus it proves that TiO2 in the F-free powdercan be used to replace F in the conventional powder forcontrolling heat transfer between the mold and the shell.

3.3. Effect of Basicity on the Heat Flux

It can be seen in Fig. 6 that the heat flux for the F-freepowder with TiO2 decreases from 0.49 to 0.35 MW·m�2

with the increasing of basicity from 0.6 to 1.2 (especially,when the range of basicity is 0.9 to 1.1, there is a signifi-cant change of the heat flux), and the heat flux for the F-bearing powder varies from 0.48 to 0.35 MW·m�2 with the


1121 © 2007 ISIJ

Fig. 4. Effect of a variety of components on melting temperature and the viscosity of F-free powders.

Fig. 5. The influence of TiO2 in the F-free powder and F in the F-bearing powder on the heat flux.

increasing of basicity from 0.8 to 1.4, when the basicity islarger than 1.4, the heat flux increases a little as well, whichimplies why the maximum binary basicity is about 1.4 foractual selection of “mild cooling” F-bearing mold powder.The control range of the heat flux for the F-free powderwith TiO2 and the F-bearing powder is 28.6% and 27.1%,respectively. Therefore, it is clear that the effect of the ba-sicity on the heat flux is the same whether the F-free pow-der with TiO2 or the conventional powder with fluoride isused. These are consistent with the expected trend that thehigh basicity powder with more network breakers has boththe higher break temperature and the crystallization ratio.

3.4. Effect of Other Components on the Heat Flux

Figure 7 shows the effect of individual components inthe F-free powder with TiO2 on the heat flux, which isNa2O, Li2O, MgO, MnO and B2O3, respectively. The heatflux changes significantly with more than 8.0 mass% Na2Oand about 4.0 mass% MnO. The effects of Li2O, B2O3 andMgO in the F-free powder on heat flux are not significant.The effect of Na2O content on the heat flux may be that toomuch Na2O in the slag makes the break temperature lower.When MnO content is about 4.0 mass%, there is a largechange of the heat flux which is similar to the one of TiO2

on the heat flux at about 6.0 mass% in Fig. 5.

It is surprising that Li2O, B2O3 and MgO have no obvi-ous influence on the heat flux. Although the break tempera-ture and crystallization tendency are related to the amountof network formers and network breakers, the relationshipbetween chemical composition/break temperature and crys-talline content is more complex. Indeed, each component ofthe powder may influence the crystallization by controllingthe nature of the crystalline phases and by interacting withother components as well.13)

4. Plant Trials

According to the results above, the range of main compo-nents of F-free powder for peritectic steel slab casting wasproposed in Table 3. Two types of F-free mold powders, 1#and 2#, were made. The F-free mold powders were used at2# slab caster of Steelmaking Plant of Chongqing Iron &Steel Co. The steel grades in the trial were the peritecticsteel grades, which were classified as two types, namelyplain steel (A, B, Q234B and 20 g (AR)) and low alloy steel(A32, Q295A and Q345A, which contain high Mn content),and their chemical compositions were shown in Table 4.The parameters of the CC slab caster and correspondingcast steel grades were shown in Table 5. The physical andchemical properties of two F-free mold powders were de-


1122© 2007 ISIJ

Table 3. Main components of the F-free powder casting peritectic steel for slab (mass%).

Fig. 6. The effect of basicity on the heat flux (left: F-free powder with TiO2, right: F-bearing powder).

Fig. 7. The effect of individual components of F-free powder on the heat flux.

tailed in Table 6. The reagents are made up of Na2O, Li2Oand B2O3 in Table 6, Tm, h1 300°C, Tc and h represent thehemisphere point temperature, the viscosity at 1 300°C, thecrystallization temperature, and the crystallization fractionof the F-free mold powders, respectively. The crystalliza-tion temperature of the slag is determined by the method ofDifferential Thermal Analysis (DTA), and the crystalliza-tion fraction of the slag is evaluated through visual observa-tions in an experimental apparatus based on the single hotthermocouple technique (SHTT).14) In addition, the heatfluxes of two F-free mold powders and two F-bearing moldpowders used were measured. The heat flux through theslag film is 0.411 MW·m�2 for 1# F-free mold powder and0.384 MW·m�2 for 2# F-free mold powder, and the corre-sponding value for 1# and 2# F-bearing mold powders is0.417 MW·m�2 and s 0.382 MW·m�2, respectively. Themeasured heat flux can be transformed into the integral heatflux of mold based on the following relation15)

qint�k ·Vc ·q.................................(1)

Where Vc represents casting speed in m ·min�1, qint and qrepresent the calculated integral heat flux at a certain cast-ing speed and measured heat flux in MW·m�2, respectively,and k is a coefficient related to casting speed. Its regressionformula for slab caster is as follow:

k�0.0131Vc3�0.0455Vc2�0.9933Vc�3.4032......(2)

For 170�1 200 mm2 section at casting speed of

1.4 m·min�1, the integral heat flux value is 1.233 MW·m�2

for 1# F-free mold powder and 1.251 MW·m�2 for 1# F-bearing mold powder; for 240�1 400 mm2 section at cast-ing speed of 0.75 m·min�1, the integral heat flux value is 0.774 MW·m�2 for 2# F-free mold powder and0.770 MW·m�2 for 2# F-bearing mold powder. The resultsshow that the heat flux value of the F-free mold powder de-veloped is close to that of the corresponding F-bearingmold powder used.

The results from the industrial trials are listed in Table 7.The F-free mold powders were uniformly melted in themold, and there were no stick phenomenon, also no forma-tion of lumps and thick slag rims in the mold. No breakoutaccidents occurred during the trials of 446 continuous cast-ing heats. The consumption of the F-free mold powders was0.71–0.77 kg/tonne steel. The temperature difference of thecooling water between inlet and outlet in the slab mold wasin the normal range of 7.8–8.6°C. The morphology of theslag film taken from plant mold for 1# powder is shown inFig. 8. The slag film contains two layers: a glassy zoneclose to the steel shell and a crystalline zone in contact withthe mold copper. The apparent “bar” or “cross” shapedcrystals were shown to be perovskite crystals when ana-lyzed by X-ray diffraction (XRD), and the crystalline frac-tion of the slag film is about 42% by the determination ofimage analysis. The morphology of the F-containing slagfilm with the crystalline fraction of 39% corresponding tothe F-free powder is shown in Fig. 9, the left crystalline


1123 © 2007 ISIJ

Table 4. Classification and chemical compositions for cast steel grades.

Table 5. Slab caster conditions and cast steel grades.

Table 6. Physical and chemical parameters of F-free mold powder.

Table 7. The result used the F-free powders in mold.

side (mold side) is composed of cuspidine crystals. It isnoteworthy that when comparing Figs. 8 and 9, the thick-ness of the crystalline layers are roughly similar.

The surface quality of the continuous casting slab pro-duced using the F-free mold powders is shown in Fig. 10. Itis reasonably good, but some degree of longitudinal crack-

ing does appear to exist. Nevertheless, F-free mold powdersproduce smaller surface crack indexes, defined as a ratio ofthe crack length to the slab length, than mold powders bearing fluoride (Fig. 11). By switching the mold powdersbearing fluoride to the F-free mold powders, for 170�1 200 mm2 slabs, the surface crack index of the plain steel


1124© 2007 ISIJ

Fig. 8. Morphology (left) and XRD result (right) of the slag film taken from plant mold for 1# powder.

Fig. 9. Morphology of the slag film taken from plant mold for F-containing powder.

Fig. 10. The cast slabs of the 170�1 200 mm2 slab (left) and of 240�1 400 mm2 slab (right) using F-free mold powders.

Fig. 11. The surface quality of the 170�1 200 mm2 slab (left) and of 240�1 400 mm2 slab (right) using mold powdersbearing fluoride (1) and F-free mold powders (2).

decreased from 4.5 to 1.6%, the surface crack index of lowalloy steel decreased from 2.3 to 0.8%; and for 240�1 400 mm2 slabs, the surface crack index of the plain steeldecreased from 6.5 to 2.1%. It appears thus that the use ofthe F-free mold powders proposed in this study consistentlyresult in improved slab surface quality compared to thoseobtained when using conventional mold powders bearingfluoride. These F-free mold powders can therefore be con-sidered as less hazardous alternatives in the continuouscasting process.

5. Conclusions

Titanium-bearing blast furnace slags (CaO–SiO2–TiO2)were used as a base material to develop F-free mold pow-ders and their properties and performance were evaluatedthrough labortoary experiments and plant trials. The con-clusions are as follows:

(1) The laboratory results show that 1) the melting tem-perature and the viscosity of the F-free powder decreasewith increasing contents of Li2O, B2O3 and Na2O, respec-tively, and the lowest viscosity is achieved with 6.0 mass%TiO2; 2) the heat flux of the F-free slag film with 1.0–6.0mass% TiO2 is close to that of the slag film with 2.0–10.0mass% F; 3) the effect of basicity of the F-free powder onthe heat flux is the same as the powder bearing fluoride; 4) the heat flux changes significantly with more than8.0 mass% Na2O and about 4.0 mass% MnO, and the effectsof Li2O and B2O3 in the F-free slag on heat flux are not sig-nificant.

(2) The suitable range of main components of the F-free powder with TiO2 is proposed for casting peritecticsteel grades (slabs), namely basicity (CaO/SiO2) 0.95–1.15,TiO2 4.0–7.0 mass%, Na2O 5.0–8.0 mass%, Li2O 1.0–2.0mass%, MnO 3.0–5.0 mass% and B2O3 4.0–8.0 mass%.

(3) The industrial trial indicates that the F-free powdercan effectively control mold heat transfer through the per-ovskite precipitated in the infiltrated slag layer instead of

the cuspidine in fluoride-bearing powder. As a result it pro-duces better slab surface quality than fluoride-bearing pow-ders in terms of crack index.

Acknowledgements

The authors wish to express their gratitude to the Na-tional Natural Science Foundation and Shanghai Baosteel(China) (Grant No.: 50374086) for funding the currentstudy. The special efforts of Henan Xixia Protective Materi-als Group and Steelmaking Works at Chongqing Iron &Steel Co., to make and use the F-free mold powders in in-dustrial trials, are also greatly appreciated.

REFERENCES

1) A. Fox, K. Mills, D. Lever, C. Bezerra, C. Valadares, I. Unamuno, J.Laraudogoitia and J. Gisby: ISIJ Int., 45 (2005), No. 7, 1051.

2) Alexander and I. Zaiter: Steel Res., 65 (1994), No. 9, 368.3) S. Choi, D. Lee, D. Shin, S. Choi, J. Cho and J. Park: J. Non-Cryst.

Solids, 345–346 (2004), 157.4) W. Han, S. Qiu and G. Zhu: Research on Iron & Steel, (2003), No. 2,

53.5) B. Harris, A. Normanton, G. Abbel, B. Barber, I. Baillie, R.

Koldewijn, A. Chown, S. Riaz, S. Higson, B. Patrick, T. Peeters, A.Smith, H. Visser and J. Kromhout: Ironmaking Steelmaking, 33(2006), No. 1, 5.

6) N. Fu, Y. Zhang and Z. Sui: Min. Metall. Eng., 17 (1997), No. 4, 36.7) H. Nakada and K. Nagata: ISIJ Int., 46 (2006), No. 3, 441.8) R. Devries, R. Roy and E. Osborn: J. Am. Ceram. Soc., 38 (1955),

No. 5, 158.9) C. Pinheiro, I. Samarasekera and J. Brimacombe: Iron Steelmaker,

22 (1995), No. 3, 76.10) K. Mills and A. Fox: ISIJ Int., 43 (2003), No. 10, 1483.11) M. Fonseca and O. Afrange: Proc. of 5th Int. Conf. on Molten Slags,

Fluxes and Salts, ISS, Warrendale, PA, (1997), 851.12) T. Mukongo, P. Pistorius and A. Garbers-Craig: Ironmaking Steel-

making, 31 (2004), No. 2, 135.13) Y. Vermeulen, E. Divry and M. Rigaud: Can. Metall. Q., 43 (2004),

No. 4, 527.14) B. Jia, J. Shi and G. Wen: J. Iron Steel Res., 18 (2006), No. 2, 21.15) G. Wen, P. Tang and X. Qi: Research Report on Fluoride-free Mold

Powder Based on Blast Furnace Slag with Titanium, ChongqingUniversity, Chongqing, (2006), 73.


1125 © 2007 ISIJ

Documents

Development of Fluoride-free Mold Powders for Peritectic