2013 Volume 53 Issue 4 Pages 648-654
Crystallization of mold flux was observed with confocal laser microscope. Crystallization temperature, CCT curve and crystallization rate were evaluated from the observed images. The evaluated results were compared between two kinds of mold flux, and influence of basicity on crystallization rate was discussed. Crystallization temperature increased with basicity. Crystallization rate also increased with basicity, but its dependency on cooling rate was differed by basicity. The difference could be explained by the viscosity of mold flux at crystallization temperature. Crystallization rate has clear relation to the viscosity at crystallization temperature, and the rate increased with decrease of the viscosity. Two kinds of mold flux were unified in this relationship. Crystallization is controlled with basicity in terms of not only equilibrium but also kinetics through viscosity.
Mold flux plays various rolls in continuous casting mold. It was introduced to the casting as lubricant in the mold initially, and also works as insulator for molten steel so that it may not be oxidized by the air and freeze on its surface, and absorbent of non-metallic inclusions in it.1) Then, as longitudinal surface cracking became obvious with increasing casting speed for the sake of productivity in steelmaking process, medium of heat transfer came to be added to its rolls, in order to prevent longitudinal cracking on the surface of casts by mild cooling of solidified shell in the mold.2)
Cooling condition can be controlled by crystallization in the film which mold flux forms along the wall of the mold. Mechanism of mild cooling with the film has been considered to be increase of thermal resistance on its surface to the mold,3) or reduction of radiation in it.4) Crystallization in the film can be controlled by designing mold flux composition suitable. Increase of basicity; mass content ratio of CaO/SiO2,5) addition of ZrO26) or decrease of Al2O3 and MgO7,8) have been conventionally considered to be effective method for promoting the crystallization.
In order to take adequate crystallization, mold flux composition should be designed in terms of metallurgical phase relation to crystal phase, so equilibrium information on the crystal phase is very important. Cuspidine; Ca4Si2O7F2 is known as common crystal phase in the conventional mold flux.9) Concerning with cuspidine, primary crystallization field was clarified in CaO–SiO2–CaF2 ternary system,10) and phase diagram of CaF2–NaF–cuspidine ternary was experimentally established.11) This author studied on phase relation between mold flux and cuspidine in CaO–SiO2–CaF2–NaF quaternary system and predicted suitable condition for the crystallization, considering affinity of F ion to Na ion rather than Ca ion.12)
In order to estimate the effect of mild cooling with the flux, there have been many studies on heat transfer or thermal resistance through the film, and on the influence of crystallization. Crystallization increases surface roughness and the air gap causes interfacial thermal resistance.13,14) There have been some trials to evaluate this resistance by both experimental and numerical simulation,8,13,14,15,16,17,18) and influence of such factors as mold flux composition,8,13,17) cooling rate or thickness of the film,15,16,18) and some casting conditions17) were pointed out. Heat transfer or thermal resistance in the film is usually treated as apparent one which includes both conduction and radiation,19,20) but recently, radiation has been extracted and its behavior has been studied in detail.21,22,23,24) It was reported that there is optimum grain size of cuspidine for reduction of radiation,22) or reduction of FeO is effective on that reduction.23) Using those obtained values of thermal properties, numerical simulations of heat transfer in the continuous casting mold are tried and their representation is being improved.25,26,27)
In spite of those information, crystallization phenomenon in the film in the continuous casting mold is still unknown, because it is difficult very much to observe it directly during actual casting. So there have been many researches on the film taken from the mold after the casting,7,9,25,28) and influence of mold flux composition or casting conditions such as casting speed or steel grade on the structure of the film. But influence of various conditions for the film such as infiltrating rate, temperature history or gradient, action from molten steel or solidified shell, on the generation or morphology of crystal grain in the actual film have never been clarified yet.
In order to understand the crystallization phenomenon in the actual film, it should be broken down into some simple factors and fundamental researches should be conduct on them from various view points.
There have been some studies on the crystallization phenomenon itself.23,29,30) Anneal of quenched glass of mold flux can be available for the observation of crystal grain.23,29,30) There have been some studies employing this method and influence of annealing time or temperature, composition of the specimen on the size or morphology of crystal grain was investigated.23,29,30) On the other hand, direct observation is useful not only for representation the phenomenon in-situ but also evaluation such behaviors that transformation temperature and growing rate. By means of employing the single or double thermo-couples method; SHTT or DHTT, TTT or CCT diagram for crystallization of the flux can be obtained and influence of cooling rate on crystallization temperature or morphology of crystal grain.31,32) Confocal laser microscope33,34) is also available to observe the crystallization. Crystallization of cuspidine was observed during heat treating, the number density or growing velocity of the crystal grain was evaluated and influence of heat treating temperature was discussed.29)
Looking back these previous studies mentioned above, it seems that there are few study on relation between crystallization behavior and some physical properties of mold flux. So, in this study, crystallization in molten flux was observed by confocal laser microscope and CCT curve and crystallization rate were tried to be evaluated. Furthermore, influence of basicity or viscosity, as a physical property of molten flux, is discussed on the basis of the results.
Confocal laser microscope was used for observation in this study. The microscope was VL2000DX-SVF17SP model made by Lasertech Corporation (now, Yonekura FMG. Co., Ltd.). This microscope was combined with an infrared image furnace, and this formation or specification in details is same as previous studies.29,33,34,35)
Conditions of observation are listed in Table 1. Mold flux as specimen was inserted in Pt crucible was heated at 200 K/min in purified Ar atmosphere in the infrared image furnace, and was melted at 1573 K for 5 min. Inner radius of the crucible was 5 mm and weight of the flux was about 0.06 g. At the state that the flux was melt, the surface of molten flux came to have meniscus, and the thickness of the flux was about 1 mm at the center of the bottom. The weight of 0.06 g was decided so as to make easy observation of crystallization on the bottom. Then it was cooled at the rate of 6 conditions; 200 K/min, 100 K/min, 50 K/min, 20 K/min, 10 K/min and 5 K/min, as shown in Fig. 1. Temperature was measured just below the bottom of crucible with thermocouples, and the temperature was controlled by this measurement. Image of the flux was captured at a rate of 30 frames per second. After observation, crystallization temperature and crystallization rate were judged on the recorded images.
| Apparatus | Confocal laser microscope |
| Specimens | Mold flux: about 0.06 g |
| Crucible | Pt: 5 mm of inner radius |
| Atmosphere | purified Ar |
| Temperature | 1573 K at maximum |
| Cooling rate | 5–200 K/min |
Temperature history of the specimen in the observation.
Mold flux for the observation was made by mixing ordinary resources such as cement, silica sand, fluorspar and soda ash. It was inserted in carbon crucible and heated at 1573 K in a furnace so as to melt. The molten flux was taken out of the furnace and quenched on the Cu block cooled by water. The quenched glass was crashed into powder.
Specifications of the flux are listed in Table 2. Two kinds of mold flux were observed in this study. Main components were T.CaO and SiO2 and basicity; mass percent ratio of T.CaO/SiO2 was differed between 1.0 and 1.2. As the other components, Na2O and F were added at 10 mass%. Al2O3 and MgO were included as impurities. Physical properties of the flux were listed in Table 3. Solidification temperature of the flux B was higher and viscosity lower than the flux A, according to its higher basicity. Their compositions are shown in Fig. 2 as phase diagram of CaO–SiO2–CaF2 ternary system. They exist out of primary crystallization field of cuspidine10) and near the tie-line between CaO·SiO2 and cuspidine.
| Sample | Basicity, - (T.CaO/SiO2) |
Contents, mass% | |||
|---|---|---|---|---|---|
| Al2O3 | MgO | Na2O | F | ||
| A | 1.0 | 3 | 0.8 | 10 | 10 |
| B | 1.2 | 3 | 0.9 | 10 | 10 |
| Sample | Basicity, - (T.CaO/SiO2) |
Solidification temperature, K |
Viscosity at 1573 K, poise |
|---|---|---|---|
| A | 1.0 | 1419 | 1.0 |
| B | 1.2 | 1475 | 0.6 |
Compositions of the flux in the phase diagram of CaO–SiO2–CaF2 ternary system.
Crystallization temperature and crystallization rate were evaluated on the basis of observation.
Observation was focused at the center on the bottom of crucible, which was under molten flux and the temperature at the moment that growing crystal phase reached at the center was defined as crystallization temperature. The temperature used for this evaluation was measured just below the bottom of crucible.
Crystallization rate was evaluated on the images. Growing distance of crystal phase was measured between two images and it was transformed into crystallization rate with the growing time between them.
2.2. Analysis after ObservationAfter observation with confocal laser microscope, specimens were whole molded in resin and cut so as to be observed solidified structure inside. The section was polished and observed with microscope. After observation with microscope, the crystal phase on the section was analyzed by X-ray diffraction, and then, the crystal composition was characterized on the basis of the detected peak pattern.
2.3. Measurement of ViscosityFor the later discussion on the influence of viscosity on crystallization behavior, viscosity of the flux was measured in this study. The measurement was conducted with oscillating-plate viscometer.36) 1 kg of mold flux was melt in carbon crucible and cooled at 2 K/min from 1723 K, and viscosity was continuously measured in this cooling state. The flux was same mixture as that for observation with confocal laser microscope.
Most of the conditions, crystallization was observed in the molten flux.
Images of observation are shown in Fig. 3, as the case of the sample B cooled at 10 K/min, for an example. Crystallization proceeded as follows;
Process of crystallization in the crucible as a result of observation, for the flux for the flux B cooled at 10 K/min. The images were focused at the center of bottom of the crucible.
(a) The center of Pt crucible bottom was focused and crystal grains of Pt were clearly observed through the transparent molten flux.
(b)–(d) Edge of crystal phase appeared from the right and grew to the left and reached at the center of the bottom of crucible.
(e)–(g) The growing crystal phase comes to reveal on the surface of the molten flux and moved as if it ran after that in the molten flux.
(h) The focus was moved from the bottom to the surface.
Crystallization could be observed in the same way in other conditions, except some cases that sample A was cooled at 100 or 200 K/min and remained glassy.
3.2. Crystallization TemperatureTemperature curves during the observation are shown in Fig. 4 for the flux A. Crystallization temperature revealed in the range of 1073–1356 K and it decreased with increase of cooling rate. White and round plots indicate the occasion of the crystallization. As mentioned above, crystallization did not occur in the cases of 100 and 200 K/min, which were shown with broken lines in the figure. In order to refer, another observation was conducted in the condition that the flux A was cooled from 1573 K to 1073 K at 200 K/min and kept at 1073 K. Then, crystallization occurred during the constant temperature. Connecting the plot of the crystallization temperature in this figure, transformation curve of crystallization arose for the flux A.
CCT diagram for the flux A.
The same curves are shown in Fig. 5 for the flux B. Crystallization occurred in all the conditions of cooling rate in the temperature range of 1264–1455 K. Black and square plots in the figure indicate the occasion of the crystallization.
CCT diagram for the flux B.
Transformation curves in Figs. 4 and 5 were shown in Fig. 6 in comparison of the flux. The curve for the flux B was upper than that for the flux A, and the difference between them corresponded to that of solidification temperature in viscosity measurement at cooling rate of 2 K/min, mentioned later.
Influence of basicity of the flux on the crystallization curve on the CCT diagram.
Comparing the curves between the fluxes, it can be considered that crystallization temperature increases with basicity.
3.3. Crystallization RateCrystallization rate is shown in Fig. 7 as a function of cooling rate, for the case that crystallization was observed. Crystallization rate of the flux B was larger than that of the flux A, in the all range of crystallization rate. It is remarkable that the tendency of crystallization rate on cooling rate differed by the fluxes. Thus, crystallization rate decreased with increase of cooling rate in the case of the flux A, but it kept constant in spite of cooling rate in the case of the flux B.
Relation between crystallization rate of the flux and cooling rate.
Solidified structure on the section of the specimen after observation is shown in Fig. 8. This figure is for the case of the flux B cooled at 10 K/min, as an example. This Image shows around the center and near the bottom in the crucible. Some broad columnar crystal lied roughly horizontally. Fine columnar crystal existed among them in right angle. This structure was common to all the specimens which gave crystallization behavior in the observation with confocal laser microscope.
Section of the specimen in the crucible after observation, for the flux B cooled at 10 K/min.
A result of X-ray diffraction for the case of the flux B cooled at 10 K/min is shown in Fig. 9, for example. The detected pattern of X-ray diffraction peak corresponded to that of CaSiO3. This crystal phase was also common to the all specimens with the crystallization.
A result of X-ray diffraction for the specimen of the flux B cooled at 10 K/min.
Viscosity of molten flux is shown in Fig. 10. Viscosity gradually increased with decrease of temperature, and rapidly increased after passing the break point. The gradual increase indicates the behavior in molten state and the rapid one in the procedure of solidification. Viscosities of both flux behaved similarly each other, but viscosity of the flux B was lower than that of the flux A. In both cases of the flux, viscosity in logarithm changed linearly with the reciprocal of absolute temperature in molten state. Both viscosities were regressed as follows;
| (1): for the flux A |
| (2): for the flux B |
Change in viscosity of the flux with temperature.
As mentioned in the paragraph of 3.3, crystallization in molten flux was observed in the same manner, except the cases of glassy solidification: the crystallization occurred from the side of crucible toward the center of the bottom. In the cooling process, heat of molten flux can be considered to be extracted through the side of Pt crucible, and temperature gradient to exist in the radius direction. So it is considered that crystallization proceeds along the direction of heat transfer.
4.2. Influence of Crystallization Temperature on Crystallization RateAs shown in 3.3, dependency of crystallization rate on cooling rate differed by basicity of the flux. This result is interesting, but the reason of the difference is not clear at the present state.
It should be noticed that difference in crystallization temperature by mold flux became larger with increase of cooling rate, as shown in Fig. 6, and the difference in crystallization rate by the fluxes behaved similarly against the cooling rate. Concerning with this point, relation between crystallization rate and crystallization temperature should be discussed, so it is shown in Fig. 11. Crystallization rate increased with crystallization temperature and the gradient of the relation is same between the fluxes.
Relation between crystallization rate of the flux and crystallization temperature.
But the results of crystallization rate differed, compared in the same range of crystallization temperature. Discussion should be preceded on this point in the next clause.
4.3. Influence of Viscosity on Crystallization RateIt is generally known that viscosity of the glass has influence on its crystallization.37) In reference on this point, viscosity at crystallization temperature is noticed and shown in Fig. 12. Viscosity of the flux A at crystallization temperature is higher than that of the flux B.
Viscosity at of the flux at crystallization temperature.
It is reasonable that the flux B of lower viscosity crystallizes more rapidly than the flux A of higher viscosity. Furthermore, relation between crystallization rate and viscosity at crystallization temperature is discussed, here. The relation is shown in Fig. 13. As a result, quite clear relation exists between crystallization rate and the temperature at crystallization temperature, and both of the flux A and B can be unified in this relation. In this relation, crystallization rate increases with decrease of the viscosity.
Crystallization rate as a function of viscosity of the flux at crystallization temperature.
It can be concluded that lower viscosity with higher basicity of the flux introduce larger crystallization rate of it. It seems that lower viscosity promotes diffusion of components of crystal phase and this is the reason of larger crystallization rate, but there is some room for discussion on this point, so further study is required.
4.4. Meaning of Crystallization Control with Basicity of Mold FluxAs mentioned in the introduction, crystallization of mold flux has conventionally been controlled with its basicity. Relation between solidification temperature and basicity for some mold fluxes in general use is shown in Fig. 14.38) Compared with the fluxes for low carbon steel casting, those for hypo-peritectic steel cast have higher solidification temperetures with higher basicities. Here, it is considered that increase of solidification temperature means a promotion of crystallization of cuspidine in terms of equilibrium.
Relation between solidification temperature and basicity of T.CaO/SiO2 for conventional mold fluxes.38)
On account of the result of this study, it seems that increase of basicity has another effect on the crystallization of cuspidine. Concerning with the mold fluxes shown in Fig. 14, relation between viscosity at 1573 K and basicity is shown in Fig. 15. As a matter of course, the viscosity decreases with increase of basicity. Decrease in viscosity can also be considered to promote crystallization, as well as increase of solidification temperature. But, in this case of viscosity, the effect is not in terms of equilibrium but kinetics.
Relation between viscosity at 1573 K and basicity of T.CaO/SiO2 for conventional mold fluxes.
It can be concluded that crystallization control with increase of basicity has two kinds of effect in terms of both equilibrium and kinetics.
Crystallization at high temperature was observed at high temperature with confocal laser microscope, and crystallization behavior such as crystallization temperature, CCT curve and crystallization rate were evaluated from the observed images. On the basis of the evaluation, influence of basicity on crystallization rate was discussed. The conclusions are as follows;
(1) Crystallization temperature increases with increase of basicity.
(2) Crystallization rate increases with increase of basicity.
(3) Crystallization rate depends on the viscosity at crystallization temperature. Increase of basicity acts on not only crystallization temperature but also viscosity.
(4) Crystallization can be controlled with basicity in terms of both equilibrium and kinetics.
