In the scope of the collaborative research center SFB 761, von Schweinichen(*) experimentally investigated the chemical interactions between a high manganese steel melt and different refractory materials employed in assembling an adequate casting system. Normally, when casting in an open atmosphere, one expects worse results considering cleanliness for a top pouring process (Figure 1) than for a bottom teeming process
(Figure 2). Nevertheless, for the case of a Fe-23%Mn-0.4%C alloy, just the opposite situation occurred: the non-metallic inclusions in the case of situation of Figure 2 were up to 10 times larger than the ones of situation of Figure 1.
Figure 1 – Scheme of the casting process used in the open furnace with top pouring
Figure 2 - Scheme of the casting process used in the open furnace with bottom teeming
An examination of the runner system after the casting procedure showed that the melt attacked the inner surface of the runner system (Figure 2). The refractory material consisted of approximately 60% SiO2 and 40% Al2O3. The presence of a green coloured layer indicated the formation of MnO in the runner system. At the same time, during the casting procedure, there was simultaneously a reduction of the [Mn] content and an increase of the [Si] content of the alloy.
According to the evidences, it seems that, among other possible reactions, the following one is taking place:
2[Mn] + SiO2 → 2(MnO) + [Si]
Nevertheless, according to the classical Ellingham diagram for oxide formation, the reaction above is not possible at the casting temperature of ca.1550 °C. What is then happening in this situation?
The important point is that Ellingham diagrams only offer a first approximation for such a situation, because they only deal with pure substances. In the present real case, one has to do with solution phases and the adequate procedure to handle this is by considering the activities of the various substances in the corresponding phases. FactSage takes this into account. Figures 3-6 shows a series of calculations done with the equilibrium module of FactSage, where the relative amounts of alloy melt and refractory material were varied. The calculation scheme is like following, one defines two sub-systems: a refractory material system and a steel melt. The chemical compositions of the individual sub-systems were explicited in the previous paragraphs. One then defines a variable <A>:
and finally, one performs a series of equilibrium calculations for series of values of <A>, 0<A<1.
The <A>-variation is schematically depicted on Figure 3.
Figure 3 – Scheme for the systematic variation of the relative amounts of steel melt and refractory material in the FactSage equilibrium simulations at 1550 °C
One notices that an increasing amount of liquid slag forms in the system when the weight percentage of refractory material in the initial system increases up to 40% (Figure 4). This slag contains considerable amounts of MnO (Figure 5). At the same time, through an increase of the amount of refractory material in the initial system, a progressive increase of the [Si] content in the melt occurs and the [Mn] content of the melt decreases (Figure 6). This proves that the oxidation of manganese by the silicon dioxide is really taking place.
Figure 4 - Relative amounts of the phases in the system as a function of the <A>-parameter
Figure 5 - Chemical composition of the slag as a function of the <A>-parameter
Figure 6 - Chemical composition of the steel bath as a function of the <A>-parameter
The problem was solved by applying a special inner ceramic coating to the runner system, so as to prevent the reaction between manganese and silica.
(*) P. von Schweinichen, Z. Chen, D. Senk, A. Lob: Metall. Mater. Trans. A, 2013, vol. 44, pp. 5416-5423