It has been demonstrated that the use of freeze lining technologies can significantly increase the working lives of pyrometallurgical furnaces that are subjected to demanding process conditions and corrosive liquid melts. By cooling the outer furnace walls, the freeze linings on the inner walls of the furnace linings can be formed. Both the thickness of the freeze lining and the heat loss through the furnace walls, are of special concern to process operators.
The aim of this thesis is to investigate the effect of transition of spinel primary phase field slag to melilite primary phase field slag and vice versa on freeze lining formations in terms of microstructure and thickness of freeze linings. Until now a lot of research has been done for the formation of steady state freeze linings and temperature at the interface. But there needs to be done intense amount of research in transient state freeze linings because this could be useful for numerous recycling smelting refineries where the secondary raw material is coming from different sources which is used as a batch feed for recycling of precious metals. Hence, there is a high chance of change in slag bath and how it effects the already formed freeze linings needs to be studied. The slag system chosen in this thesis is the SiO2-FeO-CaO-Al2O3 system. The slag compositions are selected to be in either the melilite or spinel primary phase field. FactSage simulations are done to find out suitable compositions of slag. First, the time needed to form a steady state thickness in a specific primary phase field freeze lining is determined which was found out to be probably 3 hours. After that, the transition of slag composition is done by the initial formation of a freeze lining in one specific primary phase field and then inserting it into the molten slag bath of another primary phase field composition. Microscopic characterizations are performed to investigate the freeze linings. In case of the transition from the spinel primary phase field slag to the melilite primary phase field slag, microstructure of the crystalline layer remained similar compared to the spinel primary phase field intermediate freeze lining before the transition. But there is a formation of an open crystalline layer between the crystalline layer and the slag bath. The open crystalline layer is composed of already formed undissolved anorthite and spinel crystals similar to the open crystalline layer and an entrained slag bath layer before the transition but with the solidified slag matrix where Spinel primary phase slag and melilite primary phase slags got mixed up. Also, there is a precipitation of new melilite phase after the transition in the nearby region of slag bath. In case of the transition from the melilite primary phase field slag to the spinel primary phase field slag, the microstructure of the crystalline layer is again similar to the initially formed intermediate freeze lining in a melilite primary phase field slag before the transition. But after that a closed crystalline layer is formed instead of an open crystalline layer and an entrained slag bath layer which was formed before the transition. In this closed crystalline layer, there are large sized grown crystals of Ca-pyroxene, spinel and anorthite crystals which extend up to around 1 mm of thickness from the endpoint of the crystalline layer. But in the melilite primary phase intermediate freeze lining before transition, these phases were present with a smaller size over a shorter region. Also, after the transition, there is a growth of second generation of anorthite crystals in the closed crystalline layer which was not there in the intermediate melilite primary phase freeze lining before the transition. Furthermore, in the region nearby the slag bath in the intermediate freeze lining after the transition, there is a formation of large volume of melilite crystals with already present undissolved anorthite and the spinel crystals with the solidified slag matrix where Spinel primary phase slag and melilite primary phase slags got mixed up. Whereas before the transition, only anorthite and spinel crystals were present in the region nearby the slag bath. In both experiments of transition from the spinel primary phase field slag to the melilite primary phase field slag and vice versa, there is no formation of a freeze lining of different slag composition on top of the initially formed intermediate freeze lining in a specific composition. The outcome of these experiments will be useful to understand the change in behavior of freeze linings while changing the slag bath in a quaternary system.