Formation of oxide deposits during liquid steel processing.

摘要

Tundish nozzle clogging is a problem encountered by virtually all steel manufacturers with continuous casting facilities.; Detailed observation of numerous clogged nozzles was carried out. These observations included a study of the clog bulk shape, its chemical constituents and a detailed SEM observation of the microstructure of the clog. It is revealed through the observations that the clogs are composed of a network of alumina with or without steel in the voids. These networks of alumina appear to be composed of small alumina particles joined to each other. A closer observation shows that these networks could actually be a direct result of the fashion in which alumina grows, as well as being composed of small particles sintered together. A mechanism for clogging is developed based on these observations and from the knowledge gained form the literature.; The proposed mechanism for clog growth is the precipitation of the dissolved Al and O to form Al2O3 on the surface of the nozzle refractory. The initial deposit subsequently acts as a site for further growth or agglomeration of alumina. This is in the case of aluminum-killed steels, though the clogging mechanism may be the same for other deposit chemistries.; Hot experiments were carried out in which steel was deoxidized with aluminum, among other deoxidizers. The deoxidation experiments were video recorded and show the way alumina grows across the surface of the melt as aluminum is added. A complete evaluation of the resulting ingot was done, including detailed SEM observation of the alumina particles present in the ingot and on the crucible surfaces. The alumina grows in the form of a network, which at first appears to be a clustering of numerous micron sized alumina particles, just as an alumina clog appears to be. A closer look reveals that the these alumina network actually grow as such, as supported by the video observations. In a related set of experiments the alumina from deoxidation experiments, and from other sources, was sintered at steelmaking temperatures to simulate the residence of an actual clog at casting temperatures. (Abstract shortened by UMI.)