The free meniscus problem in the continuous casting of steel: a computational model of cast surface formation

Oscillation marks appear as transverse grooves on the surface of continuously cast peritectic and Ultra Low Carbon steel products. They affect the topology of the continuously cast surface and can be the source of defects on products rolled from continuously cast material. Oscillation marks cannot be avoided with current casting technologies, and their origin is unclear. Among the many speculations on the causes of their formation, the hypothesis of a solidification and overflow of the meniscus is supported by physical evidence, but lacks a sound theoretical background. This work aims to show that the physical phenomena -- heat transfer, mass transfer and fluid flow -- involved in this hypothesis can reasonably explain the formation of oscillation makes; it also seeks to improve the continuous casting process so that oscillation marks are either minimized or eliminated.;Results of the derived model indicate that the standard conditions found in the meniscus area of a continuous caster are sufficient to explain the formation of oscillation marks: the meniscus can freeze during the oscillation cycle, and mold oscillation and strand withdrawal provokes a flow of molten steel over the frozen meniscus, which can lead to an oscillation mark. The predicted heat flux through the mold and surface profile respond to changes in casting parameters in the manner that an actual caster would respond to the same changes in casting parameters.;A numerical model of the meniscus area is designed such that all the phenomena explaining the formation of oscillation marks according to this theory are included. The heat transfer (including conduction, advection, solidification, radiation) and mass transfer (including fluid flow, moving boundary) equations are solved in a transient way that reflects caster operation. Given the complexity of the model and the difficulty of handling moving boundaries, particular consideration is given to the finite element solution of these differential equations. The model is carefully designed as a reasonable simulation of the continuous casting process particularly focused on the area of the liquid steel meniscus in the mold of a continuous caster; it does not suffer from assumptions made only to ease the solution of those differential equations.