Prognostic models for electroslag remelting process and slag engineering

Abstract

Classical electroslag remelting of consumable electrodes is a drop-by-drop melt feeding through a slag layer into a renewing molten metal bath, slowly solidifying into a homogeneous, dense, defect-free ingot in a copper water-cooled mould. The new look at the additive nature of the electroslag remelting (ESR) process in a protective atmosphere allowed us to formulate the principles of a comprehensive thermodynamic-based model that can predict dynamic changes in slag and metal composition at certain ingot remelting. The model considers drastically different slag-to-metal mass ratios at the beginning and end of remelting and predicts gas, slag and metal composition in a chain of thermodynamic subsystems. Despite consisting of calcium fluoride and stable oxides, the ESR slag can oxidise main and active elements from steel and alloys (primarily aluminium, titanium and silicon) due to chemical reactions between slag and metal. ESR is not an electrochemical process in its nature. However, slags are ionic melts and deviation in their composition causes a change in their properties, affecting both operation mode and ingot quality. Another important understanding derived from the modern metallurgy technological route is that the ability to refine metal from impurities is not a priority for the ESR because consumable electrode has already passed all stages of refining and deep degassing at ladle treatment that changes ESR slags engineering principles. The critical importance became slag’s ability to generate process heat and keep the melting composition in the metal bath unchanged (except for non-metallic inclusions assimilation). Slag engineering for ESR required a compromise between chemical inertness to a metal composition and desired physical properties deriving from technological reasons. The Directed Chemical Bonds Concept (DCBC) in a multicomponent oxide system was used to build predictive models of electric conductivity and the melting temperature of fluoride-oxide slag based on their chemical composition. Both models help to design a customised composition of effective slags for steel and alloy groups or individual grades, and they are significant steps in the development of a comprehensive model of electroslag remelting.