An Interdisciplinary Approach towards Optimal Continuous Casting of Steel

This paper describes main elements and practical use of the computational system for automated setting of optimized casting conditions in the continuous casting of steel. The system consists of the heat transfer model of the process, empirical metallurgical criteria, and optimization algorithm. The physical model of the, strand temperature distribution relies on the classical Bennon-Incropera mixture continuum formulation for the convective-diffusive heat transport. Fluid flow effects are considered through enhanced thermal conductivity of the liquid phase. The model is solved by the operator-splitting Crank-Nicolson finite volume version of the Voller-Swaminathan scheme. The solution procedure is coupled with the IDS solidification analysis package. Boundary conditions for the strand cooling rely on combination of the generic and plant-specific mold, spray, and roll heat transfer coefficients ' by coping the effects of the casting speed, casting temperature, casting powder, mold oscillation mode, steel grade, water or air-mist spray type, running water, trapped water on the rolls, and radiation. The caster safety and productivity and the quality of the strand are related to nine empirical metallurgical criteria. The criteria ensure the caster safety and economy as well limit the internal, longitudinal, and transversal surface cracking possibility. The problem of finding optimum casting parameter values is stated as a minimization problem based on the analytical form of the cooling criteria evaluated from the steady-state temperature distribution of the strand. To solve the optimization problem, a genetic algorithm is employed. The algorithm explores process parameters settings heuristically by applying the principles of biological evolution. It starts with a random set of solutions and iteratively improves them through selection and variation. Solutions are evaluated by running the heat transfer model of the process and extracting the values of the empirical metallurgical criteria.