Thermal Behaviour of Steel Plate during Accelerated Cooling

Abstract

Recent trends in the production of high strength steel plate call for increasingly sophisticated thermo-mechanical treatment schedules, including the use of high rate accelerated cooling after finish rolling in order to achieve the desired microstructure and mechanical properties. Achieving the necessary cooling process control accuracy in such cases requires a sound understanding and description of the interactions between external heat transfer processes and changes in internal energy due to phenomena such as solid-state phase transformations. The thermal physical properties of the evolving microstructures of complex phase and martensitic steels vary greatly, and are strongly dependent on temperature and constituent phases. As a result, critical parameters such as thermal diffusivity cannot be accurately estimated without appropriate linkage to both phase transformation kinetics and temperature. In the present study, a numerical simulation has been developed to investigate the unsteady heat transfer and phase transformation behaviour of a moving steel plate during accelerated cooling. The simulation includes semi-empirical microstructure evolution sub-models, fitted to measured CCT data using non-linear regression. These are coupled to thermal-physical properties sub-models and thermal conduction calculations. A comprehensive suite of thermal boundary condition models which account for direct water cooling, forced convection film boiling, air cooling, radiation and heat transfer between plate and transport rollers are also included. The required equations for the plate temperature and microstructure evolution are solved numerically using a cell centred finite volume method, and the model has been validated by comparing simulated cooling stop temperatures with measurements obtained on the plate cooling section of an industrial plate mill. The predicted cooling stop temperatures of steel plates for different thicknesses, velocities and water flow rates are in good agreement with plant operational data.