Microstructural Modelling Applied to the Design and Control of the Steel Plate Rolling Process

Modern steel plate manufacture relies on the manipulation of recrystallisation and grain growth in order to obtain maximised and precise mechanical properties from low-alloyed feedstock. The models that describe this metallurgical process nowadays inform the design of the plant itself. They are also capable of application to the on-line control of rolling, as well as to several of the ancillary mill operations. This paper will consist of a series of four case-study examples showing how the models are being used, and how future refinements can be expected to contribute further. The first example concerns slab thickness and its influence on thick plate properties. Slab thickness selection is a trade-off between sufficiency of strain (and hence austenite grain refinement) and manufacturing economics (time in process and energy consumption). At fuller thicknesses, strain homogeneity considerations also come into play. The second example concerns thermomechanical controlled rolling, by which plates are rolled in (usually) two phases. The 'hold' or break in rolling is to ensure finishing with recrystallisation suppressed, but the penalty for doing this is lost utilisation of the mill. Utilisation can be recovered by rolling several plates at once in a marshalled sequence, but this practice has to be anticipated in the mill layout and requires a sophisticated on-line prediction of the influence of temperature-strain-path on tensile strength and toughness. The third example takes the process forward into post-rolling cooling. Modern systems can apply high rates of cooling without degradation of flatness, but their full exploitation calls for accurate modelling of both grain refinement and precipitation mechanisms. All three of these areas of the process are nowadays overseen by an integrated model-based controller. The same metallurgical models also feed forward into the area of the fourth example, the prediction of final properties. Model-based accreditation is already a reality in strip rolling. The virtual test-house is still a future goal for the plate mill, not because the metallurgical processes are too complex to model, but because the application safety standards in such products as ship-plate, linepipe and construction are necessarily more exacting. The paper will contain results demonstrating that the 'microstructure monitor' for the plate mill is nonetheless already more accurate than the physical testing process, suggesting that its adoption by the industry will only be a matter of time. In addition to these specific technical topics, the paper will also review how Siemens VAI has developed its microstructural modelling capability over time, through university and research institute liaison as well as field experimentation and simulation, and how the models have been assimilated and refined in two rather different engineering environments of design office and on-site process commissioning.