Optimisation of final plate flatness by set-up coordination for subsequent manufacturing process (FinalPlateFlatness) : final report

Nowadays the flatness performance of plate manufacturing lines is still tackled in terms of enhanced local flatness of the plates in the line. The aims of this project had been to optimise local flatness approaches and to progress from this basis by taking the flatness requirements of the subsequent production quantitatively into account in form of a co-ordinated flatness set-up system. This system should integrate all sub models including models of transfer conditions from process stage to process stage and perform a line through process flatness prediction. Based on this a line set-up optimiser connecting local stage rules by a line optimisation strategy should generate the best fitting actions relevant to the evolution of flatness. The plate processing comprises rolling, cooling, hot and cold levelling quenching and stacking, whereas the process stages in each individual process route may differ, being product related. The development, enhancement and extension of process models to predict the flatness produced at each processing stage in dependence of plate data, entry flatness, main process characteristics and setup values have been investigated comprehensively. Analyses of the gained plate flatness in the various process routes showed that the reduction / evolution of flatness defects is nearly independent from the considered process route. Furthermore it has to be confirmed that flatness defects are not passed through the processing chain. This is the main reason which makes it impossible to create global flatness predictions and coordinated optimisations. As a fundamental prerequisite for global flatness control is that various inputs (e.g. changes of rolling process) would have a strongly correlated impact on final flatness. However, this correlations could not been identified on the basis of the comprehensive process data analysis. Therefore the development and implementation of the 'Global through-process flatness predictor and coordinated optimiser' is obsolete by physical reasons.