Heavy Plate Plan View Pattern Control Model Study And Optimization

The heavy plate produces are widely used in many fields such as pipelines of the petroleum and natural gas, shipping, bridge, container, construction and so on. Increasing product yields by reducing crop losses is one of the most importance ways to save energy consumption and reach cost down purpose, which largely and directly affects the business interests. The objective of the plan view control is to achieve the final plate pattern as rectangular as possible, thus improving plate yields. For the new heavy plate project of Baosteel, it is the essential task to learn and optimize the plan view control to improve the rolling yields of the Baosteel Heavy Plant.Based on the study of the preliminary plan view control model of Baosteel, physical rolling simulation is done via the simulation laboratorial heavy plate mill of the Baosteel Technology Center. According to the one pass rolling experiment, the plate formation on the flat rolling is shown, exponential function is used to describe the plate edge outline, subsection parabola function is used to describe the plate head outline, the relative between spread parameters and rolling draft is determined. According to the edging experiment, the characterized formation parameters such as the“dog bone”height, the head/tail width losses, the head/tail length losses, the“fish tail”length and so on, are analyzed. According to the MAS rolling experiment, the volume distribution of plate head/tail and edge in MAS rolling case is determined, considering the thickness overspread at the next flat rolling, the affect of the extreme uneven edging thickness is clarity. According the edging and the next flat rolling experiment, the affect of plate plan view in the flat rolling after edging pass is clarity, the edging spread is validated. These all jobs show plate volume distribution theory in Baosteel plan view control models validated. Based on elasto-plastic finite element method, the plate produce of the whole process is simulated, checking the stress, strain distribution, and geometry formation, to draw the similar conclusion with the physical simulation rolling.On the plate site, some disadvantage of the online plan view pattern control model are discovered, the main problem is about width accuracy. So objective to improve the plate width accuracy, some jobs are done to optimize the online models, including spread formula improvement, designing the spread parameters’subsets, new slab dimension checking function, new spread adaptation control. To stratify the precondition of the spread formula that the volume distribution factor is stable in one deformation process, the one big draft pass is divided into many tiny draft passes, then the formerly spread formula is used to calculate each tiny pass spread, and at finally joint the all tiny spreads into one spread calculation result. In some special aspect ratio case, the volume distribution parameters are not very suitable, designing the reasonable aspect ration subsets to set difference parameters, based on the rolling process data statistic. Using POG measure data and measure weight in front of the reheat furnaces, the slab dimension checking function is added, to correct the slab PDI information, as a forward feedback method, to improve the accuracy. A new spread adaptation control is given, and secant method is used to get available spread correction factor, while the exponential smoothing for this factor is implemented. After the optimized jobs are done, the affect is very evident, that the width accuracy is improved highly.